Issuu on Google+

feature. When DDC communication is available, the monitor can send signals to the operating system that indicate which refresh rates it supports, as well as other display information; this data is reflected by the Display Properties sheet for that monitor. Monitors that don't support PnP configuration via DDC can be configured with an .INF (information) file, just as with other Windows-compatible devices. This might be supplied with a setup disk or can be downloaded from the monitor vendor's Web site.


Note Because monitors are redrawing the screen many times per second, the change in a noninterlaced screen display is virtually invisible to the naked eye, but it is very obvious when computer screens are photographed, filmed, or videotaped. Because these cameras aren't synchronized to the monitor's refresh cycle, it's inevitable that the photo, film, or videotape will show the refresh in progress as a line across the picture. If you need to capture moving images from a monitor to videotape, use a video card with a TV-out option to send your picture to a VCR.

In my experience, a 60Hz vertical scan frequency (frame rate) is the minimum anybody should use, and even at this frequency, most people notice a flicker. Especially on a larger display, onscreen flicker can cause eyestrain and fatigue. If you can select a frame rate (vertical scan frequency) of 72Hz or higher, most people are not able to discern any flicker; 72Hz is the minimum refresh rate I recommend. Most modern mid-range or better displays easily handle vertical frequencies up to 85Hz or more at resolutions up to 1024x768. This greatly reduces the flicker a user sees. However, note that increasing the frame rate, although it improves the quality of the image, can also slow down the video hardware because it now needs to display each image more times per second. If you're a gamer, slower frame rates can reduce your score. In general, I recommend that you set the lowest frame rate you find comfortable. To adjust the video card's refresh rate with Windows 9x/Me/2000/XP, use the Display icon in Control Panel. Depending on your flavor of Windows, the refresh rates supported by the video card will appear on one of the Display tabs. Optimal is the default setting, but this really is a "safe" setting for any monitor. Select a refresh rate of at least 72Hz or higher to reduce or eliminate flicker. Click Apply for the new setting to take effect. If you choose a refresh rate other than Optimal, you might see a warning about possible monitor damage. This is a warning you should take seriously, especially if you don't have detailed information about your monitor available. You can literally smoke a monitor if you try to use a refresh rate higher than the monitor is designed to accept. Before you try using a custom refresh rate, do the following: Make sure Windows has correctly identified your monitor as either a Plug and Play monitor or by brand and model. Check the manual supplied with the monitor (or download the statistics) to determine which refresh rates are supported at a given resolution. As in the example listed earlier, low-cost monitors often don't support high refresh rates at higher resolutions. Click OK to try the new setting. The screen changes to show the new refresh rate.


If the screen display looks scrambled, wait a few moments and the screen will be restored to the previous value; you'll see a dialog box asking whether you want to keep the new setting. If the display was acceptable, click Yes; otherwise, click No to restore your display. If the screen is scrambled and you can't see your mouse pointer, just press the Enter key on your keyboard because No is the default answer. With some older video drivers, this refresh rate dialog box is not available. Get an updated video driver, or check with the video card vendor for a separate utility program that sets the refresh rate for you. If you have a scrambled display with a high refresh rate, but you think the monitor should be capable of handling the refresh rate you chose, you might not have the correct monitor selected. To check your Windows 9x/Me/2000/XP monitor selection, check the Display Properties dialog box. If your monitor is listed as Standard VGA, Super VGA, or Default Monitor, Windows is using a generic driver that will work with a wide variety of monitors. However, this generic driver doesn't support refresh rates above 75Hz because some monitors could be damaged by excessive refresh rates. In some cases, you might need to manually select the correct monitor brand and model in the Windows Display Properties dialog box. If you don't find your brand and model of monitor listed, check with your monitor vendor for a driver specific for your model. After you install it, see whether your monitor will safely support a higher refresh rate.

Horizontal Frequency Different video resolutions use different horizontal frequencies. For example, the standard VGA resolution of 640x480 requires a horizontal resolution of 31.5KHz, whereas the 800x600 resolution requires a vertical frequency of at least 72Hz and a horizontal frequency of at least 48KHz. The 1024x768 image requires a vertical frequency of 60Hz and a horizontal frequency of 58KHz, and the 1280x1024 resolution requires a vertical frequency of 60Hz and a horizontal frequency of 64KHz. If the vertical frequency increases to 75Hz at 1280x1024, the horizontal frequency must be 80KHz. For a super-crisp display, look for available vertical frequencies of 75Hz or higher and horizontal frequencies of up to 90KHz or more. My favorite 17'' NEC monitor supports vertical resolutions of up to 75Hz at 1600x1200 pixels, 117Hz at 1024x768, and 160Hz at 640x480! Virtually all the analog monitors on the market today are, to one extent or another, multiple-frequency. Because literally hundreds of manufacturers produce thousands of monitor models, it is impractical to discuss the technical aspects of each monitor model in detail. Suffice it to say that before investing in a monitor, you should check the technical specifications to ensure that the monitor meets your needs. If you are looking for a place to start, check out some of the


magazines that periodically feature reviews of monitors. If you can't wait for a magazine review, investigate monitors at the Web sites run by any of the following vendors: IBM, Sony, NEC-Mitsubishi, and ViewSonic. Each of these manufacturers creates monitors that set the standards by which other monitors can be judged. Although you typically pay a bit more for these manufacturers' monitors, they offer a known, high level of quality and compatibility, as well as service and support. Note that most monitor companies sell several lines of monitors, varying by refresh rates, CRT type, antiglare coatings, energy efficiency, and warranties. For best results at resolutions of 1024x768 and above, avoid the lowest-cost 17'' monitors because these models tend to produce fuzzy onscreen displays with low refresh rates.

Controls Most of the newer CRT monitors and LCD panels use digital controls instead of analog controls. This has nothing to do with the signals the monitor receives from the computer, but only the controls (or lack of them) on the front panel that enable you to adjust the display. Monitors with digital controls have a built-in menu system that enables you to set parameters such as brightness (which adjusts the black level of the display), contrast (which adjusts the luminance of the display), screen size, vertical and horizontal shifts, color, phase, and focus. A button brings the menu up onscreen, and you use controls to make menu selections and vary the settings. When you complete your adjustments, the monitor saves the settings in nonvolatile RAM (NVRAM) located inside the monitor. This type of memory provides permanent storage for the settings with no battery or other power source. You can unplug the monitor without losing your settings, and you can alter them at any time in the future. Digital controls provide a much higher level of control over the monitor and are highly recommended.


Tip Digital video engineer Charles Poynton's notes on adjusting brightness and contrast controls provide an excellent tutorial on the use of these often misunderstood monitor adjustments. Find them online at http://www.vision.ee.ethz.ch/~buc/brechbuehler/mirror/color/Poynton-color.html.

Digital controls make adjusting CRT monitors suffering from any of the geometry errors shown in Figure 15.7 easy. Before making these adjustments, be sure the vertical and horizontal size and position are correct.

Figure 15.7. Typical geometry errors in CRT monitors; these can be corrected on most models that have digital picture controls.


Tip Get a monitor with positioning and image controls that are easy to reach, preferably on the front of the case. Look for more than just basic contrast and brightness controls; a good monitor should enable you to adjust the width and height of your screen images and the placement of the image on the screen. The monitor should also be equipped with a tilt-swivel stand so you can adjust the monitor to the best angle for your use.

Although LCD panels aren't affected by geometry errors as CRT monitors can be, they can have their own set of image-quality problems, especially if they use the typical 15-pin analog VGA video connector. Pixel jitter and pixel swim (in which adjacent pixels turn on and off) are relatively common problems that occur when using an LCD monitor connected to your PC with an analog VGA connector.

Environment One factor you might not consider when shopping for a monitor is the size and strength of the desk on which you intend to put it. Although many 17'' CRT monitors use less desk space than before, reducing the 18''​24'' depth used by older models to a more reasonable 16''​17'' depth, these monitors are still relatively heavy at around 35​40 lbs. 21'' and larger monitors can be truly huge, weighing in at 60 lbs. or more! Some of the rickety computer stands and mechanical arms used to keep monitors off the desktop might not be capable of safely holding or supporting a large unit. Check the weight rating of the computer stand or support arm you're planning to use before you put a CRT monitor on it. It's tragic to save a few dollars on these accessories, only to watch them crumple under the weight of a large monitor and wipe out your monitor investment.


Tip If you are using a relatively narrow computer table but don't want to use an LCD panel, look for a monitor that uses the so-called short-neck or short-depth CRTs. These short-neck CRTs are used in many recent 17'' and 19'' models and allow the monitor to take up less space front to back; this often is referred to as a smaller footprint. Some 17'' short-neck models use no more front-to-back space than typical 15'' models and weigh less.

Another important consideration is the lighting in the room in which you will use the monitor. The appearance of a CRT display in the fluorescent lighting of an office is markedly different from that in your home. The presence or absence of sunlight in the room also makes a big difference. Office lighting and sunlight can produce glare that becomes incredibly annoying when you are forced to stare at it for hours on end. You can reduce onscreen glare by choosing monitors equipped with flat-square or other flat CRT technologies or LCD panels and antiglare coatings. You can retrofit aftermarket filters to monitors that lack these features to help reduce glare.


Tip If you are extremely short of space or have relatively light-duty computer furniture, consider 15'' LCD display panels, which weigh only 10 lbs. or so and feature a much narrower front-to-back footprint than 17'' CRTs.

Testing a Display Unlike most of the other peripherals you can connect to your computer, you can't really tell whether a monitor suits you by examining its technical specifications. Price might not be a reliable indicator either. Testing monitors is a highly subjective process, and it is best to "kick the tires" of a few at a dealer showroom or in the privacy of your home or office (if the dealer has a liberal return policy). Testing should also not be simply a matter of looking at whatever happens to be displayed on the monitor at the time. Many computer stores display movies, scenic photos, or other flashy graphics that are all but useless for a serious evaluation and comparison. If possible, you should look at the same images on each monitor you try and compare the manner in which they perform a specific series of tasks. Before running the tests listed here, set your display to the highest resolution and refresh rate allowed by your combination of display and graphics card. One good series of tasks is as follows: Draw a perfect circle with a graphics program. If the displayed result is an oval, not a circle, this monitor will not serve you well with graphics or design software. Using a word processor, type some words in 8- or 10-point type (1 point equals 1/72''). If the words are fuzzy or the black characters are fringed with color, select another monitor. Display a screen with as much white space as possible and look for areas of color variance. This can indicate a problem with only that individual unit or its location, but if you see it on more than one monitor of the same make, it might indicate a manufacturing problem; it could also indicate problems with the signal coming from the graphics card. Move the monitor to another system equipped with a different graphics card model and retry this test to see for certain whether it's the monitor or the video card. Display the Microsoft Windows desktop to check for uniform focus and


brightness. Are the corner icons as sharp as the rest of the screen? Are the lines in the title bar curved or wavy? Monitors usually are sharply focused at the center, but seriously blurred corners indicate a poor design. Bowed lines can be the result of a poor video adapter, so don't dismiss a monitor that shows those lines without using another adapter to double-check the effect. Adjust the brightness up and down to see whether the image blooms or swells, which indicates the monitor is likely to lose focus at high brightness levels. You can also use diagnostics that come with the graphics card or thirdparty system diagnostics programs to perform these tests. With LCD displays in particular, change to a lower resolution from the panel's native resolution using the Microsoft Windows Display properties settings. Because LCD panels have only one native resolution, the display must use scaling to handle other resolutions full-screen. If you are a Web designer, are a gamer, or must capture screens at a particular resolution, this test will show you whether the LCD panel produces acceptable display quality at resolutions other than normal. You can also use this test on a CRT, but CRTs, unlike LCD panels, are designed to handle a wide variety of resolutions. A good monitor is calibrated so that rays of red, green, and blue light hit their targets (individual phosphor dots) precisely. If they don't, you have bad convergence. This is apparent when edges of lines appear to illuminate with a specific color. If you have good convergence, the colors are crisp, clear, and true, provided there isn't a predominant tint in the phosphor. If the monitor has built-in diagnostics (a recommended feature), try them as well to test the display independently of the graphics card and system to which it's attached.


Maintaining Your Monitor Because a good 17'' or larger monitor can be used for several years on more than one computer, proper care is essential to extend its life to the fullest extent. Use the following guidelines for proper care of your monitors: Although phosphor burn (in which an image left onscreen eventually leaves a permanent shadow onscreen) is next-to-impossible with VGA-type displays​unlike the old TTL displays​screensavers are still useful for casual security. You can password-protect your system with both the standard Windows screensaver and third-party programs (although a determined snoop can easily thwart screensaver password protection). Windows includes several screensavers that can be enabled via the Display Control Panel. A bevy of free and inexpensive screensavers is available for download from the Internet. Keep in mind, though, that add-on screensavers can cause crashes and lockups if they're poorly written or out-of-date. Use screensavers written for your particular operating system version to minimize problems. To prevent premature failure of the monitor's power supply, use the powermanagement feature of the Display Properties or Power (Management) sheet to put the monitor into a low-power standby mode after a reasonable period of inactivity (10​15 minutes) and to turn it off after about 60 minutes. Using the power management feature is far better than using the on/off switch when you are away from the computer for brief periods. Turn off the monitor only at the end of your computing "day." How can you tell whether the monitor is really off or in standby mode? Look at the power LCD on the front of the monitor. A monitor that's in standby mode usually has a blinking green or solid amber LCD in place of the solid green LCD displayed when it's running in normal mode. Because monitors in standby mode still consume some power, they should be shut off at the end of the computing day. If the monitor will not go into standby when the PC isn't sending signals to it, make sure the monitor is properly defined in Windows's Display Properties sheet. In addition, the Energy Star check box should be selected for any monitor that supports power management, unless the monitor should be left on at all times (such as when used in a retail kiosk or self-standing display). Make sure the monitor has adequate ventilation along the sides, rear, and top. Because monitors use passive cooling, a lack of adequate airflow caused by


piling keyboards, folders, books, or other office debris on top of the monitor will cause it to overheat and considerably shorten its life. If you're looking at a monitor with a partly melted grille on the top of the case, you're looking at a victim of poor cooling. If you need to use a monitor in an area with poor airflow, use an LCD panel instead of a CRT because LCDs run much cooler than CRTs. The monitor screen and case should be kept clean. Turn off the power, spray a cleaner such as Endust for Electronics onto a soft cloth (never directly onto the monitor!), and wipe the screen and the case gently. If your CRT monitor has a degaussing button or feature, use it periodically to remove stray magnetic signals. Keep in mind that CRTs have powerful magnets around the picture tube, so keep magnetic media away from them.


Video Display Adapters A video adapter provides the interface between your computer and your monitor and transmits the signals that appear as images on the display. Throughout the history of the PC, there have been a succession of standards for video display characteristics that represent a steady increase in screen resolution and color depth. The following list of standards can serve as an abbreviated history of PC video-display technology: MDA (Monochrome Display Adapter)

VGA (Video Graphics Array)

HGC (Hercules Graphics Card)

SVGA (Super VGA)

CGA (Color Graphics Adapter)

XGA (Extended Graphics Array)

EGA (Enhanced Graphics Adapter)

IBM pioneered most of these standards, but other manufacturers of compatible PCs adopted them as well. Today, IBM is no longer the industry leader it once was (and hasn't been for some time), and many of these standards are obsolete. Those that aren't obsolete seldom are referred to by these names anymore. The sole exception to this is VGA, which is a term that is still used to refer to a baseline graphics display capability supported by virtually every video adapter on the market today. When you shop for a video adapter today, you are more likely to see specific references to the screen resolutions and color depths that the device supports than a list of standards such as VGA, SVGA, XGA, and UVGA. However, reading about these standards gives you a good idea of how video-display technology developed over the years and prepares you for any close encounters you might have with legacy equipment from the dark ages. Today's VGA and later video adapters can also display most older color graphics software written for CGA, EGA, and most other obsolete graphics standards. This enables you to use older graphics software (such as games and educational programs) on your current system. Although not a concern for most users, some older programs wrote directly to hardware registers that are no longer found on current video cards.

Obsolete Display Adapters


Although many types of display systems were at one time considered to be industry standards, few of these are viable standards for today's hardware and software.


Note If you are interested in reading more about MDA, HGC, CGA, EGA, or MCGA display adapters, see Chapter 8 of Upgrading and Repairing PCs, 10th Anniversary Edition, included on the DVD with this book.

Current Display Adapters When IBM introduced the PS/2 systems on April 2, 1987, it also introduced the VGA display. On that day, in fact, IBM also introduced the lower-resolution MCGA and higher-resolution 8514 adapters. The MCGA and 8514 adapters did not become popular standards like the VGA did, and both were discontinued. All current display adapters that connect to the 15-pin VGA analog connector or the DVI analog/digital connector are based on the VGA standard.

Digital Versus Analog Signals Unlike earlier video standards, which are digital, the VGA is an analog system. Why have displays gone from digital to analog when most other electronic systems have gone digital? Compact disc players (digital) have replaced most turntables (analog), mini DV camcorders are replacing 8MM and VHS-based analog camcorders, and TiVo and UltimateTV digital video recorders are performing time-shifting in place of analog VCRs for many users. With a digital television set, you can watch several channels on a single screen by splitting the screen or placing a picture within another picture. Most personal computer displays introduced before the PS/2 are digital. This type of display generates different colors by firing the RGB electron beams in on-or-off mode, which allows for the display of up to eight colors (23). In the IBM displays and adapters, another signal doubles the number of color combinations from 8 to 16 by displaying each color at one of two intensity levels. This digital display is easy to manufacture and offers simplicity with consistent color combinations from system to system. The real drawback of the older digital displays such as CGA and EGA is the limited number of possible colors. In the PS/2 systems, IBM went to an analog display circuit. Analog displays work like the digital displays that use RGB electron beams to construct various colors, but each color in the analog display system can be displayed at varying levels of intensity​64 levels, in the case of the VGA. This versatility provides 262,144 possible colors (643), of which 256 could be simultaneously displayed. For realistic computer graphics, color depth is often more important than high resolution because the human eye perceives a picture that has more colors as being more


realistic. IBM moved to analog graphics to enhance the color capabilities of its systems.

Video Graphics Array PS/2 systems incorporated the primary display adapter circuitry onto the motherboard, and both IBM and third-party companies introduced separate VGA cards to enable other types of systems to enjoy the advantages of VGA. Although the IBM MicroChannel (MCA) computers, such as the PS/2 Model 50 and above, introduced VGA, it's impossible today to find a brand-new replacement for VGA that fits into the obsolete MCA-bus systems. However, surplus and used third-party cards might be available if you look hard enough. The VGA BIOS is the control software residing in the system ROM for controlling VGA circuits. With the BIOS, software can initiate commands and functions without having to manipulate the VGA directly. Programs become somewhat hardware independent and can call a consistent set of commands and functions built into the system's ROM-control software. See "Video Adapter BIOS," p. 487.

Other implementations of the VGA differ in their hardware but respond to the same BIOS calls and functions. New features are added as a superset of the existing functions, and VGA remains compatible with the graphics and text BIOS functions built into the PC systems from the beginning. The VGA can run almost any software that originally was written for the CGA or EGA, unless it was written to directly access the hardware registers of these cards. A standard VGA card displays up to 256 colors onscreen, from a palette of 262,144 (256KB) colors; when used in the 640x480 graphics or 720x400 text mode, 16 colors at a time can be displayed. Because the VGA outputs an analog signal, you must have a monitor that accepts an analog input. VGA displays originally came not only in color, but also in monochrome VGA models, which use color summing. With color summing, 64 gray shades are displayed instead of colors. The summing routine is initiated if the BIOS detects a monochrome display when the system boots. This routine uses an algorithm that takes the desired color and rewrites the formula to involve all three color guns, producing varying intensities of gray. Users who preferred a monochrome display,


therefore, could execute color-based applications.


Note For a listing of the VGA display modes supported by the original IBM VGA card (and thus all subsequent VGA-type cards), see "VGA Display Modes" in Chapter 15 of Upgrading and Repairing PCs, 12th Edition, available in electronic form on the DVD-ROM supplied with this book.

Even the least-expensive video adapters on the market today can work with modes well beyond the VGA standard. VGA, at its 16-color, 640x480 graphics resolution, has come to be the baseline for PC graphical display configurations. VGA is accepted as the least common denominator for all Windows systems and must be supported by the video adapters in all systems running Windows. The installation programs of all Windows versions use these VGA settings as their default video configuration. In addition to VGA, virtually all adapters support a range of higher screen resolutions and color depths, depending on the capabilities of the hardware. If a Windows 9x/Me or Windows XP/2000 system must be started in Safe Mode because of a startup problem, the system defaults to VGA in the 640x480, 16-color mode. Windows 2000 and Windows XP also offer a VGA Mode startup that also uses this mode (Windows XP uses 800x600 resolution) but doesn't slow down the rest of the computer the way Safe Mode (which replaces 32-bit drivers with BIOS services) does. IBM introduced higher-resolution versions of VGA called XGA and XGA-2 in the early 1990s, but most of the development of VGA standards has come from the third-party video card industry and its trade group, the Video Electronic Standards Association (VESA).


Note If you are interested in reading more about the XGA and XGA-2 display adapters, see "XGA and XGA-2" in Chapter 8 of Upgrading and Repairing PCs, 10th Anniversary Edition, included on the DVD with this book.

Super VGA When IBM's XGA and 8514/A video cards were introduced, competing manufacturers chose not to attempt to clone these incremental improvements on their VGA products. Instead, they began producing lower-cost adapters that offered even higher resolutions. These video cards fall into a category loosely known as Super VGA (SVGA). SVGA provides capabilities that surpass those offered by the VGA adapter. Unlike the display adapters discussed so far, SVGA refers not to an adapter that meets a particular specification, but to a group of adapters that have different capabilities. For example, one card might offer several resolutions (such as 800x600 and 1024x768) that are greater than those achieved with a regular VGA, whereas another card might offer the same or even greater resolutions but also provide more color choices at each resolution. These cards have different capabilities; nonetheless, both are classified as SVGA. The SVGA cards look much like their VGA counterparts. They have the same connectors, but because the technical specifications from different SVGA vendors vary tremendously, it is impossible to provide a definitive technical overview in this book. The connector is shown in Figure 15.8; the pinouts are shown in Table 15.5.

Figure 15.8. VGA connector used for SVGA and other VGA-based standards.

Table 15.5. Standard 15-Pin VGA Connector Pinout


Table 15.5. Standard 15-Pin VGA Connector Pinout Pin #

Function

Direction

1

Red video

Out

2

Green video

Out

3

Blue video

Out

4

Monitor ID 2

In

5

TTL Ground (monitor self-test)

6

Red analog ground

7

Green analog ground

8

Blue analog ground

9

Key (plugged hole)

10

Synch Ground

11

Monitor ID 0

In

12

Monitor ID 1

In

13

Horizontal Synch

Out

14

Vertical Synch

Out

15

Monitor ID 3

In

On the VGA cable connector that plugs into your video adapter, pin 9 is often pinless. Pin 5 is used only for testing purposes, and pin 15 is rarely used; these are often pinless as well. To identify the type of monitor connected to the system, some manufacturers use the presence or absence of the monitor ID pins in various combinations.

VESA SVGA Standards The Video Electronics Standards Association includes members from various companies associated with PC and computer video products. In October 1989, VESA recognized that programming applications to support the many SVGA cards


on the market was virtually impossible and proposed a standard for a uniform programmer's interface for SVGA cards; it is known as the VESA BIOS extension (VBE). VBE support might be provided through a memory-resident driver (used by older cards) or through additional code added to the VGA BIOS chip itself (the more common solution). The benefit of the VESA BIOS extension is that a programmer needs to worry about only one routine or driver to support SVGA. Various cards from various manufacturers are accessible through the common VESA interface. Today, VBE support is a concern primarily for real-mode DOS applications, usually older games, and for non-Microsoft operating systems that need to access higher resolutions and color depths. VBE supports resolutions up to 1280x1024 and color depths up to 24-bit (16.8 million colors), depending on the mode selected and the memory on the video card. VESA compliance is of virtually no consequence to Windows versions 95 and up. These operating systems use custom video drivers for their graphics cards.


Note For a listing of VESA BIOS modes by resolution, color depth, and scan frequency, see "VESA SVGA Standards" in the Technical Reference portion of the DVD accompanying this book.

Integrated Video/Motherboard Chipsets Although built-in video has been a staple of low-cost computing for a number of years, until the late 1990s most motherboard-based video simply moved the standard video components discussed earlier in this chapter to the motherboard. Many low-cost systems​especially those using the semiproprietary LPX motherboard form factor​have incorporated standard VGA-type video circuits on the motherboard. The performance and features of the built-in video differed only slightly from add-on cards using the same or similar chipsets, and in most cases the built-in video could be replaced by adding a video card. Some motherboardbased video also had provisions for memory upgrades. See "LPX," p. 201.

However, in recent years the move toward increasing integration on the motherboard has led to the development of chipsets that include 3D accelerated video and audio support as part of the chipset design. In effect, the motherboard chipset takes the place of most of the video card components listed earlier and uses a portion of main system memory as video memory. The use of main system memory for video memory is often referred to as unified memory architecture (UMA), and although this memory-sharing method was also used by some built-in video that used its own chipset, it has become much more common with the rise of integrated motherboard chipsets. The pioneer of integrated chipsets containing video (and audio) features was Cyrix. While Cyrix was owned by National Semiconductor, it developed a two-chip product called MediaGX. MediaGX incorporated the functions of a CPU, memory controller, sound, and video and made very low-cost computers possible (although with performance significantly slower than that of Pentium-class systems with similar clock speeds). National Semiconductor retained the MediaGX after it sold Cyrix to VIA Technologies. National Semiconductor went on to develop improved versions of the MediaGX, called the Geode GX1 and Geode GX2, for use in thin clients (a terminal that runs Windows and has a high-res display), interactive settop boxes, and other embedded devices.


Intel became the next major player to develop integrated chipsets, with its 810 chipset for the Pentium III and Celeron processors. The 810 (codenamed Whitney before its official release) heralded the beginning of widespread industry support for this design. Intel later followed the release of the 810 series (810 and 810E) with the 815 series, most of which also feature integrated video. The now-obsolete 810 chipset family supports relatively low-performance (PCIequivalent) integrated graphics, whereas the 815 series chipsets with integrated video (815, 815E, 815EG, and 815G) support AGP-equivalent integrated 3D graphics with i740-class performance. Motherboards with the 815 or 815E chipset can also be equipped with an optional AGP slot for video card upgrades, whereas the 815P and 815EP chipsets lack integrated video and are used with separate AGP cards. Both the 810 and 815 series chipsets are designed to support recent and current Intel processors, such as the Pentium III and Celeron in their Socket 370 form factors. Both the 810 family and 815 family chipsets are two-piece sets: One chip contains the Graphics Memory Controller Hub, which replaces the traditional North Bridge chip and adds integrated video, and the other chip contains the I/O Controller Hub, which replaces the traditional South Bridge. Intel has continued to produce two-piece chipsets with integrated video for the Pentium 4 processor with the 845G series. All 845G-series chipsets integrate AGP 2x/4x Intel Extreme Graphics video into the Graphics Controller Hub, whereas the 845G and GE chipsets also feature support for an external AGP 4x graphics adapter. See "Intel 810, 810E, and 810E2," p. 251, "Intel 815 Family," p. 254, and "Intel 845 Family," p. 274.

Intel is far from alone in its support for integrated chipsets; most other major chipset vendors also have developed similar integrated chipsets for use in lowcost computers and motherboards using both Intel and AMD CPUs, as shown in Table 15.6. Table 15.6. Major Non-Intel Integrated Chipsets

Vendor

Chipset

Processors Supported

Pentium VIA VIA Apollo III/Celeron/VIA C3 Technologies PLE133 (Socket 370)

Number of Chips in Chipset

2

Notes

Incorporates Trident Blade3D graphics


Pentium VIA VIA Apollo III/Tualatin/Celeron/VIA 2 Technologies PLE133T C3 (Socket 370)

Incorporates Trident Blade3D graphics

VIA ProSavage Pentium III, Celeron, Technologies PL133 VIA C3

2

Incorporates S3 Savage 4 3D graphics; PC133 SDRAM

VIA ProSavage Pentium III/Tualatin, Technologies PL133T Celeron, VIA C3

2

Incorporates S3 Savage 4 3D graphics with optional LCD, TV-out; PC133 SDRAM

VIA ProSavage Pentium III, Celeron, Technologies PM133 VIA C3

2

Incorporates S3 Savage 4 3D graphics with optional external AGP 4x; PC133 SDRAM

VIA ProSavage AMD Athlon, Duron Technologies KM133 (Socket A)

2

Incorporates S3 Savage 4 3D graphics, with optional LCD and external AGP 4x; PC133 SDRAM

VIA ProSavage Pentium 4 Technologies P4M266

2

Incorporates S3 Savage 8 3D graphics with optional external AGP 4x; DDR or standard SDRAM

VIA ProSavage AMD Athlon XP, Duron 2 Technologies KM266

Incorporates S3 ProSavage8 3D graphics with optional external AGP 4x; DDR or standard SDRAM

Silicon Integrated Systems (SiS)

SiS Pentium II/III/Celeron 2 620/5595 (Slot 1 or Socket 370)

Incorporates SiS6320 3D graphics

Silicon Integrated Systems (SiS)

SiS 630

Pentium III/Celeron (Socket 370)

1

Incorporates SiS300 3D graphics; compatible with SiS Video Bridge

Silicon Integrated Systems (SiS)

SiS 630E

Pentium III/Celeron (Socket 370)

1

Incorporates SiS300 3D graphics

Silicon Integrated Systems (SiS)

SiS 630S

Pentium III/Celeron (Socket 370)

1

Incorporates SiS300 3D graphics; also supports external AGP 4X

Silicon Integrated Systems (SiS)

SiS 730S

AMD Athlon/Duron (Socket A)

1

Incorporates 3D graphics; also supports external AGP 4X

Silicon Integrated Systems (SiS)

SiS 740

AMD Athlon/Duron (Socket A)

2

Multiple monitor and TV-out; PC133 SDRAM or DDR SDRAM

Silicon Integrated


Systems (SiS)

SiS 650

Pentium 4

2

Supports external AGP 4x; multiple monitor and TV-out

Silicon Integrated Systems (SiS)

SiS 651

Pentium 4

2

Supports external AGP 4x; multiple monitor and TV-out

Acer Labs

Aladdin TNT2

Pentium II/III/Celeron 2 (Slot A and Socket A)

Based on NVIDIA's RIVA TNT2 3D video chipset

NVIDIA

nForce

AMD Athlon/Duron (Socket A)

2

Based on GeForce 2 GPU; supports DVI and TV-out and external AGP 4x

NVIDIA

nForce2

AMD Athlon/Duron (Socket A)

2

Based on GeForce 4 MX GPU; supports DVI and TV-out and external AGP 4x

ATI

RADEON IGP 330/340

Pentium 4

2

Based on RADEON VE; 340 also supports TV-out and FSB up to 533MHz; supports ATI's IXP 200/250 (USB 2.0, 10/100 Ethernet, ATA 100; IXP 250 adds manageability features) South Bridge or third-party South Bridge

ATI

RADEON IGP 320

AMD Athlon/Duron (Socket A)

2

Based on RADEON VE supports ATI's IXP 200/250 (USB 2.0, 10/100 Ethernet, ATA 100; IXP 250 adds manageability features) South Bridge or third-party South Bridge

Although a serious 3D gamer will not be satisfied with the performance of most integrated chipsets (NVIDIA's nForce2 and ATI's RADEON IGPs being notable exceptions), business, home/office, and casual gamers will find that integrated chipset-based video​particularly those that support advanced AGP 2x and faster 3D functions​are satisfactory in performance and provide cost savings compared with separate video cards. If you decide to buy a motherboard with an integrated chipset, I recommend that you select one that also supports external AGP 4x video. This enables you to add a faster video card in the future if you decide you need it.

Video Adapter Components All video display adapters contain certain basic components, such as the following: Video BIOS. Video processor/video accelerator. Video memory.


Digital-to-analog converter (DAC). Formerly a separate chip, the DAC is usually incorporated into the video processor/accelerator chip on recent chipsets. The DAC is not necessary on a purely digital subsystem (digital video card and display), but because most display subsystems have an analog VGA video card, an analog display, or both, video cards will continue to have DAC features for some time to come. Bus connector. Video driver. Figure 15.9 indicates the locations of many of these components on a typical video card. Note that the acronym GPU refers to the graphics processing unit.

Figure 15.9. The ATI RADEON 9700 PRO is a typical example of a high-end video card optimized for gaming and dual-display support. Like most recent graphics cards, it uses a nonremovable flash BIOS (not visible in this photo). Photo courtesy of ATI Technologies, Inc.

Virtually all video adapters on the market today use chipsets that include 3D acceleration features. The following sections examine these components and features in greater detail.

The Video BIOS Video adapters include a BIOS that is similar in construction but completely separate from the main system BIOS. (Other devices in your system, such as SCSI adapters, might also include their own BIOS.) If you turn on your monitor first and look quickly, you might see an identification banner for your adapter's


video BIOS at the very beginning of the system startup process. Similar to the system BIOS, the video adapter's BIOS takes the form of a ROM (read-only memory) chip containing basic instructions that provide an interface between the video adapter hardware and the software running on your system. The software that makes calls to the video BIOS can be a standalone application, an operating system, or the main system BIOS. The programming in the BIOS chip enables your system to display information on the monitor during the system POST and boot sequences, before any other software drivers have been loaded from disk. See "BIOS Basics," p. 366.

The video BIOS also can be upgraded, just like a system BIOS, in one of two ways. The BIOS uses a rewritable chip called an EEPROM (electrically erasable programmable read-only memory) that you can upgrade with a utility the adapter manufacturer provides. On older cards, you might be able to completely replace the chip with a new one​again, if supplied by the manufacturer and if the manufacturer did not hard solder the BIOS to the printed circuit board. Most recent video cards use a surface-mounted BIOS chip rather than a socketed chip. A BIOS you can upgrade using software is referred to as a flash BIOS, and most current-model video cards that offer BIOS upgrades use this method. Video BIOS upgrades (sometimes referred to as firmware upgrades) are sometimes necessary in order to use an existing adapter with a new operating system, or when the manufacturer encounters a significant bug in the original programming. Occasionally, a BIOS upgrade is necessary because of a major revision to the video card chipset's video drivers. As a general rule, the video BIOS is a component that falls into the "if it ain't broke, don't fix it" category. Try not to let yourself be tempted to upgrade just because you've discovered that a new BIOS revision is available. Check the documentation for the upgrade, and unless you are experiencing a problem the upgrade addresses, leave it alone. See "Video Adapter BIOS," p. 487.

The Video Processor


The video processor, or chipset, is the heart of any video adapter and essentially defines the card's functions and performance levels. Two video adapters built using the same chipset often have many of the same capabilities and deliver comparable performance. Also, the software drivers that operating systems and applications use to address the video adapter hardware are written primarily with the chipset in mind. You often can use a driver intended for an adapter with a particular chipset on any other adapter using the same chipset. Of course, cards built using the same chipset can differ in the amount and type of memory installed, so performance can vary. Since the first VGA cards were developed, several main types of processors have been used in video adapters; these technologies are compared in Table 15.7. Table 15.7. Video Processor Technologies Processor Type Framebuffer

Where Video Processing Takes Place

Relative Speed

Relative Cost

How Used Today

Very slow

Very low

Obsolete; mostly ISA video cards

Graphics Video card's own processor. coprocessor

Very fast

Very high

CAD and engineering workstations

Graphics Video chip draws lines, circles, shapes; accelerator CPU sends commands to draw them.

Fast

Low to moderate

All mainstream video cards; is combined with 3D GPU on current cards

3D graphics Video card's 3D GPU (in accelerator processor chipset) renders polygons and adds (GPU) lighting and shading effects as needed.

Most price ranges Fast 2D depending on chipset, and 3D memory, and RAMDAC display speed

Computer's CPU.

All gaming optimized video cards and almost all mainstream video cards

Identifying the Video and System Chipsets Before you purchase a system or a video card, you should find out which chipset the video card or video circuit uses or, for systems with integrated chipset video, which integrated chipset the system uses. This allows you to have the following: Better comparisons of card or system to others Access to technical specifications Access to reviews and opinions


Better buying decisions Choice of card manufacturer or chipset manufacturer support and drivers Because video card performance and features are critical to enjoyment and productivity, find out as much as you can before you buy the system or video card by using the chipset or video card manufacturer's Web site and third-party reviews. Poorly written or buggy drivers can cause several types of problems, so be sure to check periodically for video driver updates and install any that become available. With video cards, support after the sale can be important. So, you should check the manufacturer's Web site to see whether it offers updated drivers and whether the product seems to be well supported. The Vendor List on the DVD has information on most of the popular video chipset manufacturers, including how to contact them. You should note that NVIDIA (the leading video chipset vendor) makes only chipsets, whereas ATI (the #2 video vendor) makes branded video cards and supplies chipsets to vendors. This means a wide variety of video cards use the same chipset; it also means that you are likely to find variations in card performance, software bundles, warranties, and other features between cards using the same chipset.

Video RAM Most video adapters rely on their own onboard memory that they use to store video images while processing them; although the AGP specification supports the use of system memory for 3D textures, this feature is seldom supported now that video cards routinely ship with 32MB, 64MB, or more of onboard memory. Many low-cost systems with onboard video use the universal memory architecture (UMA) feature to share the main system memory. In any case, the memory on the video card or borrowed from the system performs the same tasks. The amount of memory on the adapter or used by integrated video determines the maximum screen resolution and color depth the device can support. You often can select how much memory you want on a particular video adapter; for example, 32MB, 64MB, and 128MB are common choices today. Although adding more memory is not guaranteed to speed up your video adapter, it can increase the speed if it enables a wider bus (from 64 bits wide to 128 bits wide) or provides nondisplay memory as a cache for commonly displayed objects. It also enables the card to generate more colors and higher resolutions and, for AGP cards (see the following), allows 3D textures to be stored and processed on the card, rather than in slower main memory. Many types of memory have been used with video adapters. These memory types


are summarized in Table 15.8. Table 15.8. Memory Types Used in Video Display Adapters Memory Type

Definition

Relative Speed

Usage

FPM DRAM

Fast Page-Mode RAM

Slow

Low-end ISA cards; obsolete

VRAM[1]

Video RAM

Fast

Expensive; obsolete

WRAM[1]

Window RAM

Fast

Expensive; obsolete

EDO DRAM

Extended Data Out DRAM

Moderate

Low-end PCI-bus

SDRAM

Synchronous DRAM

Fast

Low-end PCI/AGP

MDRAM

Multibank DRAM

Fast

Little used; obsolete

SGRAM

Synchronous Graphics DRAM

Very fast

High-end PCI/AGP; replaced by DDR SDRAM

DDR SDRAM

Double-Data Rate SDRAM

Very fast

High-end AGP

DDR-II SDRAM

DDR SDRAM, 4-bit per

Extremely fast

High-end AGP cycle memory fetch

[1] VRAM and WRAM are dual-ported memory types that can read from one port and write data through

the other port. This improves performance by reducing wait times for accessing the videoRAMcompared to FPMDRAMandEDO DRAM.


Note To learn more about memory types used on older video cards (FPD DRAM, VRAM, WRAM, EDO DRAM, and MDRAM), see Chapter 15 of Upgrading and Repairing PCs, 12th Edition, available in electronic form on the DVD packaged with this book.

SGRAM, SDRAM, DDR, and DDR-II SDRAM​which are derived from popular motherboard memory technologies​have replaced VRAM, WRAM, and MDRAM as high-speed video RAM solutions. Their high speeds and low production costs have enabled even inexpensive video cards to have 16MB or more of high-speed RAM onboard.

SDRAM Synchronous DRAM (SDRAM) is the same type of RAM used on many current systems based on processors such as the Pentium III, Pentium 4, Athlon, and Duron. The SDRAMs found on video cards are usually surface-mounted individual chips; on a few early models, a small module containing SDRAMs might be plugged into a proprietary connector. This memory is designed to work with bus speeds up to 200MHz and provides performance just slightly slower than SGRAM. SDRAM is used primarily in current low-end video cards and chipsets such as NVIDIA's GeForce2 MX and ATI's RADEON VE.

SGRAM Synchronous Graphics RAM (SGRAM) was designed to be a high-end solution for very fast video adapter designs. SGRAM is similar to SDRAM in its capability to be synchronized to high-speed buses up to 200MHz, but it differs from SDRAM by including circuitry to perform block writes to increase the speed of graphics fill or 3D Z-buffer operations. Although SGRAM is faster than SDRAM, most video card makers have dropped SGRAM in favor of even faster DDR SDRAM in their newest products.

DDR SDRAM Double Data Rate SDRAM (also called DDR SDRAM) is the most common video RAM technology on recent video cards. It is designed to transfer data at speeds twice that of conventional SDRAM by transferring data on both the rising and falling parts of the processing clock cycle. Today's high-end and mid-range video cards based on chipsets such as NVIDIA's GeForce 4 and GeForce 3 Ti and ATI's RADEON 8000 and 7000 series use DDR SDRAM for video memory.


DDR-II SDRAM The second generation of DDR SDRAM fetches 4 bits of data per cycle, instead of 2 as with DDR SDRAM. This doubles performance at the same clock speed. The first video chipset to support DDR-II was NVIDIA's GeForce FX, which became the top of NVIDIA's line of GPUs in late 2002. For more information about DDR and DDR-II, see Chapter 6, "Memory," p. 421.

Video RAM Speed Video cards with the same type of 3D graphics processor chip (GPU) onboard might use different speeds of memory. For example, two cards that use the NVIDIA GeForce 4 Ti 4200, the ABIT Siluro and the Gainward Ultra/650XP, use different memory speeds. The ABIT card uses 3.6ns memory, whereas the Gainward card uses 4ns memory. In this case, the difference exists because the ABIT card has 64MB on board, unlike the Gainward card, which uses 128MB. Although 4ns memory is a bit slower than 3.6ns memory, the larger amount of memory on the Gainward card results in faster performance. Sometimes, video card makers also match different memory speeds with different versions of the same basic GPU, as with ATI's 8500 Pro and 8500LE. The LE-series chip has a slower core speed, enabling vendors to use slower memory and create a less expensive graphics card. Unless you dig deeply into the technical details of a particular 3D graphics card, determining whether a particular card uses SDRAM, DDR SDRAM, or SGRAM can be difficult. Because none of today's 3D accelerators feature upgradeable memory, I recommend that you look at the performance of a given card and choose the card with the performance, features, and price that's right for you.

RAM Calculations The amount of memory a video adapter needs to display a particular resolution and color depth is based on a mathematical equation. A location must be present in the adapter's memory array to display every pixel on the screen, and the resolution determines the number of total pixels. For example, a screen resolution


of 1024x768 requires a total of 786,432 pixels. If you were to display that resolution with only two colors, you would need only 1 bit of memory space to represent each pixel. If the bit has a value of 0, the dot is black, and if its value is 1, the dot is white. If you use 24 bits of memory space to control each pixel, you can display more than 16.7 million colors because 16,777,216 combinations are possible with a 4-digit binary number (224=16,777,216). If you multiply the number of pixels necessary for the screen resolution by the number of bits required to represent each pixel, you have the amount of memory the adapter needs to display that resolution. Here is how the calculation works: 1024x768

=

786432 pixelsx24 bits per pixel

=

18,874,368 bits

=

2,359,296 bytes

=

2.25MB

As you can see, displaying 24-bit color (16,777,216 colors) at 1024x768 resolution requires exactly 2.25MB of RAM on the video adapter. Because most adapters support memory amounts of only 256KB, 512KB, 1MB, 2MB, or 4MB, you would need to use a video adapter with at least 4MB of RAM onboard to run your system using that resolution and color depth. To use the higher-resolution modes and greater numbers of colors common today, you would need much more memory on your video adapter than the 256KB found on the original IBM VGA. Table 15.9 shows the memory requirements for some of the most common screen resolutions and color depths used for 2D graphics operations, such as photo editing, presentation graphics, desktop publishing, and Web page design. Table 15.9. Video Display Adapter Minimum Memory Requirements for 2D Operations Resolution

Color Depth

Max. Colors

Memory Required

Memory Used

640x480

16-bit

65,536

1MB

614,400 bytes

640x480

24-bit

16,777,216

1MB

921,600 bytes

640x480

32-bit

4,294,967,296

2MB

1,228,800 bytes

800x600

16-bit

65,536

1MB

960,000 bytes


800x00

24-bit

16,777,216

2MB

1,440,000 bytes

800x600

32-bit

4,294,967,296

2MB

1,920,000 bytes

1024x768

16-bit

65,536

2MB

1,572,864 bytes

1024x768

24-bit

16,777,216

4MB

2,359,296 bytes

1024x768

32-bit

4,294,967,296

4MB

3,145,728 bytes

1280x1024

16-bit

65,536

4MB

2,621,440 bytes

1280x1024

24-bit

16,777,216

4MB

3,932,160 bytes

1280x1024

32-bit

4,294,967,296

8MB

5,242,880 bytes

1400x1050

16-bit

65,536

8MB

2,940,000 bytes

1400x1050

24-bit

16,777,216

8MB

4,410,000 bytes

1400x1050

32-bit

4,294,967,296

16MB

5,880,000 bytes

1600x1200

16-bit

65,536

8MB

3,840,000 bytes

1600x1200

24-bit

16,777,216

8MB

5,760,000 bytes

1600x1200

32-bit

4,294,967,296

16MB

7,680,000 bytes

From this table, you can see that a video adapter with 4MB can display 65,536 colors in 1600x1200 resolution mode, but for a true color (16.8 million colors) display, you would need to upgrade to 8MB. In most cases you can't add memory to your video card​you would need to replace your current video card with a new one with more memory. 3D video cards require more memory for a given resolution and color depth because the video memory must be used for three buffers: the front buffer, back buffer, and Z-buffer. The amount of video memory required for a particular operation varies according to the settings used for the color depth and Z-buffer. Triple buffering allocates more memory for 3D textures than double-buffering but can slow down performance of some games. The buffering mode used by a given 3D video card usually can be adjusted through its properties sheet. Table 15.10 lists the memory requirements for 3D cards in selected modes. For memory sizes used by other combinations of color depth and Z-buffer depth, see the eTesting Labs' Memory Requirements for 3D Applications Web site at the


following address: www.etestinglabs.com/benchmarks/3dwinbench/d5memfor3d.asp Table 15.10. Video Display Adapter Memory Requirements for 3D Operations Resolution

640x480

800x600

Color Depth

16-bit

24-bit

32-bit

16-bit

24-bit

32-bit

1024x768 16-bit

24-bit

32-bit 1

1280x1024 16-bit

24-bit

32-bit

Z-Buffer Depth

Buffer Mode

Actual Memory Used

Onboard Video Memory Size Required

Double

1.76MB

2MB

Triple

2.34MB

4MB

Double

2.64MB

4MB

Triple

3.52MB

4MB

Double

3.52MB

4MB

Triple

4.69MB

8MB

Double

2.75MB

4MB

Triple

3.66MB

4MB

Double

4.12MB

8MB

Triple

5.49MB

8MB

Double

5.49MB

8MB

Triple

7.32MB

8MB

Double

4.12MB

8MB

Triple

5.49MB

8MB

Double

6.75MB

8MB

Triple

9.00MB

16MB

Double

9.00MB

16MB

Triple

12.00MB

16MB

Double

7.50MB

8MB

Triple

10.00MB

16MB

Double

11.25MB

16MB

Triple

15.00MB

16MB

Double

15.00MB

16MB

Triple

20.00MB

32MB

16-bit

24-bit

32-bit

16-bit

24-bit

32-bit

16-bit

24-bit

32-bit

16-bit

24-bit

32-bit


1600x1200 16-bit

24-bit

32-bit

Double

10.99MB

16MB

Triple

14.65MB

16MB

Double

16.48MB

32MB

Triple

21.97MB

32MB

Double

21.97MB

32MB

Triple

29.30MB

32MB

16-bit

24-bit

32-bit


Note Although 3D adapters typically operate in a 32-bit mode (refer to Table 15.9), this does not necessarily mean they can produce more than the 16,277,216 colors of a 24-bit true-color display. Many video processors and video memory buses are optimized to move data in 32-bit words, and they actually display 24-bit color while operating in a 32-bit mode, instead of the 4,294,967,296 colors you would expect from a true 32-bit color depth.

If you spend a lot of time working with graphics and want to enjoy 3D games, you might want to invest in a 24-bit (2D) or 32-bit (3D) video card with at least 32MB or more of RAM. Although 2D operations can be performed with as little as 4MB of RAM, 32-bit color depths for realistic 3D operation with large Z-buffers use most of the RAM available on a 16MB card at 1024x768 resolution; higher resolutions use more than 16MB of RAM at higher color depths. Today's video cards provide more RAM and more 2D/3D performance for less money than ever before. Note that recent and current-model video cards don't use socketed memory anymore, so you must be sure to buy a video card with all the memory you might need now and in the future. Otherwise, you must replace a card with inadequate memory.


Windows Can't Display More Than 256 Colors If you have a video card with 1MB or more of video memory, but the Windows Display Settings properties sheet won't allow you to select a color depth greater than 256 colors, Windows might have misidentified the video card during installation or the video driver installation might be corrupted. To see which video card Windows recognizes, click the Advanced Properties button, and then click the Adapter tab if necessary. Your adapter type might be listed either by the video card's brand name and model or by the video card's chipset maker and chipset model. If the card model or chipset appears to be incorrect or not specific enough, click Change and see what other drivers your system lists that appear to be compatible, or use a utility program provided by your video card/chipset maker to identify your card and memory size. Then, manually select the correct driver if necessary. If the video card model or chipset appears to be correct, open the System properties sheet, locate the Device Manager, and remove the display adapter listing. Restart the computer and Windows will redetect the video card/chipset and install the correct driver.

Video Bus Width Another issue with respect to the memory on the video adapter is the width of the bus connecting the graphics chipset and memory on the adapter. The chipset is usually a single large chip on the card that contains virtually all the adapter's functions. It is wired directly to the memory on the adapter through a local bus on the card. Most of the high-end adapters use an internal memory bus that is 64 bits or even 128 bits wide. This jargon can be confusing because video adapters that take the form of separate expansion cards also plug into the main system bus, which has its own speed rating. When you read about a 64-bit or 128-bit video adapter, you must understand that this refers to the local video bus and that the bus connecting the adapter to the system is actually the 32- or 64-bit PCI or AGP bus on the system's motherboard. See "System Bus Types, Functions, and Features," p. 308.

The Digital-to-Analog Converter The digital-to-analog converter on a video adapter (commonly called a RAMDAC) does exactly what its name describes. The RAMDAC is responsible for converting the digital images your computer generates into analog signals the monitor can display. The speed of the RAMDAC is measured in MHz; the faster the conversion process, the higher the adapter's vertical refresh rate. The speeds of the RAMDACs used in today's high-performance video adapters range from 300MHz to 500MHz. Most of today's video card chipsets include the RAMDAC function inside the 3D accelerator chip, but some dual-display-capable video cards use a separate


RAMDAC chip to allow the second display to work at different refresh rates than the primary display. The benefits of increasing the RAMDAC speed include higher vertical refresh rates, which allows higher resolutions with flicker-free refresh rates (72Hz​85Hz or above). Typically, cards with RAMDAC speeds of 300MHz or above display flickerfree (75Hz or above) at all resolutions up to 1920x1200. Of course, as discussed earlier in this chapter, you must ensure that any resolution you want to use is supported by both your monitor and video card.

The Bus You've learned in this chapter that certain video adapters were designed for use with certain system buses. Earlier bus standards, such as the IBM MCA, ISA, EISA, and VL-Bus, have all been used for VGA and other video standards. Because of their slow performances, all are now obsolete; current video cards are made exclusively for either the PCI or AGP bus standard. In July 1992, Intel Corporation introduced Peripheral Component Interconnect (PCI) as a blueprint for directly connecting microprocessors and support circuitry. It then extended the design to a full expansion bus with Release 2 in 1993; the current standard is Release 2.1. Popularly termed a mezzanine bus, PCI combines the speed of a local bus with microprocessor independence. PCI video adapters, similar to VL-Bus adapters, can dramatically increase video performance when compared with ISA adapters. PCI video adapters, by their design, are meant to be Plug and Play (PnP), which means they require little configuration. The PCI standard virtually replaced the older VL-Bus standard overnight and has dominated Pentium-class video until recently. Although the PCI bus remains the best general-purpose standard for expansion cards, the current king of high-speed video is AGP. See "Accelerated Graphics Port," p. 333.

See "The PCI Bus," p. 329.

The most recent system bus innovation is the Accelerated Graphics Port (AGP), an Intel-designed dedicated video bus that delivers a maximum bandwidth up to 16 times larger than that of a comparable PCI bus. AGP is essentially an


enhancement to the existing PCI bus; it's intended for use with only video adapters and provides them with high-speed access to the main system memory array. This enables the adapter to process certain 3D video elements, such as texture maps, directly from system memory rather than having to copy the data to the adapter memory before the processing can begin. This saves time and eliminates the need to upgrade the video adapter memory to better support 3D functions.


Note Although the earliest AGP cards had relatively small amounts of onboard RAM, most recent and all current implementations of card-based AGP use large amounts of on-card memory and use a memory aperture (a dedicated memory address space above the area used for physical memory) to transfer data more quickly to and from the video card's own memory. Integrated chipsets featuring built-in AGP do use system memory for all operations, including texture maps. Ironically, the memory aperture used by AGP cards can actually cause out-of-memory errors with Windows 9x and Windows Me on systems with more than 512MB of RAM. See Microsoft Knowledge Base document #Q253912 for details.

Although it was designed with the Pentium II in mind, AGP is not processor dependent. However, it does require support from the motherboard chipset, and AGP cards require a special expansion slot, which means you can't upgrade an existing non-AGP system to use AGP without replacing the motherboard. All recent nonintegrated motherboard chipsets from major vendors (Intel, Acer Labs, VIA Technologies, and SiS) for Pentium II and newer Intel processors and AMD Athlon and Duron processors support some level of AGP. However, some of the integrated chipsets don't support an AGP slot. See "Socket 7 (and Super7)," p. 90.

Even with the proper chipset, however, you can't take full advantage of AGP's capabilities without the proper operating system support. AGP's Direct Memory Execute (DIME) feature uses main memory instead of the video adapter's memory for certain tasks to lessen the traffic to and from the adapter. Windows 98/Me and Windows 2000/XP support this feature, but Windows 95 and Windows NT 4 do not. However, with the large amounts of memory found on current AGP video cards, this feature is seldom implemented. Four speeds of AGP are available: 1X, 2X, 4X, and 8X (AGP 3.0). AGP 1X and 2X are part of the original AGP Specification 1.0; AGP 4X is part of AGP Specification 2.0; and AGP 8X is part of AGP Specification 3.0. AGP 1X is the original version of AGP, running at 66MHz clock speed and a maximum transfer rate of 266MBps (twice the speed of 32-bit PCI video cards). AGP 1X is obsolete, having been replaced by AGP 2X, which runs at 133MHz and offers transfer rates up to 533MBps. The most common version of AGP supports 4X mode, which offers maximum transfer rates in excess of 1GBps; AGP 4X also can be used with AGP 2X-compatible motherboards (although its performance then is limited to 2X). Intel's AGP 8X mode, which provides transfer rates in excess of 2GBps, is now known as AGP Specification 3.0; motherboards using AGP 3.0 can handle either AGP 3.0 or 1.5V AGP 2.0 video cards. Even though AGP 3.0 (better known as AGP


8X) is the fastest version of AGP, if you plan to use an AGP video card with an older system first, you might want to specify a 1.5V AGP 4X video card instead. Such cards will work in both AGP 4X (AGP 2.0) and AGP 8X (AGP 3.0) systems. Because of the bandwidth required by AGP 3.0, systems featuring this version of AGP also support DDR333 memory, which is significantly faster than DDR266 (also known as PC2100 memory). AGP 3.0 was announced in 2000, but support for the standard requires the development of motherboard chipsets that were not introduced until mid-2002. By late 2002, AGP 3.0 video cards were widely available, and some chipset vendors, such as NVIDIA, were releasing revised versions of existing AGP 4X chipsets that supported AGP 3.0 (AGP 8X). The NVIDIA 8X version of the GeForce 4 Ti 4600 GPU is erroneously referred to as the "Ti 4800" by some graphics card makers.


Caution Should you check motherboard-video card compatibility before you buy? Yes, because early AGP cards (especially those based on Intel's i740 chipset) were designed strictly for the original AGP Pentium IIbased motherboards. Some AGP users have Super Socket 7 motherboards instead, and some early AGP cards don't work with these motherboards. Check with both the video card and motherboard vendors before you make your purchase. Intel's 845-series chipsets for the Pentium 4 work with 1.5V AGP 4X video cards but not with AGP 2X/4X cards running at 3.3V.

See "Socket 7 (and Super7)," p. 90.

The high and mid-range sectors of the video card market have shifted almost completely to AGP 4X or AGP 8X; PCI video cards are being relegated to replacements for older systems. Ironically, because of the high popularity of AGP today, you might pay more for a slower PCI video card than for an AGP card with similar features.


Note Many low-cost systems implement AGP video on the motherboard without providing an AGP expansion slot for future upgrades. This prevents you from changing to a faster or higher memory AGP video card in the future, although you might be able to use slower PCI video cards. For maximum flexibility, get your AGP video on a card.

VL-Bus, PCI, and AGP have some important differences, as Table 15.11 shows. Table 15.11. Local Bus Specifications Feature

VL-Bus

PCI

AGP 533MBps throughput (2X)

Theoretical maximum

132MBps

132MBps [1]

1.06GBps throughput (4X) 2.12GBps throughput (8X)

Slots [2]

3 (typical)

4/5 (typical)

1

Plug and Play support

No

Yes

Yes

Cost

Inexpensive Slightly higher

Ideal use

Low-cost 486

High-end 486, Pentium, Pentium II/III/Celeron/4, K6, Athlon, Duron, Athlon XP, P6 Athlon 64

Status

Obsolete

Current

Similar to PCI

Current [2]

[1] At the 66MHz bus speed and 32 bits. Throughput is higher on the 100MHz system bus. [2] Most current systems supportAGP4X/8X only.

The Video Driver The software driver is an essential, and often problematic, element of a video display subsystem. The driver enables your software to communicate with the video adapter. You can have a video adapter with the fastest processor and the most efficient memory on the market but still have poor video performance because of a badly written driver. Video drivers generally are designed to support the processor on the video adapter. All video adapters come equipped with drivers the card manufacturer supplies, but often you can use a driver the chipset maker created as well.


Sometimes you might find that one of the two provides better performance than the other or resolves a particular problem you are experiencing. Most manufacturers of video adapters and chipsets maintain Web sites from which you can obtain the latest drivers; drivers for chipset-integrated video are supplied by the system board or system vendor. A driver from the chipset manufacturer can be a useful alternative, but you should always try the adapter manufacturer's driver first. Before purchasing a video adapter, you should check out the manufacturer's site and see whether you can determine how up-to-date the available drivers are. At one time, frequent driver revisions were thought to indicate problems with the hardware, but the greater complexity of today's systems means that driver revisions are a necessity. Even if you are installing a brand-new model of a video adapter, be sure to check for updated drivers on the manufacturer's Web site for best results. The video driver also provides the interface you can use to configure the display your adapter produces. On a Windows 9x/Me/2000/XP system, the Display Control Panel identifies the monitor and video adapter installed on your system and enables you to select the color depth and screen resolution you prefer. The driver controls the options that are available for these settings, so you can't choose parameters the hardware doesn't support. For example, the controls would not allow you to select a 1024x768 resolution with 24-bit color if the adapter had only 1MB of memory. When you click the Advanced button on the Settings page, you see the Properties dialog box for your particular video display adapter. The contents of this dialog box can vary, depending on the driver and the capabilities of the hardware. Typically, on the General page of this dialog box, you can select the size of the fonts (large or small) to use with the resolution you've chosen. Windows 98/Me/2000 (but not Windows XP) also add a control to activate a convenient feature. The Show Settings Icon on Task Bar check box activates a tray icon that enables you to quickly and easily change resolutions and color depths without having to open the Control Panel. This feature is often called QuickRes. The Adapter page displays detailed information about your adapter and the drivers installed on the system, and it enables you to set the Refresh Rate for your display; with Windows XP, you can use the List All Modes button to view and choose the resolution, color depth, and refresh rate with a single click. The Monitor page lets you display and change the monitor's properties and switch monitor drivers if necessary. In Windows XP, you can also select the refresh rate on this screen. If your adapter includes a graphics accelerator, the Performance page contains a Hardware Acceleration slider you can use to control the degree of graphic display assistance provided by your adapter hardware. In Windows XP, the Performance page is referred to as the Troubleshoot page.


Setting the Hardware Acceleration slider to the Full position activates all the adapter's hardware acceleration features. The necessary adjustments for various problems can be seen in Table 15.12 for Windows XP and in Table 15.13 for other versions of Windows. Table 15.12. Using Graphics Acceleration Settings to Troubleshoot Windows XP Acceleration Setting

When to Use

Effect of Setting

Left [*]

The display works in Safe or VGA mode but is corrupted in other modes.

One click from left [*]

2D and 3D graphics driver problems; It disables all but basic acceleration. mouse driver problems.

Two clicks from left [*]

3D acceleration problems.

Long-Term Solution

Update display, DirectX, and mouse drivers.

There's no acceleration.

Update display, DirectX, and mouse drivers.

It disables DirectX, DirectDraw, and Direct Update DirectX 3D acceleration (mainly used by 3D games). drivers.

Two clicks Display driver problems. from right [*]

It disables cursor and drawing accelerations.

Update display drivers.

One click Mouse pointer corruption. from right [*]

It disables mouse and pointer acceleration.

Update mouse drivers.

Right

It enables full acceleration.

N/A

Normal operation.

[*] Disable write combining, which is a method for speeding up screen display, whenever you select any

setting other than full acceleration to improve stability. Reenable write combining after you install updated drivers and retry.

Table 15.13. Using Graphics Acceleration Settings to Troubleshoot Other Windows Versions Mouse Pointer Location

When to Use

Effects of Setting

Long-Term Solution

Left

Display works in Safe or VGA mode, but it is corrupted in other modes.

It disables all acceleration.

Update display and mouse drivers.

One click from left

2D and 3D graphics driver problems, mouse driver problems.

Basic acceleration only.

Update display and mouse drivers.

One click from right

Mouse pointer corruption.

It disables mouse pointer Update mouse drivers. acceleration.

Right

Normal operation.

Full acceleration.

N/A


If you're not certain of which setting is the best for your situation, use this procedure: Move the slider one notch to the left to address mouse display problems by disabling the hardware's cursor support in the display driver. This is the equivalent of adding the SWCursor=1 directive to the [Display] section of the System.ini file in Windows 9x/Me. If you are having problems with 2D graphics in Windows XP only, but 3D applications work correctly, move the slider to the second notch from the right to disable cursor drawing and acceleration. Moving the slider another notch (to the third notch from the right in Windows XP or the second notch from the right in earlier versions) prevents the adapter from performing certain bit-block transfers; it disables 3D functions of DirectX in Windows XP. With some drivers, this setting also disables memory-mapped I/O. This is the equivalent of adding the Mmio=0 directive to the [Display] section of System.ini and the SafeMode=1 directive to the [Windows] section of Win.ini (and the SWCursor directive mentioned previously) in Windows 9x/Me. Moving the slider to the None setting (the far left) adds the SafeMode=2 directive to the [Windows] section of the Win.ini file in Windows 9x/Me. This disables all hardware acceleration support on all versions of Windows and forces the operating system to use only the device-independent bitmap (DIB) engine to display images, rather than bit-block transfers. Use this setting when you experience frequent screen lockups or receive invalid page fault error messages.


Note If you need to disable any of the video hardware features listed earlier, this often indicates a buggy video or mouse driver. If you download and install updated video and mouse drivers, you should be able to revert to full acceleration. You should also download an updated version of DirectX for your version of Windows.

In most cases, another tab called Color Management is also available. You can select a color profile for your monitor to enable more accurate color matching for use with graphics programs and printers. Video cards with advanced 3D acceleration features often have additional properties; these are discussed later in this chapter.

Multiple Monitors Windows 98/Me and Windows 2000/XP include a video display feature that Macintosh systems have had for years: the capability to use multiple monitors on one system. Windows 98/Me support up to nine monitors (and video adapters), each of which can provide a different view of the desktop. Windows 2000 and Windows XP support up to ten monitors and video adapters. When you configure a Windows 98/Me or Windows 2000/XP system to use multiple monitors, the operating system creates a virtual desktop​that is, a display that exists in video memory that can be larger than the image actually displayed on a single monitor. You use the multiple monitors to display various portions of the virtual desktop, enabling you to place the windows for different applications on separate monitors and move them around at will. Unless you use multiple-head video cards, each monitor you connect to the system requires its own video adapter. So, unless you have nine bus slots free, the prospects of seeing a nine-screen Windows display are slim, for now. However, even two monitors can be a boon to computing productivity. On a multimonitor Windows 98/Me system, one display is always considered to be the primary display. The primary display can use any PCI or AGP VGA video adapter that uses a Windows 98/Me minidriver with a linear frame buffer and a packed (nonplanar) format, meaning that most of the brand-name adapters sold today are eligible. Additional monitors are called secondaries and are much more limited in their hardware support. To install support for multiple monitors, be sure you have only one adapter installed first; then reboot the system, and install each additional adapter one at a time. For more information about multiple-monitor support for Windows 98/Me, including a list of supported adapters, see the Microsoft Knowledge Base article #Q182708. It's important that the computer correctly identifies which one of the video


adapters is the primary one. This is a function of the system BIOS, and if the BIOS on your computer does not let you select which device should be the primary VGA display, it decides based on the order of the PCI slots in the machine. You should, therefore, install the primary adapter in the highest-priority PCI slot. In some cases, an AGP adapter might be considered secondary to a PCI adapter. Depending on the BIOS used by your system, you might need to check in various places for the option to select the primary VGA display. For example, the Award BIOS used by the ASUS A7V133 motherboard for Socket A processors lists this option, which it calls Primary VGA BIOS, in the Boot menu. In contrast, the Phoenix BIOS used by the Intel DB815-EEA motherboard lists this option, which it calls Default Primary Video Adapter, in the Video Configuration menu. See "The PCI Bus," p. 329.

After the hardware is in place, you can configure the display for each monitor from the Display control panel's Settings page. The primary display is always fixed in the upper-left corner of the virtual desktop, but you can move the secondary displays to view any area of the desktop you like. You can also set the screen resolution and color depth for each display individually. For more information about configuring multiple-monitor support in Windows 98/Me, see Microsoft Knowledge Base article #Q179602. The multiple-monitor support included with Windows 2000 and Windows XP is somewhat different from that of Windows 98/Me. These versions of Windows support ten monitors, rather than nine as with Windows 98/Me. In addition, because Windows 2000 and XP use different display drivers than Windows 98/Me, some configurations that work with 98/Me might not work with Windows 2000/XP. For more information about configuring multiple-monitor support in Windows 2000, see Microsoft Knowledge Base article #Q238886. For details of the display cards compatible with Windows XP in multiple-display modes, see Microsoft Knowledge Base article #Q307397. Windows XP also supports DualView, an enhancement to Windows 2000's multiple-monitor support. DualView supports the increasing number of dual-head video cards as well as notebook computers connected to external displays. With systems supporting DualView, the first video port is automatically assigned to the primary monitor. On a notebook computer, the primary display is the built-in LCD display. Even if your BIOS enables you to specify the primary video card and you use video cards that are listed as compatible, determining exactly which display cards will work successfully in a multimonitor configuration can be difficult. Microsoft provides a list of compatible display cards in the Hcl.txt file located on the


Windows 2000 CD-ROM, but this list does not contain the latest video cards and chipsets from NVIDIA, ATI, or other companies, nor does it take into account changes in supported chipsets caused by improved drivers. Unfortunately, the online version of the Microsoft Windows Hardware Quality Labs Hardware Compatibility List (www.Microsoft.com/hcl) doesn't list information about multiplemonitor support for any version of Windows. Consequently, you should check with your video card or chipset maker for the latest information on Windows 2000 or Windows XP and multiple-monitor support issues. Because new chipsets, updated drivers, and combinations of display adapters are a continuous issue for multiple-monitor support, I recommend the following online resources: http://www.realtimesoft.com/ultramon. Home of the UltraMon multiplemonitor support enhancement program ($24.95) and an extensive database of user-supplied multiple-monitor configurations for Windows 98/Me, Windows 2000/XP, and Linux http://www.digitalroom.net (click Tech Articles and then Multiple Monitor Guide). Excellent tips on multiple-monitor setups and links to other resources Multiple-monitor support can be enabled through either of the following: Installation of a separate AGP or PCI graphics card for each monitor you want to use Installation of a single AGP or PCI graphics card that can support two or more monitors A card that supports multiple monitors (also called a multiple-head or dual-head card) saves slots inside your system. Most recent video cards with multiple-monitor support feature a 15-pin analog VGA connector for CRTs, a DVI-I digital/analog connector for digital LCD panels, and a TV-out connector for S-video or composite output to TVs and VCRs. Thus, you can connect any of the following to these cards: One analog LCD or CRT display and one digital LCD display Two analog LCD or CRT displays (when the DVI-I​to​VGA adapter is used)


One analog LCD or CRT display and one TV One digital LCD display and one TV The major video chipsets that support multiple displays are listed in Table 15.14. Table 15.14. Major Multiple-Head Video Chipsets and Cards Brand

ATI[1]

Matrox [2]

NVIDIA [5]

Chipset

Bus Type(s) Supported

Number of Monitors Supported

RADEON VE

AGP

2

RADEON 7500

AGP, PCI

2

RADEON 8500

AGP

2

RADEON 8500LE

AGP

2

RADEON 9000

AGP

2

RADEON 9000 PRO

AGP

2

RADEON 9500 PRO

AGP

2

RADEON 9700 PRO

AGP

2

M200MMS

PCI

2,4[3]

Millennium G450

AGP, PCI

2[3]

G450MMS

PCI

2,4[4]

G550

AGP

2[3]

Parhelia

AGP

3[3]

GeForce 2 MX

AGP, PCI

2

GeForce 2 MX 200

AGP, PCI

2

GeForce 2 MX 400

AGP, PCI

2

GeForce 4 MX 440

AGP

2

GeForce 4 Ti 4200

AGP

2

GeForce 4 Ti 4400

AGP

2

GeForce 4 Ti 4600

AGP

2

GeForce FX

AGP

2

[1] ATI sells video cards using these chipsets under the ATI brand and also supplies chipsets to third-

party vendors. [2] Matrox is the only vendor using its chipsets; this table lists Matrox card models.


[3] Features a separate accelerator chip for each display, enabling the independent selection of the

refresh rate and the resolution under Windows 2000. [4] Features a separate RAMDAC chip for each display, enabling the independent selection of the refresh

rate and the resolution under Windows 2000. [5] NVIDIA does not manufacture video cards; it sells chipsets only.


3D Graphics Accelerators Since the late 1990s, 3D acceleration​once limited to exotic add-on cards designed for hardcore gameplayers​has become commonplace in the PC world. Although business software has yet to embrace 3D imaging, full-motion graphics are used in sports, first-person shooters, team combat, driving, and many other types of PC gaming. Because even low-cost integrated chipsets offer some 3D support and 3D video cards are now in their sixth generation of development, virtually any user of a recent-model computer has the ability to enjoy 3D lighting, perspective, texture, and shading effects in her favorite games. The latest 3D sports games provide lighting and camera angles so realistic that a casual observer could almost mistake the computer-generated game for an actual broadcast, and the latest 3D accelerator chips enable fast PCs to compete with high-performance dedicated game machines, such as Sony's PlayStation 2, Nintendo's GameCube, and Microsoft's Xbox, for the mind and wallet of the hard-core gameplayer.

How 3D Accelerators Work To construct an animated 3D sequence, a computer can mathematically animate the sequences between keyframes. A keyframe identifies specific points. A bouncing ball, for example, can have three keyframes: up, down, and up. Using these frames as a reference point, the computer can create all the interim images between the top and bottom. This creates the effect of a smoothly bouncing ball. After it has created the basic sequence, the system can then refine the appearance of the images by filling them in with color. The most primitive and least effective fill method is called flat shading, in which a shape is simply filled with a solid color. Gouraud shading, a slightly more effective technique, involves the assignment of colors to specific points on a shape. The points are then joined using a smooth gradient between the colors. A more processor-intensive, and much more effective, type of fill is called texture mapping. The 3D application includes patterns​or textures​in the form of small bitmaps that it tiles onto the shapes in the image, just as you can tile a small bitmap to form the wallpaper for your Windows desktop. The primary difference is that the 3D application can modify the appearance of each tile by applying perspective and shading to achieve 3D effects. When lighting effects that simulate fog, glare, directional shadows, and others are added, the 3D animation comes very close indeed to matching reality. Until the late 1990s, 3D applications had to rely on support from software routines to convert these abstractions into live images. This placed a heavy


burden on the system processor in the PC, which has a significant impact on the performance not only of the visual display, but also of any other applications the computer might be running. Starting in the period from 1996 to 1997, chipsets on most video adapters began to take on many of the tasks involved in rendering 3D images, greatly lessening the load on the system processor and boosting overall system performance. Because there have now been six generations of 3D accelerators (depending on what you count as a "generation"), and more and more memory is standard to enable high-resolution 3D animations, most high-quality 3D accelerator cards cost at least $200, and some models packed with 128MB of DDR SDRAM and the latest accelerator technology can sell for as much as $350​$400. Both video games and 3D animation programs are taking advantage of their capability to render smooth, photorealistic images at high speeds and in real time. Fortunately, users with less-demanding 3D performance requirements often can purchase low-end products based on the previous generation of 3D accelerator chips for less than $100. These cards typically provide more-than-adequate performance for 2D business applications. Most current mid-range and high-end 3D accelerators also support dual-display and TV-out capabilities, enabling you to work and play at the same time. 3D technology has added an entirely new vocabulary to the world of video display adapters. Before purchasing a 3D accelerator adapter, you should familiarize yourself with some of the terms and concepts involved in the 3D image generation process. The basic function of 3D software is to convert image abstractions into the fully realized images that are then displayed on the monitor. The image abstractions typically consist of the following elements: Vertices. Locations of objects in three-dimensional space, described in terms of their x, y, and z coordinates on three axes representing height, width, and depth. Primitives. The simple geometric objects the application uses to create more complex constructions, described in terms of the relative locations of their vertices. This serves not only to specify the location of the object in the 2D image, but also to provide perspective because the three axes can define any location in three-dimensional space. Textures. Two-dimensional bitmap images or surfaces designed to be mapped onto primitives. The software enhances the 3D effect by modifying the appearance of the textures, depending on the location and attitude of the primitive. This process is called perspective correction. Some applications use


another process, called MIP mapping, which uses different versions of the same texture that contain varying amounts of detail, depending on how close the object is to the viewer in the three-dimensional space. Another technique, called depth cueing, reduces the color and intensity of an object's fill as the object moves farther away from the viewer. Using these elements, the abstract image descriptions must then be rendered, meaning they are converted to visible form. Rendering depends on two standardized functions that convert the abstractions into the completed image that is displayed onscreen. The standard functions performed in rendering are Geometry. The sizing, orienting, and moving of primitives in space and the calculation of the effects produced by the virtual light sources that illuminate the image Rasterization. The converting of primitives into pixels on the video display by filling the shapes with properly illuminated shading, textures, or a combination of the two A modern video adapter that includes a chipset capable of 3D video acceleration has special built-in hardware that can perform the rasterization process much more quickly than if it were done by software (using the system processor) alone. Most chipsets with 3D acceleration perform the following rasterization functions right on the adapter: Scan conversion. The determination of which onscreen pixels fall into the space delineated by each primitive Shading. The process of filling pixels with smoothly flowing color using the flat or Gouraud shading technique Texture mapping. The process of filling pixels with images derived from a 2D sample picture or surface image Visible surface determination. The identification of which pixels in a scene are obscured by other objects closer to the viewer in three-dimensional space Animation. The process of switching rapidly and cleanly to successive frames of motion sequences Antialiasing. The process of adjusting color boundaries to smooth edges on rendered objects


Common 3D Techniques Virtually all 3D cards use the following techniques: Fogging. Fogging simulates haze or fog in the background of a game screen and helps conceal the sudden appearance of newly rendered objects (buildings, enemies, and so on). Gouraud shading. Interpolates colors to make circles and spheres look more rounded and smooth. Alpha blending. One of the first 3D techniques, alpha blending creates translucent objects onscreen, making it a perfect choice for rendering explosions, smoke, water, and glass. Alpha blending also can be used to simulate textures, but it is less realistic than environment-based bump mapping (see the section "Environment-Based Bump Mapping and Displacement Mapping," later in this chapter). Because they are so common, data sheets for advanced cards frequently don't mention them, although these features are present.

Advanced 3D Techniques The following are some of the latest techniques that leading 3D accelerator cards use. Not every card uses every technique.

Stencil Buffering Stencil buffering is a technique useful for games such as flight simulators, in which a static graphic element​such as a cockpit windshield frame, which is known as a HUD (heads up display) and used by real-life fighter pilots​is placed in front of dynamically changing graphics (such as scenery, other aircraft, sky detail, and so on). In this example, the area of the screen occupied by the cockpit windshield frame is not re-rendered. Only the area seen through the "glass" is re-rendered, saving time and improving frame rates for animation.

Z-Buffering A closely related technique is Z-buffering, which originally was devised for


computer-aided drafting (CAD) applications. The Z-buffer portion of video memory holds depth information about the pixels in a scene. As the scene is rendered, the Z-values (depth information) for new pixels are compared to the values stored in the Z-buffer to determine which pixels are in "front" of others and should be rendered. Pixels that are "behind" other pixels are not rendered. This method increases speed and can be used along with stencil buffering to create volumetric shadows and other complex 3D objects.

Environment-Based Bump Mapping and Displacement Mapping Environment-based bump mapping introduces special lighting and texturing effects to simulate the rough texture of rippling water, bricks, and other complex surfaces. It combines three separate texture maps (for colors, for height and depth, and for environment​including lighting, fog, and cloud effects). This creates enhanced realism for scenery in games and could also be used to enhance terrain and planetary mapping, architecture, and landscape-design applications. This represents a significant step beyond alpha blending. However, a feature called displacement mapping produces even more accurate results. Special grayscale maps called displacement maps have long been used for producing accurate maps of the globe. Microsoft DirectX 9 supports the use of grayscale hardware displacement maps as a source for accurate 3D rendering. The Matrox Parhelia, the ATI Radeon 9500, 9700, and 9800 series, and the GeForce FX all support displacement mapping.

Texture Mapping Filtering Enhancements To improve the quality of texture maps, several filtering techniques have been developed, including MIP mapping, bilinear filtering, trilinear filtering, and anisotropic filtering. These techniques and several others are explained here: Bilinear filtering. Improves the image quality of small textures placed on large polygons. The stretching of the texture that takes place can create blockiness, but bilinear filtering applies a blur to conceal this visual defect. MIP mapping. Improves the image quality of polygons that appear to recede into the distance by mixing low-res and high-res versions of the same texture; a form of antialiasing. Trilinear filtering. Combines bilinear filtering and MIP mapping, calculating the most realistic colors necessary for the pixels in each polygon by comparing the values in two MIP maps. This method is superior to either MIP mapping or


bilinear filtering alone.


Note Bilinear and trilinear filtering work well for surfaces viewed straight-on but might not work so well for oblique angles (such as a wall receding into the distance).

Anisotropic filtering. Some video card makers use another method, called anisotropic filtering, for more realistically rendering oblique-angle surfaces containing text. This technique is used when a texture is mapped to a surface that changes in two of three spatial domains, such as text found on a wall down a roadway (for example, advertising banners at a raceway). The extra calculations used take time, and for that reason, it can be disabled. T-buffer. This technology eliminates aliasing (errors in onscreen images due to an undersampled original) in computer graphics, such as the "jaggies" seen in onscreen diagonal lines; motion stuttering; and inaccurate rendition of shadows, reflections, and object blur. The T-buffer replaces the normal frame buffer with a buffer that accumulates multiple renderings before displaying the image. Unlike some other 3D techniques, T-buffer technology doesn't require rewriting or optimization of 3D software to use this enhancement. The goal of T-buffer technology is to provide a movie-like realism to 3D rendered animations. The downside of enabling antialiasing using a card with T-buffer support is that it can dramatically impact the performance of an application. This technique originally was developed by now-defunct 3dfx. However, this technology is incorporated into Microsoft DirectX 8.0 and above, enabling other brands of video cards to use it. Integrated transform and lighting. The 3D display process includes transforming an object from one frame to the next and handling the lighting changes that result from those transformations. Many 3D cards put the CPU in charge of these functions, but most recent graphics accelerators from NVIDIA and ATI integrate separate transform and lighting engines into the accelerator chip for faster 3D rendering, regardless of CPU speed. NVIDIA started to use this feature in its GeForce 2 series, whereas ATI began to use this feature starting with the original RADEON. For more information, see the NVIDIA and ATI Web sites. Full-screen antialiasing. This technology reduces the jaggies visible at any resolution by adjusting color boundaries to provide gradual, rather than abrupt, color changes. Whereas early 3D products used antialiasing for certain objects only, the latest accelerators from NVIDIA and ATI use this technology for the entire display. The NVIDIA GeForce 4 Ti 4xxx series, GeForce FX, and ATI RADEON 9xxx series use highly optimized FSAA methods that allow high visual quality at high frame rates.


Vertex skinning. Also referred to as vertex blending, this technique blends the connection between two angles, such as the joints in an animated character's arms or legs. NVIDIA's GeForce2, 3, and 4 series cards use a software technique to perform blending at two matrices, whereas the ATI RADEON series chips use a more realistic hardware-based technique called 4-matrix skinning. Keyframe interpolation. Also referred to as vertex morphing, this technique animates the transitions between two facial expressions, allowing realistic expressions when skeletal animation can't be used or isn't practical. See the ATI Web site for details. Programmable vertex and pixel shading. Both NVIDIA and ATI have embraced various methods of programmable vertex and pixel shading in recent versions. The NVIDIA GeForce3's nfiniteFX technology enables software developers to customize effects such as vertex morphing and pixel shading (an enhanced form of bump mapping for irregular surfaces that enables perpixel lighting effects), rather than applying a narrow range of predefined effects. The NVIDIA GeForce4 Ti's nfiniteFXII pixel shader is DirectX 8 compatible and supports up to four textures, whereas its dual vertex shaders provide high-speed rendering up to 50% faster than the GeForce3. The ATI RADEON 8500 and 9000's version is called SmartShader, supports more complex programs than nfiniteFX, and provides comparable quality to nfiniteFXII. SmartShader is also supported by DirectX 8.1. ATI 9700 and 9500 support DirectX 9's floating-point pixel shaders and more complex vertex shader. NVIDIA GeForce FX also supports DirectX 9 pixel and vertex shaders, but it also adds more features. Floating-point calculations. Microsoft DirectX 9 supports floating-point data for more vivid and accurate color and polygon rendition. ATI Radeon 9500 and 9700 and NVIDIA GeForce FX are the first 3D accelerator chips to have full DirectX 9 support, with GeForce FX providing additional precision. The Matrox Parhelia supports this, but not all, DirectX 9 features.

Single- Versus Multiple-Pass Rendering Various video card makers handle application of these advanced rendering techniques differently. The current trend is toward applying the filters and basic rendering in a single pass rather than multiple passes. Video cards with singlepass rendering and filtering typically provide higher frame-rate performance in 3D-animated applications and avoid the problems of visible artifacts caused by errors in multiple floating-point calculations during the rendering process.


Hardware Acceleration Versus Software Acceleration Compared to software-only rendering, hardware-accelerated rendering provides faster animation. Although most software rendering would create more accurate and better-looking images, software rendering is too slow. Using special drivers, these 3D adapters can take over the intensive calculations needed to render a 3D image that software running on the system processor formerly performed. This is particularly useful if you are creating your own 3D images and animation, but it is also a great enhancement to the many modern games that rely extensively on 3D effects. Note that motherboard-integrated video solutions, such as Intel's 810 and 815 series, typically have significantly lower 3D performance because they use the CPU for more of the 3D rendering than 3D video adapter chipsets do. To achieve greater performance, many of the latest 3D accelerators run their accelerator chips at very high speeds, and some even allow overclocking of the default RAMDAC frequencies. Just as CPUs at high speeds produce a lot of heat, so do high-speed video accelerators. Both the chipset and the memory are heat sources, so most mid-range and high-end 3D accelerator cards feature a fan to cool the chipset. Also, some high-end 3D accelerators such as the Gainward GeForce 4 Ti 4200 Golden Sample (based on the NVIDIA GeForce 4 Ti4200 chipset) use finned passive heatsinks to cool the memory chips and make overclocking the video card easier, as shown in Figure 15.10.

Figure 15.10. The Gainward GeForce 4 Ti 4200 Golden Sample features a heatsink fan to keep its high-speed NVIDIA GeForce Ti 4200 accelerator chip cool and passive heatsinks to dissipate heat away from the memory chips.

Software Optimization


It's important to realize that the presence of an advanced 3D-rendering feature on any given video card is meaningless unless game and application software designers optimize their software to take advantage of the feature. Although various 3D standards exist (OpenGL, Glide, and Direct 3D), video card makers provide drivers that make their games play with the leading standards. Because some cards do play better with certain games, you should read the reviews in publications such as Maximum PC to see how your favorite graphics card performs with them. It's important to note that, even though the latest video cards based on recent ATI and NVIDIA chips support DirectX 8.0, 8.1, and 9.0, many games still support only DirectX 7. As with previous 3D features, it takes time for the latest hardware features to be supported by game vendors. Some video cards allow you to perform additional optimization by adjusting settings for OpenGL, Direct 3D, RAMDAC, and bus clock speeds, as well as other options.


Note If you want to enjoy the features of your newest 3D card immediately, be sure to purchase the individual retail-packaged version of the card from a hardware vendor. These packages typically come with a sampling of games (full and demo versions) designed or compiled to take advantage of the card with which they're sold. The lower-cost OEM or "white box" versions of video cards are sold without bundled software, come only with driver software, and might differ in other ways from the retail advertised product. Some even use modified drivers, use slower memory or RAMDAC components, or lack special TV-out or other features. Some 3D card makers use different names for their OEM versions to minimize confusion, but others don't. Also, some card makers sell their cards in bulk packs, which are intended for upgrading a large organization with its own support staff. These cards might lack individual documentation or driver CDs and also might lack some of the advanced hardware features found on individual retail-pack video cards.

Application Programming Interfaces Application programming interfaces (APIs) provide hardware and software vendors a means to create drivers and programs that can work quickly and reliably across a wide variety of platforms. When APIs exist, drivers can be written to interface with the API rather than directly with the operating system and its underlying hardware. Currently, the leading game APIs include SGI's OpenGL and Microsoft's Direct 3D. OpenGL and Direct 3D (part of DirectX) are available for virtually all leading graphics cards. A third popular game API is Glide, an enhanced version of OpenGL that is restricted to graphics cards that use 3Dfx chipsets, which are no longer on the market. Although the video card maker must provide OpenGL support, Microsoft provides support for Direct3D as part of a much larger API called DirectX. The latest version of DirectX is DirectX 9, which enhances 3D video support, enhances DirectPlay (used for Internet gaming), and provides other advanced gaming features. For more information about DirectX or to download the latest version, see Microsoft's DirectX Web site at www.microsoft.com/windows/directx.


Note DirectX 9.0 is for Windows 98 and later versions (98SE, Me, 2000, and XP) only. However, Microsoft still provides DirectX 8.0a for Windows 95 users.

3D Chipsets Virtually every mainstream video adapter in current production features a 3D acceleration-compatible chipset. With several generations of 3D adapters on the market from the major vendors, keeping track of the latest products can be difficult. Table 15.15 lists the major 3D chipset manufacturers, the various chipsets they make, and the video adapters that use them. The following manufacturers' products are not included in Table 15.15 for the reasons listed: 3Dfx Interactive. Out of business; obtain last-available official and third-party drivers from www.voodoofiles.com. 3Dlabs. Now manufacturers Open GL-compatible workstation cards only. Intel. No longer manufactures graphics boards. Micron. No longer manufactures chipsets. VideoLogic. No longer manufactures graphics boards.


Note See Chapter 15 in both Upgrading and Repairing PCs, 12th Edition and 13th Edition on this book's DVD for more information about these companies' products and other older chipsets.


Note In each manufacturer's section, the following symbols are used to provide a ranking within that manufacturer's chipsets only: CURRENT​RECENT​OLD. Generally, CURRENT chipsets provide the fastest 3D performance and advanced 3D rendering features. RECENT chipsets, although lacking some of the speed or features of CURRENT chipsets, are also worthwhile, especially for those on a budget. OLD chipsets are generally superseded by RECENT and CURRENT chipsets and therefore are not recommended. OLD chipsets are listed by chipset only, whereas CURRENT and RECENT chipsets list selected video cards using those chipsets. Be sure to use this information in conjunction with application-specific and game-specific tests to help you choose the best card/chipset solution for your needs. Only products believed to be currently available are listed; consult the chipset vendors' Web sites for the latest information about third-party video card sources using a specific chipset.

Table 15.15. 3D Video Chipset Manufacturers Manufacturer

Chipset

Available Boards

RAGE I (OLD) RAGE II+ (OLD) RAGE IIC (OLD) 3D RAGE II+DVD (OLD) ATI RAGE PRO (OLD)

RAGE PRO TURBO (OLD) RAGE 128 (OLD) RAGE 128 Pro (OLD) ATI ALL-IN-WONDER RADEON ATI RADEON 64MB DDR version ATI RADEON 32MB DDR version

RADEON (RECENT) ATI RADEON 32MB SDRAM version Club3D RADEON Jetway R6A1 ATI RADEON VE Dual Display Edition Club3D RADEON VE Dual Display Edition

RADEON VE (RECENT) Jetway RV100 PowerMagic RADEON VE series ATI RADEON 7000 FIC 1st Graphics AT007V FIC A70L

RADEON 7000 (CURRENT)

Club3D CGA-1064TVD Connect 3D Radeon 7000


PowerColor RV6 series Sapphire Radeon 7000 series ATI RADEON 7200

RADEON 7200 (CURRENT)

FIC 1st Graphics AT007 Club3D CGA-72xx series ATI RADEON 7500 Club3D RADEON 7500 Club3D CGA-73xx series Connect 3D Radeon 7500 Elsa Winner 7500 FIC 1st Graphics AT008V

RADEON 7500 (CURRENT) FIC A75L Hercules 3D Prophet 7500 PowerMagic RADEON 7500 PowerColor RV2 series Sapphire Radeon 7500 series Sapphire The Beast All in Wonder 7500 series

RADEON 7500LE

Jetway RV200LE ATI RADEON 8500 ATI ALL-IN-WONDER RADEON 8500DV Club3D CGA-86xx series

RADEON 8500 (CURRENT) Hercules 3D Prophet 8500 Jetway R200A10D Sapphire Radeon 8500 series Club3D CGA-85xx series FIC 1st Graphics AT008

RADEON 8500LE (CURRENT)

Hercules 3D Prophet 8500 LE Hercules 3D Prophet FXD 8500 LE PowerMagic RADEON 8500 Hercules 3D Prophet 9000 Club3D CGA-90xx series Club3D CGA-92xx series FIC AT009LE


RADEON 9000 (CURRENT)

FIC A91L Connect 3D Radeon 9000 Giga-byte GV-R9000 series PowerColor RV25 series Sapphire Atlantis 9000

ATI Radeon 9000 Pro Connect 3D Radeon 9000 Pro FIC AT009 Hercules 3D Prophet 9000 PRO

RADEON 9000 PRO (CURRENT)

Giga-byte GV-R9000 PRO series PowerColor RV25 series Tyan Tachyon G9000 Pro Sapphire Atlantis 9000 Pro Sapphire The Beast All in Wonder 9000 Pro Club3D CGA-93xx series Connect 3D Radeon 9100 Giga-byte GV-R9100 series

RADEON 9100 (CURRENT) PowerColor AR2 series Sapphire Atlantis 9100 VisionTek Xtacy 9100 series FIC A95 Connect 3D Radeon 9500 Elsa Gladiac 9500

RADEON 9500 (CURRENT) Giga-byte GV-R9500 PowerColor XR95 series Sapphire Atlantis 9500 ATI Radeon 9500 Pro Club3D CGA-96xx series Connect 3D Radeon 9500 Pro Elsa Gladiac 9500 PRO

RADEON 9500 PRO (CURRENT) FIC A95P PowerColor XR95-C3 Sapphire Atlantis 9500 PRO VisionTek Xtasy 9500 PRO


FIC A97 Connect 3D Radeon 9700 Giga-byte GV-R9700

RADEON 9700 (CURRENT) PowerColor XR97-C3L PowerColor XF97-C3/B Sapphire Atlantis 9700 ATI Radeon 9700 Pro ATI All in Wonder 9700 Pro Club3D CGA-97xx series Elsa Gladiac 9700 PRO FIC A97P Giga-byte GV-R9700 PRO

RADEON 9700 PRO (CURRENT) Hercules 3D Prophet 9700 PRO Jetway R300A2D PowerColor XF97-C3G Sapphire Atlantis 9700 PRO series Sapphire The Beast All in Wonder 9700 Pro Tyan Tachyon G9700 PRO MGA-200 (OLD)

Matrox MGA-400 (OLD)

Matrox Millennium G450

MGA-450 (RECENT)

Matrox Marvel G450 eTV Matrox G450 MMS Matrox Millennium G550

MGA-550 (CURRENT)

Parhelia-512

Matrox Millennium G550 Dual-DVI Matrox Parhelia 128MB Matrox Parhelia 256MB RIVA 128(2D/3D OLD) RIVA 128ZX (OLD) RIVA TNT (OLD) RIVA TNT2 (OLD)

NVIDIA VANTA (OLD)


GeForce256 (OLD) GeForce2 series (OLD) GeForce2 Ti (OLD) Abit Saluro GF3 series Elsa Gladiac 920 Gainward CardExpert GeForce3 PowerPack series Hercules 3D Prophet III

GeForce3 (RECENT)

Leadtek WinFast GeForce 3 series Leadtek WinFast GeForce 3 TD series PNY Verto GeForce3 VisionTek GeForce3 XFX PVT20KMA Abit Saluro GF3 Ti series Albatron Ti200 series Aopen GF3Ti series Club3D GeForce 3 Ti200 Chaintech A-G3xx series Gainward GeForce3 PowerPack Ti series Jaton 3DForce III Ti series

GeForce3 Ti 200/500 (RECENT)

Leadtek WinFast Titanium 200, 500 series Palit GF3 Ti200 Prolink PixelView GeForce3 Ti series PNY Verto GeForce3 Ti series Visiontek Xtasy 6564 Visiontek Xtasy 6964 XFX PVT20FMA XFX PVT20AMA Abit Saluro GF4 MX series Albatron MX4xx series Aopen GF4MX series Aopen Aeolis MX440 series ASUS V8170 series ASUS V9180 series BFG Technologies GeForce 4 MX series


Chaintech A-G4xx series Club3D CGN-17xx series eVGA 064P series eVGA 064A series eVGA 64-A4 series

GeForce4 MX series (CURRENT) eVGA 128-A4 series Gainward GeForce4 PowerPack! Pro 6xx, Pro 4xx series Jaton 3DForce4 MX series Leadtek WinFast A1xx series MSI G4MX series Palit GF4 MX series Prolink PixelView GeForce4 MX series PNY Verto GeForce4 MX series VisionTek Xtasy GeForce4 MX 440, MX 420 XFX PVT17 series XFX PVT18 series XFX PVT97 series Abit Saluro GF4 Ti series Albatron Ti4xxx series Aopen Aeolis TI4200 series Aopen GF4TI series ASUS V84xx series ASUS V9280 series BFG Technologies GeForce 4 Ti series Club3D CGN-2xxx series Chaintech A-GXxx, A-GTxx series eVGA 64-A8 series eVGA 128-A8 series

GeForce4 Ti series (CURRENT)

Gainward GeForce4 PowerPack! Ultra/750 series Gainward GeForce4 PowerPack! Ultra/650 series Jaton 3DForce4 Ti series Leadtek WinFast A2xx series MSI G4Ti series Palit GF4Ti series PNY Verto GeForce4 Ti series


Prolink PixelView GeForce4 Ti series VisionTek Xtasy GeForce4 Ti 4600 XFX PVT25 series XFX PVT28 series

Albatron FX5800 series Asus V9900

GeForce FX (CURRENT)[1]

BFG Technologies Asylum 5800 series PNY GeForce FX 5800 Prolink PixelView GeForceFX 5800 series

S3 Graphics [2]

SavageXP

Tyan Tachyon G3300

SiS

6326 (OLD)

SiS300 (RECENT)

Aopen PA300 VR Pine PT-5985 Aopen PA305 DCS WS305

SiS305 (RECENT) Pine Technology 3D Phantom XP-2800 Chaintech AGP SI40 CP Technology S315 series CP Technology CS315 DV Chaintech SIS151 Elitegroup AG315 series Gainward CardEXPERT SiS315 Hightech MX315 series Jaton 3DForce S-64 series Jaton 3DForce S-128 series

SiS315 (RECENT)

Jetway Magic 315 series Jetway 315B series Joytech Apollo 3D Thrill 315 series Pine Technology 3D Phantom XP-3800 series, XP2800 Transcend TS32MVDS3 USI VP-315S1/S2 Yuan AGP-125


Yuan AGP-130

Joytech Xabre 80 series

Xabre 80

Max Diligent X800 series Triplex Millenium Silver Xabre Lite Chaintech A-S420 series Explorer Xabre 200 Joytech Xabre 200 series

Xabre 200

Max Diligent X820 series PowerColor XP200-B1 Triplex Millenium Silver Xabre Plus Vinix VX-3320 series Aopen Xabre 400 series Chaintech A-S440 series Club3D CGS-4xx series DFI X-400 series ECS AG400 series Explorer Xabre 400 Honnie Triplex Xabre 400

Xabre 400 Jaton 3DForce Xabre 400 series Joytech Xabre 400 series Max Diligent X840 series Pine Technology 3D Phantom XP-8200 Power Color XP400-B3 Triplex Millenium Silver Xabre Pro Vinix VX-3340 series VideoLogic Vivid!

ST Microelectronics/PowerVR

KYRO PowerVR Kyro (RECENT)

InnoVISION Inno3D KYRO 2000 Hercules 3D Prophet 4000 XT Club3D Kyro II

KYRO PowerVR Kyro II (CURRENT)

Hercules 3D Prophet 4500 series InnoVISION Kyro II 4500 VideoLogic Vivid!XS, XS Elite


[1] GeForce FX-based products began to ship in quantities in the second quarter of 2003. See the Web

sites of the vendors' partners listed for other NVIDIA chipsets for the latest graphics cards to use with this chipset. [2] The desktop graphics assets of the former S3 company (now known as SONICblue) have been

transferred to this VIA/SONICblue joint venture. S3 sold its Diamond Fire GL workstation graphics product line to ATI (which continues to develop and sell it), but it continues to support its Diamondbranded desktop and workstation cards at http://www.sonicblue.com. For the latest vendors to use SavageXP, see the S3 Graphics Web site at http://www.s3graphics.com.


Upgrading or Replacing Your Video Card With current developments in rock-bottom video card pricing, improvements in 3D display technology, and massive amounts of high-speed memory available on new video cards (up to 256MB!), it makes little sense today to add most upgrades to an existing video card. The component-level upgrades that can be added generally include: TV tuners, permitting you to watch cable or broadcast TV on your monitor Video capture devices, allowing you to capture still or moving video to a file If you need better 3D performance, more memory, or support for DVI digital displays, you need to replace your video card.

TV Tuner and Video Capture Upgrades Most video cards don't have TV tuner and video capture upgrade features built in. New cards with these features tend to be either in the middle to high range of their manufacturers' price structures or less expensive but of poor quality. These features are exciting if you are already into video editing, want to add video to your Web site, or are trying to create CD-R/CD-RW archives of your home video. If you have an up-to-date video card with acceptable 2D and 3D performance and at least 16MB of video RAM, compare the price of the add-ons to the price of a new card with these features. You'll probably find the add-ons to be less expensive. If your card has 8MB of video RAM or less, I'd recommend replacing it with a new card with these features. Look at sample video captures before making your decision because all video capture solutions require image compression with at least some loss of quality. If you have a digital camcorder with IEEE-1394 (also called FireWire or i.Link) ports, you should purchase an IEEE-1394 interface board instead to use high-quality pure digital video that needs no conversion. Note that ATI's All-in-Wonder RADEON 8500 DV includes an IEEE-1394 port as well as TV-in capability, allowing you to use a single card for high-performance business and gaming graphics as well as digital and analog video capture. The USB port found on recent systems can be used to connect TV tuner and video-capture options compatible with any manufacturer's video card, such as the ATI TV-Wonder USB Edition or Hauppauge's WinTV-USB. Because the wide variety of TV and computer hardware on the market can cause compatibility problems


with USB TV/capture devices, be sure to check review sites such as http://computers.cnet.com or http://www.epinions.com.


Note If you're planning to upgrade to a new video card in the next year or two, find out whether the TV or video add-on you want to buy for your present video card will work with the new model. If it's a TV tuner, you might need to consider other cards from the same vendor your current video card came from. If you're unsure of future compatibility, hold off on your purchase or consider buying a card with the desired features included.

Warranty and Support Because a video card can go through several driver changes during its effective lifetime (about three years or two operating-system revisions), buying a video card from a major manufacturer usually assures you of better support during the card's lifetime. If you buy a card that uses widely available chipsets (such as NVIDIA's), you might be able to try a different vendor's version of drivers or use the chipset vendor's "generic" drivers if you don't get satisfactory support from your card vendor. Keep in mind that using generic drivers (chipset level) or a different brand of drivers can cause problems if your original card's design was tweaked from the chipset maker's reference design. Look at the vendor's technical support forums or third-party discussions on newsgroups, computer information Web sites such as ZDNet, or magazine Web sites to get a feel for the stability, reliability, and usefulness of a vendor's support and driver services. These sources generally also provide alternatives in case of difficulties with a specific brand or chipset. If you use Windows Me, Windows 2000, or Windows XP, make sure you use WHQLcertified drivers for best results. These drivers have been passed by Microsoft's Windows Hardware Quality Labs and might be available through Windows Update or from the vendor's own Web site.

Comparing Video Cards with the Same Chipset Many manufacturers create a line of video cards with the same chipset to sell at different pricing points. Why not save some dollars and get the cheapest model? Why not say "price is no object" and get the most expensive one? When you're faced with various cards in the "chipsetX" family, look for differences such as those shown in Table 15.16. Table 15.16. Comparing Video Cards with the Same Features Feature

Effect on You

Most current 3D accelerator cards use a 300MHz or faster RAMDAC, which provides flicker-free resolutions RAMDAC beyond 1024x768. However, less-expensive cards in older designs often used a slower RAMDAC, which


speed

reduces maximum and flicker-free resolutions. If you use a 17'' or larger monitor, this could be an eyestraining problem.

Although AGP video cards can use AGP memory (a section of main memory borrowed for texturing), performing as much work as possible on the card's own memory is still faster. PCI cards must perform all Amount functions within their own memory. Less expensive cards in a chipset family often have lower amounts of of RAM memory onboard, and most current-model cards aren't expandable. Buy a card with enough memory for your games or applications​t oday and tomorrow; at least 32MB or more for business and 64MB or more for gaming, 3D graphics, and video-related work. Virtually all video cards on the market today use SDRAM or its faster variants (SGRAM, DDR SDRAM, or Memory DDR-II SDRAM). Any of these provides you with high performance in business applications, although DDR or type DDR-II SDRAM is preferable when running high-resolution, high-quality 3D games faster.

Core clock speed

Many suppliers adjust the recommended speed of graphics controllers in an effort to provide users with maximum performance. Sometimes the supplier can choose to exceed the rated specification of the graphics chip. Be cautious: Current controller chips are large and can overheat. An overclocked device in an open system with great airflow might work, or it might fail in a period of months from overstress of the circuitry. If you have questions about the rated speed of a controller, check the chip supplier's Web site. Many reputable companies do use overclocked parts, but the best vendors supply large heatsinks or powered fans to avoid overheating. Some vendors even provide on-card temperature monitoring.

RAM Speed (ns rating)

Just as faster system RAM improves overall computer performance, faster video card RAM improves video card performance. Some high-performance 3D video cards now use DDR SDRAM memory chips with a 2.8ns access time.

TV-out

Once a rare feature, many mid-range and most high-end video cards now feature TV-out, enabling you to display DVD movies or video games on a big-screen TV. Some of the latest models have hardware-based MPEG-2 compression for higher video quality in less disk space. Some of the latest video cards are now using a VIVO-out port to support either RCA or S-video inputs on VCRs and TVs.


Video Cards for Multimedia Multimedia​including live full-motion video feeds, videoconferencing, and animations​has become an important element of the personal computing industry and is helping to blur the once-solid lines between computer and broadcast media. As the demand for multimedia content increases, so do the capabilities of the hardware and software used to produce the content. Video is just one, albeit important, element of the multimedia experience, and the graphics adapters on the market today reflect the demand for these increased capabilities. Producing state-of-the-art multimedia content today often requires that the PC be capable of interfacing with other devices, such as cameras, VCRs, and television sets, and many video adapters are now equipped with these capabilities. Other multimedia technologies, such as 3D animation, place an enormous burden on a system's processing and data-handling capabilities, and many manufacturers of video adapters are redesigning their products to shoulder this burden more efficiently. The following sections examine some of the video adapter components that make these technologies possible and practical, including VFC, VAFC, VMC, and VESA VIP. Because none of these specifications for internal video feature connectors has become a true industry standard, some manufacturers of auxiliary video products​such as dedicated 3D accelerator boards and MPEG decoders​have taken an alternative route through the standard VGA connector.

Video Feature Connectors To extend the capabilities of the VGA standard beyond direct connections to a monitor, several auxiliary connector standards have been devised, first by individual card makers and later by VESA. Four early attempts to create a common connector were the Video Feature Connector (VFC) that IBM devised in 1987, the VESA Advanced Feature Connector (VAFC), the VESA Media Channel (VMC), and the VESA Video Interface Port (VESA VIP). These connector designs were not widely used, though.


Note If you are interested in reading more about VFC, VAFC, VMC, and VESA VIP, see "Video Feature Connectors," in the Technical Reference section of the DVD with this book.

Replacements for the VESA VIP and Other Video Connectors Currently, most systems interface with video devices through their USB or IEEE1394 ports. Add-on TV tuner cards and USB devices such as the ATI TV Wonder VE can be used with most DirectX-compatible video chipsets from ATI and NVIDIA for video capture. The ATI All-in-Wonder series provides TV-in, TV-out, and fullpower graphics support in a single slot. The latest version, the ATI All-in-Wonder 9700 Pro, has S-video and composite input and output ports and a 125-channel stereo TV tuner.

Video Output Devices When video technology first was introduced, it was based on television. However, a difference exists between the signals used by a television and those used by a computer. In the United States, the National Television System Committee (NTSC) established color TV standards in 1953. Some other countries, such as Japan, followed this standard. Many countries in Europe, though, developed more sophisticated standards, including Phase Alternate Line (PAL) and Sequential Couleur Avec MĂŠmoire (SECAM). Table 15.17 shows the differences among these standards. Table 15.17. Television Versus Computer Monitors Standard

Year Est.

Country

Lines

Rate

Television 1953 (color) NTSC

U.S., Japan

525

60 fields/sec

1941 (b&w) PAL

1941

Europe[1]

625

50 fields/sec

SECAM

1962

France

625

25 fields/sec

640x480[2]

72Hz

Computer VGA

1987

U.S.


Field = 1/2 (.5 frame)

[1] England, Holland, and West Germany.

[2] VGAis based on more lines and uses pixels (480) versus lines; genlocking is used to lock pixels into

lines and synchronize computers with TV standards.

A video-output (or VGA-to-NTSC) adapter enables you to display computer screens on a TV set or record them onto videotape for easy distribution. These products fall into two categories: those with genlocking (which enables the board to synchronize signals from multiple video sources or video with PC graphics) and those without. Genlocking provides the signal stability necessary to obtain adequate results when recording to tape, but it isn't necessary for using a television as a video display. VGA-to-NTSC converters are available as internal expansion boards, external boxes that are portable enough to use with a laptop for presentations on the road, and TV-out ports on the rear of many mid-range and high-end video cards using chipsets from NVIDIA, ATI, and others. Indeed, many laptop and notebook systems these days come equipped with a built-in VGA-to-NTSC converter. The converter does not replace your existing video adapter but instead connects to the adapter using an external cable. In addition to VGA input and output ports, a video output board has a video output interface for S-video and composite video. Most VGA-to-TV converters support the standard NTSC television format and might also support the European PAL format. The resolution these devices display on a TV set or record on videotape often is limited to straight VGA at 640x480 pixels, although some TV-out ports on recent video cards can also display 800x600 resolution. The converter also might contain an antiflicker circuit to help stabilize the picture because VGA-to-TV products, as well as TV-to-VGA solutions, often suffer from a case of the jitters.

Video Capture Devices You can capture individual screen images or full-motion video for reuse in several ways, including 3D accelerator cards with TV-in ports TV tuner cards


USB or parallel port-based devices such as TV tuner/capture devices discussed earlier or a dedicated device such as the SnapMAGIC (available from www.snapnsend.com) Webcams with video input ports These units capture still or moving images from NTSC video sources, such as camcorders and VCRs. Although image quality is limited by the input signal, the results are still good enough for presentations and desktop publishing applications. These devices work with 8-, 16-, and 24-bit VGA cards and usually accept video input from VHS, Super VHS, and Hi-8 devices. As you might expect, however, Super VHS and Hi-8 video sources give better results, as do configurations using more than 256 colors. For the best results, use DV camcorders equipped with IEEE-1394 (i.Link/FireWire) connectors; these can output high-quality digital video direct to your computer without the need to perform an analog-to-digital conversion. Although a few computers feature builtin IEEE-1394 ports, you must install an IEEE-1394 add-in card into most computers if you want to capture output from a DV camcorder.

Desktop Video Boards You can also capture NTSC (television) signals to your computer system for display or editing. In other words, you can literally watch TV in a window on your computer. When capturing video, you should think in terms of digital versus analog. The biggest convenience of an analog TV signal is efficiency; it is a compact way to transmit video information through a low bandwidth pipeline. The disadvantage is that although you can control how the video is displayed, you can't edit it. Actually capturing and recording video from external sources and saving the files onto your PC requires special technology. To do this, you need a device called a video capture board (also called a TV tuner, video digitizer, or video grabber).


Note In this context, the technical nomenclature again becomes confusing because the term video here has its usual connotation; that is, it refers to the display of full-motion photography on the PC monitor. When evaluating video hardware, be sure to distinguish between devices that capture still images from a video source and those that capture full-motion video streams.

Today, video sources come in two forms: Analog Digital Analog video can be captured from traditional sources such as broadcast or cable TV, VCRs, and camcorders using VHS or similar tape standards. This process is much more demanding of storage space and system performance than still images are. Here's why. The typical computer screen was designed to display mainly static images. The storing and retrieving of these images requires managing huge files. Consider this: A single, full-screen color image in an uncompressed format can require as much as 2MB of disk space; a 1-second video would therefore require 45MB. Likewise, any video transmission you want to capture for use on your PC must be converted from an analog NTSC signal to a digital signal your computer can use. On top of that, the video signal must be moved inside your computer at 10 times the speed of the conventional ISA bus structure. You need not only a superior video card and monitor, but also an excellent expansion bus, such as PCI or AGP. Considering that full-motion video can consume massive quantities of disk space, it becomes apparent that data compression is all but essential. Compression and decompression apply to both video and audio. Not only does a compressed file take up less space, it also performs better simply because less data must be processed. When you are ready to replay the video/audio, the application decompresses the file during playback. In any case, if you are going to work with video, be sure that your hard drive is large enough and fast enough to handle the huge files that can result. Compression/decompression programs and devices are called codecs. Two types of codecs exist: hardware-dependent codecs and software (or hardwareindependent) codecs. Hardware codecs typically perform better; however, they require additional hardware​either an add-on card or a high-end video card with hardware codecs built in. Software codes do not require hardware for compression or playback, but they typically do not deliver the same quality or compression ratio. Two of the major codec algorithms are


JPEG (Joint Photographic Experts Group). Originally developed for still images, JPEG can compress and decompress at rates acceptable for nearly full-motion video (30fps). JPEG still uses a series of still images, which makes editing easier. JPEG is typically lossy (meaning that a small amount of the data is lost during the compression process, slightly diminishing the quality of the image), but it can also be lossless. JPEG compression functions by eliminating redundant data for each individual image (intraframe). Compression efficiency is approximately 30:1 (20:1​40:1). MPEG (Motion Picture Experts Group). MPEG by itself compresses video at approximately a 30:1 ratio, but with precompression through oversampling, the ratio can climb to 100:1 and higher, while retaining high quality. Thus, MPEG compression results in better, faster videos that require less storage space. MPEG is an interframe compressor. Because MPEG stores only incremental changes, it is not used during editing phases. If you will be capturing or compressing video on your computer, you'll need software based on standards such as Microsoft's DirectShow (the successor to Video for Windows and ActiveMovie), Real Network's Real Producer series, or Apple's QuickTime Pro. Players for files produced with these technologies can be downloaded free from the vendors' Web sites. To play or record video on your multimedia PC (MPC), you need some extra hardware and software: Video system software, such as Apple's QuickTime for Windows or Microsoft's Windows Media Player. A compression/digitization video adapter that enables you to digitize and play large video files. An NTSC-to-VGA adapter that combines TV signals with computer video signals for output to a VCR. Video can come from a variety of sources: TV, VCR, video camera, laserdisc player, or DVD player. When you record an animation file, you can save it in a variety of file formats: AVI (Audio Video Interleave), MOV (Apple QuickTime format), or MPG (MPEG format). Depending on the video-capture product you use, you have several choices for capturing analog video. The best option is to use component video. Component video uses three RCA-type jacks to carry the luminance (Y) and two chrominance (PR and PB) signals; this type of connector commonly is found on DVD players and high-end conventional and HDTV television sets. However, home-market video capture devices usually don't support component video. A typical


professional capture device designed for component video, such as Pinnacle Systems' DC2000DV, retails for about $2,000. The next best choice, and one that is supported by many home-market videocapture devices, is the S-video (S-VHS) connector. This cable transmits separate signals for color (chroma) and brightness (luma). Otherwise, you must use composite video, which mixes luma and chroma. This results in a lower-quality signal, and the better your signal, the better your video quality will be. You also can purchase devices that display only NTSC (TV) signals on your computer. The built-in digital movie editing features found in Windows Me and Windows XP, the increasing popularity of computer/TV solutions, and broadband Internet connections make onscreen full-motion video an increasingly common part of the computing experience. Because of the growing importance of onscreen full-motion video, more and more recent CPUs have added features to enhance playback​including MMX and SSE instructions found in the Pentium II, Pentium III, Celeron, and AMD Athlon and Duron and the instruction set found in the Intel Pentium 4's NetBurst microarchitecture and SSE2. Table 15.18 provides a breakdown of some common video cards and capture devices supporting key features. This table is not inclusive and is meant to serve only as a reference example. Table 15.18. Video Capture Devices Device Type Video card with TV tuner/capture

Example

ATI All-in-Wonder 9700 PRO

TV-tuner/capture breakout NVIDIA Personal Cinema (works with certain video cards with GeForce2, GeForce3, and box GeForce4 MX GPUs) PCI TV-tuner and capture card

ATI-TV Wonder VE

USB port video capture

Dazzle Digital Video Creator series

Parallel port still image capture

Invisco SnapMagic

PCI video capture card

Broadway Pro

IEEE-1394 (FireWire)

AVerMedia AverDV

Figure 15.11 shows a typical video adapter incorporating TV tuner and video-in and video-out features: the ATI All-in-Wonder 9700 Pro.


Figure 15.11. ATI's All-in-Wonder 9700 Pro is a high-end video accelerator with integrated TV tuner and video-capture features. Photos courtesy of ATI Technologies.

Each type of device has advantages and potential disadvantages. Table 15.19 provides a summary that will help you decide which solution is best for you. Table 15.19. Multimedia Device Comparison Device Type

Pros

Cons

Graphics card with built-in TV tuner and capture

Convenience; single-slot solution

TV-tuner card

Allows upgrade to existing graphics Might not work with all current chipsets. cards; might be movable to newer models

Upgrading requires card replacement.

Parallel port Universal usage on desktop or notebook Frame rate limited by speed of port; best for still-image attachment computer; inexpensive capture.

USB port attachment

Might not work on Windows 95B OSR 2.x with USB; requires Easy installation on late-model USBactive USB port; not all devices may be compatible with equipped computers with Windows 98/Me Windows 2000/XP; low bandwidth; not suitable for high-res and Windows 2000/XP or full-motion applications.

Dedicated Fast frame rate for realistic video; works PCI interface with any graphics card card

High resource requirements (IRQ and so on) on some models; requires internal installation.

IEEE-1394 (FireWire) connection to digital

Requires IEEE-1394 interface card, IEEE-1394 digital video source; card requiresvideo internal installation; some cards don't include capture/editing software;verify that editing software purchased separately works with card.

No conversion from analog to digital needed; all-digital image is very high quality without compression artifacts (blocky areas) in video; fast throughput


Troubleshooting Video Capture Devices Table 15.20 provides some advice for troubleshooting problems with video capture devices. Note that IRQ conflicts can be an issue with both parallel port and add-on card devices and that low-bandwidth devices such as parallel port or USB devices might not be capable of supporting full-motion video capture except in a small window. Table 15.20. Troubleshooting Video Capture Devices Device Type

Parallel port attachment

Problem Can't detect device, but printers work okay.

TV tuners (built-in No picture. graphics card or add-on)

Solutions

Check port settings; device might require IEEE-1284 settings (EPP and ECP); change in BIOS; ensure device is connected directly to port; avoid daisy-chaining devices unless device specifically allows it; check Windows Device Manager for IRQ conflicts.

Check cabling; set signal source correctly in software; update software.

All devices

Frame rate is too low. Increasing it might require capturing video in a smaller window; Video capture is use fastest parallel port setting you can; use faster CPU and increase RAM to jerky. improve results.

All devices

Video playback Hard disk might be pausing for thermal recalibration; use AV-rated SCSI hard drives has pauses, or UDMA EIDE drives; install correct busmastering EIDE drivers for motherboard dropped frames. chipset to improve speed.

USB devices

Device can't be detected or doesn't work properly.

Card can't be Interface cards detected or (all types) doesn't work.

Use Windows 98/Me/2000/XP; late versions of Windows 95 have USB drivers, but they often don't work. If you use a USB hub, be sure it's powered.

Check for IRQ conflicts in Windows Device Manager; consider setting card manually if possible.

IEEE-1394 cards

Card can't be detected or doesn't work.

Make sure power connector is attached to card if card has 4-pin power jack. Make sure correct drivers are installed.

All devices

Capture or installation problems.

Use the newest drivers available; check manufacturers' Web site for updates, FAQs, and so on.


Adapter and Display Troubleshooting Solving most graphics adapter and monitor problems is fairly simple, although costly, because replacing the adapter or display is the normal procedure. However, before you take this step, be sure that you have exhausted all your other options. One embarrassingly obvious fix to monitor display problems that is often overlooked by many users is to adjust the controls on the monitor, such as the contrast and brightness. Although most monitors today have a control panel on the front of the unit, other adjustments might be possible as well. Some NEC monitors, for example, have a focus adjustment screw on the left side of the unit. Because the screw is deep inside the case, the only evidence of its existence is a hole in the plastic grillwork on top of it. To adjust the monitor's focus, you must stick a long-shanked screwdriver about 2'' into the hole and feel around for the screw head. This type of adjustment can save you both an expensive repair bill and the humiliation of being ridiculed by the repair technician. Always examine the monitor case, documentation, and manufacturer's Web site or other online services for the locations of adjustment controls. A defective or dysfunctional adapter or display usually is replaced as a single unit rather than being repaired. Except for specialized CAD or graphics workstation​oriented adapters, virtually all of today's adapters cost more to service than to replace, and the documentation required to service the hardware properly is not always available. You usually can't get schematic diagrams, parts lists, wiring diagrams, and other documents for most adapters or monitors. Also, virtually all adapters now are constructed with surface-mount technology that requires a substantial investment in a rework station before you can remove and replace these components by hand. You can't use a $25 pencil-type soldering iron on these boards! Servicing monitors is a slightly different proposition. Although a display often is replaced as a whole unit, some displays​particularly 20'' or larger CRTs or most LCD panels​might be cheaper to repair than to replace. If you decide to repair the monitor, your best bet is to either contact the company from which you purchased the display or contact one of the companies that specializes in monitor depot repair. If your monitor has a 15'' diagonal measurement or less, consider replacing it with a unit that is 17'' or larger because repair costs on small monitors come close to replacement costs and large monitors aren't much more expensive these days. Depot repair means you send in your display to repair specialists who either fix your particular unit or return an identical unit they have already repaired. This


usually is accomplished for a flat-rate fee; in other words, the price is the same no matter what they have done to repair your actual unit. Because you usually get a different (but identical) unit in return, they can ship out your repaired display immediately on receiving the one you sent in, or even in advance in some cases. This way, you have the least amount of downtime and can receive the repaired display as quickly as possible. In some cases, if your particular monitor is unique or one they don't have in stock, you must wait while they repair your specific unit. Troubleshooting a failed monitor is relatively simple. If your display goes out, for example, a swap with another monitor can confirm that the display is the problem. If the problem disappears when you change the display, the problem is almost certainly in the original display or the cable; if the problem remains, it is likely in the video adapter or PC itself. Many of the better quality, late-model monitors have built-in self-diagnostic circuitry. Check your monitor's manual for details. Using this feature, if available, can help you determine whether the problem is really in the monitor, in a cable, or somewhere else in the system. If self diagnostics produce an image onscreen, look to other parts of the video subsystem for your problem. The monitor cable can sometimes be the source of display problems. A bent pin in the DB-15 connector that plugs into the video adapter can prevent the monitor from displaying images, or it can cause color shifts. Most of the time, you can repair the connector by carefully straightening the bent pin with sharp-nosed pliers. If the pin breaks off or the connector is otherwise damaged, you can sometimes replace the monitor cable. Some monitor manufacturers use cables that disconnect from the monitor and video adapter, whereas others are permanently connected. Depending on the type of connector the device uses at the monitor end, you might have to contact the manufacturer for a replacement. If you narrow down the problem to the display, consult the documentation that came with the monitor or call the manufacturer for the location of the nearest factory repair depot. Third-party depot repair service companies are also available that can repair most displays (if they are no longer covered by a warranty); their prices often are much lower than factory service. Check the Vendor List on the DVD for several companies that do depot repair of computer monitors and displays.


Caution You should never attempt to repair a CRT monitor yourself. Touching the wrong component can be fatal. The display circuits can hold extremely high voltages for hours, days, or even weeks after the power is shut off. A qualified service person should discharge the cathode ray tube and power capacitors before proceeding.

For most displays, you are limited to making simple adjustments. For color displays, the adjustments can be quite formidable if you lack experience. Even factory service technicians often lack proper documentation and service information for newer models; they usually exchange your unit for another and repair the defective one later. Never buy a display for which no local factory repair depot is available. If you have a problem with a display or an adapter, it pays to call the manufacturer, who might know about the problem and make repairs available. Sometimes, when manufacturers encounter numerous problems with a product, they might offer free repair, replacements, or another generous offer that you would never know about if you did not call. Remember, also, that many of the problems you might encounter with modern video adapters and displays are related to the drivers that control these devices rather than to the hardware. Be sure you have the latest and proper drivers before you attempt to have the hardware repaired; a solution might already be available.

Troubleshooting Monitors Problem No picture. Solution If the LED on the front of the monitor is yellow or flashing green, the monitor is in power-saving mode. Move the mouse or press Alt+Tab on the keyboard and wait up to 1 minute to wake up the system if the system is turned on. If the LED on the front of the monitor is green, the monitor is in normal mode (receiving a signal), but the brightness and contrast are set incorrectly; adjust them. If no lights are lit on the monitor, check the power and power switch. Check the surge protector or power director to ensure that power is going to the monitor. Replace the power cord with a known-working spare if necessary. Retest. Replace


the monitor with a known-working spare to ensure that the monitor is the problem. Check data cables at the monitor and video card end. Problem Jittery picture quality. Solution LCD monitors. Use display-adjustment software or onscreen menus to reduce or eliminate pixel jitter and pixel swim. All monitors. Check cables for tightness at the video card and the monitor (if removable): Remove the extender cable and retest with the monitor plugged directly into the video card. If the extended cable is bad, replace it. Check the cables for damage; replace as needed. If problems are intermittent, check for interference. (Microwave ovens near monitors can cause severe picture distortion when turned on.) CRT monitors. Check refresh-rate settings; reduce them until acceptable picture quality is achieved: Use onscreen picture adjustments until an acceptable picture quality is achieved. If problems are intermittent and can be "fixed" by waiting or gently tapping the side of the monitor, the monitor power supply is probably bad or has loose connections internally. Service or replace the monitor.

Troubleshooting Video Cards and Drivers Problem Display works in DOS but not in Windows. Solution


If you have an acceptable picture quality in MS-DOS mode (system boot) but no picture in Windows, most likely you have an incorrect or corrupted video driver installed in Windows. Boot Windows 9x/Me in Safe Mode (which uses a VGA driver), boot Windows 2000/XP in Enable VGA mode, or install the VGA driver and restart Windows. If Safe Mode or VGA Mode works, get the correct driver for the video card and reinstall. If you have overclocked your card with a manufacturer-supplied or third-party utility, you might have set the speed too high. Restart the system in Safe Mode, and reset the card to run at its default speed. Problem Can't replace built-in video card with add-on PCI video card. Solution Check with the video card and system vendor for a list of acceptable replacement video cards. Try another video card with a different chipset. Check the BIOS or motherboard for jumper or configuration settings to disable built-in video. Place the add-on card in a different PCI slot. Problem Can't select desired color depth and resolution combination. Solution Verify that the card is properly identified in Windows and that the card's memory is working properly. Use diagnostic software provided by the video card or chipset maker to test the card's memory. If the hardware is working properly, check for new drivers. Problem Can't select desired refresh rate. Solution Verify that the card and monitor are properly identified in Windows. Obtain updated drivers for the card and monitor.


DisplayMate DisplayMate is a unique diagnostic and testing program designed to thoroughly test your video adapter and display. It is somewhat unique in that most conventional PC hardware diagnostics programs do not emphasize video testing the way this program does. I find it useful not only in testing whether a video adapter is functioning properly, but also in examining video displays. You easily can test the image quality of a display, which allows you to make focus, centering, brightness and contrast, color level, and other adjustments much more accurately than before. If you are purchasing a new monitor, you can use the program to evaluate the sharpness and linearity of the display and to provide a consistent way of checking each monitor you are considering. If you use projection systems for presentations​as I do in my PC hardware seminars​you will find it invaluable for setting up and adjusting the projector. DisplayMate also can test a video adapter thoroughly. It sets the video circuits into each possible video mode so you can test all its capabilities. It even helps you determine the performance level of your card, both with respect to resolution and colors as well as speed. You can then use the program to benchmark the performance of the display, which enables you to compare one type of video adapter system to another. See the Vendor List on the CD for more information on DisplayMate (formerly Sonera) Technologies, or visit www.displaymate.com. Some video-card vendors supply a special version of DisplayMate for use in diagnostics testing.


Chapter 16. Audio Hardware Since the first edition of this book was published in 1988, a lot has happened to audio hardware. Although rudimentary audio capabilities were part of the original IBM PC of 1981 and its many successors, audio was used on early computers for troubleshooting rather than for creative tasks. Computers used beeps for little other than to signal problems such as a full keyboard buffer or errors during the power on self test (POST) sequence. The Macintosh, first introduced in 1984, included high-quality audio capabilities in its built-in hardware, but PCs did not gain comparable audio capabilities until the first add-on sound cards from companies such as Ad Lib and Creative Labs were developed in the late 1980s. Thanks to competition among many companies, we now enjoy widely supported de facto hardware and software standards for audio. Audio hardware has gone from being an expensive, exotic add-on to being an assumed part of virtually any system configuration. Today's PC audio hardware might take one of the following forms: An audio adapter on a PCI expansion card that you install into a bus slot in the computer. A sound chip located on the motherboard, using sound chips from companies such as Crystal, Analog Devices, Sigmatel, ESS, or others. Hardware that's integrated into the motherboard's main chipset, as with some of the most recent chipsets for computers developed by Intel, SiS, NVIDIA, and VIA Technologies. A unique series of motherboards from Aopen combines this with a vacuum tube in the amplifier circuit catering to the audio purist. Regardless of their location, the audio features use jacks for speakers and a microphone. In addition, many of them provide dedicated jacks for MIDI hardware (and some also provide an analog game port for joysticks). As you will see later in this chapter, many mid-range and high-end audio adapters also support sophisticated digital audio input and output. On the software side, the audio adapter requires the support of a driver that you install either directly from an application or in your computer's operating system. This chapter focuses on the audio products found in today's PCs, their uses, and how you install and operate them.


Early PC Audio Adapters When the first audio adapters were introduced in the late 1980s by companies such as AdLib, Roland, and Creative Labs, they were aimed squarely at a gaming audience. The 1989-vintage Creative Labs Game Blaster card cost more than $100 but supported only a handful of games. And, because no sound standards existed at the time, the sound card you selected could turn out to be useless with some games if the game manufacturer chose not to specifically support it.


Note About the same time as the release of the Game Blaster, hardware supporting the Musical Instrument Digital Interface (MIDI) became available for the PC. At this time, however, such hardware was used only in very specialized recording applications. As MIDI support became a more common feature in musical instruments, though, it also became a more affordable PC add-on.

The Game Blaster was soon replaced by the Sound Blaster, which was compatible with the AdLib sound card and the Creative Labs Game Blaster card, enabling it to support games that specified one sound card or the other. The Sound Blaster included a built-in microphone jack, stereo output, and a MIDI port for connecting the PC to a synthesizer or other electronic musical instrument. This established a baseline of features that would be supported by virtually all other sound cards and onboard sound features up to the present. Finally, the audio adapter had the potential for uses other than games. The follow-up Sound Blaster Pro featured improved sound when compared to the original Sound Blaster. The Sound Blaster Pro and its successors eventually triumphed over earlier rivals to become de facto standards for PC sound reproduction.


Note Unlike de jure standards such as the IEEE-1394 port, which is an official standard of the IEEE organization, de facto standards are those that develop informally due to the widespread acceptance of the market leader's products in a particular segment of the marketplace. The Sound Blaster Pro is just one of many examples of a de facto standard: IBM's VGA card became a de facto baseline standard for video, and HP and Apple's different printer languages (HP PCL and Adobe PostScript) became de facto standards for printers.

Limitations of Sound Blaster Pro Compatibility Through the mid-1990s, while MS-DOS was the standard gaming platform, many users of non-Creative Labs sound cards struggled with the limitations of their hardware's imperfect emulation of the Sound Blaster Pro. Ideally, a Sound Blaster Pro​compatible card would be capable of using the same IRQ, DMA, and I/O port addresses as a Sound Blaster Pro card from Creative Labs and would be used by an application program in the same way as an actual Sound Blaster Pro. Unfortunately, some cards required two separate sets of hardware resources, using one set of IRQ, DMA, and I/O port addresses for native mode and a second set for Sound Blaster Pro compatibility. Others worked well within Windows or within an MS-DOS session running with Windows in the background but required the user to install a DOS-based Terminate and Stay Resident (TSR) driver program to work in MS-DOS itself. If you've never needed to configure a game for a particular sound card, you're probably playing all 32-bit Windows games. Windows applications use the operating system's drivers to interface with hardware, relieving the software developer from needing to write different code for different sound cards, 3D graphics cards, and so on. For 3D sound and gaming graphics, Microsoft Windows uses a technology called DirectX, which was first introduced in December 1995; the current version is DirectX 9.

DirectX and Audio Adapters Microsoft's DirectX is a series of application program interfaces (APIs) that sit between multimedia applications and hardware. Unlike MS-DOS applications that required developers to develop direct hardware support for numerous models and brands of audio cards, video cards, and game controllers, Windows uses DirectX to "talk" to hardware in a more direct manner than normal Windows drivers do. This improves program performance and frees the software developer from the need to change the program to work with different devices. Instead, a game developer must work with only the DirectX sound engine, DirectX 3D renderer, and DirectX modem or network interface routines.


For more information about DirectX and sound hardware, see "3D Audio," p. 930.

Thanks to DirectX, sound card and chipset developers are assured that their products will work with recent and current versions of Windows. However, if you still enjoy playing MS-DOS​based games, current audio adapters, chipsets, and integrated audio solutions still might present a compatibility challenge to you because of fundamental hardware differences between the ISA expansion slots used by classic Creative Labs and other sound cards and PCI slots, chipsets, and integrated audio. For more information about using PCI sound hardware with MS-DOS games, see "Legacy (MS-DOS) Game Support Issues," p. 918.


PC Multimedia History Virtually every computer on the market today is equipped with some type of audio adapter and a CD-ROM or CD-ROM​compatible drive such as a CD-RW or DVD drive. Computers equipped with an audio adapter and a CD-ROM​compatible drive are often referred to as multimedia PCs after the old MPC-1, MPC-2, and MPC-3 standards that were used to rate early multimedia computers. Since 1996, all computers with onboard sound and a CD-ROM or compatible optical drive have exceeded MPC-3 standards by increasingly huge margins.


Note For more information about the MPC series of multimedia standards, see the section "Multimedia" in Chapter 20 of Upgrading and Repairing PCs, 11th Edition, available in electronic form on the DVD-ROM accompanying this book.

Suggested Multimedia Minimums The following list of specifications gives you a comfortable multimedia experience today; in fact, even the lowest-cost computer currently available at retail should meet or exceed these specifications: CPU: 700MHz Pentium III, Celeron, Athlon, Duron, or other Pentium-class processor RAM: 128MB Hard disk: 20GB CD-ROM drive: 24X read or DVD-ROM drive​10X read (equals about 27X CDROM read rate) Audio sampling rate: 16-bit VGA video resolution: 1024x768 Color depth: 16.8 million colors (24-bit color) I/O devices: MIDI and USB Minimum operating system: Windows 98, Windows Me, or Windows XP Home Edition Note that although speakers or headphones are technically not a part of the MPC specification or the updated list, they are certainly required for sound reproduction. Additionally, sound input for voice recording or voice control also requires a microphone. Keep in mind that many prebuilt PCs include low-cost powered or unpowered speakers that you will likely want to replace with speakers or headphones that provide the mixture of size, frequency response, and overall sound quality you need.


Because the MPC specifications reflect multimedia's past, users who want to know what comes next need to turn somewhere else for guidance. Microsoft and Intel jointly manage the PC Design Guide Web site at www.pcdesguide.org, where the older PC99 and latest PC2001 specifications and addenda can be read and reviewed. These standards are widely followed by the industry and point the way to many other enhancements for audio hardware and software in the Windows environment. Although virtually every computer is a "multimedia PC" today, the features of the audio adapter in your system will help determine how satisfied you will be with the wide range of specialized uses for multimedia-equipped systems. Later in this chapter, you learn more about the features you need to specify to ensure your audio adapter​regardless of type​is ready to work for you.


Audio Adapter Features To make an intelligent purchasing decision, you should be aware of some audio adapter basic components and the features they provide, as well as the advanced features you can get on better audio adapters. The following sections discuss the features you should consider while evaluating audio adapters for your PC.

Basic Connectors Most audio adapters have the same basic external connectors. These 1/8'' minijack connectors provide the means of passing sound signals from the adapter to speakers, headphones, and stereo systems, and of receiving sound from a microphone, CD player, tape player, or stereo. The four types of connectors your audio adapter should have at a minimum are shown in Figure 16.1.

Figure 16.1. The basic input and output connectors that most audio adapters have in common.

The jacks shown in Figure 16.1 are usually labeled, but when setting up a computer on or under a desk, the labels on the back of the PC can be difficult to


see. One of the most common reasons a PC fails to produce any sound is that the speakers are plugged into the wrong socket. To avoid this problem, many consumer-oriented audio cards color-code the jacks according to specifications found in the PC99 Design Guide (see www.pcdesguide.org/documents/pc99icons.htm). The color-coding can vary on some audio adapters (or not be present at all). Regardless, the basic set of connections included on most audio cards is as follows: Stereo line, or audio, out connector (lime green). The line-out connector is used to send sound signals from the audio adapter to a device outside the computer. You can hook up the cables from the line-out connector to stereo speakers, a headphone set, or your stereo system. If you hook up the PC to your stereo system, you can have amplified sound. Stereo line, or audio, in connector (light blue). With the line-in connector, you can record or mix sound signals from an external source, such as a stereo system or VCR, to the computer's hard disk. Rear out or speaker/headphone connector (no standard color). Older sound cards often provided an amplified jack supplying up to 4 watts of power for use with unpowered speakers or headphones along with the line-out connector. Today, you are more likely to find this jack used for rear speakers in four-speaker setups. The rear out jack often is disabled by default; check your audio adapter properties or setup program to see whether you need to enable this port when you connect rear speakers.


Note If you have only a single speaker/line-out connector, you must carefully adjust your mixer volume control and the volume control on your amplified speakers to find the best quality sound. Don't use powered speakers with an already-amplified sound if you can avoid it.

Microphone, or mono, in connector (pink). The mono-in connector is used to connect a microphone for recording your voice or other sounds to disk. This microphone jack records in mono​not in stereo​and is therefore not suitable for high-quality music recordings. Many audio adapter cards use Automatic Gain Control (AGC) to improve recordings. This feature adjusts the recording levels on-the-fly. A 600ohm​10,000ohm dynamic or condenser microphone works best with this jack. Some inexpensive audio adapters use the line-in connector instead of a separate microphone jack. Game port (gold). The game port (also called the joystick connector) is a 15pin, D-shaped connector that can connect to any standard joystick or game controller. Sometimes the joystick port can accommodate two joysticks if you purchase an optional Y-adapter. Many computers already contain a joystick port as part of a multifunction I/O circuit on the motherboard or an expansion card. If this is the case, you must note which port your operating system or application is configured to use when connecting the game controller. Some of the latest sound cards and systems with onboard sound omit this connector because most recent game controllers support USB connectors. MIDI connector (gold). Audio adapters typically use the same joystick port as their MIDI connector. Two of the pins in the connector are designed to carry signals to and from a MIDI device, such as an electronic keyboard. In most cases, you must purchase a separate MIDI connector from the audio adapter manufacturer that plugs into the joystick port and contains the two round, 5pin DIN connectors used by MIDI devices, plus a connector for a joystick. However, high-end sound cards might use 5-pin MIDI ports connected to a daughtercard or a breakout box (see Figure 16.3, later in this chapter). Because their signals use separate pins, you can connect the joystick and a MIDI device at the same time. You need this connector only if you plan to connect your PC to external MIDI devices. You can still play the MIDI files found on many Web sites by using the audio adapter's internal synthesizer.

Figure 16.3. The Sound Blaster Audigy 2 Platinum (left) includes the Audigy 2 Drive internal, front-panel breakout box (right) to support the Audigy 2's many features.


Note Some recent systems with onboard audio lack the joystick/MIDI connector, but you can use digital game controllers that feature a USB connector to attach your game controllers to the USB port. And, as mentioned earlier, you can attach MIDI devices through the same USB connector with a suitable interface box.

In addition to the external connections, most sound cards feature at least one (and possibly multiple) internal CD-audio connectors. Most audio adapters have an internal 4-pin connector you can use to plug an internal CD-ROM drive directly into the audio adapter, using a small round cable. This connection enables the drive to send analog audio signals from the CD-ROM directly to the audio adapter, so you can play the sound through the computer's speakers. Some sound cards use a different connector than the CD-ROM, whereas others use the same type of connector as the CD-ROM (see Figure 16.2).

Figure 16.2. Typical CD-ROM analog audio cables.

To play audio CDs, you have two choices: The playback can be either analog or digital. Analog playback is supported via an analog audio cable connected between the drive and sound card. This cable does not carry data from the CD-ROM to the system bus; it connects the analog audio output of the CD-ROM drive directly to the audio amplifier in the sound card. With many drives and sound cards, if you don't connect an analog audio cable between the drive and card, you might be unable to play music CDs or hear some game audio. Most newer drives and cards support digital playback as well as the standard direct analog connection. If your drive and card support digital playback, open the Control Panel in Windows, click the Multimedia icon, and then select the CD Music tab. You should see settings for changing the default drive for playing CD music and a box you can check to enable digital CD audio for that drive. If the box is grayed out (meaning you can't check it), digital audio is not supported for that drive or card. Using digital audio enables multiple drives to play audio CDs. Typically, a sound


card has only a single analog audio connector, so if you have multiple optical drives, only one can have an analog audio connection to the sound card for playing audio CDs. If you want to play audio CDs on multiple drives, you must either enable digital CD audio for those drives or purchase a CD audio Y-cable. By either enabling digital audio or installing an analog audio cable, you should be able to play audio CDs with any given CD/DVD drive.

Connectors for Advanced Features Many of the newest sound cards are designed for advanced gaming, DVD audio playback, and sound production uses and have additional connectors to support these uses, such as the following: MIDI in and MIDI out. Some advanced sound cards don't require you to convert the game port (joystick port) to MIDI interfacing by offering these ports on a separate external connector. This permits you to use a joystick and have an external MIDI device connected at the same time. Typical location: external device. SPDIF (also called SP/DIF) in and SPDIF out. The Sony/Philips Digital Interface receives digital audio signals directly from compatible devices without converting them to analog format first. Typical location: external device. (SPDIF interfaces are also referred to by some vendors as "Dolby Digital" interfaces.)


Note SPDIF connectors use cables with the standard RCA jack connector but are designed to work specifically at an impedance of 75ohms​t he same as composite video cables. Thus, you can use RCA-jack composite video cables with your SPDIF connectors. Although audio cables are also equipped with RCA jacks, their impedance is different, making them a less desirable choice.

CD SPDIF. Connects compatible CD-ROM drives with SPDIF interfacing to the digital input of the sound card. Typical location: side of audio card. The cable used resembles the cable shown in Figure 16.2, but it uses only two wires. TAD in. Connects internal modems with Telephone Answering Device support to the sound card for sound processing of voice messages. Typical location: side of audio card. Digital DIN out. This supports multispeaker digital speaker systems, such as those produced by Cambridge for use with the SoundBlaster Live! series. Typical location: external device. Aux in. Provides input for other sound sources, such as a TV tuner card. Typical location: side of audio card. I2S in. This enables the sound card to accept digital audio input from an external source, such as two-channel decoded AC-3 from DVD decoders and MPEG-2 Zoom Video. Typical location: side of audio card. USB port. This enables the sound card to connect to USB speakers, game controllers, and other types of USB devices. The Hercules Game Theater XP series, the first sound card with built-in USB ports, supports USB 1.1 only. Typical location: external breakout box. IEEE-1394. This enables the sound card to connect to IEEE-1394-compatible DV camcorders, scanners, hard drives, and other devices. The Creative Labs Sound Blaster Audigy, Audigy 2, and Hercules Digifire 7.1 all feature one or more IEEE-1394 ports. Typical location: card bracket or external cable or breakout box. Sometimes, these additional connectors are found on the card itself, or sometimes they are attached to an internal or external breakout box, daughtercard, or external rack. For example, the Sound Blaster Audigy Platinum 2 EX, Audigy 2 Platinum, and Hercules Game Theater XP 6.1 and 7.1 are two-piece units. The Audigy audio adapter itself plugs into a PCI slot, but some additional connectors are routed to an internal breakout box on the Platinum, which fits into an unused 5 1/4'' drive bay (see Figure 16.3). The top-of-the-line Platinum EX uses an


external breakout box with the same connection options. Both models feature a remote control (not shown) and external volume controls (visible in Figure 16.3). The Hercules Game Theater XP series also features an external breakout box, which Hercules refers to as an audio rack. Figure 16.4 shows Voyetra Turtle Beach's Santa Cruz audio adapter card with the internal connectors common on today's 3D sound cards.

Figure 16.4. Voyetra Turtle Beach's Santa Cruz is a typical example of an advanced 3D sound card.

Adding Advanced Sound Features Without Replacing Onboard Audio Traditionally, audio card upgrades have been designed specifically for users of desktop computers, leaving the growing number of notebook computer users out in the cold if they decided that the bare-bones features of their integrated audio weren't sufficient. However, the Creative Labs Sound Blaster Extigy allows both notebook and desktop computer users to add advanced I/O features without opening the system. The front and rear panels of the Sound Blaster Extigy are shown in Figures 16.5 and 16.6, respectively. The Extigy connects to any system that has a USB 1.x or 2.x port, enabling the Extigy to be used on virtually any recent PC with any manufacturer's sound chip, sound card, or audio integrated chipset.


Figure 16.5. The Sound Blaster Extigy expands bare-bones onboard audio by adding advanced I/O features similar to those found on the Sound Blaster Audigy series. Photo courtesy of Creative Labs.

Figure 16.6. The rear panel of the Sound Blaster Extigy adds ports for stereo, rear, and subwoofer speakers, as well as the USB input to the computer, MIDI keyboard/device support, and digital input/output.

Volume Control


With virtually all recent sound cards, the volume is controlled through a Windows Control Panel speaker icon that can also be found in the system tray (near the onscreen clock). If you're switching from a bare-bones stereo sound card to a more sophisticated one featuring Dolby Digital 5.1, 6.1, or 7.1 output or input, you will need to use the mixing options in the volume control to select the proper sources and appropriate volume levels for incoming and outgoing audio connected to the card or a breakout box. Keep in mind that if you are sending sound to an external audio receiver, you will need to adjust the volume on that device as well. By contrast, some older audio adapters included a thumbwheel volume control next to the input/output jacks. The volume wheel can be troublesome; if you aren't aware of its existence and it is turned all the way down, you might be puzzled by the adapter's failure to produce sufficient sound. If the PC speakers are amplified but you aren't hearing any sound, remember to check that the power is on, the volume control on the speakers is turned up, and the correct speakers are selected and properly connected.

MIDI Support Features At one time, when evaluating audio adapters, you had to decide whether to buy a monophonic or stereophonic card. Today, all audio adapters are stereophonic and can play music using the MIDI standard, which plays scores using either synthesized instruments or digital samples stored on the audio adapter or in RAM. Stereophonic cards produce many voices concurrently and from two sources. A voice is a single sound produced by the adapter. A string quartet uses four voices, one for each instrument. On the other hand, a polyphonic instrument, such as a piano, requires one voice for each note of a chord. Thus, fully reproducing the capabilities of a pianist requires 10 voices​one for each finger. The more voices an audio adapter is capable of producing, the better the sound fidelity. The best audio adapters on the market today can produce up to 1,024 simultaneous voices. Early audio adapters used FM synthesis for MIDI support; the Yamaha OPL2 (YM3812) featured 11 voices, whereas the OPL3 (YMF262) featured 20 voices and stereophonic sound. However, virtually all audio adapters today use recorded samples for MIDI support; audio adapters using this feature are referred to as wavetable adapters. Wavetable audio adapters use digital recordings of real instruments and sound effects instead of imitations generated by an FM chip. When you hear a trumpet in a MIDI score played on a wavetable sound card, you hear the sound of an actual trumpet, not a synthetic imitation of a trumpet. The first cards featuring wavetable support stored 1MB of sound clips embedded in ROM chips on the card.


However, with the widespread use of the high-speed PCI bus for sound cards and large amounts of RAM in computers, most soundcards now use a so-called "soft wavetable" approach, loading 2MB​8MB of sampled musical instruments into the computer's RAM.


Note Some vendors, such as Creative Labs, Voyetra Turtle Beach, and Yamaha, have produced daughtercards to add wavetable support to FM-synthesis or soft wavetable sound cards. Most of these products have been discontinued due to the popularity and high quality of PCI audio adapters with large sample sets of soft wavetable sounds.

While early games supported only digitized audio samples (because most early sound cards had very poor MIDI support), late DOS games such as DOOM began to exploit the widespread wavetable-based MIDI support found on most mid-1990s and more recent sound cards. With all current sound hardware supporting wavetable MIDI and the improvements in DirectX 8.x and above for MIDI support, MIDI sound has become far more prevalent for game soundtracks. Many Web sites also offer instructions for patching existing games to allow MIDI support. Whether you play the latest games or like music, good MIDI performance is likely to be important to you. The most important factor for high-performance MIDI is the number of hardware voices. Even the best sound cards, such as Creative Labs' Sound Blaster Audigy 2 series, support only 64 voices in hardware; the remainder of the voices required by a MIDI soundtrack must come from software. If your sound card supports only 32 MIDI voices in hardware or uses software synthesis only, consider replacing it with a newer model. Many of the models currently on the market support more than 500 simultaneous voices and 64 hardware voices for under $100.

Data Compression Virtually all audio adapters on the market today can easily produce CD-quality audio, which is sampled at 44.1KHz. At this rate, recorded files (even of your own voice) can consume more than 10MB for every minute of recording. To counter this demand for disk space, many audio adapters include their own datacompression capability. For example, the Sound Blaster series includes on-the-fly compression of sound files in ratios of 2:1, 3:1, and 4:1. Most manufacturers of audio adapters use an algorithm called Adaptive Differential Pulse Code Modulation (ADPCM) compression (it's also called IMAADPCM), which was developed by the Interactive Multimedia Association (IMA) to reduce file size by more than 50%. IMA-ADPCM compresses 16-bit linear samples down to 4 bits per sample. However, a simple fact of audio technology is that when you use such compression, you lose sound quality. Unfortunately, no standard exists for the use of ADPCM. For example, although both Apple and Microsoft support IMA-ADPCM compression, they implement it in different ways. Apple's standard AIFF and Microsoft's standard WAV file formats are incompatible with each other unless you use a media player that can play both.


When you install an audio adapter, several codecs (programs that perform compression and decompression) are installed. Typically, some form of ADPCM is installed along with many others. To see which codecs are available on your system, open the Windows Control Panel and open the Multimedia icon (Windows 9x), the Sounds and Multimedia icon (Windows 2000), or the Sounds and Audio Devices icon (Windows XP). In Windows 9x, click the Devices tab followed by the plus sign next to Audio Compression to see the installed codecs. In Windows 2000 and Windows XP, click the Hardware tab, followed by Audio Codecs and Properties. The codecs are listed in order of priority, highest to lowest. You can also change the priority if you prefer a different order of priority. If you create your own recorded audio for use on another computer, both computers must use the same codec. You can select which codec you want to use for recording sounds with most programs, including the Windows Sound Recorder. The most popular compression standard is the Motion Pictures Experts Group (MPEG) standard, which works with both audio and video compression and is gaining support in the non-PC world from products such as DVD players. MPEG by itself provides a potential compression ratio of 30:1, and largely because of this, full-motion-video MPEG DVD and CD-ROM titles are now available. The popular MP3 sound compression scheme is an MPEG format, and it can be played back on recent versions of the Windows Media Player, as well as by various other audio player programs and devices.

Multipurpose Digital Signal Processors Many audio adapters use digital signal processors (DSPs) to add intelligence to the adapter, freeing the system processor from work-intensive tasks, such as filtering noise from recordings or compressing audio on-the-fly. The Sound Blaster Audigy's E-MU10K2 programmable DSP, for example, supports hardware sound acceleration needed by the latest version of Microsoft DirectX/DirectSound 3D, which enables multiple sounds to be played at the same time to synchronize with the onscreen action in a video game. The DSP can be upgraded with software downloads to accommodate more simultaneous audio streams. The widespread use of DSPs in better-quality audio adapters enables you to upgrade them through software instead of the time-consuming, expensive process of physical replacement. For additional examples of DSPs, see the section "Who's Who in Audio," later in this chapter.

Sound Drivers


As with many PC components, a software driver provides a vital link between an audio adapter and the application or operating system that uses it. Operating systems such as Windows 9x/Me and Windows 2000/XP include a large library of drivers for most of the audio adapters on the market (Windows NT 4.0 also supports some sound hardware but not as much as other versions of Windows). In most cases, these drivers are written by the manufacturer of the audio adapter and distributed only by Microsoft. You might find that the drivers that ship with the adapter are more recent than those included with the operating system. Although traditionally the best place to find the most recent drivers for a piece of hardware has been the manufacturer's own Web site or other online service, Windows Me, 2000, and XP prefer digitally signed drivers that have been certified by the Microsoft Hardware Quality Labs. You might find these drivers available at the vendor's own Web site, but you can also download and install them automatically through Windows Update. Any DOS applications you might still use do not typically include as wide a range of driver support as an operating system, but you should find that most games and other programs support the Sound Blaster adapters. If you are careful to buy an adapter that is compatible with Sound Blaster, you should have no trouble finding driver support for all your applications. Older ISA Sound Blaster cards provided hardware support for DOS games, but recent models (including the latest Sound Blaster Audigy and Audigy 2 series), as well as most comparable sound cards, require you to run software drivers to obtain Sound Blaster compatibility for DOS games. This software must be run before the game starts. If your game program locks up when you try to detect the sound card during configuration, set the card type and settings manually. This is often a symptom of inadequate emulation for Sound Blaster by a third-party card. If you have problems, check the game developer's or audio adapter's Web site for patches or workarounds.


Choosing the Best Audio Adapter for Your Needs Although sound features in computers have become commonplace, the demand for sophisticated uses for sound hardware have grown and demanded more and more powerful hardware. If your idea of a perfect multimedia PC includes any of the following, the plain-vanilla multimedia hardware found in many of today's PCs won't be sufficient: Realistic 3D and 360° sound for games Theater-quality audio for DVD movies Voice dictation and voice command Creating and recording MIDI, MP3, CD-Audio, and WAV audio files Table 16.1 summarizes the additional hardware features and software you'll need to achieve the results you want with your high-performance audio adapter. The following sections examine in detail these advanced uses and the features you'll need for each. Table 16.1. Audio Adapter Intended Uses and Features Comparison Intended Use

Features You Need

Additional Hardware

Additional Software

Gaming

Game or USB port; 3D sound; audio acceleration

Gaming controller; rear speakers Games

DVD movies

Dolby 5.1, 6.1, 7.1 decoding

Dolby 5.1, 6.1, 7.1 speakers compatible with audio adapter

MPEG/DVD decoding program

Audio adapter on software's Voice dictation and compatibility list or equal to SB16 in voice command quality

Voice-recognition microphone

Voice-dictation software

Creating MIDI files MIDI In adapter

MIDI-compatible musical keyboard

MIDI composing program

Creating MP3 files Digital audio extraction

CD-R or CD-RW drive

MP3 ripper

Creating WAV files Microphone

CD-R or CD-RW drive

Sound recording program

Creating CD audio External sound source files

CD-R or CD-RW drive

WAV or MP3 to CD audio conversion program


The following sections discuss many of these special uses in detail.

Gaming Thanks to the widespread availability of audio adapters, game playing has taken on a new dimension. Support for 3D and surround digitized sound and realistic MIDI music in current games has added a level of realism that would otherwise be impossible even with today's sophisticated graphics hardware. Mere stereo playback isn't good enough for hardcore gamers who want to be able to hear monsters behind them or feel the impact of a car crash. These users should choose sound cards with support for four or more speakers and some form of directional sound, such as the Creative Labs EAX technology used in Sound Blaster Live! and the Audigy/Audigy 2 series or Sensaura 3D Positional Audio (3DPA) used by ESS, VideoLogic, Cirrus Crystal Logic, Analog Devices, C-Media, and NVIDIA. Many sound cards feature support for these standards, either through direct hardware support or through software emulation and conversion. As with 3D video cards (see Chapter 15, "Video Hardware"), most cards today merely need to work with the 3D audio APIs included in the current revision of Microsoft's DirectX technology. Any audio adapter built in the last few years will still work with today's games, thanks in large part to the Hardware Emulation Layer (HEL) built into DirectX. HEL emulates the features of newer hardware, such as 3D sound, on older hardware. However, as you can imagine, the task of emulating advanced performance on older hardware can slow down gameplay and doesn't produce sounds as realistic as those available with today's best audio adapters.

Sound Card Minimums for Gameplay The replacement of the old ISA Sound Blaster Pro standard by PCI sound card standards has helped improve performance a great deal, but for the best gameplay with current and forthcoming titles, you need to consider sound cards with the following features: 3D audio support in the chipset. 3D audio means you'll be able to hear sounds appear to move toward you, away from you, and at various angles corresponding to what's happening onscreen. Microsoft's DirectX, version 9, includes support for 3D audio, but you'll have faster 3D audio performance if you use an audio adapter with 3D support built in. DirectX 9 works along with proprietary 3D audio APIs, such as Creative's EAX and EAX 2.0, Sensaura's 3D


Positional Audio, and the A3D technology from now-defunct Aureal. 3D sound acceleration. Sound cards using chipsets with this feature require very little CPU utilization, which speeds up overall gameplay. For best results, use a chipset that can accelerate a large number of 3D streams; otherwise, the CPU will be bogged down with managing 3D audio. This can slow down gameplay, particularly on systems with processors running under 1GHz or that are running at a high-resolution, high-color depth setting (1,024x768/32-bit). Game ports with support for force-feedback game controllers if your game controllers aren't USB compatible. If your games don't work properly with USB controllers, or your force-feedback game controllers use the game port only, check the game port features to make sure you'll feel as well as hear and see the action. Note that many recent game cards no longer include a game port or might use an additional slot for the game port connector. Features such as these don't necessarily cost a ton of money; many of the midrange audio adapters on the market ($50​$100 at retail) support at least the first two features. With new 3D audio chipsets available from a number of vendors, it might be time for you to consider an upgrade if you're heavily into 3D gaming.

Legacy (MS-DOS) Game Support Issues Support for the classic Sound Blaster Pro standard was once the primary requirement for a good gaming audio adapter, but with the rise of great Windowsbased games and the development of DirectX, this is no longer the case for many users. If you want to play both MS-DOS and Windows games on the same computer, you need to understand the implications of the changes in the newer audio adapters. Emulating the Sound Blaster with a PCI-based sound card is difficult because today's PCI audio adapters and motherboard-based audio solutions don't use separate DMA channels the same way that ISA cards and motherboard resources do. DMA support is vital to achieving a high degree of compatibility with software written for the older Sound Blaster Pro or Sound Blaster 16 cards. Four methods are used by PCI-based sound cards to emulate Sound Blaster at the DMA hardware level demanded by older DOS and some early Windows games: DDMA (Distributed DMA), developed by Compaq TDMA (Transparent DMA), developed by ESS Technology


PC/PCI interface, developed by Intel and Creative Labs TSR (memory-resident) utility programs Distributed DMA and Transparent DMA are similar; both convert calls to the 8237 DMA controller (no longer present on PCI-based motherboards) into calls to a master DMA central resource, which converts the signals into a form compatible with the way PCI-based systems handle DMA transfers. No external connector is required for these features; just enable the Sound Blaster compatibility feature in your audio adapter's setup to activate these features. This is the easiest way for you to achieve Sound Blaster compatibility if you still need it. The PC/PCI interface, which uses a 6-pin ribbon cable to connect the sound card to a special motherboard header, is supported by several recent Intel and thirdparty motherboards and some recent sound cards made by Creative Labs and others. However, it is no longer supported by Creative Labs' latest products (the Audigy and Audigy 2 series). If you want to learn more about PC/PCI (also called the SB-Link connector), see the section "SB-Link Connectors" in Chapter 12 of Upgrading and Repairing PCs, 12th Edition, included on the DVD with this book. All these methods require that the feature be built into both the motherboard and the sound card. Many recent PCI-based sound cards support the first two of these options, easing the issue of finding a match between your motherboard and your sound card's method of emulating the Sound Blaster. The least desirable method for emulating a Sound Blaster under DOS has made a comeback after fading from view a few years ago: requiring the user to load a Terminate and Stay Resident (TSR) program into RAM before starting the DOSbased game or educational program. For example, the Sound Blaster Audigy and Audigy 2 series modifies the DOS-mode Autoexec.bat and Config.sys files to use the memory-resident SBEINIT.COM driver along with the Windows HIMEM.SYS and EMM386.EXE memory managers to enable DOS games to run. A poorly written driver could also cause problems with sound card detection by your game and compatibility problems with some games. For these reasons, if you have an older sound card that works well with your DOS-based games, you might want to keep it installed if you can spare the hardware resources it uses. Additionally, if you bought an early PCI-based sound card that doesn't work well with DOS games, consider upgrading to a model that uses a high-quality method of emulating the Sound Blaster Pro.

Avoiding Game Port Conflicts


On older systems that use multi-I/O type cards to provide serial and parallel ports and on low-cost systems based on integrated audio chipsets such as the Intel 810, an audio card's game port could be a potential area of conflict because these cards sometimes also include a game port. You must disable one of the game interfaces if you have duplicate game ports. To learn how to disable a duplicate game port, see the section "Resolving Resource Conflicts," later in this chapter. In many cases, you might not need to use the game port/MIDI port for gaming controllers. Many current game controllers often offer both USB and the traditional game port connectors, allowing you to use the control with either type of connector, although some features might work only when the USB port is used. The game port is still useful for older games that might not recognize a USBattached game controller or for specialized controllers such as rudder pedals for some flight sims. If you have invested in a high-performance game port, such as the Thrustmaster ACM Game card (www.thrustmaster.com) designed for fast PCs, and sophisticated controllers, such as driving wheels, force-feedback joysticks, and so on, you should disable the sound card's game port and use the separate one you installed. If you have a plain-vanilla standard game port, remove it or disable it and use the game adapter port on your audio adapter, especially if you plan to use the MIDI port.


Tip If your sound card uses a header cable and a second expansion slot for the game port, as is the case with the Sound Blaster Audigy and Audigy 2 series, but you don't use the game port for anything, you can disconnect the game port from your sound card and free up a PCI slot.

DVD Movies on Your Desktop You don't need a dedicated DVD player to enjoy the clarity, control, extra features, and excitement of DVD movies. DVD-ROM drives help bring the DVD movie experience to your PC, but having a DVD-ROM and a DVD movie player program is only part of what you need to bring the big screen to your desktop. To get the most out of your desktop DVD experience, you need the following: DVD playback software that supports Dolby Digital 5.1 or better output. One of the best choices is Cyberlink's PowerDVD4.x, available from www.gocyberlink.com. An audio adapter that supports Dolby Digital input from the DVD drive and will output to Dolby Digital 5.1​compatible audio hardware. Some will remix Dolby 5.1 to work on four-speaker setups if you don't have Dolby 5.1 hardware or will accept S/PDIF AC3 (Dolby Surround) input designed for a four-speaker system; some can also pass through Dolby Digital audio to speakers that can perform the Dolby Digital 5.1 decoding. Some high-end audio adapters now support 6.1 and 7.1 speaker configurations; these also work with Dolby Digital 5.1 sound. Dolby Digital 5.1-compatible stereo receiver and speakers. Most high-end sound cards with Dolby Digital 5.1 support connection to analog-input Dolby Digital 5.1 receivers, but some, such as the Creative Labs Sound Blaster Live!, Audigy, and Audigy 2 Platinum series and the Hercules Digifire 7.1, Game Theater XP 7.1, and Fortissimo III 7.1 support digital-input speaker systems. Depending on which types of speakers you are using and how they are attached, you might need to switch your mixer settings in Windows from analog to digital to hear sounds from your applications (movies, games, and so on). To learn more about speaker terminology and how to ensure your speaker configuration is correct, see the section "Speakers," later in this chapter.

Voice Dictation and Control


Some audio adapters are equipped with software capable of voice recognition that can be used to control some of your computer's operations. You also can get voice recognition for your current adapter in the form of add-on software. Voice recognition, as the name implies, is when your computer is "taught" to recognize spoken word forms and react to them. Voice recognition products generally take two forms: those that are designed to provide a simple voice interface to basic computer functions and those that can accept vocal dictation and insert the spoken text into an application, such as a word processor. The minimum standard for most voice-recognition software is a Sound Blaster 16 or equivalent sound card.

Voice Command Software The voice interface application is clearly the simpler of the two because the software has to recognize only a limited vocabulary of words. With this type of software, you can sit in front of your computer and say the words "file open" to access the menu in your active Windows application. For the average user, this type of application is of dubious value. For a time, Compaq was shipping computers to corporate clients with a microphone and an application of this type at little or no additional cost. The phenomenon of dozens of users in an office talking to their computers was interesting, to say the least. The experiment resulted in virtually no increased productivity, a lot of wasted time as users experimented with the software, and noisier offices. However, for users with physical handicaps that limit their ability to use a keyboard, this type of software can represent a whole new avenue of communication. For this reason alone, continued development of voicerecognition technology is essential.


Note Voice-recognition applications, by necessity, have to be somewhat limited in their scope. For example, it quickly becomes a standard office joke for someone to stick his head into another person's cubicle and call out "Format C!" as though this command would erase the user's hard drive. Obviously, the software must be incapable of damaging a user's system because of a misinterpreted voice command.

Voice Dictation Software The other type of voice-recognition software is far more complex. Converting standard speech into text is an extraordinarily difficult task, given the wide variation in human speech patterns. For this reason, nearly all software of this type (and some of the basic voice command applications, as well) must be "trained" to understand a particular user's voice. You do this training by reading prepared text samples supplied with the software to the computer. Because the software knows what you're supposed to be saying beforehand, it can associate certain words with the manner in which you speak them. Users' results with this type of application vary widely, probably due in no small part to their individual speech patterns. I've heard people rave about being able to dictate pages of text without touching the keyboard, whereas others claim that correcting the many typographical errors is more trouble than typing the text manually.


Note For more information about voice-recognition, see "Voice Dictation Software" in Chapter 20 of Upgrading and Repairing PCs, 12th Edition, included on the DVD-ROM accompanying this book. Current vendors of voice-dictation software include ScanSoft. Dragon NaturallySpeaking; www.scansoft.com IBM. ViaVoice; www.ibm.com Microsoft. Microsoft Office XP English, simplified Chinese, and Japanese editions; www.microsoft.com Corel. WordPerfect Suite 2002 Professional; www.corel.com

Sound Producers Sound producers are people who intend to create their own sound files. These can range from casual business users recording low-fidelity voice annotations to professional musicians and MIDI maniacs. These users need an adapter that can perform as much of the audio processing as possible itself, so as not to place an additional burden on the system processor. Adapters that use DSPs to perform compression and other tasks are highly recommended in this case. Musicians will certainly want an adapter with as many voices as possible and a wavetable synthesizer. Adapters with expandable memory arrays and the capability to create and modify custom wavetables are also preferable. Many of the best sound cards for hardcore gamers also are suitable for sound producers by adding the appropriate sound-editing programs, such as Sound Forge, and by equipping the card with the appropriate connectors for SPDIF digital audio and MIDI interfaces. The latest Sound Blaster Audigy 2 Platinum and Platinum EX include internal (Platinum) and external (Platinum 2) breakout boxes with these features. Hercules' Game Theater XP 7.1 also includes a breakout box. The Creative Labs Extigy provides features similar to those found on the Audigy 2 Platinum series, but it can be added to any system with a USB port. Most other audio cards designed for sound production features add jacks to the traditional trio of connectors on the rear card bracket.


Playing and Creating Digitized Sound Files You can use two basic types of files to store audio on your PC. One type is generically called a sound file and uses formats such as WAV, VOC, AU, and AIFF. Sound files contain waveform data, which means they are analog audio recordings that have been digitized for storage on a computer. Just as you can store graphic images at different resolutions, you can have sound files that use various resolutions, trading off sound quality for file size. The default sound resolution levels used in Windows are shown in Table 16.2. Table 16.2. Windows Default Sound File Resolutions Format

Resolution

Frequency

Bandwidth

File Size

PCM

Telephone quality

11,025Hz

8-bit mono

11KBps

Radio quality

22,050Hz

8-bit mono

22KBps

CD quality

44,100Hz

16-bit stereo

172KBps

If you have a sound card that supports DVD-quality (48,000Hz, 16-bit stereo, 187KBps), you can also save sounds at that frequency, but you must select it manually if you are using the Windows Sound Recorder to digitize sounds. Note that the Windows Sound Recorder applet uses the default Pulse Code Modulation (PCM) method for storing sounds. PCM produces the highest quality of sound, but because it doesn't use any type of data compression, file sizes can be enormous. As you can see, the difference in file sizes between the highest and lowest audio resolution levels is substantial. CD-quality sound files can occupy enormous amounts of disk space. At this rate, just 60 seconds of audio would require more than 10MB of storage. For applications that don't require or benefit from such high resolution, such as voice annotation, telephone-quality audio is sufficient and generates much smaller files. To achieve a balance between high quality and smaller file sizes, you can convert conventional WAV files into compressed formats, such as MP3 or WMA audio files. The other type of file is a MIDI file, which consists of a musical score that is played back by synthesized or sampled musical instruments incorporated into the sound card's MIDI support.


Note To learn more about the differences between MP3, WMA, and MIDI files, see "Audio Compression and MIDI Files" in the Technical Reference located on the DVD-ROM packaged with this book.

On a multimedia PC, it is often possible for two or more sound sources to require the services of the audio adapter at the same time. Any time you have multiple sound sources you want to play through a single set of speakers, a mixer is necessary. Most audio adapters include a mixer that enables all the different audio sources, MIDI, WAV, line in, and CD to use the single line-out jack. Starting with Windows 95 through the latest Windows versions (XP Pro/XP Home), Windows uses a single mixer for both recording and playback features, instead of using separate mixers as with Windows 3.x. Normally, the adapter ships with software that displays visual sliders like you would see on an actual audio mixer in a recording studio. With these controls, you can set the relative volume of each of the sound sources.


Tip Whenever you change from analog to digital speakers or add speakers to a two-speaker configuration, you must adjust the mixer controls to match your current speaker configuration. If you don't, you will be unable to hear anything through your speakers.


Audio Adapter Concepts and Terms To fully understand audio adapters and their functions, you need to understand various concepts and terms. Terms such as 16-bit, CD quality, and MIDI port are just a few. Concepts such as sampling and digital-to-audio conversion (DAC) are often sprinkled throughout stories about new sound products. You've already learned about some of these terms and concepts; the following sections describe many others.

The Nature of Sound To understand an audio adapter, you must understand the nature of sound. Every sound is produced by vibrations that compress air or other substances. These sound waves travel in all directions, expanding in balloon-like fashion from the source of the sound. When these waves reach your ear, they cause vibrations that you perceive as sound. Two of the basic properties of any sound are its pitch and intensity. Pitch is the rate at which vibrations are produced. It is measured in the number of hertz (Hz), or cycles per second. One cycle is a complete vibration back and forth. The number of Hz is the frequency of the tone; the higher the frequency, the higher the pitch. Humans can't hear all possible frequencies. Very few people can hear sounds with frequencies less than 16Hz or greater than about 20KHz (kilohertz; 1KHz equals 1,000Hz). In fact, the lowest note on a piano has a frequency of 27Hz, and the highest note has a frequency a little higher than 4KHz. Frequency-modulation (FM) radio stations can broadcast notes with frequencies as high as 15KHz. The amazing compression ratios possible with MP3 files, compared to regular CDquality WAV files, is due in part to the discarding of sound frequencies that are higher or lower than normal hearing range during the ripping process. The intensity of a sound is called its amplitude. This intensity determines the sound's volume and depends on the strength of the vibrations producing the sound. A piano string, for example, vibrates gently when the key is struck softly. The string swings back and forth in a narrow arc, and the tone it sends out is soft. If the key is struck more forcefully, however, the string swings back and forth in a wider arc, producing a greater amplitude and a greater volume. The loudness of sounds is measured in decibels (db). The rustle of leaves is rated at 20db, average street noise at 70db, and nearby thunder at 120db.


Evaluating the Quality of Your Audio Adapter The quality of an audio adapter is often measured by three criteria: frequency response (or range), total harmonic distortion, and signal-to-noise ratio. The frequency response of an audio adapter is the range in which an audio system can record or play at a constant and audible amplitude level. Many cards support 30Hz​20KHz. The wider the spread, the better the adapter. The total harmonic distortion measures an audio adapter's linearity and the straightness of a frequency response curve. In layman's terms, the harmonic distortion is a measure of accurate sound reproduction. Any nonlinear elements cause distortion in the form of harmonics. The smaller the percentage of distortion, the better. This harmonic distortion factor might make the difference between cards that use the same audio chipset. Cards with cheaper components might have greater distortion, making them produce poorer-quality sound. The signal-to-noise ratio (S/N or SNR) measures the strength of the sound signal relative to background noise (hiss). The higher the number (measured in decibels), the better the sound quality. For example, the top-of-the-line Sound Blaster Audigy 2 sound card features an SNR of 106db, whereas the older Sound Blaster Audigy is rated at 100db and the AWE64 series has an SNR of 90db. These factors affect all types of audio adapter use, from WAV file playback to speech recognition. Keep in mind that low-quality microphones and speakers can degrade the performance of a high-quality sound card.

Sampling With an audio adapter, a PC can record waveform audio. Waveform audio (also known as sampled or digitized sound) uses the PC as a recording device (like a tape recorder). Small computer chips built into the adapter, called analog-todigital converters (ADCs), convert analog sound waves into digital bits that the computer can understand. Likewise, digital-to-analog converters (DACs) convert the recorded sounds to an audible analog format. Sampling is the process of turning the original analog sound waves into digital (binary) signals that the computer can save and later replay (see Figure 16.7). The system samples the sound by taking snapshots of its frequency and amplitude at regular intervals. For example, at time X the sound might be measured with an amplitude of Y. The higher (or more frequent) the sample rate, the more accurately the digital sound replicates its real-life source and the larger the amount of disk space needed to store it.


Figure 16.7. Sampling turns a changing sound wave into measurable digital values.

Originally, sound cards used 8-bit digital sampling that provided for only 256 values (28), which could be used to convert a sound. More recently, sound cards have increased the quality of digitized sound by using 16-bit (216) sampling to produce 65,536 distinct values. Today's highest-quality sound cards feature 24-bit sampling (224), which translates into more than 16.8 million possible digital values that can be matched to a given sound.


Note For more information on the differences between 8-bit and 16-bit audio sampling, see "8-Bit Versus 16Bit" in Chapter 16 of Upgrading and Repairing PCs, 13th Edition, included on the DVD-ROM accompanying this book.

You can experiment with the effects of various sampling rates (and compression technologies) by recording sound with the Windows Sound Recorder or a thirdparty application set to CD-quality sound. Save the sound and play it back at that highest quality setting. Then convert the file to a lower-quality setting, and save the sound file again with a different name. Play back the various versions, and determine the lowest quality (and smallest file size) you can use without serious degradation to sound quality.


Who's Who in Audio Because audio adapters have become common features in systems, many vendors have produced audio adapters, audio chips, integrated motherboard chipsets with audio features, and even specialized vacuum tube audio. This section examines some of these companies and their products. As you've learned in other chapters, I believe it is very important to get all the technical information you can about your computer and its components. By knowing who makes the audio chip your computer depends on, you can find out what the hardware can do and be better able to find upgrades to the software drivers you need to get the most out of your audio hardware.

Chipset Makers Who Make Their Own Audio Adapters Just as graphics card vendors are divided into two camps, chipset makers are divided into these two categories: Card makers who produce their own chipsets Card makers who use chipsets from other vendors Audio adapter vendors fall into the same categories. One of the pioneers of the audio adapter business, Creative Labs, has also been among the leaders in developing audio chips. Creative Labs develops audio chips primarily for its own Sound Blaster​branded products, but it has sold some of its older Sound Blaster 16 products into OEM markets. Creative's major audio chips have included the following: Vibra-16. This was used in the later Sound Blaster 16 cards; it doesn't support wavetable or 3D effects. Ensoniq ES1370 series (ES1370/71/73). These were used in the Sound Blaster PCI64 and PCI 128 series as well as the Ensoniq Audio PCI and Vibra PCI series. They support soft wavetable features, four speakers on some models and Microsoft Direct 3D but don't support 3D acceleration or EAX positional audio. EMU-8000. This audio chip was used by the AWE32/64 series and features


32-voice wavetable synthesis but no 3D acceleration. The AWE64 used software to generate 32 additional voices for a total of 64 voices. EMU10K1. This audio chip was at the heart of the Live! and Live 5.1 series sound cards as well as the PCI 512; it features 3D acceleration, EAX positional audio for one audio stream, a reprogrammable DSP, and soft wavetable support. EMU10K2 (also known as Audigy). This is the audio chip at the heart of Creative Labs' Sound Blaster Audigy series sound cards; it features 3D acceleration, EAX HD positional audio for up to four audio streams, a reprogrammable DSP, and soft wavetable support. This chip supports professional-level 24-bit sampling at 96KHz and real-time sampling at Dolby Digital​quality 24-bit samples at 48KHz. CA0102 (also known as Audigy 2). This audio chip is the one used by the Creative Labs Audigy 2 series. It's a development of the EMU10K2 chip, adding support for 24-bit 96KHz output, Dolby Digital EX 6.1 decoding and 6.1 sound in DirectX games, and 64 hardware polyphonic voices. Another major player is Philips, which bought chipset maker VLSI and integrated it into its Philips Semiconductor operation in mid-1999. Philips introduced its line of audio adapters in the fall of 2000, using the ThunderBird chipsets it jointly developed with Qsound Labs, Inc. These include the following: ThunderBird Q3D (SAA7780). Features 3D audio acceleration of up to 64 3D streams in hardware, positional 3D audio supporting EAX and Qsound standards, quadraphonic speaker support with virtual surround sound, wavetable, and DOS Sound Blaster emulation; it is used by the original Philips Rhythmic Edge (PSC702) and Seismic Edge (PSC704) audio adapters. Thirdparty audio cards using this chip include the Aztech PCI-368DSP, I/O Magic MagicQuad 8, and Labway Thunder 3D. ThunderBird Avenger (SAA7785). Features 3D audio acceleration of up to 96 3D streams in hardware, positional 3D audio supporting EAX and Qsound standards, support for Dolby Digital 5.1, wavetable, and DOS Sound Blaster emulation; it is used by the Philips Seismic Edge 5.1 (PSC705), Rhythmic Edge Surround (PSC703), Acoustic Edge (PSC706), PSC605 Sonic Edge (PSC605), and Dynamic Edge (PSC604) audio adapters. Both of these highly regarded chipsets are offered as OEM products, but so far only the Q3D/SAA7780 chip has been used by third-party audio card vendors.


Various other companies have produced their own sound chips in the past but no longer do so. The two primary makers that fit in this category are Aureal. Its A3D technology was regarded by many as superior to Creative Labs' original EAX 3D positional audio, but the company was absorbed by Creative Labs in mid-2000. Because Creative's new EAX HD is superior to A3D, there will be no further development of this technology. Yamaha. Its OPL2 and OPL3 chips were among the best FM-synthesis chips used on older sound cards, and its MIDI performance in later models was outstanding. However, its emphasis is now on MIDI daughtercards and professional sound-recording cards such as the SW1000XG. Some of its retail and OEM products might still be available, though. Yamaha maintains drivers and support links for its branded and some OEM cards at www.yamaha.co.jp/english/lsi/us. Should you panic if your favorite audio adapter is an "orphan"? Not necessarily. If the audio adapter vendor provides good technical support and up-to-date drivers, you're okay for now. But, the next time an operating system update or new audio API shows up, you probably won't be able to take advantage of it unless you replace your audio adapter.

Major Sound Chip Makers Most companies other than Creative Labs and Philips depend on third parties to make their audio chips. Some of the major vendors include: Cirrus Logic/Crystal Semiconductors. The top-of-the-line Sound Fusion CS4630, an enhanced version of the CS4624, features 3D acceleration, support for both EAX and Sensaura positional audio, unlimited-voice wavetable synthesis, and S/PDIF support for AC3 and Dolby 5.1 input and output at rates up to 48KHz. The CS4630 is used in the popular and highly rated Hercules Game Theater XP, Voyetra Turtle Beach Santa Cruz, TerraTec SiXPack 5.1, Video Logic Sonic Fury, and SonicXplosion audio adapters. The CS4624 is used by the Hercules GameSurround Fortissimo II/III 7.1, Hercules DigiFire 7.1, TerraTec DMX Xfire, and Hoontech SoundTrack I-Phone Digital CS audio adapters. Other Sound Fusion series chips include the CS4614 (obsolete) and CS4624 (added), both of which feature support for 3D Direct Sound positional audio, DOS Sound Blaster emulation, and wavetable synthesis.


ESS Technology. The Canyon3D-2 (ESS1990/1992) is ESS Technology's flagship audio chip, featuring four-channel analog output, support for Dolby and THX digital sound, SPDIF input and output, and Sensaura 3D positional audio, and it is optimized for use with DirectX 8.0. It is used by the I/O Magic Hurricane Extreme, Diamond Monster Sound 3D, Hercules MaxiSound Fortissimo, and TerraTec DMX audio adapters. The Maestro-2 series features wavetable, positional 3D from Sensaura, and 3D audio acceleration; the Maestro 2E and 2EM also support S/PDIF output for DVD movie support. The Maestro series chips are optimized for notebook computers, and Maestro chips are used in recent models of Dell, Toshiba, Gateway, Compaq, and HP portables. The Allegro series (ESS-1989 for desktops and ESS-1988 for notebook computers) features DirectSound, Direct3D, S/PDIF output, and Sensaura 3D positional sound. The ESS-1989 is used by the Philips Harmonic Edge (PSC602) sound card as well as models sold by Pine Technologies and others. ESS's earlier AudioDrive series was popular with many notebook computers and second-tier sound card makers in the mid-1990s. C-Media Electronics. The CMI 8738 features 4.1 and 5.1 speaker support for quadraphonic and Dolby Digital output, Direct Sound 3D and A3D positional audio, and wavetable and is available for desktop or notebook computers; some versions also integrate a software modem and SDPIF port. It is used by sound cards such as Guillemot's MUSE and Leadtek's WinFast 4x Sound; some generic sound cards; and motherboards made by Asus, Soyo, and others. ForteMedia, Inc. The FM-801 is the first audio chip to feature Dolby Digital 5.1 output to analog speakers for both DVD movies and games. The FM-801 also features Qsound's Q3D 2.0 3D API and optional support for SPDIF input/output. The FM-801 is used by many smaller sound card makers and some motherboard makers, such as Shuttle. For a review of the sound chip and a feature comparison of some sound cards using this chip, see www.3dsoundsurge.com/reviews/FM801/FM801.html. Realtek. Although it's better known for low-cost Ethernet network card chipsets, Realtek also offers audio chipsets, including the ALC650, which was introduced in March 2002. The ALC650 is the first motherboard integrated chipset to support AC '97 audio along with six-channel output, Dolby Digital 5.1, stereo, and surround sound. It is featured on high-performance motherboards made by MSI, Giga-byte, Asus, AOpen, and others.

Discontinued and Orphan Sound Chips and Sound Card


Producers The following sound chips are no longer being sold, and ongoing support is limited or no longer available. If you use an audio adapter based on one of these products, you might need to upgrade if you can't get drivers for new and forthcoming operating systems. Discontinued products include Oak Technology OTI-601 series. Oak left the audio chipset business in early 1998. Trident 4DWave-NX series. This 3D audio chipset is still available on cards from smaller audio adapter vendors, such as Aztech, Jaton, and Hoontech. Diamond Multimedia, which used sound chips from several vendors, no longer produces sound cards since its parent company, S3, relaunched itself in the fall of 2000 as an Internet appliance and MP3 audio-focused company called SONICblue. SONICblue still offers limited support for Diamond-brand audio cards, but ongoing driver development is no longer taking place.

Motherboard Chipsets with Integrated Audio The Intel 810 chipset was the first mainstream chipset for a major CPU to integrate audio; it works with Celeron CPUs. Its inspiration might have been the Cyrix/National Semiconductor Media GX series, which used a trio of chips to substitute for the CPU, VGA video, onboard audio, memory, and I/O tasks. Thanks to improvements in chipset design and faster CPU performance, today's best integrated chipsets can provide solid mid-range performance. Almost all recent chipsets from Intel, VIA, ALi, and SiS have integrated audio (see Chapter 4, "Motherboards and Buses," for details). In almost every case, integrated audio supports the AC'97 audio standard.

AC'97 Integrated Audio The phrase AC'97 integrated audio can be found in the descriptions of most recent systems. Because AC'97 can replace the need for a separate audio card but might not be a satisfactory replacement, you need to understand what it is and how it works.


AC'97 (often referred to as AC97) is an Intel specification that connects an audio codec (compression/decompression) architecture to a section of a South Bridge or an I/O Communications Hub chip called the AC-Link control. The AC-Link control works with the CPU and an AC '97 digital signal processor (DSP) to create audio. The AC'97 audio codec could be a physical chip on the motherboard, a chip on a small daughterboard called a communications and networking riser (CNR), or a software program. Thus, a motherboard with AC '97 integrated sound support doesn't require the use of a separate audio card for sound playback. Sometimes AC'97 is also used to refer to audio chips on a sound card, but in this discussion we will use it to refer only to integrated audio. Sometimes motherboards also integrate an analog modem through an MC '97 codec chip, or they might have an AMC '97 (audio/modem) codec chip to combine both functions. It's important to realize that, although most recent chipsets support AC'97 audio, this does not mean that every motherboard built on a particular chipset uses the same AC'97 codec, or even the same method of creating sound. In most cases, AC'97 is implemented through a small AC'97 codec chip on the motherboard (see Figure 16.8). It can be surface-mounted as shown in Figure 16.8, but many vendors use a small socket instead.

Figure 16.8. The VIA VT1612A is a typical AC'97 2.2​compliant codec chip (foreground), seen here providing integrated audio for the VIA EPIA-V Mini-ITX motherboard (background). Photos courtesy of VIA Technologies, Inc.

For various reasons, including features and price, different motherboard vendors might use different AC'97 codec chips on motherboards that use the same chipset. For example, compare the following motherboards based on the Intel 815E-series chipsets as listed in Table 16.3. Table 16.3. AC'97 Codecs Used with Intel 815E​Based Motherboards


Table 16.3. AC'97 Codecs Used with Intel 815E​Based Motherboards Vendor

Motherboard Model

AC'97 Codec Chip

Intel

D815EPFV

Analog Devices AD1885 (SoundMax)

Giga-Byte

GA-6IEM

Realtek (probably ALC-650)

Kontron

786Flex-8145

Crystal CS4299

The drivers for a particular AC'97 codec chip are supplied by your motherboard vendor because they must be customized to the combination of codec and South Bridge/ICH chip your motherboard uses. Although the AC'97 specification recommends a standard pinout, differences do exist between AC'97 codec chips. Some vendors of AC'97 chips provide technical information to help motherboard builders design sockets that can be used with different models of the AC'97 codec chip. The four versions of the AC'97 codec are as follows: AC'97 1.0. Has fixed 48KHz sampling rate and stereo output AC'97 2.1. Has options for variable sampling rate and multichannel output AC'97 2.2. Has AC'97 2.1 features plus optional S/PDIF digital audio and enhanced riser card support; released in September 2000 AC'97 2.3. Has AC'97 2.1/2.2 features plus support for true Plug and Play detection of audio devices; released in July 2002 Most motherboards with integrated audio support AC'97 2.1 or 2.2 at this time. To learn more about the AC'97 specifications, see the Intel ​ Research and Development, Audio Codec site at www.intel.com/labs/media/audio/index.htm. To determine whether a particular motherboard's implementation of AC'97 audio will be satisfactory, follow these steps: 1. Determine which codec chip the motherboard uses. Read the motherboard manual or see which driver the motherboard uses for audio. Look up the chip's features and specifications. If you are not sure of the chip manufacturer, look up the part number with a search engine such as Google.


Use a search engine to find reviews of the chip's sound quality and performance (typically found as part of a motherboard review). The Web site 3D Sound Surge (www.3dsoundsurge.com) reviews both sound cards and motherboards/audio codecs. Look at the motherboard's features to determine whether it uses the full capabilities of the codec chip. Chips that support AC'97 2.1 can offer up to sixchannel analog audio; those that support AC'97 2.2 can also offer S/PDIF digital audio. However, motherboard makers don't always provide the proper outputs. Analyze how you use audio. If you play a lot of 3D games, you're not likely to be satisfied with the performance of any integrated audio solution, no matter what its features might be. You can disable onboard audio with a BIOS setting if you prefer to install your own audio card. For details on how to enable and disable onboard audio, see "Peripheral Configuration," p. 401.

AOpen TubeSound The Taiwan-based motherboard maker AOpen, part of the Acer Group, came up with a very interesting gimmick in June 2002 when it introduced the world's first PC motherboard with a vacuum tube​based audio amplifier​the AOpen AX4B-533 Tube. The motherboard was based on the Intel 845E chipset, and uses a Realtek ALC650 AC97 sound chip. At first, many PC users wondered whether this was an April Fool's joke that showed up late. Why a vacuum tube? AOpen engineers pointed out that serious audiophiles have continued to use vacuum-tube amplifiers because of their rich sound. They felt that audiophiles would pay a premium price for similar technology in the sound circuitry of a PC. AOpen used the following design features to bring the vacuum tube into the twenty-first century: A switching mode power supply to provide adequate tube power. Tubes fell out of favor in the late 1950s because they require more power than transistors and integrated circuits. A dual-triode. This design has one tube with two front stereo channels and is modeled after the design used by classic pre-amp circuits, which can also


accept input from standard sound cards. Frequency isolation wall (FIW) noise reduction. This shields the tube circuitry from the normal EFI/RFI interference inside the computer. High mean time between failure (MTBF) design for motherboard and tube circuitry. The AX4B-533 Tube is among the most expensive motherboards using the 845E chipset, selling for about $160​$190 compared to about $100 for other models (see Figure 16.9). However, it has received rave reviews from many computer publications and users for audio quality, performance, and (not least) the snob appeal of having the first motherboard on the block like it.

Figure 16.9. A close look at the A4XB-533's vacuum tube sound system.

The AX4B-533 Tube's audio quality is optimized for classical and jazz music listening. AOpen has now released two additional vacuum-tube-based motherboards: the AX4GE Tube and AX4PE Tube, which are optimized for rock and pop music thanks to a slightly revised tube and amplifier design.


Tip To compare these motherboards in more detail, see the AOpen TubeSound technology Web site at www.aopen.com/tech/techinside/Tube.htm.


3D Audio One of the biggest issues for serious game players when audio adapters are considered is how well they perform 3D audio tasks. This has been complicated by several factors, including the following: Differing standards for positional audio Hardware versus software processing of 3D audio DirectX support issues

Positional Audio The underlying issue common to all 3D sound cards is that of positional audio, which refers to adjusting features such as reverberation; balance; and apparent sound "location" to produce the illusion of sound coming from in front of, beside, or even behind the user. One very important element in positional audio is HRTF (Head Related Transfer Function), which refers to how the shape of the ear and the angle of the listener's head changes the perception of sound. Because HRTF factors mean that a "realistic" sound at one listener's head angle might sound artificial when the listener turns to one side or the other, the addition of multiple speakers that "surround" the user, as well as sophisticated sound algorithms that add controlled reverberation to the mix, are making computer-based sound more and more realistic. The technology might be promising, but one of the most frustrating problems for PC game players in particular has been the continued rivalry between various APIs designed to perform essentially the same task, such as the battle between the now-defunct 3dfx Glide standard and OpenGL for graphics, or the battle over which 3D audio standard to support. The original version of Microsoft's Direct3D for DirectX didn't support third-party 3D software, but recent and current versions do, enabling 3D audio adapters to improve the normal positional audio available with Direct3D. During 1999 and the first part of 2000, the major rivals for the most popular gaming 3D standard were Aureal's A3D and Creative Labs' EAX (Environmental Audio Extensions) technology. A3D, especially in version 2.0, was frequently regarded as superior to its Creative Labs rival. However, most developer support went to EAX. When Aureal closed in mid-2000 and was later absorbed by Creative, this spelled the end for A3D as a viable game API.


Virtually all new sound cards on the market support Creative's EAX, while Creative's own Audigy and Audigy 2 series has gone onto the next level with EAX Advanced HD, which supports up to four streams of accelerated 3D audio for amazing environmental effects. However, many other audio adapter vendors are enhancing the effects of EAX by adopting Sensaura's Virtual Ear engine. Virtual Ear enables the user to adjust the apparent position of sound by adjusting the size and shape of the "ear" used to listen to sound. Virtual Ear is currently available on audio adapters from AOpen, Yamaha, Voyetra Turtle Beach, Guillemot/Hercules, and others. Virtual Ear can be purchased as an upgrade for existing audio adapters that use the ADI SoundMax 2, SoundMax 3, and Yamaha YMF 724 and 744 chips.

3D Audio Processing A second important issue for game players is how the sound cards produce 3D audio. As with 3D video, there are two major methods: Host-based processing (which uses the CPU to process 3D, which can slow down overall system operation) Processing on the audio adapter (referred to as 3D acceleration) Some 3D audio cards perform some or all of the processing necessary for 3D using the host's CPU, whereas others use a powerful DSP that performs the processing on the audio adapter itself. Cards that use host-based processing for 3D can cause major drops in frame rate (frames per second of animation displayed onscreen by a 3D game) when 3D sound is enabled if you use under 1GHz processors, whereas cards with their own 3D audio processors onboard have little change in frame rate whether 3D sound is enabled or disabled. Many of the latest chips from major audio adapter and chipset vendors support 3D acceleration, but the number of 3D audio streams supported varies greatly by chip​and it can sometimes be limited by problems with software drivers. A good rule of thumb for realistic gaming is to have an overall average frame rate of at least 30fps (frames per second). With CPUs running at 1GHz or above, this is easy to achieve with any recent 3D card. However, gamers using older CPUs, such as those running slower than 1GHz, will find that cards using the host CPU for some of the 3D processing will have frame rates that fall below the desired average of 30fps, making for clumsy gameplay. To see the effect of enabling 3D sound on the speed of popular games, you can use the built-in frame-rate tracking feature found in many games or check online game-oriented hardware review sources, such as www.anandtech.com. Frame rates are closely related to CPU utilization; the more CPU attention your 3D audio card requires, the slower


the frame rate will be. As with 3D video, the main users of 3D sound are game developers, but business uses for ultra-realistic sound will no doubt follow.

DirectX Support Issues The latest version of DirectX, DirectX 9, is designed to give all sound cards with 3D support a major boost in performance. Previous versions of DirectX supported 3D with DirectSound3D, but the performance of DirectSound3D was limited. Game programmers needed to test the audio adapter to see whether it supported DirectSound3D acceleration and then would either enable or disable 3D sounds based on the host hardware. Starting with DirectX 5.0, DirectSound3D works with third-party 3D acceleration features. Compared to DirectX 8, DirectX 9 improves 3D audio quality and performance. You can download it from the Microsoft DirectX Web site at www.microsoft.com/windows/directx.

Installing the Sound Card Before you can install a sound card, you must open your computer. In almost all cases today, you will install a PCI audio adapter that supports Plug and Play configuration. Compared to the previous generation of ISA audio adapters, PCI audio adapters use fewer hardware resources, feature a lower CPU utilization rate, and provide better support for advanced 3D gaming APIs. If you need to install an ISA audio adapter, see "Installing the Sound Card (Detailed Procedure)" in Chapter 20 of Upgrading and Repairing PCs, 11th Edition, supplied in electronic form on the DVD packaged with this book. If your computer has integrated audio, in most cases you should disable it. You could have audio conflicts with AC'97 codec-based solutions and resource conflicts with solutions that emulate the Creative Labs Sound Blaster. See Chapter 5, "BIOS," for details. If you have several empty bus slots from which to choose, install the audio adapter in the slot that is as far away as possible from the other cards in the computer. This reduces any possible electromagnetic interference; that is, it reduces stray radio signals from one card that might affect the sound card. The analog components on audio adapters are highly susceptible to interference, and even though they are shielded, they should be protected as well as is possible. Next, you must remove the screw that holds the metal cover over the empty expansion slot you've chosen. Remove your audio adapter from its protective


packaging. When you open the bag, carefully grab the card by its metal bracket and edges. Do not touch any of the components on the card because any static electricity you might transmit can damage the card. Also, do not touch the goldedge connectors. You might want to invest in a grounding wrist strap, which continually drains you of static build-up as you work on your computer. Before you make your final decision about which slot to use for your audio adapter, take a careful look at the external cables you must attach to the card. Front and rear speakers, microphone, game controller, line in, S/PDIF, and other cables that attach to your system can interfere with (or be interfered by) existing cables already attached to your system. It's usually best to choose a slot that allows you to route the audio cables away from other cables. If you're installing a Sound Blaster Live! or Audigy series card that uses an internal 5 1/4'' breakout box (Live! Drive or Audigy Drive), be sure the ribbon cable from the drive bay used for the breakout box can comfortably reach the connector on the sound card. You might have to move a CD-ROM, CD-RW, or DVD drive to a different drive bay to free up a drive bay needed by the breakout box. If your system has an internal CD-ROM drive with an analog audio cable, connect the audio cable to the adapter's CD Audio In connector. This connector is a fourpin connector and is keyed so that you can't insert it improperly. Note that no true standard exists for this audio cable, so be sure you get the correct one that matches your drive and adapter. If you need to purchase one, you can find cables with multiple connectors designed for various brands of CD-ROM drives. This will allow you to play music CDs through the sound card's speakers and to use analog ripping if you want to create MP3 files from your CDs. Many recent CD-ROM and DVD drives also have a digital audio connector that supports a two-wire connector. Attach one end of the digital audio cable to the rear of the drive and the other end to the CD SPDIF or CD Digital Audio connector on the sound card. This enables you to perform digital ripping if you want to create MP3 files from your CDs. Next, insert the adapter's edge connector in the bus slot, but first touch a metal object, such as the inside of the computer's cover, to drain yourself of static electricity. When the card is firmly in place, attach the screw to hold the expansion card and then reassemble your computer.

Connecting PC Speakers and Completing the Installation After the adapter card is installed, you can connect small speakers to the external speaker jack(s). Typically, sound cards provide 4 watts of power per channel to drive bookshelf speakers. If you are using speakers rated for less than 4 watts, do not turn up the volume on your sound card to the maximum; your speakers might


burn out from the overload. You'll get better results if you plug your sound card into powered speakers​that is, speakers with built-in amplifiers. If your sound card supports a four-speaker system, check the documentation to see which jack is used for the front speakers and which for the rear speakers. To use the rear speakers for 3D audio, adjust the properties with the mixer control software supplied with your sound card.


Tip If you have powered speakers but don't have batteries in them or have them connected to an AC adapter, don't turn on the speakers! Turning on the speakers without power will prevent you from hearing anything at all. Leave the speakers turned off and use the volume control built into your sound card's mixer software instead. Powered speakers sound better, but most small models can run without power in an emergency. Some computer power supplies feature small jacks to provide power for computer speakers.

When the sound card installation is finished, you should have a speaker icon in the Windows System Tray. If the speaker icon (indicating the Volume Control) isn't visible, you can install it through the Control Panel's Add/Remove Programs icon. With Windows 9x/Me, select the Windows Setup tab and open the Multimedia section. Then, check the box labeled Volume Control. With Windows XP, open the Sounds and Audio Devices icon in Control Panel, click the Volume tab, and click the Place Volume icon in the taskbar box. In some cases you might be asked to insert the Windows CD-ROM if additional drivers are required to complete the installation. If you use digital sound sources or output such as Dolby 5.1, CD digital, or S/PDIF, open the properties sheet for your mixer device and enable display of these volume controls. Use the Volume Control to ensure your speakers are receiving a sound signal. The mixer sometimes defaults to Mute. You can usually adjust volume separately for wave (WAV) files, MIDI, microphone, and other components.

Using Your Stereo Instead of Speakers Another alternative is to patch your sound card into your stereo system for greatly amplified sound and for support of advanced Dolby Digital sound for DVD playback. Check the plugs and jacks at both ends of the connection. Most stereos use pin plugs, also called RCA or phono plugs, for input. Although pin plugs are standard on some sound cards, most use miniature 1/8'' phono plugs, which require an adapter when connecting to your stereo system. For example, from Radio Shack you can purchase an audio cable that provides a stereo 1/8'' miniplug on one end and phono plugs on the other (Cat. No. 42-2481A). If you want to attach your sound card to Dolby 5.1 speakers, be sure you use cabling designed for the S/PDIF connectors on your sound card. Some might use RCA-type plugs, whereas others use an optical cable with a square end. Make sure that you get stereo​not mono​plugs, unless your sound card supports only mono. To ensure that you have enough cable to reach from the back of your PC to your stereo system, get a 6-ft. long cable.


Hooking up your stereo to an audio adapter is a matter of sliding the plugs into the proper jacks. If your audio adapter gives you a choice of outputs​speaker/headphone and stereo line-out​choose the stereo line-out jack for the connection. This will give you the best sound quality because the signals from the stereo line-out jack are not amplified. The amplification is best left to your stereo system. In some cases, you'll attach a special DIN plug to your audio adapter that has multiple connections to your stereo system. Connect this cable output from your audio adapter to the auxiliary input of your stereo receiver, preamp, or integrated amplifier. If your stereo doesn't have an auxiliary input, other input options include​in order of preference​tuner, CD, or Tape 2. (Do not use phono inputs, however, because the level of the signals will be uneven.) You can connect the cable's single stereo miniplug to the sound card's stereo line-out jack, for example, and then connect the two RCA phono plugs to the stereo's Tape/VCR 2 Playback jacks. The first time you use your audio adapter with a stereo system, turn down the volume on your receiver to prevent blown speakers. Barely turn up the volume control and then select the proper input (such as Tape/VCR 2) on your stereo receiver. Finally, start your PC. Never increase the volume to more than threefourths of the way up. Any higher and the sound might become distorted.


Note If your stereo speakers are not magnetically shielded, you might hear a lot of crackling if they are placed close to your computer. Try moving them away from the computer, or use magnetically shielded speakers.


Tricks for Using the Tape Monitor Circuit of Your Stereo Your receiver might be equipped with something called a tape monitor. This outputs the sound coming from the tuner, tape, or CD to the tape-out port on the back; it then expects the sound to come back in on the tape-in port. These ports, in conjunction with the line-in and line-out ports on your audio adapter, enable you to play computer sound and the radio through the same set of speakers. Here's how you do it: 1. Turn off the tape monitor circuit on your receiver. Turn down all the controls on the sound card's mixer application. Connect the receiver's tape-out ports to the audio adapter's line-in port. Connect the audio adapter's line-out port to the receiver's tape-in ports. Turn on the receiver, select some music, and set the volume to a medium level. Turn on the tape monitor circuit. Slowly adjust the line-in and main-out sliders in the audio adapter's mixer application until the sound level is about the same as before. Disengage and re-engage the tape monitor circuit while adjusting the output of the audio adapter so that the sound level is the same regardless of whether the tape monitor circuit is engaged. Start playing a WAV file. Slowly adjust up the volume slider for the WAV file in the audio adapter's mixer application until it plays at a level (slightly above or below the receiver) that is comfortable. Now you can get sounds from your computer and the radio through the receiver's speakers.

Different connectors might be needed if you have digital surround speakers and newer PCI-based sound cards. Check your speakers and sound card before you start this project.


Troubleshooting Sound Card Problems To operate, an audio adapter needs hardware resources, such as IRQ numbers, a base I/O address, and DMA channels that don't conflict with other devices. Most adapters come preconfigured to use the standard Sound Blaster resources that have come to be associated with audio adapters. However, problems occasionally arise even with Plug and Play adapters. Troubleshooting might mean that you have to change the settings used by your system BIOS for PnP devices, move the sound card to another slot, or even reconfigure the other devices in your computer. No one said life was fair.

Hardware (Resource) Conflicts The most common problem for audio adapters (particularly if you still use ISA cards) is that they might conflict with other devices installed in your PC. You might notice that your audio adapter simply doesn't work (no sound effects or music), repeats the same sounds over and over, or causes your PC to freeze. This situation is called a device, or hardware, conflict. What are they fighting over? Mainly the same bus signal lines or channels (called resources) used for talking to your PC. The sources of conflict in audio adapter installations are generally threefold: Interrupt Requests (IRQs). Hardware devices use IRQs to "interrupt" your PC's CPU and get its attention. PCI cards can share IRQs, but ISA cards and onboard legacy ports such as serial, parallel, and PS/2 mouse ports can't. Direct Memory Access (DMA) channels. DMA channels move information directly to your PC's memory, bypassing the system processor. DMA channels enable sound to play while your PC is doing other work. ISA sound cards and PCI sound cards emulating the Sound Blaster standard require DMA settings; PCI sound cards running in native mode don't use DMA channels. Input/output (I/O) port addresses. Your PC uses I/O port addresses to channel information between the hardware devices on your audio adapter and PC. The addresses usually mentioned in a sound card manual are the starting or base addresses. An audio adapter has several devices on it, and each one uses a range of addresses starting with a particular base address. Most audio adapters include installation software that analyzes your PC and attempts to notify you if any of the standard settings are in use by other devices.


The Windows Device Manager (accessed from the System Control Panel) can also help you to resolve conflicts. Although these detection routines can be fairly reliable, unless a device is operating during the analysis, it might not always be detectable. Some of the newer PCI-based sound cards and Intel chipset motherboards might not properly support ISA-type I/O addresses used by Sound Blaster​compatible software to communicate with the card. If you have problems getting older games to work with your system, see the tips listed earlier for emulation methods, check with your sound card and system board/computer supplier for help, and check with the game developer for possible patches and workarounds. Table 16.4 shows the default resources used by the components on a typical Sound Blaster 16 card, which uses a 16-bit ISA slot. Other sound cards used to emulate a Sound Blaster typically use these same settings. Table 16.4. Default Sound Blaster Resource Assignments Device

Interrupt

Audio

IRQ 5

I/O Ports 220h​233h

MIDI Port

330h​331h

FM Synthesizer

388h​38Bh

Game Port

200h​207h

16-Bit DMA DMA 5

8-Bit DMA DMA 1

All these resources are used by a single sound card in your system. No wonder so many people have had conflicts and problems with audio adapter installations! In reality, working out these conflicts is not all that hard, as you will see. You can change most of the resources that audio adapters use to alternative settings, if conflicts with other devices occur; even better, you can change the settings of the other device to eliminate the conflicts. Note that some devices on the audio adapter, such as the MIDI port, FM synthesizer, and game port, do not use resources such as IRQs or DMA channels. Compare the resource settings used by a traditional ISA sound card (refer to Table 16.4) to those used by a typical PCI-based sound card (see Table 16.5). Table 16.5. Typical Sound Blaster Audigy Resource Assignments Device Audio

Interrupt IRQ 3

I/O Ports 4000​401F

16-Bit DMA

8-Bit DMA


MIDI Port

(included in above configuration)

FM Synthesizer

(included in above configuration)

Game Port

1400​1407

If you don't need MS-DOS game support, the PCI card uses far fewer resources than the ISA card. However, if you are using MS-DOS games, you will need to set up a number of "virtual" IRQ, I/O port, and DMA port settings for use with those games only. When using a typical PCI-based sound card's Sound Blaster Legacy settings, you must verify that MS-DOS software using the Sound Blaster is set for the appropriate configuration (see Figure 16.10).

Figure 16.10. One advantage of PCI cards is visible here, even in emulation mode: All emulations can use the same IRQ, rather than up to three different IRQs, as is the case with most thirdparty ISA cards that emulate the Sound Blaster.

If you are still using MS-DOS games (or even some Windows programs designed for Windows 95 or earlier), I recommend that you install your audio adapter (or configure a notebook computer's built-in sound adapter circuits) to use the default settings whenever possible. This is mainly because of poorly written software that can't work properly with alternative settings, even if they do not cause conflicts. In other words, if you are having a conflict with another type of adapter, modify the settings of the other device, rather than those of the audio adapter. Take this from experience; otherwise, you will have to explain to your five-year-old why the new Dinosaur program you just installed does not make any sounds! This problem is primarily associated with DOS-based game programs, but some older


Windows-based programs have also been incapable of working with alternative hardware settings.


Note Some of the chipsets for "legacy-free" systems, such as the Intel 810 and 820 chipsets, don't decode addresses below 1000h to the PCI bus. All the Sound Blaster I/O port addresses are well below 1000h, and thus sound card manufacturers will need to adjust driver software or use TSR programs to reroute calls to these I/O port addresses by software to the correct I/O port addresses. This design choice is not a problem for typical Windows software that uses sound, but it can cause legacy (DOS-based) applications to fail.

Resolving Resource Conflicts Resource conflicts are quite rare with current PCI-based audio adapters, thanks to their support of Windows Plug and Play technology and the PCI IRQ sharing/steering feature supported by Windows 95 OSR 2.x and above and recent chipsets. However, if you are still using the original version of Windows 95 or an ISA-based audio adapter (including some motherboard-based chipsets on older systems), you can have significant problems because of IRQ and DMA conflicts with other devices. See Upgrading and Repairing PCs, 12th Edition on the DVD-ROM for details.

Other Sound Card Problems Like the common cold, audio adapter problems have common symptoms. Use the following sections to diagnose your problem.

No Sound If you don't hear anything from your audio adapter, consider these solutions: Make sure the audio adapter is set to use all default resources and that all other devices using these resources have been either reconfigured or removed. Use the Device Manager to determine this information. Are the speakers connected? Check that the speakers are plugged into the sound card's stereo line-out or speaker jack (not the line-in or microphone jack). Are the speakers receiving power? Check that the power "brick" or power cord


is plugged in securely. Are the speakers stereo? Check that the plug inserted into the jack is a stereo plug, not mono. Are the mixer settings correct? Many audio adapters include a sound mixer application. The mixer controls the volume settings for various sound devices, such as the microphone or the CD player. There might be separate controls for both recording and playback. Increase the master volume or speaker volume when you are in the play mode. If the Mute option is selected in your sound mixer software, you won't hear anything. Depending on the speaker type and sound source type, you might need to switch from analog to digital sound for some types of sound output. Make sure that the correct digital audio volume controls are enabled in your audio device's mixer control. Use your audio adapter's setup or diagnostic software to test and adjust the volume of the adapter. Such software usually includes sample sounds used to test the adapter. Turn off your computer for 1 minute and then turn it back on. A hard reset (as opposed to pressing the Reset button or pressing Ctrl+Alt+Delete) might clear the problem. If your computer game lacks sound, check that it is designed to work with your audio adapter. For example, some legacy and early Windows games might require the exact settings of IRQ 7 (or IRQ 5), DMA 1, and I/O address 220 to be Sound Blaster compatible. You also might need to load DOS drivers to enable some recent sound cards to work with DOS games.

One-Sided Sound If you hear sound coming from only one speaker, check out these possible causes: Are you using a mono plug in the stereo jack? A common mistake is to use a mono plug in the sound card's speaker or stereo-out jacks. Seen from the side, a stereo connector has two darker stripes. A mono connector has only one stripe. If you're using amplified speakers, are they powered on? Check the strength of the batteries or the AC adapter's connection to the electrical outlet. If each


speaker is powered separately, be sure that both have working batteries. Are the speakers wired correctly? When possible, use keyed and color-coded connectors to avoid mistakes. Is the audio adapter driver loaded? Some sound cards provide only leftchannel sound if the driver is not loaded correctly. Rerun your adapter's setup software or reinstall it in the operating system. Are both speakers set to the same volume? Some speakers use separate volume controls on each speaker. Balance them for best results. Separate speaker volume controls can be an advantage if one speaker must be farther away from the user than the other. Is the speaker jack loose? If you find that plugging your speaker into the jack properly doesn't produce sound but pulling the plug half-way out or "jimmying" it around in its hole can temporarily correct the problem, you're on the road to a speaker jack failure. There's no easy solution; buy a new adapter or whip out your soldering iron and spend a lot more time on the test bench than most audio adapters are worth. To avoid damage to the speaker jack, be sure you insert the plug straight in, not at an angle.

Volume Is Low If you can barely hear your sound card, try these solutions: Are the speakers plugged into the proper jack? Speakers require a higher level of drive signal than headphones. Again, adjust the volume level in your mixer application. Are the mixer settings too low? Again, adjust the volume level in your mixer application. If your mixer lets you choose between speakers and headphones, be sure to select the correct speaker configuration. Is the initial volume too low? If your audio adapter has an external thumbwheel volume control located on the card bracket, check to ensure that it is not turned down too low. Are the speakers too weak? Some speakers might need more power than your audio adapter can produce. Try other speakers or put a stereo amplifier between your sound card and speakers.


Scratchy Sound Scratchy or static-filled sound can be caused by several problems. Improving the sound can be as simple as rearranging your hardware components. The following list suggests possible solutions to the problem of scratchy sound: Is your audio adapter near other expansion cards? The adapter might be picking up electrical interference from other expansion cards inside the PC. Move the audio card to an expansion slot as far away as possible from other cards. An ISA-based audio adapter requires a lot of CPU attention. Frequent hard disk access can cause dropouts due to the CPUs switching between managing the sound card and the hard drive. Are your speakers too close to your monitor? The speakers can pick up electrical noise from your monitor. Move them farther away. Subwoofers should never be placed near the monitor because their powerful magnets can interfere with the picture. They should be on the floor to maximize lowfrequency transmission. Are you experiencing compatibility problems between particular games and your sound card? If you notice sound problems such as stuttering voices and static on some games but not others, check with the game vendor for a software patch or with the sound card vendor for updated drivers. If the game uses DirectX, run the DXDIAG diagnostics program (select Start, Run; type DXDIAG; and click OK) and click the Sound tab. Adjust the slider for Hardware Sound Acceleration Level down one notch from Full (the default) to Standard, click Save All Information, and exit. Retry the game. If the problem persists, adjust the Hardware Sound Acceleration Level to Basic. If other games have performance problems after you adjust the Hardware Sound Acceleration Level, be sure to reset it to Full before playing those games.

Your Computer Won't Start If your computer won't start at all, you might not have inserted the audio adapter completely into its slot. Turn off the PC and then press firmly on the card until it is seated correctly. If you can't start your computer after installing a new sound card and its drivers, you can use the Windows "bootlog" feature to record every event during startup; this file records which hardware drivers are loaded during startup and indicates


whether the file loaded successfully, didn't load successfully, or froze the computer. See the documentation for your version of Windows for details on how to create a bootlog when necessary.

Parity Errors or Other Lockups Your computer might display a memory parity error message or simply crash. This is usually caused by resource conflicts in one of the following areas: IRQ DMA I/O ports If other devices in your system are using the same resources as your audio adapter, crashes, lockups, or parity errors can result. You must ensure that multiple devices in your system do not share these resources.

Advanced Features If you are having problems playing DVD audio, playing MP3 files, or using SPDIF connections, make sure that You have enabled the hardware resources on the sound card. You are using the correct playback program. Your mixer has the correct volume control setting for the device. Your cabling is correct for the device.

Other Problems Sometimes sound problems can be difficult to solve. Due to quirks and problems with the way DMA is implemented in some motherboard chipsets, problems interacting with certain cards or drivers can occur. Sometimes altering the Chipset Setup options in your CMOS settings can resolve problems. These types of problems can take a lot of trial and error to solve.


The PC standard is based loosely on the cooperation among a handful of companies. Something as simple as one vendor's BIOS or motherboard design can make the standard nonstandard. A good way to solve problems of all types with Plug and Play cards, a PnP BIOS, and a PnP operating system (Windows 9x/Me/2000/XP) is to use the Device Manager to remove the sound card, restart the system, and allow the card's components to be redetected. This installs a "fresh" copy of the software and reinserts Registry entries. If you are using a motherboard with a VIA chipset, be sure to download and install the latest versions of VIA drivers.


Speakers Successful business presentations, multimedia applications, and MIDI work demand external high-fidelity stereo speakers. Although you can use standard stereo speakers, they are often too big to fit on or near your desk. Smaller bookshelf speakers are better. Sound cards offer little or none of the amplification needed to drive external speakers. Although some sound cards have small 4-watt amplifiers, they are not powerful enough to drive quality speakers. Also, conventional speakers sitting near your display can create magnetic interference, which can distort colors and objects onscreen or jumble the data recorded on nearby floppy disks or other magnetic media. To solve these problems, computer speakers need to be small, efficient, and selfpowered. Also, they should be provided with magnetic shielding, either in the form of added layers of insulation in the speaker cabinet or electronic cancellation of the magnetic distortion.


Caution Although most computer speakers are magnetically shielded, do not leave recorded tapes, watches, credit cards, or floppy disks in front of the speakers for long periods of time.

Quality sound depends on quality speakers. A 16-bit audio adapter might provide better sound to computer speakers, but even an 8-bit adapter sounds good from a good speaker. Conversely, an inexpensive speaker makes both 8-bit and 16-bit adapter cards sound tinny. Now dozens of models of PC speakers are on the market, ranging from inexpensive minispeakers from Sony, Creative, and LabTech to larger selfpowered models from prestigious audio companies such as Bose, Cambridge Sound Works, Klipsch, Monsoon, and Altec Lansing. Many of the medium- to higher-end speaker systems even include subwoofers to provide additional bass response. To evaluate speakers, it helps to know the jargon. Speakers are measured by three criteria: Frequency response. A measurement of the range of high and low sounds a speaker can reproduce. The ideal range is 20Hz​20KHz, the range of human hearing. No speaker system reproduces this range perfectly. In fact, few people hear sounds above 18KHz. An exceptional speaker might cover a range of 30Hz​23,000Hz, and lesser models might cover only 100Hz​20,000Hz. Frequency response is the most deceptive specification because identically rated speakers can sound completely different. Total Harmonic Distortion (THD). An expression of the amount of distortion or noise created by amplifying the signal. Simply put, distortion is the difference between the sound sent to the speaker and the sound you hear. The amount of distortion is measured in percentages. An acceptable level of distortion is less than .1% (one-tenth of 1%). For some CD-quality recording equipment, a common standard is .05%, but some speakers have a distortion of 10% or more. Headphones often have a distortion of about 2% or less. Watts. Usually stated as watts per channel, this is the amount of amplification available to drive the speakers. Check that the company means "per channel" (or RMS) and not total power. Many audio adapters have built-in amplifiers, providing up to 8 watts per channel (most provide 4 watts). This wattage is not enough to provide rich sound, however, which is why many speakers have built-in amplifiers. With the flick of a switch or the press of a button, these speakers amplify the signals they receive from the audio adapter. If you do not want to amplify the sound, you typically leave the speaker switch set to "direct." In most cases, you'll want to amplify the signal.


Inexpensive PC speakers sometimes use batteries to power the amplifiers. Because these speakers require so much power, you might want to invest in an AC adapter or purchase speakers that use AC power. With an AC adapter, you won't have to buy new batteries every few weeks. If your speakers didn't come with an AC adapter, you can pick one up from your local Radio Shack or hardware store. Be sure that the adapter you purchase matches your speakers in voltage and polarity; most third-party adapters are multiple voltage, featuring interchangeable tips and reversible polarity. You can control the volume and other sound attributes of your speakers in various ways, depending on their complexity and cost. Typically, each speaker has a volume knob, although some share a single volume control. If one speaker is farther away than the other, you might want to adjust the volume accordingly. Many computer speakers include a dynamic bass boost (DBB) switch. This button provides a more powerful bass and clearer treble, regardless of the volume setting. Other speakers have separate bass and treble boost switches or a threeband equalizer to control low, middle, and high frequencies. When you rely on your audio adapter's power rather than your speakers' built-in amplifier, the volume and dynamic bass boost controls have no effect. Your speakers are at the mercy of the adapter's power. For best audio quality, adjust the master volume on the sound card near the high end and use the volume control on powered speakers to adjust the volume. Otherwise, your speakers will try to amplify any distortions coming from the lowpower input from the PC's audio adapter. A 1/8'' stereo minijack connects from the audio adapter's output jack to one of the speakers. The speaker then splits the signal and feeds through a separate cable from the first speaker to the second one (often referred to as the "satellite speaker"). Before purchasing a set of speakers, check that the cables between the speakers are long enough for your computer setup. For example, a tower case sitting alongside your desk might require longer speaker wires than a desktop computer. Beware of speakers that have a tardy built-in sleep feature. Such speakers, which save electricity by turning themselves off when they are not in use, might have the annoying habit of clipping the first part of a sound after a period of inactivity. Speakers that are USB based will not be capable of playing CD music unless the CD-ROM drive can perform digital audio extraction. Check your drive's specifications for information. Headphones are an option when you can't afford a premium set of speakers. Headphones also provide privacy and enable you to play your PC audio as loud as you like.


For best results with newer sound cards that support four speakers or more, check the properties sheet for the audio adapter and set whether you're using headphones, stereo speakers, or a larger number of speakers. Make sure that speakers are placed properly. If you use a subwoofer, put it on the floor for better bass sound and to reduce EMI interference with other devices. How can you tell whether wireless satellite speakers are causing interference? Watch your monitor; frequencies as high as 2KHz can interfere with your video display. Move the speakers away from the monitor and check the display again.

Theater and Surround Sound Considerations If you're a serious gamer or DVD movie lover, you won't be content with ordinary stereophonic sound. Most audio adapters now support front and rear speakers, and many of the best audio adapters also support Dolby-compatible 4.1 and 5.1 speaker setups. To ensure you get the sound you expect from four or more speakers, check the following: Use the properties sheet for your audio adapter to properly describe your speaker setup. This includes selecting the number of speakers you are using, setting options for 3D environmental audio and positional sound such as reverb, and setting up your subwoofer if present. Make sure you use the correct cabling between your speakers and audio adapter. If you are planning to use AC3/Dolby speaker setups, such as 4.1, 5.1, 6.1, or 7.1, be sure you use the correct S/PDIF connection and configuration. This varies from audio adapter to audio adapter; check the vendor's Web site for details. Make sure you have placed your speakers correctly. In some cases you can adjust the audio adapter's properties to improve sound quality, but sometimes you might need to move the speakers themselves. Make sure you have connected your speakers to the proper jacks. Mixing up left and right or front and rear causes poor sound quality.

Typical Speaker Setups


The simplest audio configuration available today is stereo, which uses two speakers placed to overlap sound. Most audio adapters now support at least four speakers, but depending on the audio adapter, settings, and sound output options in the program, the rear speakers might simply mirror the front speakers' output, or you might have four distinct sound streams. 4-point surround sound uses four speakers plus a subwoofer to surround you with music and gaming sound effects; the four speakers are placed around the listener, and the subwoofer is usually placed near a wall or in the corner to amplify its lowfrequency sound. The subwoofer in such setups is not on a separate circuit but is controlled by the same signals sent to the other speakers. 5.1 Surround sound, also referred to as Dolby Digital or DTS Surround sound, uses five speakers plus a subwoofer. The fifth speaker is placed between the front two speakers to fill in any missing sound caused by incorrect speaker placement. The subwoofer is independently controlled. This is the preferred sound system for use with DVD movies. Most lower-cost audio adapters lack support for 5.1 Surround sound. Some of the latest sound cards support 6.1 and 7.1 Surround sound. The 6.1 configuration resembles the 5.1 Surround setup but adds a middle speaker along with a subwoofer. 7.1 Surround sound uses left-middle and right-middle speakers to flank the listener, along with a subwoofer.


Microphones Some audio adapters come complete with a microphone, but most do not. You'll need one to record your voice to a WAV file. Selecting a microphone is quite simple. You need one that has a 1/8'' minijack to plug into your audio adapter's microphone, or audio in, jack. Most microphones have an on/off switch. Like speakers, microphones are measured by their frequency ranges. This is not an important buying factor, however, because the human voice has a limited range. If you are recording only voices, consider an inexpensive microphone that covers a limited range of frequencies. An expensive microphone's recording capabilities extend to frequencies outside the voice's range. Why pay for something you won't be needing? If you are recording music, invest in an expensive microphone, but be sure that your audio adapter can do justice to the signal produced by the microphone. A high-quality microphone can produce mediocre results when paired with a cheap 8-bit audio adapter. Your biggest decision is to select a microphone that suits your recording style. If you work in a noisy office, you might want a unidirectional microphone that will prevent extraneous noises from being recorded. An omnidirectional microphone is best for recording a group conversation. Some audio adapters include a microphone. This can be a small lapel microphone, a handheld microphone, or one with a desktop stand. If you want to keep your hands free, you might want to shun the traditional handheld microphone for a lapel or desktop model. If your audio adapter does not come with a microphone, see your local stereo or electronics parts store. Be sure that any microphone you purchase has the correct impedance to match the audio adapter's input. If you're using software such as Dragon Naturally Speaking, IBM Via Voice, Philips FreeSpeech, or other voice-recognition software, use the microphone supplied with the software or choose from alternative models the software vendor recommends. Run the microphone setup program again if your software has trouble recognizing your voice. Some newer models feature a battery pack to boost sound quality; be sure to check the batteries and replace them to keep recognition quality high. If you're talking but your voice-recognition or recording software isn't responding, check the following: Incorrect jack. It's easy to plug the microphone into the wrong jack. Try using a magic marker to color-code the microphone wire and jack to make matching


up easier. Check the recording volume in the mixer control. This usually defaults to Mute to avoid spurious noise. Make sure the microphone is turned on in the voice-recognition or recording software. You must click the Record button in recording software, and many voice-recognition programs let you "pick up" the microphone for use or "put it down" when you need to answer the phone. Look for an onscreen microphone icon in the Windows System Tray for fast toggling between modes.


Chapter 17. I/O Interfaces from Serial and Parallel to IEEE-1394 and USB Introduction to Input/Output Ports USB and IEEE-1394 (i.Link or FireWire) Standard Serial and Parallel Ports Serial Ports Parallel Ports


Introduction to Input/Output Ports This chapter covers the primary peripheral input/output ports on a modern PC system. This includes a discussion of both the so-called "legacy" serial and parallel ports that have been standard on PCs since the beginning, as well as a discussion of the more current Universal Serial Bus (USB), which is replacing both serial and parallel ports, and IEEE-1394 (i.Link or FireWire) interfaces. (IEEE stands for the Institute of Electrical and Electronic Engineers.) Although SCSI and IDE are also I/O interfaces, they are mainly used as internal interfaces and are important or complicated enough to warrant their own chapters for more specific and detailed coverage.


USB and IEEE-1394 (i.Link or FireWire) The two most popular high-speed serial-bus architecture families for desktop and portable PCs are Universal Serial Bus (USB) and IEEE-1394, which is also called i.Link or FireWire. Each interface type is available in two versions: USB 1.1 and USB 2.0; IEEE-1394a and IEEE-1394b (also called FireWire 800). The USB and IEEE-1394 port families are high-speed communications ports that far outstrip the capabilities of older standard serial and parallel ports. They can also be used as an alternative to SCSI for high-speed external peripheral connections. In addition to performance, these newer ports offer I/O device consolidation, which means that all types of external peripherals can connect to these ports.

Why Serial? The recent trend in high-performance peripheral bus design is to use a serial architecture, in which 1 bit at a time is sent down a wire. Because parallel architecture (used by SCSI, ATA, and LPT ports) uses 8, 16, or more wires to send bits simultaneously, the parallel bus is actually much faster at the same clock speed. However, increasing the clock speed of a serial connection is much easier than increasing that of a parallel connection. Parallel connections in general suffer from several problems, the biggest being signal skew and jitter. Skew and jitter are the reasons high-speed parallel buses such as SCSI (small computer systems interface) are limited to short distances of 3 meters or less. The problem is that, although the 8 or 16 bits of data are fired from the transmitter at the same time, by the time they reach the receiver, propagation delays have conspired to allow some bits to arrive before the others. The longer the cable, the longer the time between the arrival of the first and last bits at the other end! This signal skew, as it is called, prevents you from running a high-speed transfer rate or a longer cable​or both. Jitter is the tendency for the signal to reach its target voltage and float above and below for a short period of time. With a serial bus, the data is sent 1 bit at a time. Because there is no worry about when each bit will arrive, the clocking rate can be increased dramatically. For example, the top transfer rate possible with EPP/ECP parallel ports is 2MBps, whereas IEEE-1394a ports (which use high-speed serial technology) support transfer rates as high as 400Mbps (about 50MBps)​25 times faster than parallel ports. USB 2.0 supports transfer rates of 480Mbps (about 60MBps), which is about 30 times faster than parallel ports, and the new IEEE-1394b (FireWire 800) ports reach transfer rates as high as 800Mbps (or about 100MBps), which is about 50 times faster than parallel ports!


At high clock rates, parallel signals tend to interfere with each other. Serial again has an advantage because, with only one or two signal wires, crosstalk and interference between the wires in the cable are negligible. In general, parallel cabling is more expensive than serial cabling. Besides the many additional wires needed to carry the multiple bits in parallel, the cable also must be specially constructed to prevent crosstalk and interference between adjacent data lines. This is one reason external SCSI cables are so expensive. Serial cabling, by comparison, is very inexpensive. For one thing, it has significantly fewer wires. Furthermore, the shielding requirements are far simpler, even at very high speeds. Because of this, transmitting serial data reliably over longer distances is also easier, which is why parallel interfaces have shorter recommended cable lengths than do serial interfaces. For these reasons​in addition to the need for new Plug and Play external peripheral interfaces and the elimination of the physical port crowding on portable computers​these high-performance serial buses were developed. USB is a standard feature on virtually all PCs today; is used for most general-purpose, high-speed external interfacing; and is the most compatible, widely available, and fastest general-purpose external interface. In addition, IEEE-1394 (more commonly known as FireWire), although mainly used in certain niche markets​such as connecting DV (digital video) camcorders​is also spreading into other highbandwidth uses, such as high-resolution scanners, external hard drives, and networking.

Universal Serial Bus Universal Serial Bus (USB) is an external peripheral bus standard designed to bring Plug and Play capability for attaching peripherals externally to the PC. USB eliminates the need for special-purpose ports, reduces the need to use specialpurpose I/O cards, (thus reducing the need to reconfigure the system with each new device added), and saves important system resources such as interrupts (IRQs); regardless of the number of devices attached to a system's USB ports, only one IRQ is required. PCs equipped with USB enable peripherals to be automatically recognized and configured as soon as they are physically attached, without the need to reboot or run setup. USB allows up to 127 devices to run simultaneously on a single bus, with peripherals such as monitors and keyboards acting as additional plug-in sites, or hubs. USB cables, connectors, hubs, and peripherals can be identified by icons, as shown in Figure 17.1. Note the "plus" symbol added to the upper icon, which indicates that port supports USB 2.0 (HiSpeed USB) in addition to the standard 1.x support.

Figure 17.1. These icons identify USB cables, connectors, hubs,


and peripherals.

Intel has been the primary proponent of USB, and all its PC chipsets starting with the PIIX3 South Bridge chipset component (introduced in February 1996) have included USB support as standard. Other chipset vendors have followed suit, making USB as standard a feature of today's desktop and notebook PCs as the serial and parallel ports once were. Six other companies initially worked with Intel in co-developing the USB, including Compaq, Digital, IBM, Microsoft, NEC, and Northern Telecom. Together, these companies have established the USB Implementers Forum (USB-IF) to develop, support, and promote USB architecture. See "Chipsets," p. 229.

The USB-IF formally released USB 1.0 in January 1996, USB 1.1 in September 1998, and USB 2.0 in April 2000. The 1.1 revision was mostly a clarification of some issues related to hubs and other areas of the specification. Most devices and hubs should be 1.1 compliant, even if they were manufactured before the release of the 1.1 specification. The biggest change was USB 2.0, which is 40 times faster than the original USB and yet fully backward compatible. USB ports can be retrofitted to older computers that lack built-in USB connectors through the use of either an add-on PCI card (for desktop computers) or a PC Card on Cardbuscompatible notebook computers. You can also use USB add-on cards to update an older system that has only USB 1.1 on the motherboard. As of mid-2002, virtually all motherboards include four or more USB 2.0 ports as standard. Notebook computers were slower to catch on​it wasn't until early 2003 that most notebook or laptop computers included USB 2.0 ports as standard.

USB Technical Details USB 1.1 runs at 12Mbps (1.5MBps) over a simple four-wire connection. The bus supports up to 127 devices connected to a single root hub and uses a tiered-star topology, built on expansion hubs that can reside in the PC, any USB peripheral,


or even standalone hub boxes. Note that although the standard allows up to 127 devices to be attached, they all must share the 1.5MBps bandwidth, meaning that for every active device you add, the bus will slow down some. In practical reality, few people will have more than 8 devices attached at any one time. For low-speed peripherals, such as pointing devices and keyboards, the USB also has a slower 1.5Mbps subchannel. The subchannel connection is used for slower interface devices, such as keyboards and mice. USB employs what is called Non Return to Zero Invert (NRZI) data encoding. NRZI is a method of encoding serial data in which 1s and 0s are represented by opposite and alternating high and low voltages where there is no return to a zero (or reference) voltage between the encoded bits. In NRZI encoding, a 1 is represented by no change in signal level, and a 0 is represented by a change in level. A string of 0s causes the NRZI data to toggle each bit time; a string of 1s causes long periods with no transitions in the data. This is an efficient transfer encoding scheme because it eliminates the need for additional clock pulses that would otherwise waste time and bandwidth. USB devices are considered either hubs or functions, or both. Functions are the individual devices that attach to the USB, such as a keyboard, mouse, camera, printer, telephone, and so on. Hubs provide additional attachment points to the USB, enabling the attachment of more hubs or functions. The initial ports in the PC system unit are called the root hub, and they are the starting point for the USB. Most motherboards have two, three, or four USB ports, any of which can be connected to functions or additional hubs. Some systems place one or two of the USB ports in the front of the computer, which is very convenient for devices you use only occasionally, such as digital cameras or flash memory card readers. External hubs (also called generic hubs) are essentially wiring concentrators, and through a star-type topology they allow the attachment of multiple devices. Each attachment point is referred to as a port. Most hubs have either four or eight ports, but more are possible. For more expandability, you can connect additional hubs to the ports on an existing hub. The hub controls both the connection and distribution of power to each of the connected functions. A typical hub is shown in Figure 17.2.

Figure 17.2. A typical USB hub with four ports.


Besides providing additional sockets for connecting USB peripherals, a hub provides power to any attached peripherals. A hub recognizes the dynamic attachment of a peripheral and provides at least 0.5W of power per peripheral during initialization. Under control of the host PC driver software, the hub can provide more device power, up to a maximum of 2.5W, for peripheral operation.


Tip For the most reliable operation, I recommend that you use self-powered hubs, which plug into an AC adapter. Bus-powered hubs pull power from the PC's USB root hub connector and aren't always capable of providing adequate power for high-power requirement devices, such as optical mice.

A newly attached hub is assigned a unique address, and hubs can be cascaded up to five levels deep (see Figure 17.3). A hub operates as a bidirectional repeater and repeats USB signals as required both upstream (toward the PC) and downstream (toward the device). A hub also monitors these signals and handles transactions addressed to itself. All other transactions are repeated to attached devices. A USB 1.1 hub supports both 12Mbps (full-speed) and 1.5Mbps (lowspeed) peripherals.

Figure 17.3. A typical PC with USB devices can use multiple USB hubs to support a variety of peripherals, connected to whichever hub is most convenient.

Maximum cable length between two full-speed (12Mbps) devices or a device and a hub is 5 meters using twisted-pair shielded cable with 20-gauge wire. Maximum cable length for low-speed (1.5Mbps) devices using non-twisted-pair wire is 3 meters. These distance limits are shorter if smaller-gauge wire is used (see Table 17.1). Table 17.1. Maximum Cable Lengths Versus Wire Gauge Gauge

Resistance (in Ohms/Meter

Length (Max.)

/m)

28

0.232

/m

0.81m

26

0.145

/m

1.31m


24

0.091

/m

2.08m

22

0.057

/m

3.33m

20

0.036

/m

5.00m

Although USB 1.1 is not as fast at data transfer as FireWire or SCSI, it is still more than adequate for the types of peripherals for which it is designed. USB 2.0 operates a surprising 40 times faster than USB 1.1 and allows transfer speeds of 480Mbps or 60MBps. Because it is fully backward-compatible and supports older 1.1 devices, I recommend purchasing only motherboards and add-in USB cards that conform to the faster USB 2.0 (Hi-Speed USB) standard. One of the additional benefits of USB 2.0 is the capability to handle concurrent transfers, which enables your USB 1.1 devices to transfer data at the same time without tying up the USB bus. USB 2.0 drivers were not provided with the initial launch of Windows XP but are available through system update downloads or service packs. Use the Windows Update feature to connect to the Microsoft site and download any updates as necessary. Add-on USB 2.0 cards might include their own drivers, which should be installed.

USB Connectors Four main styles of connectors are specified for USB, called Series A, Series B, Mini-A, and Mini-B connectors. The A connectors are used for upstream connections between a device and the host or a hub. The USB ports on motherboards and hubs are usually Series A connectors. Series B connectors are designed for the downstream connection to a device that has detachable cables. In all cases, the mini connectors are simply smaller versions of the larger ones, in a physically smaller form factor for smaller devices. The physical USB plugs are small (especially the mini plugs) and, unlike a typical serial or parallel cable, the plug is not attached by screws or thumbscrews. There are no pins to bend or break, making USB devices very user friendly to install and remove. The USB plug shown in Figure 17.4 snaps into place on the USB connector.

Figure 17.4. USB Series A and Series B plugs and receptacles.


Note that a Mini-A/B socket is a dual-purpose socket that can accept either Mini-A or Mini-B plugs. The newer mini plugs and sockets have plastic portions inside the connectors that are required to be color-coded as shown in Table 17.2. Table 17.2. Color-Coding for USB Mini-Plugs and Sockets Connector

Color

Mini-A socket

White

Mini-A plug

White

Mini-B socket

Black

Mini-B plug

Black

Mini-A/B socket

Gray

Tables 17.3 and 17.4 show the pinouts for the USB connectors and cables. Most systems with USB connectors feature one or two pairs of Series A plugs on the rear of the system. Some also feature one or two pairs on the front of the system for ease of use with items that are not permanently connected. Table 17.3. USB Connector Pinouts for Series A/B Connectors Pin 1

Signal Name Vbus

Wire Color Red

Comment Bus power


2

- Data

White

Data transfer

3

+ Data

Green

Data transfer

4

Ground

Black

Cable ground

Shell

Shield

Drain wire

Table 17.4. USB Connector Pinouts for Mini-A/B Connectors Pin

Signal Name

Wire Color

Comment

1

Vbus

Red

Bus power

2

- Data

White

Data transfer

3

+ Data

Green

Data transfer

4

ID

4

Ground

Shell

Shield

A/B identification[*] Black

Cable ground Drain wire

[*] Used to identify a Mini-A from a Mini-B connector to the device. ID is connected to Ground in a Mini-

A plug and not connected (open) in a Mini-B plug.

USB conforms to Intel's Plug and Play (PnP) specification, including hot plugging, which means that devices can be plugged in dynamically without powering down or rebooting the system. Simply plug in the device, and the USB controller in the PC detects the device and automatically determines and allocates the required resources and drivers. Microsoft has developed USB drivers and included them automatically in Windows 98 and later. Windows 95B and 95C have very limited support for USB 1.1; the necessary drivers are not present in the original Windows 95 or 95A. With Windows 95B, the USB drivers are not automatically included; they are provided separately, although a late release of Windows 95​Windows 95C​includes USB support. Many USB devices will not work with any Windows 95 release, including those that have the USB support files included. Windows 98 and later have USB 1.1 support built in; however, additional drivers are required for USB 2.0 or later. In most cases, these drivers can be downloaded from Microsoft using the Windows Update feature.


USB support is also required in the BIOS for devices such as keyboards and mice. This is included in all newer systems with USB ports built in. Aftermarket PCI and PC Card boards also are available for adding USB to systems that don't include it as standard on the motherboard. USB peripherals include printers, CD-ROMs, modems, scanners, telephones, joysticks, keyboards, and pointing devices such as mice and trackballs. A free utility called USBready is available from http://www.usb.org; it examines your PC's hardware and software and informs you of your PC's USB capabilities. Most PCs built in 1995 or earlier don't support USB. During 1996 most PC motherboards began supporting USB, and if your system dates from 1997 to 1998 or later, USB support is almost a certainty. One interesting feature of USB is that, with certain limitations, attached devices can be powered by the USB bus. The PnP aspects of USB enable the system to query the attached peripherals as to their power requirements and issue a warning if available power levels are being exceeded. This is important for USB when it is used in portable systems because the battery power that is allocated to run the external peripherals might be limited. You can determine the amount of power available to each port in a USB root or generic hub and the amount of power required by a USB peripheral with the Windows Device Manager (see Figure 17.5).

Figure 17.5. The Power tab of the properties sheet for a USB generic hub lists available power and power usage by device.

Devices that use more than 100mA, such as the Webcam shown in Figure 17.5, must be connected to a root hub or a self-powered generic hub. Devices that use 100mA or less can be connected to bus-powered hubs, such as those built in to


some keyboards and monitors.


Tip If a device plugged in to a self-powered hub stops working, check the power source for the self-powered hub​it might have failed or been disconnected. In such cases, a self-powered hub becomes a buspowered hub, providing only 100mA per port instead of the 500mA per port available in self-powered mode.

To avoid running out of power when connecting USB devices, use a self-powered hub. Another of the benefits of the USB specification is the self-identifying peripheral, a feature that greatly eases installation because you don't have to set unique IDs or identifiers for each peripheral​the USB handles that automatically. Also, USB devices can be "hot" plugged or unplugged, meaning that you do not have to turn off your computer or reboot every time you want to connect or disconnect a peripheral. However, to prevent data loss with USB drives and storage devices, you need to use the Eject Hardware or Safely Remove Hardware feature in the Windows system tray. Click the device, select Stop, click OK, and wait for the system to indicate that the device has been stopped before you remove it.

Enabling USB Support Many systems shipped before Windows 98 was introduced in mid-1998 have onboard USB ports that were disabled at the factory. In some cases, especially with Baby-AT motherboards, there is no way to tell from the outside which systems have USB support built in. This is because many of these same systems were not shipped with the USB header cables necessary to bring the USB root hub connectors from the motherboard to the rear of the system. If USB support is disabled in the system BIOS, restart your system and locate the BIOS setup screen that refers to the USB ports. Enable the USB feature. If you see a separate entry for USB IRQ, enable this as well. After you restart the computer with a USB-aware operating system, your "new" USB root hub will be detected and the drivers will be installed if you are using Windows 98 or newer; you might need to manually install drivers with late releases of Windows 95. If your system has USB connectors present, you also will be able to use the "new" USB ports as soon as the system is rebooted after the USB drivers are installed. However, if your motherboard vendor didn't provide USB connectors, you must buy USB header cables. Before you order them, check the configuration of your motherboard's USB header pins. The standard is two rows of five pins each. Companies such as Belkin, CyberGuys, and Cables To Go sell header cables that are compatible with standard USB header pins if your motherboard supplier doesn't have the header cable in stock. Figure 17.6 shows a typical USB header cable set.


Figure 17.6. A typical USB header cable set; plug it into your motherboard to connect devices to the additional onboard USB ports (if present).

One of the biggest advantages of an interface such as USB is that it requires only a single interrupt (IRQ) from the PC. Therefore, you can connect up to 127 devices and they will not use separate interrupts, as they might if each were connected over a separate interface. This is a major benefit of the USB interface. The USB interface can also be adapted to older peripherals. See the section "USB Adapters," later in this chapter, for details.

USB 2.0/Hi-Speed USB USB 2.0 (also called Hi-Speed USB) is a backward-compatible extension of the USB 1.1 specification that uses the same cables, connectors, and software interfaces, but it runs 40 times faster than the original 1.0 and 1.1 versions. The higher speed enables higher-performance peripherals, such as higher-resolution Web/videoconferencing cameras, scanners, and faster printers, to be connected externally with the same easy plug-and-play installation of current USB peripherals. From the end-user point of view, USB 2.0 works exactly the same as 1.1​only faster and with more interesting, higher-performance devices available. All existing USB 1.1 devices work in a USB 2.0 bus because USB 2.0 supports all the slower-speed connections. USB data rates are shown in Table 17.5. Table 17.5. USB Data Rates Interface

Megabits per Second

Megabytes per Second

USB 1.1 low speed

1.5Mbps

0.1875MBps

USB 1.1 full speed

12Mbps

1.5MBps


USB 2.0 high speed

480Mbps

60MBps

If your motherboard or system features USB 2.0​compatible (Hi-Speed USB) ports, you might need to enable USB 2.0/Hi-Speed USB support in the system BIOS and install an appropriate driver. Otherwise, USB 2.0/Hi-Speed USB ports will be used as USB 1.1 ports. See "USB Configuration Submenu," p. 406, for details.

The support of higher-speed USB 2.0 peripherals requires using a USB 2.0 hub. You can still use older USB 1.1 hubs on a 2.0 bus, but any peripherals or additional hubs connected downstream from a 1.1 hub will operate at the slower 1.5MBps USB 1.1 maximum speed. Devices connected to USB 2.0 hubs will operate at the maximum speed of the device, up to the full USB 2.0 speed of 60MBps. The higher transmission speeds through a 2.0 hub are negotiated on a device-by-device basis, and if the higher speed is not supported by a peripheral, the link operates at a lower USB 1.1 speed. As such, a USB 2.0 hub accepts high-speed transactions at the faster USB 2.0 frame rate and must deliver them to high-speed USB 2.0 peripherals as well as USB 1.1 peripherals. This data rate matching responsibility requires increased complexity and buffering of the incoming high-speed data. When communicating with an attached USB 2.0 peripheral, the 2.0 hub simply repeats the high-speed signals; however, when communicating with USB 1.1 peripherals, a USB 2.0 hub buffers and manages the transition from the high speed of the USB 2.0 host controller (in the PC) to the lower speed of a USB 1.1 device. This feature of USB 2.0 hubs means that USB 1.1 devices can operate along with USB 2.0 devices and not consume any additional bandwidth. Some manufacturers of add-on USB 2.0 cards are equipping the cards with both external and internal USB 2.0 ports. How can you tell which devices are designed to support USB 1.1 and which support the emerging USB 2.0 standard? The USB Implementer's Forum (USB-IF), which owns and controls the USB standard, introduced new logos in late 2000 for products that have passed its certification tests. The logos are shown in Figure 17.7.

Figure 17.7. The USB-IF USB 1.1​compliant logo (left) compared to the USB-IF USB 2.0​compliant logo (right).


As you can see from Figure 17.7, USB 1.1 is also known simply as USB, and USB 2.0 is also known as Hi-Speed USB. Also note the icons shown earlier, where the addition of the plus symbol to the standard USB trident is used to identify ports that support USB 2.0.

USB On-The-Go In December 2001, the USB-IF released a supplement to the USB 2.0 standard called USB On-The-Go. It was designed to address the one major shortcoming of USB: the fact that a PC was required to transfer data between two devices. In other words, you couldn't connect two cameras together and transfer pictures between them without a PC orchestrating the transfer. With USB On-The-Go, however, devices that conform to the specification still work normally when they are connected to a PC, but they also have additional capabilities when connected to other devices supporting the standard. Although this capability can also work with PC peripherals, it was mainly added to address issues using USB devices in the consumer electronics area, where a PC might not be available. Using this standard, devices such as digital video recorders can connect to other recorders to transfer recorded movies or shows, items such as personal organizers can transfer data to other organizers, and so on. The addition of the On-The-Go supplement to USB 2.0 greatly enhances the use and capabilities of USB both in the PC and consumer electronics markets. The first products using USB On-The-Go technologies are expected sometime in 2003; ATI's Imageon display coprocessor for PDAs and smart phones and Qualcomm's next-generation wireless chipsets are among early adopters of this technology.

USB Adapters If you still have a variety of older peripherals and yet you want to take advantage of the USB connector on your motherboard, several signal converters or adapters are available. Companies such as Belkin and others currently have adapters in the following types: USB-to-parallel (printer)


USB-to-serial USB-to-SCSI USB-to-Ethernet USB-to-keyboard/mouse USB-to-TV/video These adapters usually look just like a cable, with a USB connector at one end (which you plug into your USB port) and various other interface connectors at the other end. In some cases, you attach standard USB and device cables to a standalone adapter, such as with the USB-to-Ethernet adapter shown in Figure 17.8.

Figure 17.8. A typical USB-to-Ethernet adapter from D-Link.

There is more to these devices than just a cable: If the unit is a one-piece device, active electronics are hidden in a module along the cable or are sometimes packed into one of the cable ends. The electronics are powered by the USB bus and convert the signals to the appropriate other interface. If you cannot install a native adapter card for your device, converting it to use the USB port through an adapter is much better than not using the device at all. For example, a USB-toEthernet adapter such as the one shown in Figure 17.8 can enable a computer without expansion slots to connect to a broadband Internet device such as a cable or DSL modem.


However, some drawbacks do exist to these adapters. One is cost: They typically cost $30​$60 or more. It can be tough to spend $40 on a USB-to-parallel adapter to drive a printer that barely cost twice that amount. In addition, other limitations might apply. For example, USB-to-parallel converters work only with printers and not other parallel-connected devices, such as scanners, cameras, external drives, and so on. Before purchasing one of these adapters, ensure that it will work with the device or devices you have in mind. If you need to use more than one nonUSB device with your system, consider special USB hubs that also contain various combinations of other port types; these are sometimes referred to as multifunction USB hubs, USB port replicators, or USB docking stations. These special hubs are more expensive than USB-only hubs but are less expensive than the combined cost of a comparable USB hub and two or more USB adapters. Another type of adapter available is a direct-connect cable, which enables you to connect two USB-equipped PCs directly together using USB as a network. These are popular for people playing two-player games, with each player on his own system. Another use is for transferring files because this connection usually works as well or better than the direct parallel connection that otherwise might be used. Also available are USB switchboxes that enable one peripheral to be shared among two or more USB buses. Note that both the direct connect cables and USB switchboxes are technically not allowed as a part of the USB specification, although they do exist.

Legacy-Free PCs USB adapters might find more use in the future as more and more legacy-free PCs are shipped. A legacy-free PC is one that lacks any components that were connected to or a part of the traditional ISA bus. This especially includes the otherwise standard Super I/O chip, which integrated serial, parallel, keyboard, mouse, floppy, and other connections. A legacy-free motherboard therefore does not have the standard serial, parallel, and keyboard/mouse connectors on the back and lacks an integrated floppy controller. The devices previously connected to those ports must instead be connected via USB, ATA/IDE, PCI, and other interfaces. Legacy-free systems are primarily found on the low-end, consumer-oriented systems. For those systems, USB will likely be one of the only external connections provided. To compensate for the loss of the other external interfaces, most legacy-free motherboards feature four or more integrated USB connectors on one or two buses.

IEEE-1394


The Institute of Electrical and Electronic Engineers Standards Board introduced IEEE-1394 (or just 1394 for short) in late 1995. The number comes from the fact that this happened to be the 1,394th standard they published. It is the result of the large data-moving demands of today's audio and video multimedia devices. The key advantage of 1394 is that it's extremely fast; the popular 1394a standard supports data transfer rates up to an incredible 400Mbps.

1394 Standards The most common version of the 1394 standard is actually referred to as 1394a, or sometimes as 1394a-2000 for the year it was adopted. The 1394a standard was introduced to solve interoperability and compatibility issues in the original 1394 standard; it uses the same connectors and supports the same speeds as the original 1394 standard. The first products to use the 1394b standard were introduced in early 2003. Initially, 1394b supports 800Mbps transfer rates, but future versions of the standard might reach speeds of up to 3,200Mbps. 1394b will be capable of reaching much higher speeds than the current 1394/1394a standard because it will also support network technologies such as glass and plastic fiber-optic cable and Category 5 UTP cable, increased distances when Category 5 cabling is used between devices, and improvements in signaling. 1394b will also be fully backward-compatible with 1394a devices. 1394 is also known by two other common names: i.Link and FireWire. i.Link is an IEEE-1394 designation initiated by Sony in an effort to put a more user-friendly name on IEEE-1394 technology. Most companies that produce 1394 products for PCs have endorsed this new name initiative. Originally, the term FireWire was an Apple-specific trademark that Apple licensed to vendors on a fee basis. However, in May 2002, Apple and the 1394 Trade Association announced an agreement to allow the trade association to provide no-fee licenses for the FireWire trademark on 1394-compliant products that pass the trade association's tests. Apple continues to use FireWire as its marketing term for IEEE-1394 devices. FireWire 400 refers to Apple's IEEE1394a-compliant products, whereas FireWire 800 refers to Apple's IEEE-1394bcompliant products.

1394a Technical Details The IEEE-1394a standard currently exists with three signaling rates​100Mbps, 200Mbps, and 400Mbps (12.5MBps, 25MBps, and 50MBps). Most PC adapter cards support the 400Mbps (50MBps) rate, although device speeds can vary. A maximum of 63 devices can be connected to a single IEEE-1394 adapter card by way of daisy-chaining or branching. 1394 devices, unlike USB devices, can be


used in a daisy-chain without using a hub, although hubs are recommended for devices that will be hot-swapped. Cables for IEEE-1394/1394a devices use Nintendo GameBoy​derived connectors and consist of six conductors: Four wires transmit data, and two wires conduct power. Connection with the motherboard is made either by a dedicated IEEE-1394 interface or by a PCI adapter card. Figure 17.9 shows the 1394/1394a cable, socket, and connector.

Figure 17.9. IEEE-1394 port, 6-pin cable, and 4-pin cable.

The 1394 bus was derived from the FireWire bus originally developed by Apple and Texas Instruments, and it is also a part of a new Serial SCSI standard. 1394a uses a simple six-wire cable with two differential pairs of clock and data lines, plus two power lines; the four-wire cable end shown in Figure 17.9 is used with self-powered devices, such as DV camcorders. Just as with USB, 1394 is fully PnP, including the capability for hot-plugging (insertion and removal of components without powering down). Unlike the much more complicated parallel SCSI bus, 1394 does not require complicated termination, and devices connected to the bus can draw up to 1.5 amps of electrical power. 1394 offers equal or greater performance compared to ultra-wide SCSI, with a much less expensive and less complicated connection. 1394 is built on a daisy-chained and branched topology, and it allows up to 63 nodes, with a chain of up to 16 devices on each node. If this is not enough, the standard also calls for up to 1,023 bridged buses, which can interconnect more than 64,000 nodes! Additionally, as with SCSI, 1394 can support devices with various data rates on the same bus. Most 1394 adapters have three nodes, each of which can support 16 devices in a daisy-chain arrangement. Some 1394 adapters also support internal 1394 devices. The types of devices that can be connected to the PC via 1394 mainly include video cameras; editing equipment; and all forms of disk drives, including hard


disk, optical, floppy, CD-ROM, and DVD-ROM drives. Also, digital cameras, tape drives, high-resolution scanners, and many other high-speed peripherals that feature 1394 have interfaces built in. The 1394 bus appears in some desktop and portable computers as a replacement or supplement for other external high-speed buses, such as USB or SCSI. Chipsets and PCI adapters for the 1394 bus are available from a number of manufacturers, including some models that support both 1394 and other port types in a single slot. Microsoft has developed drivers to support 1394 in Windows 9x and later, including Windows XP. The most popular devices that conform to the IEEE-1394 standard are camcorders and VCRs with digital video capability. Sony was among the first to release such devices (under the i.Link name). In typical Sony fashion, however, its products have a unique four-wire connector that requires an adapter cord to be used with IEEE-1394 PC cards, and Sony doesn't even call it IEEE-1394 or FireWire​it created its own designation (i.Link) instead. DV products using 1394 also are available from Panasonic, Sharp, Matsushita, and others. Non-computer IEEE-1394 applications include DV conferencing devices, satellite audio and video data streams, audio synthesizers, DVD, and other highspeed disc drives. Because of the current DV emphasis for IEEE-1394 peripherals, many FireWire cards currently offered are bundled with DV capturing and editing software. With a DV camera or recording equipment, these items provide substantial video editing and dubbing capabilities on your PC. Of course, you need IEEE-1394 I/O connectivity, which is a growing, but still somewhat rare, feature on current motherboards.

IEEE-1394b Technical Details IEEE-1394b is the second generation of the 1394 standard, with the first products (high-performance external hard drives) introduced in January 2003. IEEE-1394b uses one of two new nine-pin cables and connectors to support speeds of 800Mbps​3200Mbps with copper or fiber-optic cabling. In addition to supporting faster transfer rates, 1394b has other new features, including Self-healing loops. If you improperly connect 1394b devices together to create a logical loop, the interface corrects the problem instead of failing as with 1394a. Continuous dual simplex. Of the two wire pairs used, each pair transmits data to the other device, so that speed remains constant. Support for fiber-optic and CAT5 network cable as well as standard 1394a and


1394b copper cable. Improved arbitration of signals to support faster performance and longer cable distances. Support for CAT5 cable, even though it uses pairs on pins 1 and 2 and 7 and 8 only for greater reliability. It also doesn't require crossover cables. The initial implementations of IEEE-1394b use a new nine-wire interface with two pairs of signaling wires. However, to enable a 1394b port to connect to 1394acompatible devices, there are two different versions of the 1394b port: Beta Bilingual Beta connectors support only 1394b devices, whereas bilingual connectors can support both 1394b and 1394a devices. As Figure 17.10 shows, the connectors and cables have the same pinout but are keyed differently.

Figure 17.10. Bilingual and beta 1394b connectors and cables. Many 1394b implementations use both types of connectors.


Note that bilingual sockets and cables have a narrower notch than beta sockets and cables. This prevents cables designed for 1394a devices from being connected to the beta socket. Figure 17.11 compares a beta-to-beta 1394b cable to bilingual-to-1394a cables.

Figure 17.11. A beta-to-beta cable (top) compared to bilingual​to​4-pin (middle) and bilingual​to​6-pin 1394a devices (bottom).


Comparing IEEE-1394 and USB Because of the similarity in both the form and function of USB and 1394 ports, there has been some confusion about the differences between them. Table 17.6 summarizes the differences between these technologies. Table 17.6. IEEE-1394 and USB Comparison

PC-host required

IEEE-1394b IEEE-1394a (also called i.Link (also called or FireWire 400) FireWire 800)

USB 1.1

USB 2.0

No

No

Yes

Yes/No [1]

Maximum number of 63 devices

63

127

127

HotYes swappable

Yes

Yes

Yes

Maximum cable length 4.5 meters between

4.5 meters (9pin copper); 100 meters 5 meters (glass optical

5 meters


fiber)[2]

devices

Transfer rate

400Mbps (50MBps)

Proposed future None transfer rates

Typical devices

800Mbps (100MBps)

12Mbps (1.5MBps)

480Mbps (60MBps)

1,600Mbps (400MBps); 3,200Mbps (800MBps)

None

None

DV camcorders; high-res digital cameras; HDTV; set-top boxes; All 1394a high-speed drives; high-res devices scanners; electronic musical instruments

Keyboards; mice; All USB 1.1 devices; DV joysticks; low-res digital camcorders; high-res digital cameras; low-speed cameras HDTV; set-top drives; ;modems; printers; boxes; high-speed drives; low-res scanners high-res scanners

[1] No with USB On-The-Go. [2] CAT-5 UTP supported for 100Mbps speeds (100 meters max.); step-index plastic optical fiber

supported for 100Mbps and 200Mbps speeds (50 meters max.).

Because the overall performance and physical specifications are similar, the main difference between USB and 1394 is popularity. The bottom line is that USB is by far the most popular external interface for PCs, eclipsing all others by comparison. This is primarily because Intel developed most of USB and has placed built-in USB support in all its motherboard chipsets and motherboards since 1996. Virtually no motherboard chipsets integrate 1394a or 1394b; in most cases, it has to be added as an extra-cost chip to the motherboard. The cost of the additional 1394 circuitry (and a $0.25 royalty paid to Apple Computer per system) and the fact that all motherboards already have USB, have limited the popularity of 1394 (FireWire) in the PC marketplace. Even with the overwhelming popularity of USB, a market for 1394 still exists. Perhaps the main reason 1394 will survive in conjunction with the USB 2.0 interface is that USB is normally PC-centric, whereas 1394 is not. In other words, USB and Hi-Speed USB require a PC as the host, whereas 1394 can connect two devices directly without a PC between them. As such, 1394 can be used to directly connect a DV camcorder to a DV-VCR for dubbing tapes or editing. Even this has changed, however, as a supplement called USB On-The-Go was added to the USB 2.0 specification in December 2001. USB On-The-Go enables the same device-to-device connections as was capable in 1394 (FireWire) and essentially nullifies the one advantage 1394 had over USB. Because of the popularity and capabilities of USB, I recommend seeking out only USB peripherals over their 1394 (FireWire) counterparts where possible. Many people like to bring any comparison of USB and 1394 down to speed, but that is a constantly changing parameter. 1394a offers a data transfer rate more


than 33 times faster than that of USB 1.1, but is only about 83% as fast as USB 2.0. However, 1394b is about 66% faster than USB 2.0. Because both USB 2.0 and 1394a (FireWire) offer relatively close to the same overall capabilities and performance, you make your choice based on which devices you intend to connect. If the digital video camera you want to connect has only a 1394 (FireWire/i.Link) connection, you will need to add a 1394 FireWire card to your system, if such a connection isn't already present on your motherboard. Most general-purpose PC storage, I/O, and other devices are USB, whereas only video devices usually have 1394 connections. However, many devices now offer both USB 1.1/2.0 and 1394a interfaces to enable use with the widest range of computers.


Standard Serial and Parallel Ports Traditionally, the most basic communications ports in any PC system have been the serial and parallel ports, and these ports continue to be important. Serial ports (also known as communication or COM ports) originally were used for devices that had to communicate bidirectionally with the system. Such devices include modems, mice, scanners, digitizers, and any other devices that "talk to" and receive information from the PC. Newer parallel port standards now allow the parallel port to perform high-speed bidirectional communications. Several companies manufacture communications programs that perform highspeed transfers between PC systems using serial or parallel ports. Versions of these file transfer programs have been included with DOS 6.0 and higher (Interlink) and with Windows 95 and newer versions (Direct Cable Connection [DCC]). Although USB continues to replace the parallel port in new computers, you can still purchase products that make nontraditional use of the parallel port. For example, external drive vendors continue to make CD-ROM, CD-RW, and DVD drives; floppy drives; print servers; and tape backups that connect to the parallel port, although many of these devices also include USB connections as well.


Serial Ports The asynchronous serial interface was designed as a system-to-system communications port. Asynchronous means that no synchronization or clocking signal is present, so characters can be sent with any arbitrary time spacing. Each character that is sent over a serial connection is framed by a standard startand-stop signal. A single 0 bit, called the start bit, precedes each character to tell the receiving system that the next eight bits constitute a byte of data. One or two stop bits follow the character to signal that the character has been sent. At the receiving end of the communication, characters are recognized by the start-andstop signals instead of by the timing of their arrival. The asynchronous interface is character oriented and has an approximate 20% overhead for the extra information that is needed to identify each character. Serial refers to data that is sent over a single wire, with each bit lining up in a series as the bits are sent. This type of communication is used over the phone system because it provides one wire for data in each direction.

Typical Locations for Serial Ports Typical systems include one or two serial ports, with connectors typically located at the rear of the system. Some recent consumer-oriented computers label a front-mounted serial port the "digital camera port." This name comes from the use of serial ports for data transfer from low-end digital cameras. On recent systems, these built-in serial ports are controlled by a highly integrated South Bridge chip in the latest motherboard designs. If you need more serial ports than your system has as standard, you can purchase single-port or multiport serial port cards or so-called multi-I/O cards that feature one or two serial ports and one or two parallel ports. Older systems based on the ISA or VL-Bus standard often have the serial ports attached to a multifunction card that also has IDE hard disk and floppy disk interfaces. Note that card-based modems also incorporate a built-in serial port on the card as part of the modem circuitry. Figure 17.12 shows the standard 9-pin connector used with most modern external serial ports. Figure 17.13 shows the original standard 25-pin version.

Figure 17.12. AT-style 9-pin serial-port connector specifications.


Figure 17.13. Standard 25-pin serial-port connector specifications. NC stands for no connect, which indicates a dead pin.


Serial ports can connect to a variety of devices, such as modems, plotters, printers, PDA docking devices, other computers, bar code readers, scales, and device control circuits. The official specification recommends a maximum cable length of 50 feet. The limiting factor is the total load capacitance of cable and input circuits of the interface. The maximum capacitance is specified as 2500pF (picofarads). Special low-capacitance cables can effectively increase the maximum cable length greatly, to as much as 500 feet or more. Also available are line drivers (amplifier/repeaters) that can extend cable length even further. Tables 17.7, 17.8, and 17.9 show the pinouts of the 9-pin (AT-style), 25-pin, and 9-pin​to​25-pin serial connectors, respectively. Table 17.7. 9-Pin (AT) Serial Port Connector Pin

Signal

Description

I/O


1

CD

Carrier detect

In

2

RD

Receive data

In

3

TD

Transmit data

Out

4

DTR

Data terminal ready

Out

5

SG

Signal ground

6

DSR

Data set ready

In

7

RTS

Request to send

Out

8

CTS

Clear to send

In

9

RI

Ring indicator

In

Table 17.8. 25-Pin (PC, XT, and PS/2) Serial Port Connector Pin

Signal

1

Description

I/O

Chassis ground

2

TD

Transmit data

Out

3

RD

Receive data

In

4

RTS

Request to send

Out

5

CTS

Clear to send

In

6

DSR

Data set ready

In

7

SG

Signal ground

8

CD

Carrier detect

In

9

+Transmit current loop return

Out

11

-Transmit current loop data

Out

18

+Receive current loop data

In

Data terminal ready

Out

20

DTR


22

RI

25

Ring indicator

In

-Receive current loop return

In

Table 17.9. 9-Pin​to​25-Pin Serial Cable Adapter Connections 9-Pin

25-Pin

Signal

Description

1

8

CD

Carrier detect

2

3

RD

Receive data

3

2

TD

Transmit data

4

20

DTR

Data terminal ready

5

7

SG

Signal ground

6

6

DSR

Data set ready

7

4

RTS

Request to send

8

5

CTS

Clear to send

9

22

RI

Ring indicator


Note Macintosh systems use a similar serial interface, defined as RS-422. Most external modems in use today can interface with either RS-232 or RS-422, but it is safest to make sure that the external modem you get for your PC is designed for a PC, not a Macintosh.

UARTs The heart of any serial port is the Universal Asynchronous Receiver/Transmitter (UART) chip. This chip completely controls the process of breaking the native parallel data within the PC into serial format and later converting serial data back into the parallel format. Several types of UART chips have been available on the market. The original PC and XT used the 8250 UART, which was used for many years in low-priced serial cards. Starting with the first 16-bit systems, the 16450 UART typically was used. The only difference between these chips is their suitability for high-speed communications. The 16450 is better suited for high-speed communications than is the 8250; otherwise, both chips appear identical to most software. The 16550 UART was the first serial chip used in the IBM PS/2 line. Other 386 and higher systems rapidly adopted it. The 16550 functioned as the earlier 16450 and 8250 chips, but it also included a 16-byte buffer that aided in faster communications. This is sometimes referred to as a FIFO (first in first out) buffer. Unfortunately, the early 16550 chips had a few bugs, particularly in the buffer area. These bugs were corrected with the release of the 16550A. The most current version of the chip is the 16550D, which was released in 1995 and is produced by National Semiconductor. Even though virtually all Pentium-class and newer systems have 16550equivalent UART functionality in their serial ports, any search for a socketed 16550 chip on most of these systems would be done in vain. Instead, the functionality of the 16550, parallel port, and other ports is included as part of the Super I/O chip or, on the newest systems, the South Bridge chip.


Tip The high-speed buffered 16550A (or newer) UART chip is pin for pin compatible with the 16450 UART. If your 16450 UART is socketed, a cheap and easy way to improve serial performance is to install a 16550 UART chip in the socket.

Because the 16550 is a faster, more reliable chip than its predecessors, it is best to ensure that your serial ports have either that chip or an equivalent. If you are in doubt about which type of UART you have in your system, you can use the Microsoft MSD program (provided with MS-DOS 6.x and Windows 9x/Me/2000) to determine the type of UART you have. Note that MSD often reports a 16450 UART as an 8250.


Note Another way to tell whether you have a 16650 UART in Windows is to click the Start menu and then select Settings, Control Panel. Next, double-click Modems, and then click the Diagnostics tab. The Diagnostics tab shows a list of all COM ports in the system, even if they don't have a modem attached to them. Select the port you want to check in the list and click More Info. Windows communicates with the port to determine the UART type, and that information is listed in the Port Information portion of the More Info box. If a modem is attached, additional information about the modem is displayed.

The original designer of these UARTs is National Semiconductor (NS). So many other manufacturers are producing clones of these UARTs that you probably don't have an actual NS brand part in your system. Even so, the part you have is compatible with one of the NS parts, hopefully the 16550. In other words, check to see that whatever UART chip you have does indeed feature the 16-byte FIFO buffer, as found in the NS 16550 part. See "Super I/O Chips," p. 304.

8250 IBM used this original chip in the PC serial port card. This chip had no transmit/receive buffer, so it was very slow. The chip also had several bugs, none of which were serious. The PC and XT ROM BIOSs were written to anticipate at least one of the bugs. The 8250B replaced this chip.

8250A Do not use the second version of the 8250 in any system. This upgraded chip fixes several bugs in the 8250, including one in the interrupt enable register, but because the PC and XT ROM BIOSs expect the bug, this chip does not work properly with those systems. The 8250A works in an AT system that does not expect the bug, but it does not work adequately at 9600bps.

8250B The last version of the 8250 fixes bugs from the previous two versions. The interrupt enable bug in the original 8250, which is expected by the PC and XT ROM BIOS software, has been put back into this chip, making the 8250B the most


desirable chip for any non-AT serial port application. The 8250B chip might work in an AT under DOS, but it does not run properly at 9600bps because, like all 8250s, it has no transmit/receive buffer.

16450 IBM selected the higher-speed version of the 8250 for the AT. The higher performance comes mainly from a 1-byte transmit/receive buffer contained within the chip. Because this chip has fixed the aforementioned interrupt enable bug, the 16450 does not operate properly in many PC or XT systems because they expect this bug to be present. OS/2 requires this chip as a minimum; otherwise, the serial ports do not function properly. It also adds a scratch-pad register as the highest register. The 16450 is used primarily in AT systems because of its increase in throughput over the 8250B.

16550 Series This chip is pin compatible with the 16450 but is much faster due to a built-in 16character transmit/receive FIFO buffer. It also allows multiple DMA channel access. The original version of this chip did not allow the buffer to work, but all 16550A or later revisions have the bug fixed. The last version produced by National Semiconductor was called the 16550D. Use a version of this UART in your serial port if you do any communications at 9600bps or higher. If your communications program uses the FIFO​and all of them do today​it can greatly increase communications speed and eliminate lost characters and data at the higher speeds. Virtually all Super I/O chips contain the equivalent of dual 16550A or later chips. Most 16550 UARTs have a maximum communications speed of 115Kbps.

16650, 16750, and 16850 Several companies have produced versions of the 16550 with larger buffers: The 16550 has a 32-byte buffer. The 16750 has a 64-byte buffer. The 16850 has a 128-byte buffer. The 16950 has a 128-byte buffer and the option to run at normal or 4x


normal baud rates. These chips are not from National Semiconductor, and the designations only imply that they are compatible with the 16550 but have a larger buffer. These largerbuffered versions allow speeds of 230Kbps (16650), 460Kbps (16750), and 920Kbps (16850 and 16950) and are recommended when running a high-speed external communications link, such as an ISDN terminal adapter or external 56Kbps modem. These are discussed more in the following section.

High-Speed Serial Port Cards If you are using external RS-232 devices designed to run at speeds higher than 115Kbps (the maximum speed of the 16550 series UARTs and equivalents), you can't achieve maximum performance unless you replace your existing serial ports with add-on cards using one of the 16650, 16750, 16850, or 16950 UARTs discussed earlier. Most cards allow raw port speed settings of 230Kbps, 460Kbps, or even higher, which is valuable when connecting a PC to a high-speed external component that is connected to a serial port, such as an ISDN terminal adapter. You can't really get the full-speed benefit of an external ISDN modem (terminal adapter) unless your serial port can go at least 230Kbps. Lava Computer Mfg. and SIIG are two of the companies that offer a complete line of high-speed serial and parallel port cards (see the Vendor List on the accompanying DVD).

Onboard Serial Ports Starting with late-model 486-based systems in the mid-1990s, a component on the motherboard called a Super I/O chip began to replace separate UART chips. This usually has two serial port UARTs as well as a multimode parallel port, floppy controller, keyboard controller, and sometimes the CMOS memory​all built into a single tiny chip. Still, this chip acts as if all these separate devices were installed: That is, from a software point of view, both the operating system and applications still act as if separate UART chips were installed on serial port adapter cards. The most recent systems integrate the functions of a Super I/O chip into the South Bridge chip. As with the Super I/O chip, South Bridge chips with integrated I/O are transparent to software. Super I/O and South Bridge chips are discussed throughout this chapter and in Chapter 4, "Motherboards and Buses."

Serial Port Configuration Each time a character is received by a serial port, it has to get the attention of


the computer by raising an interrupt request line (IRQ). Eight-bit ISA bus systems have eight of these lines, and systems with a 16-bit ISA bus have 16 lines. The 8259 interrupt controller chip or equivalent typically handles these requests for attention. In a standard configuration, COM1 uses IRQ4, and COM2 uses IRQ3. Even on the latest systems, the default COM port assignments remain the same for compatibility with older software and hardware. When a serial port is installed in a system, it must be configured to use specific I/O addresses (called ports) and interrupts (called IRQs). The best plan is to follow the existing standards for how these devices are to be set up. For configuring serial ports, use the addresses and interrupts indicated in Table 17.10. Table 17.10. Standard Serial I/O Port Addresses and Interrupts COMx

I/O Ports

IRQ

COM1

3F8​3FFh

IRQ4

COM2

2F8​2FFh

IRQ3

COM3

3E8​3EFh

IRQ4[1]

COM4

2E8​2EFh[1]

IRQ3[1]

2. This I/O address can conflict with registers in some video cards. In such cases, you cannot use COM4 unless COM4 or the video card can be configured to use a different I/O port address range.

[1] Note that although many serial ports can be set up to share IRQ3 and IRQ4 with COM1 and COM2, it

is not recommended. The best recommendation is setting COM3 to IRQ10 and COM4 to IRQ11 (if available). If ports above COM3 are required, it is recommended that you purchase a special multiport serial board, preferably a PCI-based board that supports IRQ sharing without conflicts.

Be sure that, if you are adding more than the standard COM1 and COM2 serial ports, they use unique and nonconflicting interrupts. If you purchase a serial port adapter card and intend to use it to supply ports beyond the standard COM1 and COM2, be sure it can use interrupts other than IRQ3 and IRQ4; PCI-based serial port boards take advantage of IRQ sharing features to allow COM3 and above to use a single IRQ without conflicts. Note that BIOS manufacturers never built support for COM3 and COM4 into the BIOS. Therefore, DOS can't work with serial ports above COM2 because DOS gets its I/O information from the BIOS. The BIOS finds out what is installed in your system, and where it is installed, during the power-on self test (POST). The POST checks only for the first two installed ports. This is not a problem under Windows because Windows 95 and later have built-in support for up to 128 ports.


With support for up to 128 serial ports in Windows, using multiport boards in the system is much easier. Multiport boards give your system the capability to collect or share data with multiple devices while using only one slot and one interrupt.


Caution Sharing interrupts between COM ports​o r any devices​can function properly sometimes and not others. It is recommended that you never share interrupts between multiple ISA-based serial ports, such as the COM ports built into your motherboard or found on an ISA modem. Trying to track down drivers, patches, and updates to allow this to work successfully​if it's even possible in your system​can cause you hours of frustration.

See "Resolving Resource Conflicts," p. 348.

Testing Serial Ports You can perform several tests on serial and parallel ports. The two most common types of tests are those that involve software only and those that involve both hardware and software. The software-only tests are done with diagnostic programs, such as Microsoft's MSD or the Modem diagnostics built into Windows, whereas the hardware and software tests involve using a wrap plug to perform loopback testing. See "Testing Parallel Ports," p. 978.

See "Advanced Diagnostics Using Loopback Testing," p. 969.

Microsoft Diagnostics Microsoft Diagnostics (MSD) is a diagnostic program supplied with MS-DOS 6.x, Windows 3.x, and Windows 9x/Me/2000. In Windows 95 this program can be found on the CD-ROM in the \other\msd directory, and in Windows 98/Me/2000, you can find it on the CD-ROM in the \tools\oldmsdos directory. MSD is not automatically installed when you install the operating system. To use it, you must run it from the CD-ROM directly or copy the program from the CD-ROM to your


hard disk. For the most accurate results, many diagnostics programs, such as MSD, are best run in a DOS-only environment. Because of this, you need to restart the machine in DOS mode before using them. If you use Windows 2000 or XP, you can format a floppy disk with MS-DOS startup files and use it to start your computer. Then, to use MSD, switch to the directory in which it is located. This is not necessary, of course, if the directory that contains the program is in your search path​which is often the case with the DOS 6.x or Windows-provided versions of MSD. Then, simply type MSD at the DOS prompt and press Enter. Soon you see the MSD screen. Select the Serial Ports option. Notice that you are given information about which type of serial chip you have in your system, as well as information about which ports are available. If any of the ports are in use (with a mouse, for example), that information is provided as well. MSD is helpful in at least determining whether your serial ports are responding. If MSD can't determine the existence of a port, it does not provide the report that indicates that the port exists. This sort of "look-and-see" test is the first action I usually take to determine why a port is not responding.

Troubleshooting Ports in Windows Windows 9x/Me can tell you whether your ports are functioning. First, you must verify that the required communications files are present to support the serial ports in your system: 1. Verify the file sizes and dates of both COMM.DRV (16-bit serial driver) and SERIAL.VXD (32-bit serial driver) in the SYSTEM directory, compared to the original versions of these files from the Windows 9x/Me CDROM. Confirm that the following lines are present in SYSTEM.INI: [boot] comm.drv=comm.drv [386enh] device=*vcd The SERIAL.VXD driver is not loaded in SYSTEM.INI; instead, it is loaded through the Registry. Windows 2000 and XP use the SERIAL.SYS and SERENUM.SYS drivers for handling


RS-232 devices. You can compare the file sizes and dates for these files to those on the Windows 2000 CD-ROM. If both drivers are present and accounted for, you can determine whether a particular serial port's I/O address and IRQ settings are properly defined by following these steps for Windows 9x/Me/2000: 1. Right-click the My Computer icon on the desktop and select Properties, or double-click the System icon in the Control Panel. Click the Device Manager tab, click Ports, and then select a specific port (such as COM1). Click the Properties button, and then click the Resources tab to display the current resource settings (IRQ, I/O) for that port. Check the Conflicting Devices List to see whether the port is using resources that conflict with other devices. If the port is in conflict with other devices, click the Change Setting button, and then select a configuration that does not cause resource conflicts. You might need to experiment with these settings until you find the correct one. If the resource settings can't be changed, most likely they must be changed via the BIOS Setup. Shut down and restart the system, enter the BIOS Setup, and change the port configurations there. See "Resolving Resource Conflicts," p. 348.

A common problem with non​Plug and Play modems can occur when people try to use a modem on COM3 with a serial mouse or other device on COM1. Typically, COM1 and COM3 ports use the same IRQ, meaning that they can't be used simultaneously. The COM2 and COM4 ports have the same problem sharing IRQs. If possible, change the COM3 or COM4 port to an IRQ setting that is not in conflict with COM1 or COM2. Note also that some video adapters have an automatic address conflict with COM4.

Advanced Diagnostics Using Loopback Testing One of the most useful types of diagnostic test is the loopback test, which can be used to ensure the correct function of the serial port and any attached cables. Loopback tests are basically internal (digital) or external (analog). You can run


internal tests by simply unplugging any cables from the port and executing the test via a diagnostics program. The external loopback test is more effective. This test requires that a special loopback connector or wrap plug be attached to the port in question. When the test is run, the port is used to send data out to the loopback plug, which simply routes the data back into the port's receive pins so the port is transmitting and receiving at the same time. A loopback or wrap plug is nothing more than a cable that is doubled back on itself. Most diagnostics programs that run this type of test include the loopback plug, and if not, these types of plugs easily can be purchased or even built. Following is a list of the wiring necessary to construct your own serial port loopback or wrap plugs: Standard IBM-type 25-pin serial (female DB25S) loopback connector (wrap plug). Connect the following pins: 1 to 7 2 to 3 4 to 5 to 8 6 to 11 to 20 to 22 15 to 17 to 23 18 to 25 Norton Utilities (Symantec) 25-pin serial (female DB25S) loopback connector (wrap plug). Connect the following pins: 2 to 3 4 to 5 6 to 8 to 20 to 22 Standard IBM-type 9-pin serial (female DB9S) loopback connector (wrap plug). Connect the following pins: 1 to 7 to 8 2 to 3


4 to 6 to 9 Norton Utilities (Symantec) 9-pin serial (female DB9S) loopback connector (wrap plug). Connect the following pins: 2 to 3 7 to 8 1 to 4 to 6 to 9 See "Serial Ports," p. 961.

To make these loopback plugs, you need a connector shell with the required pins installed. Then, wire wrap or solder the wires, interconnecting the appropriate pins inside the connector shell as specified in the preceding list. In most cases, purchasing a set of loopback connectors that are premade is less expensive than making them yourself. Most companies that sell diagnostics software can also sell you a set of loopback plugs. Some hardware diagnostic programs include loopback plugs with the software. One advantage of using loopback connectors is that you can plug them into the ends of a cable that is included in the test. This can verify that both the cable and the port are working properly. If you need to test serial ports further, see Chapter 23, "PC Diagnostics, Testing, and Maintenance," which describes third-party testing software.


Parallel Ports Parallel ports are normally used for connecting printers to a PC. Even though that was their sole original intention, parallel ports have become much more useful over the years as a more general-purpose, relatively high-speed interface between devices (when compared to serial ports). Today, USB 1.1​compliant ports are almost as fast as parallel ports, and USB 2.0, IEEE-1394a, and IEEE-1394b ports are significantly faster. Originally, parallel ports were one way only; however, modern parallel ports can send and receive data when properly configured. Parallel ports are so named because they have eight lines for sending all the bits that comprise 1 byte of data simultaneously across eight wires. This interface is fast and traditionally has been used for printers. However, programs that transfer data between systems always have used the parallel port as an option for transmitting data because it can do so 4 bits at a time, rather than 1 bit at a time as with a serial interface. The following section looks at how these programs transfer data between parallel ports. The only problem with parallel ports is that their cables can't be extended for any great length without amplifying the signal; otherwise, errors occur in the data. Table 17.11 shows the pinout for a standard PC parallel port. Table 17.11. 25-Pin PC-Compatible Parallel Port Connector Pin

Description

I/O

Pin

Description

I/O

1

-Strobe

Out

14

-Auto Feed

Out

2

+Data Bit 0

Out

15

-Error

In

3

+Data Bit 1

Out

16

-Initialize Printer

Out

4

+Data Bit 2

Out

17

-Select Input

Out

5

+Data Bit 3

Out

18

-Data Bit 0 Return (GND)

In

6

+Data Bit 4

Out

19

-Data Bit 1 Return (GND)

In

7

+Data Bit 5

Out

20

-Data Bit 2 Return (GND)

In

8

+Data Bit 6

Out

21

-Data Bit 3 Return (GND)

In

9

+Data Bit 7

Out

22

-Data Bit 4 Return (GND)

In


10

-Acknowledge

In

23

-Data Bit 5 Return (GND)

In

11

+Busy

In

24

-Data Bit 6 Return (GND)

In

12

+Paper End

In

25

-Data Bit 7 Return (GND)

In

13

+Select

In

IEEE-1284 Parallel Port Standard The IEEE-1284 standard, called Standard Signaling Method for a Bidirectional Parallel Peripheral Interface for Personal Computers, was approved for final release in March 1994. This standard defines the physical characteristics of the parallel port, including data-transfer modes and physical and electrical specifications. IEEE-1284 defines the electrical signaling behavior external to the PC for a multimodal parallel port that can support 4-bit modes of operation. Not all modes are required by the 1284 specification, and the standard makes some provision for additional modes. The IEEE-1284 specification is targeted at standardizing the behavior between a PC and an attached device, specifically attached printers. However, the specification is also of interest to vendors of parallel port peripherals (removablemedia drives, scanners, and so on). IEEE-1284 pertains only to hardware and line control and does not define how software is to talk to the port. An offshoot of the original 1284 standard has been created to define the software interface. The IEEE-1284.3 committee was formed to develop a standard for software that is used with IEEE-1284-compliant hardware. This standard, designed to address the disparity among providers of parallel port chips, contains a specification for supporting EPP (the Enhanced Parallel Port) mode via the PC's system BIOS. IEEE-1284 enables much higher throughput in a connection between a computer and a printer or two computers. The result is that the printer cable is no longer the standard printer cable. The IEEE-1284 printer cable uses twisted-pair technology, which results in a much more reliable and error-free connection. The IEEE-1284 standard also defines the parallel port connectors, including the two preexisting types (called Type A and Type B), as well as an additional highdensity Type C connector. Type A refers to the standard DB25 connector used on most PC systems for parallel port connections, whereas Type B refers to the standard 36-pin Centronics-style connector found on most printers. Type C is a high-density 36-pin connector that can be found on some of the newer printers on the market, such as those from HP. The three connectors are shown in Figure


17.14.

Figure 17.14. The three different types of IEEE-1284 parallel port connectors.

The IEEE-1284 parallel port standard defines five port-operating modes, emphasizing the higher-speed EPP and ECP modes. Some of the modes are input only, whereas others are output only. These five modes combine to create four types of ports, as shown in Table 17.12. Table 17.12. Types of IEEE-1284 Ports Parallel Port Type

Input Mode

Output Mode

Comments

SPP (Standard Parallel Port)

Nibble

Compatible

4-bit input, 8-bit output

Bidirectional

Byte

Compatible

8-bit I/O

EPP (Enhanced Parallel Port)

EPP

EPP

8-bit I/O

ECP (Enhanced Capabilities Port)

ECP

ECP

8-bit I/O, uses DMA

The 1284-defined parallel port modes are shown in Table 17.13, which also shows the approximate transfer speeds. Table 17.13. IEEE-1284 Parallel Port Modes Parallel Port Mode

Direction

Transfer Rate

Nibble (4-bit)

Input only

50KBps

Byte (8-bit)

Input only

150KBps


Compatible

Output only

150KBps

EPP (Enhanced Parallel Port)

Input/Output

500KBps​2MBps

ECP (Enhanced Capabilities Port)

Input/Output

500KBps​2MBps

Each of the port types and modes is discussed in the following sections.

Standard Parallel Ports Older PCs did not have different types of parallel ports available. The only available port was the parallel port that was used to send information from the computer to a device such as a printer. The unidirectional nature of the original PC parallel port is consistent with its primary use​that is, sending data to a printer. There were times, however, when it was desirable to have a bidirectional port​for example, when it was necessary to receive feedback from a printer, which was common with PostScript printers. This could not be done easily with the original unidirectional ports. Although it was never intended to be used for input, a clever scheme was devised in which four of the signal lines can be used as a 4-bit input connection. Thus, these ports can do 8-bit (byte) output (called compatible mode) and 4-bit input (called nibble mode). This is still very common on low-end desktop systems. Systems built after 1993 are likely to have more capable parallel ports, such as bidirectional, EPP, or ECP. Standard parallel ports are capable of effective transfer rates of about 150KBps output and about 50KBps input.

Bidirectional (8-Bit) Parallel Ports With the introduction of the PS/2 series of machines in 1987, IBM introduced the bidirectional parallel port. These are commonly found in PC-compatible systems today and can be designated "bidirectional," "PS/2 type," or "extended" parallel ports. This port design opened the way for true communication between the computer and the peripheral across the parallel port. This was done by defining a few of the previously unused pins in the parallel connector, and by defining a status bit to indicate in which direction information was traveling across the channel. This allowed for a true 8-bit (called byte mode) input. These ports can do both 8-bit input and output using the standard eight data lines and are considerably faster than the 4-bit ports when used with external devices.


Bidirectional ports are capable of approximately 150KBps transfer rates on both output and input. Some newer systems use this as their "standard" mode.

Enhanced Parallel Port EPP is a newer specification, sometimes referred to as the Fast Mode parallel port. Intel, Xircom, and Zenith Data Systems developed and announced the EPP in October 1991. The first products to offer EPP were Zenith Data Systems laptops, Xircom Pocket LAN adapters, and the Intel 82360 SL I/O chip. Currently, almost all systems include a multimode parallel port that supports EPP mode; it's usually built into the Super I/O chip or South Bridge chip on the motherboard. EPP operates at almost ISA bus speed and offers a tenfold increase in the raw throughput capability over a conventional parallel port. EPP is especially designed for parallel port peripherals, such as LAN adapters, disk drives, and tape backups. EPP has been included in the IEEE-1284 Parallel Port standard. Transfer rates of up to 2MBps are possible with EPP. Since the original Intel 82360 SL I/O chip in 1992, other major chip vendors (such as National Semiconductor, SMC, Western Digital, and VLSI) have also produced I/O chipsets that offer some form of EPP capability. One problem is that the procedure for enabling EPP across the various chips differs widely from vendor to vendor, and many vendors offer more than one I/O chip. EPP version 1.7 (March 1992) identifies the first popular version of the hardware specification. With minor changes, this has since been abandoned and folded into the IEEE-1284 standard. Some technical reference materials have erroneously made reference to "EPP specification version 1.9," causing confusion about the EPP standard. Note that "EPP version 1.9" technically does not exist, and any EPP specification after the original version 1.7 is more accurately referred to as a part of the IEEE-1284 specification. Unfortunately, this has resulted in two somewhat incompatible standards for EPP parallel ports: the original EPP Standards Committee version 1.7 standard and the IEEE-1284 Committee standard, usually called EPP version 1.9. The two standards are similar enough that new peripherals can be designed to support both, but older EPP 1.7 peripherals might not operate with EPP 1284 (EPP 1.9) ports. For this reason, many multimode ports allow configuration in either EPP 1.7 or 1.9 mode, normally selected via the BIOS Setup. EPP ports are now supported in virtually all Super I/O chips that are used on modern motherboards and in South Bridge chips that integrate I/O functions. Because the EPP port is defined in the IEEE-1284 standard, it also has gained software and driver support, including support in Windows NT.


Enhanced Capabilities Port AIn 1992, Microsoft and Hewlett-Packard announced another type of high-speed parallel port is the Enhanced Capabilities Port (ECP). Similar to EPP, ECP offers improved performance for the parallel port and requires special hardware logic. Since the original announcement, ECP is included in IEEE-1284​just like EPP. Unlike EPP, however, ECP is not tailored to support portable PCs' parallel port peripherals; its purpose is to support an inexpensive attachment to a very highperformance printer or scanner. Furthermore, ECP mode requires the use of a DMA channel, which EPP did not define and which can cause troublesome conflicts with other devices that use DMA, such as ISA sound cards or high-performance ISA SCSI host adapters. Most PCs with newer Super I/O chips or integrated South Bridge I/O support can support either EPP or ECP mode. Most new systems are being delivered with ECP ports that support high throughput communications. In most cases, the ECP ports can be turned into EPP or standard parallel ports via BIOS. However, it's recommended that the port be placed in ECP mode for the best throughput. Depending on the motherboard, the DMA channel assignment used for ECP mode on a built-in parallel might be performed through the BIOS setup program or through moving a jumper block on the motherboard itself.

Upgrading to EPP/ECP Parallel Ports Virtually every new or recent system today supports both the EPP and ECP modes of the IEEE-1284 parallel port standard. However, if you are working with an older system that might not have these capabilities built into the BIOS or are dealing with an add-on parallel port card, you might not know which modes the port can support.If you want to test the parallel ports in a system, especially to determine which type they are, I highly recommend a utility called Parallel (PARA14.ZIP). This is a handy parallel port information utility that examines your system's parallel ports and reports the port type, I/O address, IRQ level, BIOS name, and an assortment of informative notes and warnings in a compact and easy-to-read display. The output can be redirected to a file for tech support purposes. Parallel uses very sophisticated techniques for port and IRQ detection, and it is aware of a broad range of quirky port features. You can get it from Parallel Technologies (www.lpt.com). If you have an older system that does not include an EPP/ECP port and you want to upgrade, several companies now offer boards with the correct Super I/O chips to implement these features. I recommend that you check with Parallel


Technologies, Byterunner Technologies, Lava Computer Mfg., or SIIG; they are listed in the Vendor List on the DVD accompanying this book. Although USB ports are replacing most nontraditional parallel port uses, highspeed parallel ports such as EPP and ECP are still used for supporting external peripherals, such as Zip drives, CD-ROM drives, scanners, tape drives, and even hard disks. Most of these devices attach to the parallel port using a pass-through connection. This means your local printer can still work through the port, along with the device. The device will have its own drivers that mediate both the communications with the device and the printer pass-through. Using EPP or ECP mode, communications speeds that are as high as 2MBps can be achieved. This can enable a relatively high-speed device to function almost as if it were connected to the system bus internally.

Parallel Port Configuration The configuration of parallel ports is not as complicated as it is for serial ports. Even the original IBM PC had BIOS support for three LPT ports. Table 17.14 shows the standard I/O address and interrupt settings for parallel port use. Table 17.14. Parallel Interface I/O Port Addresses and Interrupts Standard LPTx

Alternate LPTx

LPT1

I/O Ports

IRQ

3BC​3BFh

IRQ7

LPT1

LPT2

378​37Ah

IRQ5

LPT2

LPT3

278h​27Ah

IRQ5

Because the BIOS and DOS have always provided three definitions for parallel ports, problems with older systems are infrequent. Problems can arise, however, from the lack of available interrupt-driven ports for ISA/PCI bus systems. Usually, an interrupt-driven port is not absolutely required for printing operations; in fact, many programs do not use the interrupt-driven capability. However, some programs do use the interrupt, such as network print programs and other types of background or spooler-type printer programs. Also, high-speed laser-printer utility programs often use the interrupt capabilities to enable printing. If you use these types of applications on a port that is not interrupt driven, you see the printing slow to a crawl​if it works at all. The only solution is to use an interrupt-driven port. MS-DOS and Windows 9x/Me/2000/XP now support up to 128 parallel ports.


To configure parallel ports in ISA/PCI bus systems, you normally use the BIOS Setup program for ports that are built into the motherboard, or you might need to set jumpers and switches or use a setup program for adapter-card-based ports. Because each board on the market is different, you always should consult the OEM manual for that particular card if you need to know how the card should be configured. See "BIOS Hardware/Software," p. 369.

Linking Systems with Serial or Parallel Ports You can connect two systems locally using their serial or parallel ports along with specially wired cables. This form of connection has been used over the years to enable a quick and easy mini-LAN to be set up that allows for the transfer of data between two systems. This can be especially useful when you are migrating your data to a new system you have built or purchased. Even though using serial or parallel ports to connect two systems to migrate data can be useful, a much better way to connect two systems is to use Ethernet cards and what is commonly called a crossover cable. Using this type of connection, you can establish a LAN connection between the two systems and transfer data at the full speed of the Ethernet network, which is 10Mbps, 100Mbps, or 1,000Mbps (1.25MBps, 12.5MBps, or 125MBps). Using serial or parallel ports to transfer data between two systems was more prevalent when most systems did not include some form of network interface card (NIC). Several free and commercial programs can support serial or parallel-port file transfers. MS-DOS 6.0 and later include a program called Interlink, whereas Windows 95 and later include software called either Direct Cable Connection or Direct Parallel Connection. Other commercially available software includes Laplink.com's LapLink series, Smith Micro's CheckIt Fast Move, and Symantec's PC Anywhere, among others.

Serial Port Transfers Serial ports have always been used for communications between systems, but the ports were usually connected to modems that convert the data to be transmitted over telephone lines, enabling long-distance connections. If the systems were in


the same room, you could connect the serial ports directly without using modems by instead using what is commonly referred to as a null modem cable. The name comes from the fact that there really aren't any modems present​the cable effectively replaces the modems that would be used at either end. The only drawback is that, because serial ports evolved to support relatively low-speed modem communications, even when using a null modem cable, the connection would be slow because most ports can transmit at 115.2Kbps (14.4KBps). By using a null modem cable, serial ports can be used to transfer data between systems, although the performance is much lower than when using parallel ports. A serial null modem cable has either a 9-pin or a 25-pin female connector on both ends. The cable should be wired as shown in Table 17.15. Table 17.15. Serial Null Modem Cable Wiring (9-Pin or 25-Pin Connectors) 9-Pin

25-Pin

Signal

<-to->

Signal

25-Pin

9-Pin

Pin 5

Pin 7

Ground

<->

Ground

Pin 7

Pin 5

Pin 3

Pin 2

Transmit

<->

Receive

Pin 3

Pin 2

Pin 7

Pin 4

RTS

<->

CTS

Pin 5

Pin 8

Pin 6

Pin 6

DSR

<->

DTR

Pin 20

Pin 4

Pin 2

Pin 3

Receive

<->

Transmit

Pin 2

Pin 3

Pin 8

Pin 5

CTS

<->

RTS

Pin 4

Pin 7

Pin 4

Pin 20

DTR

<->

DSR

Pin 6

Pin 6

If both systems feature infrared serial ports, you can connect the two systems via an infrared connection that uses no cable at all. This type of connection is subject to the limitations of the infrared ports, which allow only very short (a few feet at most) distances with fairly slow data rates. Because they work so slowly, I usually recommend using serial ports to transfer data between two systems only as a last resort. If possible, you should try to use Ethernet cards and a crossover cable, a USB direct connect cable, or parallel ports with an interlink cable for faster connections.

Parallel Port Transfers Although the original intention for the parallel port was for it to be used to


connect a printer, later implementations allowed for bidirectional data transfer. Bidirectional parallel ports can be used to quickly and easily transfer data between one system and another. If both systems use an EPP/ECP port, you can actually communicate at rates of up to 2MBps, which is far faster than the speeds achievable with serial port or infrared (IR) data transfers. Connecting computers with parallel ports requires a special cable known as a Direct Cable Connection (DCC), Interlink, LapLink, parallel crossover, or parallel null modem cable. Many of the commercial file transfer programs provide these cables with their software, or you can purchase them separately from most computer stores or companies that sell cables. However, if you need to make one for yourself, Table 17.16 provides the wiring diagram you need. Table 17.16. Direct Cable Connection, Interlink, Laplink, or Parallel Null Modem Cable Wiring (25-Pin Connectors) 25-Pin

Signal Description

<-to->

Signal Description

25-Pin

Pin 2

Data Bit 0

<->

-Error

Pin 15

Pin 3

Data Bit 1

<->

Select

Pin 13

Pin 4

Data Bit 2

<->

Paper End

Pin 12

Pin 5

Data Bit 3

<->

-Acknowledge

Pin 10

Pin 6

Data Bit 4

<->

Busy

Pin 11

Pin 15

-Error

<->

Data Bit 0

Pin 2

Pin 13

Select

<->

Data Bit 1

Pin 3

Pin 12

Paper End

<->

Data Bit 2

Pin 4

Pin 10

-Acknowledge

<->

Data Bit 3

Pin 5

Pin 11

Busy

<->

Data Bit 4

Pin 6

Pin 25

Ground

<->

Ground

Pin 25


Tip Even though cables usually are provided for data-transfer programs, notebook users might want to look for an adapter that makes the appropriate changes to a standard parallel printer cable. This can make traveling lighter by preventing the need for additional cables. A DB25 null modem adapter, such as Belkin's F4A602, can be used for this purpose.

Although the prebuilt parallel cables referred to in the previous tip work for connecting two machines with ECP/EPP ports, they can't take advantage of the advanced transfer rates of these ports. Special cables are needed to communicate between ECP/EPP ports. Parallel Technologies is a company that sells ECP/EPP cables for connecting to other ECP/EPP computers, as well as a universal cable for connecting any two parallel ports to use the highest speed. Windows 9x/Me include a special program called Direct Cable Connection or Direct Parallel Connection, which enables two systems to be networked together via a parallel transfer cable. With Windows 2000 and XP, use the Network Connections dialog box to set up a direct connection. Consult the Windows documentation for information on how to establish a DCC/Direct Parallel connection. Parallel Technologies has been contracted by Microsoft to supply the special DCC cables used to connect the systems, including a special type of cable that uses active electronics to ensure a reliable high-speed interconnection.

Parallel-to-SCSI Converters Parallel ports can also be used to connect SCSI peripherals to a PC. With a parallel-to-SCSI converter, you can connect any type of SCSI deviceâ&#x20AC;&#x2039;such as hard disks, CD-ROM drives, tape drives, Zip drives, or scannersâ&#x20AC;&#x2039;to your PC, all via the parallel port. To connect to a SCSI device and also continue to be able to print, most parallel-to-SCSI converters include a pass-through connection for a printer. Therefore, at one end of the converter is a connection to your parallel port, and at the other end is both SCSI and parallel-port connections. Thus, you can plug in a single SCSI device and still connect your printer as well. The drivers for the parallel-to-SCSI converter automatically pass through any information to the printer, so the printer works normally. Some of these converters handle only one SCSI device, but others can handle up to seven SCSI devices, just as normal SCSI host adapters do. SCSI devices are much slower when connected to a parallel port than when they are connected to a native SCSI host adapter.

Testing Parallel Ports Testing parallel ports is, in most cases, simpler than testing serial ports. The


procedures are effectively the same as those used for serial ports, except that when you use the diagnostics software, you (obviously) select choices for parallel ports rather than serial ports. Even though the software tests are similar, the hardware tests require the proper plugs for the loopback tests on the parallel port. Several loopback plugs are required, depending on what software you are using. Most use the IBM-style loopback, but some use the style that originated in the Norton Utilities diagnostics. You can purchase loopback plugs or build them yourself. The following wiring is needed to construct your own parallel loopback or wrap plugs to test a parallel port: IBM 25-pin parallel (male DB25P) loopback connector (wrap plug).Connect the following pins: 1 to 13 2 to 15 10 to 16 11 to 17 Norton Utilities 25-pin parallel (male DB25P) loopback connector (wrap plug). Connect the following pins: 2 to 15 3 to 13 4 to 12 5 to 10 6 to 11


Chapter 18. Input Devices Keyboards Keyboard Technology Keyboard Troubleshooting and Repair Keyboard Recommendations Pointing Devices Input Devices for Gaming Wireless Input Devices


Keyboards One of the most basic system components is the keyboard, which is the primary input device. It is used for entering commands and data into the system. This section looks at the keyboards available for PC-compatible systems, examining the various types of keyboards, how the keyboard functions, the keyboard-tosystem interface, and keyboard troubleshooting and repair. In the years since the introduction of the original IBM PC, IBM has created three keyboard designs for PC systems, and Microsoft has augmented one of them. These designs have become de facto standards in the industry and are shared by virtually all PC manufacturers. The primary keyboard types are as follows: 101-key Enhanced keyboard 104-key Windows keyboard 83-key PC and XT keyboard (obsolete) 84-key AT keyboard (obsolete) This section discusses the 101-key Enhanced keyboard and the 104-key Windows keyboard, showing the layout and physical appearance of both. Although you can still find old systems that use the 83-key and 84-key designs, these are rare today. Because all new systems today use the 101- or 104-key keyboard design, these versions are covered here.


Note If you need to learn more about the 83-key PC and XT keyboard or the 84-key AT keyboard, see Chapter 7 of Upgrading and Repairing PCs, 10th Anniversary Edition, on the DVD included with this book.

Enhanced 101-Key (or 102-Key) Keyboard In 1986, IBM introduced the "corporate" Enhanced 101-key keyboard for the newer XT and AT models. I use the word corporate because this unit first appeared in IBM's RT PC, which was a RISC (reduced instruction set computer) system designed for scientific and engineering applications. Keyboards with this design were soon supplied with virtually every type of system and terminal IBM sold. Other companies quickly copied this design, which became the standard on Intel-based PC systems until the introduction of the 104-key Windows keyboard in 1995 (discussed later in this chapter). The layout of this universal keyboard was improved over that of the 84-key unit, with perhaps the exception of the Enter key, which reverted to a smaller size. The 101-key Enhanced keyboard was designed to conform to international regulations and specifications for keyboards. In fact, other companies such as Digital Equipment Corporation (DEC) and Texas Instruments (TI) had already been using designs similar to the IBM 101-key unit. The IBM 101-key units originally came in versions with and without the status-indicator LEDs, depending on whether the unit was sold with an XT or AT system. Now many other variations are available from which to choose, including some with integrated pointing devices, such as the IBM TrackPoint II pointing stick, trackballs and touch pads, and programmable keys useful for automating routine tasks. The Enhanced keyboard is available in several variations, but all are basically the same electrically and all can be interchanged. IBMâ&#x20AC;&#x2039;with its Lexmark keyboard and printer spinoffâ&#x20AC;&#x2039;and Unicomp (which now produces these keyboards) have produced a number of keyboard models, including versions with built-in pointing devices and new ergonomic layouts. With the replacement of the Baby-AT motherboard and its five-pin DIN (an acronym for Deutsche Industrie Norm) keyboard connector by ATX motherboards, which use the six-pin mini-DIN keyboard connector, virtually all keyboards on the market today come with cables for the six-pin mini-DIN connector introduced on the IBM PS/2s. Although the connectors might be physically different, the keyboards are not, and you can either interchange the cables or use a cable adapter to plug one type into the other; some keyboards you can buy at retail include the adapter in the package. See the section "Keyboard/Mouse Interface Connectors" and Figure 18.8 later in this chapter for the physical and electronic details of these connectors.


Many keyboards now include both the standard mini-DIN as well as USB connectors for maximum flexibility when attaching to newer systems. See the section "USB Keyboards" later in this chapter for details on connecting keyboards via USB.

Figure 18.8. Keyboard and mouse connectors.

The 101-key keyboard layout can be divided into the following four sections: Typing area Numeric keypad Cursor and screen controls Function keys The 101-key arrangement is similar to the Selectric keyboard layout, with the exception of the Enter key. The Tab, Caps Lock, Shift, and Backspace keys have a larger striking area and are located in the familiar Selectric locations. Ctrl and Alt keys are on each side of the spacebar, and the typing area and numeric keypad have home-row identifiers for touch typing. The cursor- and screen-control keys have been separated from the numeric keypad, which is reserved for numeric input. (As with other PC keyboards, you can use the numeric keypad for cursor and screen control when the keyboard is not in Num Lock mode.) A division-sign key (/) and an additional Enter key have been added to the numeric keypad.


The cursor-control keys are arranged in the inverted T format that is now expected on all computer keyboards. The Insert, Delete, Home, End, Page Up, and Page Down keys, located above the dedicated cursor-control keys, are separate from the numeric keypad. The function keys, spaced in groups of four, are located across the top of the keyboard. The keyboard also has two additional function keys: F11 and F12. The Esc key is isolated in the upper-left corner of the keyboard. In addition, dedicated Print Screen/Sys Req, Scroll Lock, and Pause/Break keys are provided for commonly used functions. Foreign-language versions of the Enhanced keyboard include 102 keys and a slightly different layout from the 101-key U.S. versions. One of the many useful features of the IBM/Lexmark enhanced keyboard (now manufactured by Unicomp) is removable keycaps. This permits the replacement of broken keys and provides access for easier cleaning. Also, with clear keycaps and paper inserts, you can customize the keyboard. Keyboard templates are also available to provide specific operator instructions.

104-Key (Windows 9x/Me/2000/XP) Keyboard With the introduction of Windows 95, a modified version of the standard 101-key design (created by Microsoft) appeared, called the 104-key Windows keyboard. If you are a touch typist as I am, you probably really hate to take your hands off the keyboard to use a mouse. Windows 9x and newer versions make this even more of a problem because they use both the right and left mouse buttons (the right button is used to open shortcut menus). Many new keyboards, especially those in portable computers, include a variation of the IBM TrackPoint or the Cirque GlidePoint pointing devices (discussed later in this chapter), which enable touch typists to keep their hands on the keyboard even while moving the pointer. However, another alternative is available that can help. When Microsoft released Windows 95, it also introduced the Microsoft Natural Keyboard, which implemented a revised keyboard specification that added three new Windowsspecific keys to the keyboard. The Microsoft Windows keyboard specification, which has since become a de facto industry standard for keyboard layouts, outlines a set of additional keys and key combinations. The 104-key layout includes left and right Windows keys and an Application key (see Figure 18.1). These keys are used for operating system and application-level keyboard combinations, similar to the existing Ctrl and Alt combinations. You don't need the new keys to use Windows, but software vendors are adding specific functions to their Windows products that use the new Application key (which provides the same functionality as clicking the right mouse button). The recommended Windows keyboard layout calls for the Left and Right


Windows keys (called WIN keys) to flank the Alt keys on each side of the spacebar, as well as an Application key on the right of the Right Windows key. Note that the exact placement of these keys is up to the keyboard designer, so variations exist from keyboard to keyboard.

Figure 18.1. The 104-key Windows keyboard layout.

The WIN keys open the Windows Start menu, which you can then navigate with the cursor keys. The Application key simulates the right mouse button; in most applications, it brings up a context-sensitive pop-up menu. Several WIN key combinations offer preset macro commands as well. For example, you can press WIN+E to launch the Windows Explorer application. Table 18.1 shows a list of all the Windows key combinations used with the 104-key keyboard. Table 18.1. Windows 9x/Me/2000/XP Key Combinations Key Combination

Resulting Action

WIN+R

Runs dialog box

WIN+M

Minimize All

Shift+WIN+M

Undo Minimize All

WIN+D

Minimize All or Undo Minimize All

WIN+F1

Help

WIN+E

Starts Windows Explorer

WIN+F

Find Files or Folders

Ctrl+WIN+F

Find Computer


WIN+Tab

Cycles through taskbar buttons

WIN+Break

Displays System Properties dialog box

Application key

Displays a context menu for the selected item

The preceding keystroke combinations work with any manufacturer's 104-key keyboard, but users of certain Microsoft 104-key keyboards can enhance their keyboard use further by installing the IntelliType Pro software supplied by Microsoft with the keyboard. The Windows keys are not mandatory when running Windows. In fact, preexisting standard key combinations perform the same functions as these newer keys. I also have noticed that only power users wanting to work as efficiently as possible by keeping their hands on the keyboard (and off the mouse) primarily use these combinations. The Windows keyboard specification requires that keyboard makers increase the number of trilograms in their keyboard designs. A trilogram is a combination of three rapidly pressed keys that perform a special function, such as Ctrl+Alt+Delete. Designing a keyboard so that the switch matrix correctly registers the additional trilograms plus the additional Windows keys adds somewhat to the cost of these keyboards compared to the previous 101-key standard models. Virtually all keyboard manufacturers have standardized on 104-key keyboards that include these Windows-specific keys. Some manufacturers have added browser control or other keys that, although not standard, can make them easier to use for navigating Web pages and launching various applications. For additional keyboard combinations you can use, see "Running Windows Without a Mouse" in the Technical Reference section of the DVD included with this book.

USB Keyboards Most keyboards now on the market can connect to the PC via a USB port instead of the standard PS/2 keyboard port. Because USB is a universal bus that uses a hub to enable multiple devices to connect to a single port, a single USB port in a system can replace the standard serial and parallel ports as well as the keyboard and mouse ports. Most current systems and motherboards still include the standard ports (now called legacy ports) as well as USB, but most so-called legacy-free systems and replacement motherboards have only USB ports for


interfacing external devices. Most keyboard manufacturers now market USB keyboards, but if you want to use your keyboard with both legacy (PS/2) and legacy-free (USB) systems, the most economic way to do so is to specify a keyboard that includes both a USB connector and an adapter to permit the keyboard to work with PS/2 ports. Although Microsoft's Natural Keyboard Elite was the first widely available model to offer USB and PS/2 compatibility, other wired and wireless models from Microsoft, Logitech, Belkin, and others now offer this feature. You can also purchase thirdparty USB-to-PS/2 adapters, but these can be expensive and might not work with all keyboards. Not all systems accept USB keyboards, even those with USB ports, because the standard PC BIOS has a keyboard driver that expects a standard keyboard port interface to be present. When a USB keyboard is installed on a system that lacks USB keyboard support, the system can't use it because no driver exists in the BIOS to make it work. In fact, some systems see the lack of a standard keyboard as an error and halt the boot process until one is installed. To use a keyboard connected via the USB port, you must meet three requirements: Have a USB port in the system Run Microsoft Windows 98, Windows Me, Windows 2000, or Windows XP (all of which include USB keyboard drivers) Have a system chipset and BIOS that feature USB Legacy support USB Legacy support means your motherboard has a chipset and ROM BIOS drivers that enable a USB keyboard to be used outside the Windows GUI environment. When a system has USB Legacy support enabled, a USB keyboard can be used with MS-DOS (for configuring the system BIOS) when using a command prompt within Windows or when installing Windows on the system for the first time. If USB Legacy support is not enabled on the system, a USB keyboard will function only when Windows is running. Most recent systems include USB Legacy support, although it might be disabled by default in the system BIOS. Also, if the Windows installation fails and requires manipulation outside of Windows, the USB keyboard will not function unless it is supported by the chipset and the BIOS. Almost all 1998 and newer systems with USB ports include a chipset and BIOS with USB Legacy (meaning USB Keyboard) support.


Even though USB Legacy support enables you to use a USB keyboard in almost all situations, don't scrap your standard-port keyboards just yet. Some Windowsrelated bugs and glitches reported by users include the following: Can't log on to Windows the first time after installing a USB keyboard. The solution in some cases is to click Cancel when you are asked to log on and then allow the system to detect the keyboard and install drivers. The logon should work normally thereafter. In other cases, you might have to leave the keyboard unplugged when first booting and then plug it in after the OS desktop is up and running. This allows the keyboard to be detected and drivers loaded. Some USB keyboards won't work when the Windows Emergency Boot Disk (EBD) is used to start the system. The solution is to turn off the system, connect a standard keyboard, and restart the system. Some users of Windows 98 and Windows 98 SE have reported conflicts between Windows and the BIOS when USB Legacy support is enabled on some systems. This conflict can result in an incapability to detect the USB keyboard if you use the Windows 9x shutdown menu and choose to Restart the computer in MS-DOS mode. Check with the system or BIOS vendor for an updated BIOS or a patch to solve this conflict. If you have problems with Legacy USB support, look at these possible solutions: Microsoft's Knowledge Base might address your specific combination of hardware. Your keyboard vendor might offer new drivers. Your system or motherboard vendor might have a BIOS upgrade you can install. Connect the keyboard to the PS/2 port with its adapter (or use a PS/2 keyboard) until you resolve the problem. See "Universal Serial Bus," p. 947.


Notebook Computer Keyboards One of the biggest influences on keyboard design in recent years has been the proliferation of laptop and notebook systems. Because of size limitations, it is obviously impossible to use the standard keyboard layout for a portable computer. Manufacturers have come up with many solutions. Unfortunately, none of these solutions has become an industry standard, as the 101-key layout is. Because of the variations in design, and because a portable system keyboard is not as easily replaceable as that of a desktop system, the keyboard arrangement should be an important part of your purchasing decision. Early laptop systems often used smaller-than-normal keys to minimize the size of the keyboard, which resulted in many complaints from users. Today, the keytops on portable systems are usually comparable in size to that of a desktop keyboard, although some systems include half-sized keytops for the function keys and other less frequently used keyboard elements. In addition, consumer demand has caused most manufacturers to retain the inverted-T design for the cursor keys after a few abortive attempts at changing their arrangement. Of course, the most obvious difference in a portable system keyboard is the sacrifice of the numeric keypad. Most systems now embed the keypad into the standard alphabetical part of the keyboard, as shown in Figure 18.2. To switch the keys from their standard values to their keypad values, you typically must press a key combination involving a proprietary function key, often labeled Fn.

Figure 18.2. Most portable systems today embed the numeric keypad into an oddly shaped block of keys on the alphabetical part of the keyboard.

This is an extremely inconvenient solution, and many users abandon their use of the keypad entirely on portable systems. Unfortunately, some activitiesâ&#x20AC;&#x2039;such as the entry of ASCII codes using the Alt keyâ&#x20AC;&#x2039;require the use of the keypad numbers, which can be quite frustrating on systems using this arrangement.


To alleviate this problem, many portable system manufacturers sell external numeric keypads that plug into the external keyboard port, a serial port, or a USB port. This is a great feature for somebody performing a lot of numeric data entry. In addition to keypad control, the Fn key often is used to trigger other proprietary features in portable systems, such as toggling between an internal and external display and controlling screen brightness and sound volume. Some portable system manufacturers have gone to great lengths to provide users with adequate keyboards. For a short time, IBM marketed systems with a keyboard that used a "butterfly" design. The keyboard was split into two halves that rested on top of one another when the system was closed. When you opened the lid, the two halves separated to rest side by side, forming a keyboard that was actually larger than the computer case. Ironically, the trend toward larger-sized displays in portable systems has made this sort of arrangement unnecessary. Many manufacturers have increased the footprint of their notebook computers to accommodate 14.1'' and even 15'' display panels, leaving more than adequate room for a keyboard with full-size keys. However, even on the newest systems, there still isn't enough room for a separate numeric keypad.

Num Lock On IBM systems that support the Enhanced keyboard, when the system detects the keyboard on powerup, it enables the Num Lock feature and the light goes on. If the system detects an older 84-key AT-type keyboard, it does not enable the Num Lock function because these keyboards do not have cursor keys separate from the numeric keypad. When the Enhanced keyboards first appeared in 1986, many users (including me) were irritated to find that the numeric keypad was automatically enabled every time the system booted. Most system manufacturers subsequently began integrating a function into the BIOS setup that enabled you to specify the Num Lock status imposed during the boot process. Some users thought that the automatic enabling of Num Lock was a function of the Enhanced keyboard because none of the earlier keyboards seemed to operate in this way. Remember that this function is not really a keyboard function but instead a function of the motherboard ROM BIOS, which identifies an Enhanced 101-key unit and turns on the Num Lock as a "favor." In systems with a BIOS that can't control the status of the numeric keypad, you can use the DOS 6.0 or higher version NUMLOCK= parameter in CONFIG.SYS to turn Num Lock on or off, as desired. Some versions of Windows, particularly Windows NT and Windows 2000 (but not Windows XP), disable Num Lock by default.


Keyboard Technology The technology that makes up a typical PC keyboard is very interesting. This section focuses on all the aspects of keyboard technology and design, including the keyswitches, the interface between the keyboard and the system, the scan codes, and the keyboard connectors.

Keyswitch Design Today's keyboards use any one of several switch types to create the action for each key. Most keyboards use a variation of the mechanical keyswitch. A mechanical keyswitch relies on a mechanical momentary contact-type switch to make the electrical contact that forms a circuit. Some high-end keyboards use a more sophisticated design that relies on capacitive switches. This section discusses these switches and the highlights of each design. The most common type of keyswitch is the mechanical type, available in the following variations: Pure mechanical Foam element Rubber dome Membrane

Pure Mechanical Switches The pure mechanical type is just thatâ&#x20AC;&#x2039;a simple mechanical switch that features metal contacts in a momentary contact arrangement. The switch often includes a tactile feedback mechanism, consisting of a clip and spring arrangement designed to give a "clicky" feel to the keyboard and offer some resistance to the keypress (see Figure 18.3).

Figure 18.3. A typical mechanical switch used in older NMB keyboards. As the key is pressed, the switch pushes down on the contacts to make the connection.


Mechanical switches are very durable, usually have self-cleaning contacts, and are normally rated for 20 million keystrokes (which is second only to the capacitive switch in longevity). They also offer excellent tactile feedback. Despite the tactile feedback and durability provided by mechanical keyswitch keyboards, they have become much less popular than membrane keyboards (discussed later in this chapter). In addition, many companies that produce keyboards that use mechanical keyswitches either use them for only a few of their high-priced models or have phased out their mechanical keyswitch models entirely. With the price of keyboards nosediving along with other traditional devices, such as mice and drives, the pressure on keyboard makers to cut costs has led many of them to abandon or de-emphasize mechanical-keyswitch designs in favor of the less expensive membrane keyswitch. The Alps Electric mechanical keyswitch is used by many of the vendors who produce mechanical-switch keyboards, including Alps Electric itself. Other vendors who use mechanical keyswitches for some of their keyboard models include Adesso, Inc. (www.adessoinc.com), Avant Prime and Stellar (revivals of the classic Northgate keyboards and available from Ergonomic Resources; www.ergo2000.com), Kinesis (www.kinesis-ergo.com), SIIG (www.siig.com), and Focus (www.focustaipei.com). Many of these vendors sell through the OEM market, so you must look carefully at the detailed specifications for the keyboard to see whether it is a mechanical keyswitch model.

Foam Element Switches Foam element mechanical switches were a very popular design in some older keyboards. Most of the older PC keyboards, including models made by Key Tronic and many others, used this technology. These switches are characterized by a foam element with an electrical contact on the bottom. This foam element is mounted on the bottom of a plunger that is attached to the key (see Figure 18.4).


Figure 18.4. Typical foam element mechanical keyswitch.

When the switch is pressed, a foil conductor on the bottom of the foam element closes a circuit on the printed circuit board below. A return spring pushes the key back up when the pressure is released. The foam dampens the contact, helping to prevent bounce, but unfortunately it gives these keyboards a "mushy" feel. The big problem with this type of keyswitch design is that little tactile feedback often exists. These types of keyboards send a clicking sound to the system speaker to signify that contact has been made. Preferences in keyboard feel are somewhat subjective; I personally do not favor the foam element switch design. Another problem with this type of design is that it is more subject to corrosion on the foil conductor and the circuit board traces below. When this happens, the key strikes can become intermittent, which can be frustrating. Fortunately, these keyboards are among the easiest to clean. By disassembling the keyboard completely, you usually can remove the circuit board portionâ&#x20AC;&#x2039;without removing each foam pad separatelyâ&#x20AC;&#x2039;and expose the bottoms of all the pads. Then, you easily can wipe the corrosion and dirt off the bottoms of the foam pads and the circuit board, thus restoring the keyboard to a "like-new" condition. Unfortunately, over time, the corrosion problem will occur again. I recommend using some Stabilant 22a from D.W. Electrochemicals (www.stabilant.com) to improve the switch contact action and prevent future corrosion. Because of such problems, the foam element design is not used much anymore and has been superseded in popularity by the rubber dome design. KeyTronicEMS, the most well-known user of this technology, now uses a centerbearing membrane switch technology in its keyboards, so you are likely to encounter foam-switch keyboards only on very old systems.

Rubber Dome Switches Rubber dome switches are mechanical switches similar to the foam element type


but are improved in many ways. Instead of a spring, these switches use a rubber dome that has a carbon button contact on the underside. As you press a key, the key plunger presses on the rubber dome, causing it to resist and then collapse all at once, much like the top of an oil can. As the rubber dome collapses, the user feels the tactile feedback, and the carbon button makes contact between the circuit board traces below. When the key is released, the rubber dome re-forms and pushes the key back up. The rubber eliminates the need for a spring and provides a reasonable amount of tactile feedback without any special clips or other parts. Rubber dome switches use a carbon button because it resists corrosion and has a self-cleaning action on the metal contacts below. The rubber domes themselves are formed into a sheet that completely protects the contacts below from dirt, dust, and even minor spills. This type of switch design is the simplest, and it uses the fewest parts. This made the rubber dome keyswitch very reliable for several years. However, its relatively poor tactile feedback has led most keyboard manufacturers to switch to the membrane switch design covered in the next section.

Membrane Switches The membrane keyswitch is a variation on the rubber dome type, using a flat, flexible circuit board to receive input and transmit it to the keyboard microcontroller. Industrial versions of membrane boards use a single sheet for keys that sits on the rubber dome sheet for protection against harsh environments. This arrangement severely limits key travel. For this reason, flatsurface membrane keyboards are not considered usable for normal touch typing. However, they are ideal for use in extremely harsh environments. Because the sheets can be bonded together and sealed from the elements, membrane keyboards can be used in situations in which no other type could survive. Many industrial applications use membrane keyboards for terminals that do not require extensive data entry but are used instead to operate equipment, such as cash registers and point-of-sale terminals in restaurants. Membrane keyswitches are no longer relegated to fast food or industrial uses, though. Over the last few years, the membrane keyswitch used with conventional keyboard keytops has replaced the rubber dome keyswitch to become the most popular keyswitch used in low-cost to mid-range keyboards. Inexpensive to make, membrane switches have become the overwhelming favorite of low-cost Pacific Rim OEM suppliers and are found in most of the keyboards you'll see at your local computer store or find inside the box of your next complete PC. Although low-end membrane keyswitches have a limited life of only 5â&#x20AC;&#x2039;10 million keystrokes, some of the better models are rated to handle up to 20 million keystrokes, putting them in the range of pure mechanical switches for durability (see Figure 18.5). A few membrane switches are even more durable: Cherry Corporation's G8x-series


keyboards use Cherry's own 50-million-keystroke membrane switch design (www.cherrycorp.com).

Figure 18.5. A typical membrane keyswitch used in NMB keyboards.

Membrane keyboards provide a firmer touch than rubber dome keyboards or the old foam-element keyboards, but they are still no match for mechanical or capacitive keyswitch models in their feel. One interesting exception is the line of keyboards made by KeyTronicEMS using its center-bearing version of membrane keyswitches. Most of its keyboards feature Ergo Technology, which has five levels of force from 35 grams to 80 grams, depending on the relative strength of the fingers used to type various keys. As little as 35 grams of force is required for keys that are used by the little finger, such as Q, Z, and A, and greater levels of force are required for keys used by the other fingers. The spacebar requires the most force: 80 grams. This compares to the standard force level of 55 grams for all keys on normal keyboards (see Figure 18.6). For more information about keyboards with Ergo Technology, visit the KeyTronicEMS Web site (www.keytronic.com).

Figure 18.6. Force levels used on KeyTronicEMS keyboards with


Ergo Technology.

To find the best membrane keyboards from the vast numbers on the market, look at the lifespan rating of the keyswitches. Longer-lasting keyswitches make the keyboard cost more but will lead to a better experience over the life of the keyboard.

Capacitive Switches Capacitive switches are the only nonmechanical keyswitch in use today (see Figure 18.7). The capacitive switch is the Cadillac of keyswitches. It is much more expensive than the more common mechanical membrane switch, but it is more resistant to dirt and corrosion and offers the highest-quality tactile feedback of any type of switch. This type of keyboard is sometimes referred to as a buckling spring keyboard because of the coiled spring used to provide feedback.

Figure 18.7. A capacitive buckling spring keyswitch.


A capacitive switch does not work by making contact between conductors. Instead, two plates usually made of plastic are connected in a switch matrix designed to detect changes in the capacitance of the circuit. When the key is pressed, the plunger moves the top plate in relation to the fixed bottom plate. Typically, a buckling spring mechanism provides for a distinct overcenter tactile feedback with a resounding "click." As the top plate moves, the capacitance between the two plates changes. The comparator circuitry in the keyboard detects this change. Because this type of switch does not rely on metal contacts, it is nearly immune to corrosion and dirt. These switches are also very resistant to the key bounce problems that result in multiple characters appearing from a single strike. In addition, they are the most durable in the industryâ&#x20AC;&#x2039;rated for 25 million or more keystrokes, as opposed to 10â&#x20AC;&#x2039;20 million for other designs. The tactile feedback is unsurpassed because the switch provides a relatively loud click and a strong overcenter feel. The only drawback to the design is the cost. Capacitive switch keyboards are among the most expensive designs. The quality of the feel and their durability make them worth the price, however. Originally, the only vendor of capacitive keyswitch keyboards was IBM. Although some of IBM's older keyboards still feature capacitive keyswitches, most current IBM keyboards use rubber-dome or other lower-cost keyswitches. In 1991, IBM spun off its keyboard/printer division as Lexmark, which then spun off the keyboard division as Unicomp in 1996. Today, Unicomp still manufactures and sells "IBM" keyboards with the classic buckling spring capacitive switch ("clickety" as some would say) technology. As a bonus, it also has models with the IBM trackpoint built in. You can purchase new Unicomp (IBM) keyboards direct by calling its toll-free number (800-777-4886) or by visiting its online store (http://www.pckeyboard.com). My personal recommendations are for either the EnduraPro/104 (http://www.pckeyboard.com/ep104.html) or the Customizer 101 or 104 (http://www.pckeyboard.com/customizer.html). These are brand-new, not reconditioned or rebuilt, keyboards. The EnduraPro/104 is notable for including a built-in TrackPoint pointing device and a pass-through mini-DIN mouse port, being programmable and reconfigurable, requiring no special drivers, and of course having the famous buckling spring keyswitches. Because of the buckling spring capacitive keyswitches (and the resulting clickety feel), I've always been a huge fan of the IBM, Lexmark, and now Unicomp keyboards. In my opinion, they are the absolute best keyboards in the world and the only ones I willingly use on desktop systems. I especially like the fact that they include the IBM TrackPoint because I use a laptop system as my main


machine and therefore use only laptops that include the TrackPoint device (mainly IBM, Toshiba, and some Dell/HP/others). The feel and durability of the buckling spring capacitive keyswitches is outstanding, and with the integrated TrackPoint, I never have to move my hands off the keyboard, resulting in much greater efficiency when working with my systems.

The Keyboard Interface A keyboard consists of a set of switches mounted in a grid or an array called the key matrix. When a switch is pressed, a processor in the keyboard identifies which key is pressed by determining which grid location in the matrix shows continuity. The keyboard processor, which also interprets how long the key is pressed, can even handle multiple keypresses at the same time. A 16-byte hardware buffer in the keyboard can handle rapid or multiple keypresses, passing each one to the system in succession. When you press a key, the contact bounces slightly in most cases, meaning that several rapid on/off cycles occur just as the switch makes contact. This is called bounce. The processor in the keyboard is designed to filter this, or debounce the keystroke. The keyboard processor must distinguish bounce from a double key strike the keyboard operator intends to make. This is fairly easy, though, because the bouncing is much more rapid than a person could simulate by striking a key quickly several times. The keyboard in a PC is actually a computer itself. It communicates with the main system in one of two ways: Through a special serial data link if a standard keyboard connector is used Through the USB port The serial data link used by conventional keyboards transmits and receives data in 11-bit packets of information, consisting of 8 data bits, plus framing and control bits. Although it is indeed a serial link (in that the data flows on one wire), the keyboard interface is incompatible with the standard RS-232 serial port commonly used to connect modems. The processor in the original PC keyboard was an Intel 8048 microcontroller chip. Newer keyboards often use an 8049 version that has built-in ROM or other microcontroller chips compatible with the 8048 or 8049. For example, in its Enhanced keyboards, IBM has always used a custom version of the Motorola 6805 processor, which is compatible with the Intel chips. The keyboard's built-in processor reads the key matrix, debounces the keypress signals, converts the


keypress to the appropriate scan code, and transmits the code to the motherboard. The processors built into the keyboard contain their own RAM, possibly some ROM, and a built-in serial interface. In the original PC/XT design, the keyboard serial interface is connected to an 8255 Programmable Peripheral Interface (PPI) chip on the motherboard of the PC/XT. This chip is connected to the interrupt controller IRQ1 line, which is used to signal to the system that keyboard data is available. The data is then sent from the 8255 to the processor via I/O port address 60h. The IRQ1 signal causes the main system processor to run a subroutine (INT 9h) that interprets the keyboard scan code data and decides what to do. In an AT-type keyboard design, the keyboard serial interface is connected to a special keyboard controller on the motherboard. This controller was an Intel 8042 Universal Peripheral Interface (UPI) slave microcontroller chip in the original AT design. This microcontroller is essentially another processor that has its own 2KB of ROM and 128 bytes of RAM. An 8742 version that uses erasable programmable read-only memory (EPROM) can be erased and reprogrammed. In the past, when you purchased a motherboard ROM upgrade for an older system from a motherboard manufacturer, the upgrade included a new keyboard controller chip as well because it had somewhat dependent and updated ROM code in it. Some older systems might use the 8041 or 8741 chips, which differ only in the amount of built-in ROM or RAM. However, recent systems incorporate the keyboard controller into the main system chipset. In an AT system, the (8048-type) microcontroller in the keyboard sends data to the (8042-type) keyboard controller on the motherboard. The motherboard-based controller also can send data back to the keyboard. When the keyboard controller on the motherboard receives data from the keyboard, it signals the motherboard with an IRQ1 and sends the data to the main motherboard processor via I/O port address 60h, just as in the PC/XT. Acting as an agent between the keyboard and the main system processor, the 8042-type keyboard controller can translate scan codes and perform several other functions as well. The system also can send data to the 8042 keyboard controller via port 60h, which then passes it on to the keyboard. Additionally, when the system needs to send commands to or read the status of the keyboard controller on the motherboard, it reads or writes through I/O port 64h. These commands usually are followed by data sent back and forth via port 60h. In older systems, the 8042 keyboard controller is also used by the system to control the A20 memory address line, which provides access to system memory greater than 1MB. More modern motherboards typically incorporate this functionality directly into the motherboard chipset. The AT keyboard connector was renamed the "PS/2" port after the IBM PS/2 family of systems debuted in 1987. That was the time when the connector changed in size from the DIN to the min-DIN, and even though the signals were the same, the mini-DIN version


became known from that time forward as the PS/2 port. Keyboards connected to a USB port work in a surprisingly similar fashion to those connected to conventional DIN or mini-DIN (PS/2) ports after the data reaches the system. Inside the keyboard a variety of custom controller chips is used by various keyboard manufacturers to receive and interpret keyboard data before sending it to the system via the USB port. Some of these chips contain USB hub logic to enable the keyboard to act as a USB hub. After the keyboard data reaches the USB port on the system, the USB port routes the data to the 8042-compatible keyboard controller, where the data is treated as any other keyboard information. This process works very well after a system has booted into Windows. But what about users who need to use the keyboard at a command prompt or within the BIOS configuration routine? As discussed earlier in this chapter, USB Legacy support must be enabled in the BIOS. A BIOS with USB Legacy support is capable of performing the following tasks: Configure the host controller Enable a USB keyboard and mouse Set up the host controller scheduler Route USB keyboard and mouse input to the 8042 Keyboard Controller Systems with USB Legacy support enabled use the BIOS to control the USB keyboard until a supported operating system is loaded. At that point, the USB host controller driver in the operating system takes control of the keyboard by sending a command called StopBIOS to the BIOS routine that was managing the keyboard. When Windows shuts down to MS-DOS, the USB host controller sends a command called StartBIOS to restart the BIOS routine that manages the keyboard. When the BIOS controls the keyboard, after the signals reach the 8042 Keyboard Controller, the USB keyboard is treated just like a conventional keyboard if the BIOS is correctly designed to work with USB keyboards. As discussed previously in this chapter, a BIOS upgrade might be necessary in some cases to provide proper support of USB keyboards on some systems. The system chipset also must support USB Legacy features.

Typematic Functions


If a key on the keyboard is held down, it becomes typematic, which means the keyboard repeatedly sends the keypress code to the motherboard. In the AT-style keyboards, the typematic rate is adjusted by sending the appropriate commands to the keyboard processor. This is impossible for the earlier PC/XT keyboard types because the keyboard interface for these types is not bidirectional. AT-style keyboards have programmable typematic repeat rate and delay parameters. You can adjust the typematic repeat rate and delay parameters with settings in your system BIOS (although not all BIOS chips can control all functions) or in your operating system. In Windows you use the Keyboard icon in the Control Panel; in DOS you use the MODE command. The next section describes how to adjust the keyboard parameters in Windows because this is more convenient than the other methods and enables the user to make further adjustments at any time without restarting the system.

Adjusting Keyboard Parameters in Windows You can modify the default values for the typematic repeat rate and delay parameters in any version of Windows using the Keyboard icon in the Control Panel. The Repeat Delay slider controls the number of times a key must be pressed before the character begins to repeat, and the Repeat Rate slider controls how fast the character repeats after the delay has elapsed.


Note The increments on the Repeat Delay and Repeat Rate sliders in Keyboard Properties in the Control Panel correspond to the timings given for the MODE command's RATE and DELAY values. Each mark in the Repeat Delay slider adds about 0.25 seconds to the delay, and the marks in the Repeat Rate slider are worth about one character per second each.

The dialog box also contains a text box you can use to test the settings you have chosen before committing them to your system. When you click in the box and press a key, the keyboard reacts using the settings currently specified by the sliders, even if you have not yet applied the changes to the Windows environment. To learn how to adjust keyboard parameters in DOS, see "Adjusting Keyboard Parameters in DOS" in Chapter 17 of Upgrading and Repairing PCs, 11th Edition, which is available in electronic form on the DVD-ROM included with this book.

Keyboard Key Numbers and Scan Codes When you press a key on the keyboard, the processor built into the keyboard (8048- or 6805-type) reads the keyswitch location in the keyboard matrix. The processor then sends to the motherboard a serial packet of data containing the scan code for the key that was pressed. This is called the Make code. When the key is released, a corresponding Break code is sent, indicating to the motherboard that the key has been released. The Break code is equivalent to the Make scan code plus 80h. For example, if the Make scan code for the "A" key is 1Eh, the Break code would be 9Eh. By using both Make and Break scan codes, the system can determine whether a particular key has been held down and determine whether multiple keys are being pressed. In AT-type motherboards that use an 8042-type keyboard controller, the 8042 chip translates the actual keyboard scan codes into one of up to three sets of system scan codes, which are sent to the main processor. It can be useful in some cases to know what these scan codes are, especially when troubleshooting keyboard problems or when reading the keyboard or system scan codes directly in software. When a keyswitch on the keyboard sticks or otherwise fails, the Make scan code of the failed keyswitch usually is reported by diagnostics software, including the power on self test (POST), as well as conventional disk-based diagnostics. This means you must identify the malfunctioning key by its scan code. See the Technical Reference section of the DVD included with this book for a comprehensive listing of keyboard key numbers and scan codes for both the


101/102-key (Enhanced) keyboard and 104-key Windows keyboard. By looking up the reported scan code on these charts, you can determine which keyswitch is defective or needs to be cleaned.


Note The 101-key Enhanced keyboards are capable of three scan code sets. Set 1 is the default. Some systems, including some of the IBM PS/2 machines, use one of the other scan code sets during the POST. For example, my IBM P75 uses Scan Code Set 2 during the POST but switches to Set 1 during normal operation. This is rare, and it really threw me off in diagnosing a stuck key problem one time. It is useful to know whether you are having difficulty interpreting the scan code number, however.

IBM also assigns each key a unique key number to distinguish it from the others. This is important when you are trying to identify keys on foreign keyboards that might use symbols or characters different from what the U.S. models do. In the Enhanced keyboard, most foreign models are missing one of the keys (key 29) found on the U.S. version and have two additional keys (keys 42 and 45). This accounts for the 102-key total instead of the 101 keys found on the U.S. version.


Note See the Technical Reference section of the DVD included with this book for a comprehensive listing of keyboard key numbers and scan codes for both the 101/102-key (Enhanced) keyboard and 104-key Windows keyboard, including HID and hotkey scan codes used on the latest USB and hotkey keyboards. Knowing these key number figures and scan codes can be useful when you are troubleshooting stuck or failed keys on a keyboard. Diagnostics can report the defective keyswitch by the scan code, which varies from keyboard to keyboard on the character it represents and its location.

Many enhanced and USB keyboards now feature hotkeys that either have fixed uses​such as opening the default Web browser, sending the system into standby mode, and adjusting the speaker volume​or are programmable for user-defined functions. Each of these keys also has scan codes. USB keyboards use a special series of codes called Human Interface Device (HID), which are translated into PS/2 scan codes.

International Keyboard Layouts After the keyboard controller in the system receives the scan codes generated by the keyboard and passes them to the main processor, the operating system converts the codes into the appropriate alphanumeric characters. In the United States, these characters are the letters, numbers, and symbols found on the standard American keyboard. However, no matter which characters you see on the keytops, adjusting the scan code conversion process to map different characters to the keys is relatively simple. Windows (post 3.x) takes advantage of this capability by enabling you to install multiple keyboard layouts to support various languages. In Windows 9x/Me, open the Keyboard icon in the Control Panel and select the Language page. The Language box should display the keyboard layout you selected when you installed the operating system. In Windows XP, click the Details button found on the Languages tab in the Regional and Language Options applet (in the Windows Control Panel). By clicking the Add button, you can select any one of several additional keyboard layouts supporting other languages. These keyboard layouts map various characters to certain keys on the standard keyboard. The standard French layout provides easy access to the accented characters commonly used in that language. For example, pressing the 2 key produces the é character. To type the numeral 2, you press the Shift+2 key combination. Other French-speaking countries have different keyboard conventions for the same characters, so Windows includes support for several keyboard layout variations for some languages, based on nationality.


Note It is important to understand that this feature is not the same as installing the operating system in a different language. These keyboard layouts do not modify the text already displayed onscreen; they only alter the characters generated when you press certain keys.

The alternative keyboard layouts also do not provide support for non-Roman alphabets, such as Russian and Chinese. The accented characters and other symbols used in languages such as French and German are part of the standard ASCII character set. They are always accessible to English-language users through the Windows Character Map utility or through the use of Alt+keypad combinations. An alternative keyboard layout simply gives you an easier way to access the characters used in certain languages. If you work on documents using more than one language, you can install as many keyboard layouts as necessary and switch between them at will. When you click the Enable Indicator on Taskbar check box on the Language page of the Keyboard control panel, a selector appears in the taskbar's tray area that enables you to switch languages easily. On the same page, you can enable a key combination that switches between the installed keyboard layouts.

Keyboard/Mouse Interface Connectors Keyboards have a cable with one of two primary types of connectors at the system end. On most aftermarket keyboards, the cable is connected inside the keyboard case on the keyboard end, requiring you to open the keyboard case to disconnect or test it; different vendors use different connections, making cable interchange between brands of keyboards unlikely. When IBM manufactured its own enhanced keyboards, it used a unique cable assembly that plugged into both the keyboard and the system unit to make cable replacement or interchange easy. Current IBM keyboards, unfortunately, no longer use either the shielded data link (SDL) connector inside the keyboard or the telephone cable-style removable plug-in external keyboard connector used on some more recent models. Although the method of connecting the keyboard cable to the keyboard can vary, all PC keyboards (except those using the USB port) use either of the following two connectors to attach to the computer: 5-pin DIN connector. Used on most PC systems with Baby-AT form factor motherboards 6-pin mini-DIN connector. Used on PS/2 systems and most PCs with LPX, ATX, and NLX motherboards


Figure 18.8 and Table 18.2 show the physical layout and pinouts of all the respective keyboard connector plugs and sockets; although the 6-pin SDL connector is not used in this form by most keyboard vendors, most non-IBM keyboards use a somewhat similar connector to attach the keyboard cable to the inside of the keyboard. You can use the pinouts listed in Table 18.2 to test the continuity of each wire in the keyboard connector. Table 18.2. Keyboard Connector Signals and Specifications Signal Name

5-Pin DIN

6-Pin Mini-DIN

6-Pin SDL

Test Voltage

Keyboard Data

2

1

B

Ground

4

3

C

+5V Power

5

4

E

+2.0V to +5.5V

Keyboard Clock

1

5

D

+2.0V to +5.5V

Not Connected

2

A

Not Connected

6

F

Not Connected

+4.8V to +5.5V

3

DIN = Deutsche Industrie Norm, a committee that sets German dimensional standards SDL = Shielded data link, a type of shielded connector created by AMP and used by IBM and others for keyboard cables

Motherboard non-USB mouse connectors also use the 6-pin mini-DIN connector and have the same pinout and signal descriptions as the keyboard connector; however, the data packets are incompatible. Therefore, you can easily plug a motherboard mouse (PS/2-style) into a mini-DIN keyboard connector or plug the mini-DIN keyboard connector into a motherboard mouse port. Neither one will work properly in this situation, though.


Caution I have also seen PCs with external power supplies that used the same standard DIN connectors to attach the keyboard and power supply. Although cross-connecting the mini-DIN connectors of a mouse and keyboard is a harmless annoyance, connecting a power supply to a keyboard socket can be disastrous.

USB keyboards use the Series A USB connector to attach to the USB port built into modern computers. For more information on USB, see Chapter 17, "I/O Interfaces from Serial and Parallel to IEEE-1394 and USB."

Keyboards with Special Features Several keyboards on the market have special features not found in standard designs. These additional features range from simple things, such as built-in calculators, clocks, and volume control, to more complicated features, such as integrated pointing devices, special character layouts, shapes, and even programmable keys.


Note In 1936, August Dvorak patented a simplified character layout called the Dvorak Simplified Keyboard (DSK). The Dvorak keyboard was designed to replace the common QWERTY layout used on nearly all keyboards available today. The Dvorak keyboard was approved as an ANSI standard in 1982 but has seen limited use. For a comparison between the Dvorak keyboard and the common QWERTY keyboard you most likely use, see "The Dvorak Keyboard" in the Technical Reference section of the DVD accompanying this book.

Ergonomic Keyboards A trend that began in the late 1990s is to change the shape of the keyboard instead of altering the character layout. This trend has resulted in several socalled ergonomic designs. The goal is to shape the keyboard to better fit the human hand. The most common of these designs splits the keyboard in the center, bending the sides outward. Some designs allow the angle between the sides to be adjusted, such as the now-discontinued Lexmark Select-Ease, the Goldtouch keyboard designed by Mark Goldstein (who also designed the Select-Ease), and the Kenisis Maxim split keyboards. Others, such as the Microsoft Natural keyboard series, PC Concepts Wave, and Cirque Smooth Cat, are fixed. These split or bent designs more easily conform to the hands' natural angles while typing than the standard keyboard. They can improve productivity and typing speed and help prevent repetitive strain injuries (RSI), such as carpal tunnel syndrome (tendon inflammation). Even more radical keyboard designs are available from some vendors, including models such as the 3-part Comfort and ErgoMagic keyboards, the Kinesis concave contoured keyboard, and others. A good source for highly ergonomic keyboards, pointing devices, and furniture is Ergonomic Resources (www.ergo-2000.com). Because of their novelty and trendy appeal, some ergonomic keyboards can be considerably more expensive than traditional designs, but for users with medical problems caused or exacerbated by improper positioning of the wrists at the keyboard, they can be an important remedy to a serious problem. General users, however, are highly resistant to change, and these designs have yet to significantly displace the standard keyboard layout. If you don't want to spend big bucks on the more radical ergonomic keyboards but want to give yourself at least limited protection from RSI, consider keyboards with a built-in wrist rest or add a gel-based wrist rest to your current keyboard. These provide hand support without making you learn a modified or brand-new keyboard layout.

USB Keyboards with Hubs Some of the latest USB keyboards feature a built-in USB hub designed to add two or more USB ports to your system. Even though this sounds like a good idea,


keep in mind that a keyboard-based hub won't provide additional power to the USB connectors. Powered hubs work better with a wider variety of devices than unpowered hubs do. I wouldn't choose a particular model based solely on this feature, although if your keyboard has it and your devices work well when plugged into it, that's great. I'd recommend that you use this type of keyboard with your USB mouse or other devices that don't require much power. Buspowered devices such as scanners and Webcams should be connected to a selfpowered hub or directly to the USB ports built in to the computer. USB keyboards and mice correspond to the USB 1.1 standard but can also be connected to the faster USB 2.0 ports on the latest systems.

Multimedia and Web-Enabled Keyboards As I discussed earlier in this chapter, many keyboards sold at retail and bundled with systems today feature fixed-purpose or programmable hotkeys that can launch Web browsers, run the Microsoft Media Player, adjust the volume on the speakers, change tracks on the CD player, and so forth. You need Windows 98 or later to use these hotkeys; Windows Me, Windows 2000, and Windows XP add additional support for these keyboards. For the best results, you should download the latest drivers for your keyboard and version of Windows from the keyboard vendor's Web site.


Keyboard Troubleshooting and Repair Keyboard errors are usually caused by two simple problems. Other more difficult, intermittent problems can arise, but they are much less common. The most frequent problems are as follows: Defective cables Stuck keys Defective cables are easy to spot if the failure is not intermittent. If the keyboard stops working altogether or every keystroke results in an error or incorrect character, the cable is likely the culprit. Troubleshooting is simple, especially if you have a spare cable on hand. Simply replace the suspected cable with one from a known, working keyboard to verify whether the problem still exists. If it does, the problem must be elsewhere. If you remove the cable from the keyboard, you can test it for continuity with a digital multimeter (DMM). DMMs that have an audible continuity tester built in make this procedure much easier to perform. To test each wire of the cable, insert the DMM's red pin into the keyboard connector and touch the DMM's black pin to the corresponding wire that attaches to the keyboard's circuit board. Wiggle the ends of the cable as you check each wire to ensure no intermittent connections exist. If you discover a problem with the continuity in one of the wires, replace the cable or the entire keyboard, whichever is cheaper. Because replacement keyboards are so inexpensive, it's almost always cheaper to replace the entire unit than to get a new cable, unless the keyboard is a deluxe model. For more information about using digital multimeters for testing hardware, see Chapter 23, "PC Diagnostics, Testing, and Maintenance." Many times you first discover a problem with a keyboard because the system has an error during the POST. Many systems use error codes in a 3xx numeric format to distinguish the keyboard. If you encounter any such errors during the POST, write them down. Some BIOS versions do not use cryptic numeric error codes; they simply state something such as the following: Keyboard stuck key failure This message is usually displayed by a system with a Phoenix BIOS if a key is stuck. Unfortunately, the message does not identify which key it is! If your system displays a 3xx (keyboard) error preceded by a two-digit


hexadecimal number, the number is the scan code of a failing or stuck keyswitch. Look up the scan code in the tables provided in the Technical Reference section on the DVD to determine which keyswitch is the culprit. By removing the keycap of the offending key and cleaning the switch, you often can solve the problem. For a simple test of the motherboard keyboard connector, you can check voltages on some of the pins. Using Figure 18.8, which was shown earlier in the chapter, as a guide, measure the voltages on various pins of the keyboard connector. To prevent possible damage to the system or keyboard, turn off the power before disconnecting the keyboard. Then, unplug the keyboard and turn the power back on. Make measurements between the ground pin and the other pins according to Table 18.2, shown earlier in the chapter. If the voltages are within these specifications, the motherboard keyboard circuitry is probably okay. If your measurements do not match these voltages, the motherboard might be defective. Otherwise, the keyboard cable or keyboard might be defective. If you suspect that the cable is the problem, the easiest thing to do is replace the keyboard cable with a known good one. If the system still does not work normally, you might have to replace the entire keyboard or the motherboard. In many newer systems, the motherboard's keyboard and mouse connectors are protected by a fuse that can be replaced. Look for any type of fuse on the motherboard in the vicinity of the keyboard or mouse connectors. Other systems might have a socketed keyboard controller chip (8042-type). In that case, you might be able to repair the motherboard keyboard circuit by replacing this chip. Because these chips have ROM code in them, you should get the replacement from the motherboard or BIOS manufacturer. If the motherboard uses a soldered keyboard controller chip or a chipset that integrates the keyboard controller with other I/O chips, you'll need to replace the motherboard. See the DVD included with this book for a listing of the standard POST and diagnostic keyboard error codes used by some systems.

Keyboard Disassembly Although disassembling a keyboard is possible, most likely you won't need or want to do that given the reasonable prices of keyboards. If you do want to disassemble your keyboard, see "Keyboard Disassembly" in the Technical Reference section of the DVD accompanying this book.

Cleaning a Keyboard


One of the best ways to keep a keyboard in top condition is periodic cleaning. As preventive maintenance, you should vacuum the keyboard weekly, or at least monthly. When vacuuming, you should use a soft brush attachment; this will help dislodge the dust. Also note that many keyboards have keycaps that can come off easily. Be careful when vacuuming; otherwise, you'll have to dig them out of the vacuum cleaner. I recommend using a small, handheld vacuum cleaner made for cleaning computers and sewing machines; these have enough suction to get the job done with little risk of removing your keytops. You also can use canned compressed air to blow the dust and dirt out instead of using a vacuum. Before you dust a keyboard with the compressed air, turn the keyboard upside down so that the particles of dirt and dust collected inside can fall out. On all keyboards, each keycap is removable, which can be handy if a key sticks or acts erratically. For example, a common problem is a key that does not work every time you press it. This problem usually results from dirt collecting under the key. An excellent tool for removing keycaps on almost any keyboard is the Ushaped chip puller included in many computer tool kits. Simply slip the hooked ends of the tool under the keycap, squeeze the ends together to grip the underside of the keycap, and lift up. IBM sells a tool designed specifically for removing keycaps from its keyboards, but the chip puller works even better. After removing the cap, spray some compressed air into the space under the cap to dislodge the dirt. Then replace the cap and check the action of the key.


Caution When you remove the keycaps, be careful not to remove the spacebar on the original 83-key PC and the 84-key AT-type keyboards. This bar is difficult to reinstall. The newer 101-key units use a different wire support that can be removed and replaced much more easily.

When you remove the keycap on some keyboards, you are actually detaching the entire key from the keyswitch. Be careful during the removal or reassembly of the keyboard; otherwise, you'll break the switch. The classic IBM/Lexmark-type keyboards (now made by Unicomp) use a removable keycap that leaves the actual key in place, enabling you to clean under the keycap without the risk of breaking the switches. If your keyboard doesn't have removable keycaps, consider using cleaning wands with soft foam tips to clean beneath the keytops. Spills can be a problem, too. If you spill a soft drink or cup of coffee into a keyboard, you do not necessarily have a disaster. Many keyboards that use membrane switches are spill resistant. However, you should immediately (or as soon as possible) disconnect the keyboard and flush it out with distilled water. Partially disassemble the keyboard and use the water to wash the components. (See "Keyboard Disassembly" in the Technical Reference section of the DVD accompanying this book for disassembly instructions.) If the spilled liquid has dried, soak the keyboard in some of the water for a while. When you are sure the keyboard is clean, pour another gallon or so of distilled water over it and through the keyswitches to wash away any residual dirt. After the unit dries completely, it should be perfectly functional. You might be surprised to know that drenching your keyboard with water does not harm the components. Just make sure you use distilled water, which is free from residue or mineral content (bottled water is not distilled; the distinct taste of many bottled waters comes from the trace minerals they contain!). Also, make sure the keyboard is fully dry before you try to use it; otherwise, some of the components might short out.


Tip If spills or excessive dust or dirt are expected because of the environment or conditions in which the PC is used, several companies make thin membrane skins that mold over the top of the keyboard, protecting it from liquids, dust, and other contaminants. These skins are generally thin enough so that they don't interfere too much with the typing or action of the keys.


Keyboard Recommendations In most cases, replacing a keyboard is cheaper or more cost effective than repairing it. This is especially true if the keyboard has an internal malfunction or if one of the keyswitches is defective. Replacement parts for keyboards are almost impossible to procure, and installing any repair part is usually difficult. In addition, many of the keyboards supplied with lower-cost PCs leave much to be desired. They often have a mushy feel, with little or no tactile feedback. A poor keyboard can make using a system a frustrating experience, especially if you are a touch typist. For all these reasons, it is often a good idea to replace an existing keyboard with something better. Perhaps the highest-quality keyboards in the entire computer industry are those made by IBM, or, more accurately today, Unicomp. Unicomp maintains an extensive selection of more than 1,400 Lexmark and IBM keyboard models and continues to develop and sell a wide variety of traditional and customized models, including keyboards that match the school colors of several universities.


Note See the section "Replacement Keyboards" in Chapter 17 of Upgrading and Repairing PCs, 11th Edition on this book's DVD for a listing of IBM keyboard and cable part numbers.

Some of the classic-design IBM keyboards are available in the retail market under either the IBM or IBM Options brand name. Items under the IBM Options program are sold direct by IBM's Web site or through normal retail channels, such as CompUSA and Computer Discount Warehouse (CDW). These items are also priced much more cheaply than items purchased as spare parts. They include a full warranty and are sold as complete packages, including cables. Table 18.3 lists some of the IBM Options keyboards and part numbers; even though the IBM Web site no longer offers these models, they can be purchased from various online retailers. Models marked with an * are also available from Unicomp. Table 18.3. IBM Options Keyboards (Sold Retail) Description

Part Number

IBM Enhanced keyboard (cable w/ DIN plug)

92G7454*

IBM Enhanced keyboard (cable w/ mini-DIN plug)

92G7453*

IBM Enhanced keyboard, built-in Trackball (cable w/ DIN plug)

92G7456*

IBM Enhanced keyboard, built-in Trackball (cable w/ mini-DIN plug)

92G7455*

IBM Enhanced keyboard, integrated TrackPoint II (cables w/ mini-DIN plugs)

92G7461*

IBM TrackPoint IV keyboard, Black

01K1260

IBM TrackPoint IV keyboard, White

01K1259

IBM TrackPoint USB Space Saver keyboard (black)

22P5150

IBM USB Keyboard with two-port hub (black)

10K3849

IBM Rapid Access III (black)

22P5185

Keep in mind, though, that because IBM spun off its keyboard business some years ago, many recent and current IBM-labeled keyboards no longer have the distinct feel, quality, or durability found in the older models. Ironically, one of the best ways to get an "IBM" keyboard is to buy the model with the features you want from Unicomp, most of whose keyboards still use the capacitive buckling spring technology IBM originally made famous.


The extremely positive tactile feedback of the IBM/Lexmark/Unicomp design is also a benchmark for the rest of the industry. Although keyboard feel is an issue of personal preference, I have never used a keyboard that feels better than the IBM/Lexmark/Unicomp designs. I now equip every system I use with a Unicomp keyboard, including the many non-IBM systems I use. You can purchase these keyboards directly from Unicomp at very reasonable prices. Many models are available, including some with a built-in trackball or even the revolutionary TrackPoint pointing device. (TrackPoint refers to a small stick mounted between the G, H, and B keys.) This device was first featured on the IBM ThinkPad laptop systems, although the keyboards are now sold for use on other manufacturers' PCs. The technology is being licensed to many other manufacturers, including Toshiba. Other manufacturers of high-quality keyboards that are similar in feel to the IBM/Lexmark/Unicomp units are Alps, Lite-On, NMB Technologies, and the revived Northgate designs sold under the Avant Prime and Avant Stellar names by Creative Vision Technologies. These keyboards have excellent tactile feedback, with a positive click sound. They are my second choice, after a Unicomp unit.


Pointing Devices The mouse was invented in 1964 by Douglas Englebart, who at the time was working at the Stanford Research Institute (SRI), a think tank sponsored by Stanford University. The mouse was officially called an X-Y Position Indicator for a Display System. Xerox later applied the mouse to its revolutionary Alto computer system in 1973. At the time, unfortunately, these systems were experimental and used purely for research. In 1979, several people from Appleâ&#x20AC;&#x2039;including Steve Jobsâ&#x20AC;&#x2039;were invited to see the Alto and the software that ran the system. Steve Jobs was blown away by what he saw as the future of computing, which included the use of the mouse as a pointing device and the graphical user interface (GUI) it operated. Apple promptly incorporated these features into what was to become the Lisa computer and lured away 15â&#x20AC;&#x2039;20 Xerox scientists to work on the Apple system. Although Xerox released the Star 8010 computer that used this technology in 1981, it was expensive, poorly marketed, and perhaps way ahead of its time. Apple released the Lisa computer, its first system that used the mouse, in 1983. It was not a runaway success, largely because of its $10,000 list price, but by then Jobs already had Apple working on the low-cost successor to the Lisa: the Macintosh. The Apple Macintosh was introduced in 1984. Although it was not an immediate hit, the Macintosh has grown in popularity since that time. Many credit the Macintosh with inventing the mouse and GUI, but as you can see, this technology was actually borrowed from others, including SRI and Xerox. Certainly the Macintosh, and now Microsoft Windows and OS/2, have gone on to popularize this interface and bring it to the legion of Intel-based PC systems. Although the mouse did not catch on quickly in the PC marketplace, today the GUIs for PC systems, such as Windows, practically demand the use of a mouse. Therefore, virtually every new system sold at retail comes with a mouse. And, because the mice packaged with retail systems are seldom high-quality or up-todate designs, sooner or later most users are in the market for a better mouse or compatible pointing device. Mice come in many shapes and sizes from many manufacturers. Some have taken the standard mouse design and turned it upside down, creating the trackball. In the trackball devices, you move the ball with your hand directly rather than moving the unit itself. Trackballs were originally found on arcade video games, such as Missile Command, but have become popular with users who have limited desk space. In most cases, the dedicated trackballs have a much larger ball than would be found on a standard mouse. Other than the orientation and perhaps the size of the ball, a trackball is identical to a mouse in design, basic function, and


electrical interface. Like many recent mice, trackballs often come in ergonomic designs, and the more recent models even use the same optical tracking mechanisms used by the latest Microsoft and Logitech mice. The largest manufacturers of mice are Microsoft and Logitech; these two companies provide designs that inspire the rest of the industry and each other and are popular OEM choices as well as retail brands. Even though mice can come in different varieties, their actual use and care differ very little. The standard mouse consists of several components: A housing that you hold in your hand and move around on your desktop A method of transmitting movement to the system: either ball/roller or optical sensors Buttons (two or more, and often a wheel or toggle switch) for making selections An interface for connecting the mouse to the system; conventional mice use a wire and connector, whereas wireless mice use a radio-frequency or infrared transceiver in both the mouse and a separate unit connected to the computer to interface the mouse to the computer The housing, which is made of plastic, consists of very few moving parts. On top of the housing, where your fingers normally rest, are buttons. There might be any number of buttons, but in the PC world, two is the standard. If your mouse has additional buttons or a wheel, specialized driver software provided by the mouse vendor is required for them to operate to their full potential. Although the latest versions of Windows support scrolling mice, other features supported by the vendor still require installing the vendor's own mouse driver software.

Ball-Type Mice The bottom of the mouse housing is where the detection mechanisms or electronics are located. On traditional mice, the bottom of the housing contains a small, rubber ball that rolls as you move the mouse across the tabletop. The movements of this rubber ball are translated into electrical signals transmitted to the computer across the cable. Internally, a ball-driven mouse is very simple, too. The ball usually rests against two rollers: one for translating the x-axis movement and the other for translating the y-axis movement. These rollers are typically connected to small disks with


shutters that alternately block and allow the passage of light. Small optical sensors detect movement of the wheels by watching an internal infrared light blink on and off as the shutter wheel rotates and "chops" the light. These blinks are translated into movement along the axes. This type of setup, called an optomechanical mechanism, is still the most popular type of mouse mechanism (see Figure 18.9), although optical mice are gaining in popularity. Figure 18.10 shows a PS/2 mouse connector.

Figure 18.9. Typical opto-mechanical mouse mechanism.

Figure 18.10. Typical PS/2-type mouse connector.

Optical Mice


The other major method of motion detection is optical. Some of the early mice made by Mouse Systems and a few other vendors used a sensor that required a special grid-marked pad. Although these mice were very accurate, the need to use them with a pad caused them to fall out of favor. Microsoft's IntelliMouse Explorer pioneered the return of optical mice. The IntelliMouse Explorer and the other new-style optical mice from Logitech and other vendors use optical technology to detect movement, and they have no moving parts of their own (except for the scroll wheel and buttons on top). Today's optical mice need no pad; they can work on virtually any surface. This is done by upgrading the optical sensor from the simple type used in older optical mice to a more advanced CCD (charge coupled device). This essentially is a crude version of a video camera sensor that detects movement by seeing the surface move under the mouse. An LED is used to provide light for the sensor. The IntelliMouse Explorer revolutionized the mouse industry; first Logitech, then virtually all other mouse makers, including both retail and OEM suppliers, have moved to optical mice for most of their product lines, offering a wide variety of optical mice in most price ranges. Figure 18.11 shows the essential features of a typical optical mouse.

Figure 18.11. The bottom of the Logitech iFeel optical mouse.

Their versatility and low maintenance (not to mention that neat red or blue glow out the sides!) make optical mice an attractive choice, and the variety of models available from both vendors means you can have the latest optical technology for about the price of a good ball-type mouse. Figure 18.12 shows the interior of a typical optical mouse.


Figure 18.12. The LED inside an optical mouse illuminates the surface by blinking many times per second. The light is reflected from the mousing surface back to the sensor, which converts the information into digital form and sends it to the computer.

All optical mice have a resolution of at least 400dpi and at least one sensor. However, for better performance, some optical mice have improved on these basic features, as listed in Table 18.4. Table 18.4. Enhanced Optical Mouse Features Feature

Benefit

Example Product

800dpi optical resolution

Improves accuracy for mouse positioning

Logitech MX series, Microsoft IntelliMouse Explorer and Optical series

Larger sensor size

Improves tracking on surfaces with repetitive patterns, such as wood desktops

Logitech MX series

Faster tracking speed

Able to keep up with fast hand movements, such as when playing games

Microsoft IntelliMouse Explorer and Optical series

Dual sensors

Faster speed and accuracy, especially for gaming

Logitech MouseMan Dual Optical

Optical mice, similar to traditional ball-type mice, are available as corded or cordless models. Cordless ball-type mice are usually much larger than ordinary mice because of the need to find room for both the bulky ball mechanism and batteries, but cordless optical mice are about the same size as high-end corded mice. The cable can be any length, but it is typically between 4 and 6 feet long. Mice are also available in a cordless design, which uses either infrared or RF transceivers to replace the cable. A receiver is plugged into the mouse port, while the battery-powered mouse contains a compatible transmitter.


Tip If you have a choice on the length of cable to buy, get a longer one. This allows easier placement of the mouse in relation to your computer. Extension cables can be used if necessary.

After the mouse is connected to your computer, it communicates with your system through the use of a device driver, which can be loaded explicitly or built into the operating system software. For example, no separate drivers are necessary to use a mouse with Windows or OS/2, but using the mouse with most DOS-based programs requires a separate driver to be loaded from the CONFIG.SYS or AUTOEXEC.BAT file. Regardless of whether it is built in, the driver translates the electrical signals sent from the mouse into positional information and indicates the status of the buttons. The standard mouse drivers in Windows are designed for the traditional twobutton mouse or scroll mouse (in Windows Me/2000/XP or later), but increasing numbers of mice feature additional buttons, toggles, or wheels to make them more useful. These additional features require special mouse driver software supplied by the manufacturer.

Pointing Device Interface Types The connector used to attach your mouse to the system depends on the type of interface you are using. Three main interfaces are used for mouse connections, with a fourth option you also occasionally might encounter. Mice are most commonly connected to your computer through the following three interfaces: Serial interface Dedicated motherboard (PS/2) mouse port USB port

Serial A popular method of connecting a mouse to older PCs is through the standard serial interface. As with other serial devices, the connector on the end of the mouse cable is typically a 9-pin male connector; some very old mice used a 25pin male connector. Only a couple of pins in the DB-9 or DB-25 connector are used for communications between the mouse and the device driver, but the mouse connector typically has all 9 or 25 pins present.


Because most older PCs come with two serial ports, a serial mouse can be plugged into either COM1 or COM2. The device driver, when initializing, searches the ports to determine to which one the mouse is connected. Some mouse drivers can't function if the serial port is set to COM3 or COM4, but most newer drivers can work with any COM port 1â&#x20AC;&#x2039;4. Because a serial mouse does not connect to the system directly, it does not use system resources by itself. Instead, the resources are those used by the serial port to which it is connected. For example, if you have a mouse connected to COM2, and if COM2 is using the default IRQ and I/O port address range, both the serial port and the mouse connected to it use IRQ3 and I/O port addresses 2F8hâ&#x20AC;&#x2039;2FFh. See "Serial Ports," p. 961.

Motherboard Mouse Port (PS/2) Most computers include a dedicated mouse port built into the motherboard. This practice was introduced by IBM with the PS/2 systems in 1987, so this interface is often referred to as a PS/2 mouse interface. This term does not imply that such a mouse can work only with a PS/2; instead, it means the mouse can connect to any system that has a dedicated mouse port on the motherboard. From a hardware perspective, a motherboard mouse connector is usually exactly the same as the mini-DIN connector used for newer keyboards. In fact, the motherboard mouse port is connected to the 8042-type keyboard controller found on the motherboard. All the PS/2 computers include mini-DIN keyboard and mouse port connectors on the back. Most computers based on the semiproprietary LPX motherboards and all ATX-series motherboards use these same connectors for space reasons. Most Baby-AT motherboards have a pin-header type connector for the mouse port because most standard cases do not have a provision for the miniDIN mouse connector. If that is the case, an adapter cable is usually supplied with the system. This cable adapts the pin-header connector on the motherboard to the standard mini-DIN type connector used for the motherboard mouse.


Caution As mentioned in the "Keyboard/Mouse Interface Connectors" section earlier in this chapter, the mini-DIN sockets used for both keyboard and mouse connections on many systems are physically and electrically interchangeable, but the data packets they carry are not. Be sure to plug each device into the correct socket; otherwise, neither will function correctly. Don't panic if you mix them up, though. They are electrically identical to each other, so you can't damage the ports or the devices.

Connecting a mouse to the built-in mouse port is the best method of connection on systems that don't have USB ports because you do not sacrifice any of the system's interface slots or any serial ports, and the performance is not limited by the serial port circuitry. The standard resource usage for a motherboard (or PS/2) mouse port is IRQ12, as well as I/O port addresses 60h and 64h. Because the motherboard mouse port uses the 8042-type keyboard controller chip, the port addresses are those of this chip. IRQ12 is an interrupt that is usually free on most systems, but if you use a USB mouse, you can probably disable the mouse port to make IRQ12 available for use by another device.

Hybrid Mice Hybrid mice are those designed to plug into two types of ports. Although a few low-cost mice sold at retail are designed to plug into either the serial port or the PS/2 port, most mice on the retail market today are designed to plug into either the PS/2 port or the USB port. These combination mice are more flexible than the mice typically bundled with systems, which are designed to work only with the PS/2 or USB port to which they attach. Circuitry in a hybrid mouse automatically detects the type of port to which it is connected and configures the mouse automatically. Serial-PS/2 hybrid mice usually come with a mini-DIN connector on the end of their cable and an adapter that converts the mini-DIN to a 9- or 25-pin serial port connector, although the reverse is sometimes true on early examples of these mice. PS/2-USB mice usually come with the USB connector on the end of their cable and include a miniDIN (PS/2) adapter, as shown in Figure 18.13.

Figure 18.13. A typical USB mouse with a PS/2 adapter.


Sometimes people use adapters to try to connect a serial mouse to a motherboard mouse port or a motherboard mouse to a serial port. If this does not work, it is not the fault of the adapter. If the mouse does not explicitly state that it is both a serial and a PS/2-type mouse, it works only on the single interface for which it was designed. Most of the time, you find the designation for which type of mouse you have printed on its underside. A safe rule of thumb to follow is if the mouse didn't come with an adapter or came bundled with a system, it probably won't work with an adapter.

USB The extremely flexible USB port has become the most popular port to use for mice as well as keyboards and other I/O devices. Compared to the other interfaces, USB mice (and other USB pointing devices such as trackballs) have the following advantages: USB mice move much more smoothly than the traditional PS/2-type. This is because the frequency with which the mouse reports its position is much higher. A typical PS/2 mouse has a reporting rate of about 40Hz, whereas an average USB-wired mouse has a reporting rate of 125Hz (most USB wireless mice have a reporting rate of 40Hzâ&#x20AC;&#x2039;50Hz). Several utilities are available to test and adjust the mouse frequency. Mice with the most advanced features are usually made especially for the USB port. One example is the Logitech iFeel mouse, the first mouse with an optical sensor plus force feedback. It vibrates gently as you move the mouse over clickable buttons on Web pages, software menus, and the Windows desktop, and it's made especially for USB.


USB mice and pointing devices, similar to all other USB devices, are hotswappable. If you like to use a trackball and your computing partners prefer mice, you can just lean over and unplug the other users' pointing device and plug in your own, or move it from PC to PC. You can't do that with the other port types. USB mice can be attached to a USB hub, such as the hubs contained in some USB keyboards, as well as standalone hubs. Using a hub makes attaching and removing your mouse easy without crawling around on the floor to reach the back of the computer. Many computers now have front-mounted USB ports, letting you easily attach and remove a USB mouse without the use of an external hub. Although the early USB mice were decidedly on the premium end of the price scale, low-cost USB mice are now available for less than $20. If you want to use a USB mouse at an MS-DOS prompt, in Windows safe mode, or in some other environment outside of normal Windows 98 or later, make sure that USB Legacy mode is enabled in your PC's BIOS, as discussed earlier in this chapter. Legacy mode enables non-USB-aware systems to recognize a USB keyboard and mouse. A fourth type of connection, the bus mouse (referred to by Microsoft as the Inport mouse), used a dedicated adapter card and is now obsolete. For more information about the bus mouse, see Chapter 17 of Upgrading and Repairing PCs, 11th Edition, which is included on the DVD-ROM supplied with this book.

Mouse Troubleshooting If you are experiencing problems with your mouse, you need to look in only two general placesâ&#x20AC;&#x2039;hardware or software. Because mice are basically simple devices, looking at the hardware takes very little time. Detecting and correcting software problems can take a bit longer, however. To troubleshoot wireless mice, see "Troubleshooting Wireless Input Devices," later in this chapter.

Cleaning Your Mouse If you notice that the mouse pointer moves across the screen in a jerky fashion, it might be time to clean your mouse. For a mouse with a roller-ball, this jerkiness is caused when dirt and dust become trapped around the mouse's ball-and-roller assembly, thereby restricting its free movement.


From a hardware perspective, the mouse is a simple device, so cleaning it is easy. The first step is to turn the mouse housing over so that you can see the ball on the bottom. Notice that surrounding the ball is an access panel you can open. Sometimes instructions indicate how the panel is to be opened. (Some off-brand mice might require you to remove some screws to get at the roller ball.) Remove the panel to see more of the roller ball and the socket in which it rests. If you turn the mouse back over, the rubber roller ball should fall into your hand. Take a look at the ball. It might be gray or black, but it should have no visible dirt or other contamination. If it does, wash it in soapy water or a mild solvent, such as contact lens cleaner solution or alcohol, and dry it off. Now take a look at the socket in which the roller ball normally rests. You will see two or three small wheels or bars against which the ball usually rolls. If you see dust or dirt on or around these wheels or bars, you need to clean them. The best way is to use a compressed air duster, which can blow out any dust or dirt. You also can use some electrical contact cleaner to clean the rollers. Remember, any remaining dirt or dust impedes the movement of the roller ball and results in the mouse not working as it should. Put the mouse back together by inserting the roller ball into the socket and then securely attaching the cover panel. The mouse should look just as it did before you removed the panel, except that it will be noticeably cleaner. One of the major advantages of the new breed of optical mice is the lack of moving parts. Just wipe away dust from the optical sensor, and that's all the cleaning an optical mouse needs.

Interrupt Conflicts Interrupts are internal signals used by your computer to indicate when something needs to happen. With a mouse, an interrupt is used whenever the mouse has information to send to the mouse driver. If a conflict occurs and the same interrupt used by the mouse is used by a different device, the mouse will not work properlyâ&#x20AC;&#x2039;if at all. Interrupt conflicts caused by mice can occur when a serial or PS/2 mouse is used, but not when a USB mouse is used. Mouse ports built in to modern motherboards are almost always set to IRQ12. If your system has a motherboard mouse port, be sure you don't set any other adapter cards to IRQ12; otherwise, a conflict will result. If you are using a serial mouse, interrupt conflicts typically occur if you add third and fourth serial ports, using either an expansion card or internal serial device, such as a modem. This happens because in ISA bus systems, the odd-numbered


serial ports (1 and 3) usually are configured to use the same interrupts as the even-numbered ports (2 and 4) are; IRQ4 is shared by default between COM1 and COM3, and IRQ 2 is shared by default between COM2 and COM4. Therefore, if your mouse is connected to COM2 and an internal modem uses COM4, they both might use the same interrupt, and you can't use them at the same time. Because the mouse generates interrupts only when it is moved, you might find that the modem functions properly until you touch the mouse, at which point the modem is disconnected. Another example is when your system will run properly until you try to go online with your modem; then the conflict usually locks up the system. You might be able to use the mouse and modem at the same time by moving one of them to a different serial port. For instance, if your mouse uses COM1 and the modem still uses COM4, you can use them both simultaneously because odd and even ports use different interrupts. The best way around these interrupt conflicts is to make sure no two devices use the same interrupt. Serial port adapters are available for adding COM3 and COM4 serial ports that do not share the interrupts used by COM1 and COM2. These boards enable the new COM ports to use other normally available interrupts, such as IRQs 10, 11, 12, 15, and 5. I never recommend configuring a system with shared ISA interrupts; it is a sure way to run into problems later. However, interrupts used by PCI boards can be shared if you use Windows 95 OSR 2.x, Windows 98, Windows Me, Windows 2000, or Windows XP with recent chipsets that support a feature called IRQ steering. See "PCI Interrupts," p. 340.

If you suspect an interrupt problem with a bus-type mouse, you can use the Device Manager built into Windows (which is accessible from the System control panel). See Chapter 24, "File Systems and Data Recovery," p. 1299.

The Device Manager in Windows 9x/Me/2000/XP is part of the Plug and Play (PnP) software for the system, and it is usually 100% accurate on PnP hardware. Although some of these interrupt-reporting programs can have problems, most can easily identify the mouse IRQ if the mouse driver has been loaded. After the IRQ is identified, you might need to change the IRQ setting of the bus mouse


adapter or one or more other devices in your system so that everything works together properly. If your driver refuses to recognize the mouse at all, regardless of its type, try using a different mouse that you know works. Replacing a defective mouse with a known good one might be the only way to know whether the problem is indeed caused by a bad mouse. I have had problems in which a bad mouse caused the system to lock right as the driver loaded or when third-party diagnostics were being run on the system. If you use a DOS-based diagnostic, such as Microsoft MSD or AMIDIAG, and the system locks up during the mouse test, you have found a problem with either the mouse or the mouse port. Try replacing the mouse to see whether that helps. If it does not, you might need to replace the serial port or bus mouse adapter. If a motherboard-based mouse port goes bad, you can replace the entire motherboard​which is usually expensive​or you can just disable the motherboard mouse port via jumpers or the system BIOS setup program and install a serial mouse instead. This method enables you to continue using the system without having to replace the motherboard. On systems with Windows 98/Me/2000/XP, you also can switch to a USB mouse, using USB ports on your motherboard or by installing a PCI-based USB card​provided your system has a USB port.


Note To learn more about using the Microsoft MSD diagnostic program to test for mouse or mouse-port problems, see Chapters 17 and 25 of Upgrading and Repairing PCs, 11th Edition, which is included on the DVD-ROM packaged with this book.

Driver Software Most mice and other pointing devices in use today emulate a Microsoft mouse, enabling you to have basic two-button plus scrolling functions with current versions of Windows without loading any special drivers. However, if your mouse has additional buttons or other special features, you will need to install devicespecific drivers available from the mouse vendor. If you plan to use the mouse from a Windows 9x/Me command prompt or with DOS, you must load the driver manually. To learn more about this process, see "Mouse Driver Software" in the Technical Reference section of the DVD packaged with this book.

Scroll Wheels Late in 1996, Microsoft introduced the IntelliMouse, which differed from standard Microsoft mice by adding a small gray wheel between the mouse buttons. This was the first scrolling mouse, and since then, Logitech, IBM, and virtually all other mouse vendors have made scroll wheels or similar devices standard across almost all models, including OEM mice bundled with computer systems. The wheel has two main functions. The primary function is to act as a scrolling device, enabling you to scroll through documents or Web pages by manipulating the wheel with your index finger. The wheel also functions as a third mouse button when you press it. Although three-button mice have been available for years from vendors such as Logitech, the scrolling function provided a real breakthrough. No longer do you have to move the mouse pointer to click the scrollbar on the right side of your screen or take your hand off the mouse to use the arrow keys on the keyboard. You just push or pull on the wheel. This is a major convenience, especially when browsing Web pages or working with word processing documents or spreadsheets. Also, unlike three-button mice from other vendors, the IntelliMouse's wheelbutton doesn't seem to get in the way and you are less likely to click it by mistake. Although it took a while for software vendors to support the wheel, improvements in application software and Windows support allow today's wheel mice to be fully useful with almost any recent or current Windows program.


Each vendor's mouse driver software offers unique features to enhance the basic operation of the mouse. For example, Logitech's MouseWare 9.7 driver enables you to select many uses for all three mouse buttons (the scroll wheel is treated as a third mouse button), as well as provides various options for how to scroll with each wheel click (three lines, six lines, or one screen). Microsoft's IntelliMouse driver offers a feature called ClickLock, which allows you to drag items without holding down the primary mouse button. In addition, it offers a Universal Scroll feature that adds scrolling mouse support to applications that lack such support. To get the most from whatever scrolling or other advanced-feature mouse you have, be sure you periodically download and install new mouse drivers. Instead of the wheel used by Microsoft and Logitech, IBM and other mouse vendors frequently use various types of buttons for scrolling. Some inexpensive mice use a rocker switch, but the most elegant of the non-wheel alternatives is IBM's ScrollPoint Pro, which uses a pressure-sensitive scroll stick similar to the TrackPoint pointing device used on IBM's notebook computer and some PC keyboards made by IBM and Unicomp. The scrollpointer in the center of the mouse enables you to smoothly scroll through documents without having to lift your finger to roll the wheel, as you do on the Microsoft version, which makes it much easier and more convenient to use. Because no moving parts exist, the ScrollPoint is also more reliable.

TrackPoint II/III/IV On October 20, 1992, IBM introduced a revolutionary new pointing device called TrackPoint II as an integrated feature of its new ThinkPad 700 and 700C computers. Often referred to as a pointing stick device, TrackPoint II and its successors consist primarily of a small rubber cap that appears on the keyboard right above the B key, between the G and H keys. This was the first significant new pointing device since the mouse had been invented nearly 30 years earlier! This device occupies no space on a desk, does not have to be adjusted for lefthanded or right-handed use, has no moving parts to fail or become dirty, and (most importantly) does not require you to move your hands from the home row to use it. This is an absolute boon for touch typists. I was fortunate enough to meet the actual creator and designer of this device in early 1992 at the spring Comdex/Windows World in Chicago. He was in a small corner of the IBM booth showing off his custom-made keyboards with a small silicone rubber nub in the middle. In fact, the devices he had were hand-built prototypes installed in standard desktop keyboards, and he was there trying to get public reaction and feedback on his invention. I was invited to play with one of the keyboards, which was connected to a demonstration system. By pressing on the stick with my index finger, I could move the mouse pointer on the screen. The


stick itself did not move, so it was not a joystick. Instead, it had a silicone rubber cap that was connected to pressure transducers that measured the amount of force my finger was applying and the direction of the force and moved the mouse pointer accordingly. The harder I pressed, the faster the pointer moved. After playing around with it for just a few minutes, the movements became automaticâ&#x20AC;&#x2039;almost as though I could just think about where I wanted the pointer to go. The gentleman at the booth turned out to be Ted Selker, the primary inventor of the device. He and Joseph Rutledge created this integrated pointing device at the IBM T.J. Watson Research Center. When I asked him when such keyboards would become available, he could not answer. At the time, there were apparently no plans for production, and he was only trying to test user reaction to the device. Just over six months later, IBM announced the ThinkPad 700, which included this revolutionary deviceâ&#x20AC;&#x2039;then called the TrackPoint II Integrated Pointing Device. Since the original version came out, enhanced versions with greater control and sensitivity, called the TrackPoint III and IV, have become available.


Note The reason the device was called TrackPoint II is that IBM had previously been selling a convertible mouse/trackball device called the TrackPoint. No relationship exists between the original TrackPoint mouse/trackball, which has since been discontinued, and the TrackPoint II integrated device. Since the original TrackPoint II came out, improved versions known as TrackPoint III and TrackPoint IV have become available. In the interest of simplicity, I will refer to all the TrackPoint II, III, and successive devices as just TrackPoint.

In its final production form, the TrackPoint consists of a small, red, silicone rubber knob nestled between the G, H, and B keys on the keyboard. The primary and secondary mouse buttons are placed below the spacebar where you can easily reach them with your thumbs without taking your hands off the keyboard. IBM studies conducted by Selker found that the act of removing your hand from the keyboard to reach for a mouse and replacing the hand on the keyboard takes approximately 1.75 seconds. If you type 60wpm, that can equal nearly two lost words per minute, not including the time lost while you regain your train of thought. Almost all this time can be saved each time you use TrackPoint to move the pointer or make a selection (click or double-click). The combination of the buttons and the positioning knob also enable you to perform drag-and-drop functions easily. IBM's research also found that people can get up to 20% more work accomplished using the TrackPoint instead of a mouse, especially when the application involves a mix of typing and pointing activities, such as with word processing, spreadsheets, and other typical office applications. In usability tests with the TrackPoint III, IBM gave a group of desktop computer users a TrackPoint and a traditional mouse. After two weeks, 80% of the users had unplugged their mice and switched solely to the TrackPoint device. Selker is convinced (as am I) that the TrackPoint is the best pointing solution for both laptop and desktop systems. Another feature of the TrackPoint is that a standard mouse can be connected to the system at the same time to enable dual-pointer use. This setup not only enables a single person to use both devices, but also enables two people to use the TrackPoint and the mouse simultaneously to move the pointer on the screen. The first pointing device that moves (thus issuing a system interrupt) takes precedence and retains control over the mouse pointer on the screen until it completes its movement action. The second pointing device is automatically locked out until the primary device is stationary. This enables the use of both devices and prevents each one from interfering with the other. IBM has added various versions of the TrackPoint to its notebook computers, as well as to high-end keyboards sold under the IBM, Lexmark, and Unicomp names. Notebook computer makers, such as HP and Toshiba, also have licensed the TrackPoint device (Toshiba calls it Accupoint).


I have compared the TrackPoint device to other pointing devices for notebooks, such as the trackballs and even the capacitive touch pads, but nothing compares in terms of accuracy and controlâ&#x20AC;&#x2039;and, of course, you don't have to take your hands off the keyboard! Some notebook computer makers copied the TrackPoint instead of licensing it, but with poor results that include sluggish response to input and poor accuracy. One way of telling whether the TrackPoint device is licensed from IBM and uses the IBM technology is if it accepts IBM TrackPoint II/III/IV rubber caps. They have a square hole in them and will properly lock on to any of the licensed versions, such as those found in Toshiba systems. IBM has upgraded its pointing stick to the TrackPoint III and the current TrackPoint IV. Two main differences exist in the III/IV system, but the most obvious one is in the rubber cap. The IBM TrackPoint II and Toshiba Accupoint caps are made from silicone rubber, which is grippy and works well in most situations. However, if the user has greasy fingers, the textured surface of the rubber can absorb some of the grease and become slippery. Cleaning the cap (and the user's hands) solves the problem, but it can be annoying at times. The TrackPoint III/IV caps are made from a different type of rubber, which Selker calls "plastic sandpaper." This type of cap is much more grippy and does not require cleaning except for cosmetic purposes. I have used both types of caps and can say for certain that the TrackPoint III/IV cap is superior.


Note Because the Accupoint device used in the Toshiba notebooks is licensed from IBM, it uses the same hardware (a pressure transducer called a strain gauge) and takes the same physical caps. I ordered a set of the TrackPoint III caps and installed them on my Toshiba portable systems, which dramatically improved the grip. You can get these caps by ordering them from IBM Parts directly or from others who sell IBM parts, such as Compu-Lock (www.compu-lock.com). The cost is approximately $16 for a set of four "plastic sandpaper" red caps.

Replacing the cap is easyâ&#x20AC;&#x2039;grab the existing cap with your fingers and pull straight up; it pops right off. Then, push on the new red IBM TrackPoint III/IV cap in its place. You will thank me when you feel how you can grip the new IBM cap much more easily than you can grip the designs used by others. The other difference between the TrackPoint II and III/IV from IBM is in the control software. IBM added routines that implement a subtle technique Selker calls "negative inertia," which is marketed under the label QuickStop response. This software not only takes into account how far you push the pointer in any direction, but also how quickly you push or release it. Selker found that this improved software (and the sandpaper cap) enables people to make selections up to 8% faster. TrackPoint IV includes an extra scroll button, as well as the ability to press the TrackPoint nub to select as if using the left mouse button. These new features make the TrackPoint even better to use. The bottom line is that anyone who touch types should strongly consider only portable systems that include an IBM-licensed TrackPoint device (such as Toshiba). TrackPoints are far superior to other pointing devices, such as the touch pads, because the TrackPoint is faster to use (you don't have to take your hands off the keyboard's home row), easier to adapt to (especially for speedy touch typists), and far more precise. It takes some getting accustomed to, but the benefits are worth it. The benefits of the TrackPoint are not limited to portable systems, however. If you use a notebook computer with TrackPoint like I do, you can have the same features on your desktop keyboard. For desktop systems, I use a Lexmark keyboard with the IBM-licensed TrackPoint device built in. This makes for a more consistent interface between desktop and notebook use because I can use the same pointing device in both environments. You also can buy these keyboards directly from Unicomp (Unicomp keyboards are TrackPoint III compatible); IBM also offers TrackPoint IV in some of its high-end keyboards available at retail.

Mouse and Pointing Stick Alternatives


Because of Windows, many users spend at least as much time moving pointers around the screen as they do in typing, making pointing device choices very important. In addition to the mouse and the pointing stick choices discussed earlier in this chapter, several other popular pointing devices are available, including Track pads, such as the Cirque GlidePoint Trackballs from many vendors Upright mice, such as the 3M Renaissance Mouse All these devices are treated as mice by the operating system but offer radically different options for the user in terms of comfort. If you're not satisfied with a regular mouse and don't want to use an integrated pointing stick such as the TrackPoint II/III/IV, look into these options.

GlidePoint/Touch Pad Cirque originated the touch pad (also called a track pad) pointing device in 1994. Cirque refers to its technology as the GlidePoint and has licensed the technology to other vendors such as Alps Electric, which also uses the term Glidepoint for its touch pads. The GlidePoint uses a flat, square pad that senses finger position through body capacitance. This is similar to the capacitance-sensitive elevator button controls you sometimes encounter in office buildings or hotels. When it is used on a portable computer's keyboard, the touch pad is mounted below the spacebar, and it detects pressure applied by your thumbs or fingers. Transducers under the pad convert finger movement into pointer movement. Several laptop and notebook manufacturers have licensed this technology from Cirque and have incorporated it into their portable systems. Touch pads are also integrated into a number of mid-range to high-end keyboards from many vendors. When used on a desktop keyboard, touch pads are often offset to the right side of the keyboard's typing area. Touch pads feature mouse buttons, although the user also can tap or double-tap on the touch pad's surface to activate an onscreen button located under the touch pad's cursor. Dragging and dropping is accomplished without touching the touch pad's buttons; just move the cursor to the object to be dragged, press down on the pad, hold while moving the cursor to the drop point, and raise the finger to drop the object. Some recent models also feature additional hot buttons with functions similar to those on hot-button keyboards.


The primary use for touch pads has been for notebook computerâ&#x20AC;&#x2039; and desktop keyboardâ&#x20AC;&#x2039;integrated pointing devices, although Cirque and Alps have both sold standalone versions of the touch pad for use as a mouse alternative on desktop systems. Cirque's touch pads are now available at retail under the Fellowes brand name, as well as direct from the Cirque Web site. The Internet Touchpad (also sold by Fellowes) has enhanced software to support touch gestures, has programmable hot buttons, and includes other features to make Web surfing easier. Although it has gained wide acceptance, especially on portable computers, touch pad technology can have many drawbacks for some users. Operation of the device can be erratic, depending on skin resistance and moisture content. The biggest drawback is that to operate the touch pad, users must remove their hands from the home row on the keyboard, which dramatically slows their progress. In addition, the operation of the touch pad can be imprecise, depending on how pointy your finger or thumb is! On the other hand, if you're not a touch typist, removing your hands from the keyboard to operate the touch pad might be easier than using a TrackPoint. Even with their drawbacks, touch pad pointing devices are still vastly preferable to using a trackball or a cumbersome external mouse with portable systems. Unless you want to use a "real" mouse with a portable system, I recommend you sit down with portable computers that have both touch pad and TrackPoint pointing devices. Try them yourself for typing, file management, and simple graphics and see which type of integrated pointing device you prefer. I know what I like, but you might have different tastes.

Trackballs The first trackball I ever saw outside of an arcade was the Wico trackball, a perfect match for mid-1980s video games and computer games, such as Missile Command and others. It emulated the eight-position Atari 2600 analog joystick but was capable of much more flexibility. Unlike the mid-80s trackballs, today's trackballs are used primarily for business instead of gaming. Most trackballs use a mouse-style positioning mechanismâ&#x20AC;&#x2039;the differences being that the trackball is on the top or side of the case and is much larger than a mouse ball. The user moves the trackball rather than the input device case, but rollers or wheels inside most models translate the trackball's motion and move a cursor onscreen the same way that mouse rollers or wheels convert the mouse ball's motion into cursor movement. Trackballs come in a variety of forms, including ergonomic models shaped to fit the (right) hand, ambidextrous models suitable for both lefties and right-handers,


optical models that use the same optical sensors found in the latest mice in place of wheels and rollers, and multibutton monsters that look as if they're the result of an encounter with a remote control. Because they are larger than mice, trackballs lend themselves well to the extra electronics and battery power needed for wireless use. Logitech offers several wireless trackball models that use radio-frequency transceivers; for details of how this technology works, see the section "Wireless Input Devices," later in this chapter. Trackballs use the same drivers and connectors as conventional mice. For basic operations, the operating-system-supplied drivers will work, but you should use the latest version of the vendor-supplied drivers to achieve maximum performance with recent models.

Cleaning and Troubleshooting Trackballs Trackball troubleshooting is similar to mouse troubleshooting. For issues other than cleaning the trackball, see the section "Mouse Troubleshooting," earlier in this chapter. Because trackballs are moved by the user's hand rather than by rolling against a tabletop or desktop, they don't need to be cleaned as often as mouse mechanisms do. However, occasional cleaning is recommended, especially with trackballs that use roller movement-detection mechanisms. If the trackball pointer won't move, skips, or drags when you move the trackball, try cleaning the trackball mechanism. Trackballs can be held into place by a retaining ring, an ejection tab, or simply by gravity. Check the vendor's Web site for detailed cleaning instructions if your trackball didn't come with such instructions. Swabs and isopropyl alcohol are typically used to clean the trackball and rollers or bearings; see the trackball's instructions for details.

3M's Ergonomic Mouse Many PC users who grew up using joysticks on the older video games experienced some "interface shock" when they turned in their joysticks for mice. And even long-time mouse users nursing sore arms and elbows have wondered whether the mouse was really as "ergonomic" as it is sometimes claims to be. 3M's solution, developed late in 2000, is to keep the traditional ball-type mouse positioning mechanism but change the user interface away from the hockey


puck/soap bar design used for many years to a slanted handle that resembles a joystick (see Figure 18.14). 3M's Ergonomic Mouse (originally called the Renaissance Mouse) is available in two hand sizes and attaches to either the PS/2 port or USB port (serial ports are not supported). The single button on the top of the handle is a rocker switch; push on the left side to left-click and on the right side to right-click. The front handgrip provides scrolling support when the special Ergonomic Mouse driver software is installed.

Figure 18.14. The 3M Ergonomic Mouse combines an ergonomic shape with a standard mouse-movement mechanism.

The Ergonomic Mouse enables the user to hold the pointing device with a "handshake"-style hand and arm position. 3M's Web site provides detailed ergonomic information to encourage the proper use of the Ergonomic Mouse, which comes with software to support scrolling and other advanced functions. It's available in various colors and separate models for Windows-based PCs and Macs.


Input Devices for Gaming Originally, game players on the PC used the arrow keys or letter keys on the keyboard to play all types of games; I remember putting Larry Bird and Dr. J through their paces in the original version of Electronic Arts One-on-One basketball with an 84-key keyboard! As you can imagine, this limited the number and type of games that could be played on the PC.

Analog Joysticks and the Game Port As video standards improved, making games more realistic, input devices made especially for game play also became more and more popular. The first joysticks made for the IBM PC were similar to joysticks made for its early rival, the Apple II series. Both the IBM and Apple II joysticks were analog devices that lacked much of the positive feedback game players were accustomed to from the Atari 2600, Commodore 64, or arcade joysticks. These joysticks also required frequent recalibration to work properly and were far from satisfactory to hardcore game players. Also, these devices required their own connectorâ&#x20AC;&#x2039;the 15-pin game port. The game port found its way onto many sound cards as well as onto multi-I/O cards made for ISA and VL-Bus systems. Even though joysticks began to add better features, including spring action, video game-style gamepads, and flight control options, the analog nature and slow speed of the gameport began to restrict performance as CPU speeds climbed above 200MHz and high-speed AGP and PCI video cards made ultra-realistic flying, driving, and fighting simulators possible. USB controllers offer the additional speed necessary for more sophisticated gamers.

USB Ports for Gaming The very versatile USB port has become the preferred connector for all types of gaming controllers, including joysticks, gamepads, and steering wheels. Instead of making a single inadequate joystick work for all types of games, users can now interchange controllers using the hot-swap benefits of USB and use the best controller for each type of game. Although a few low-end game controllers still on the market can connect to either the venerable game port or the USB port, serious gamers want USB because of its higher speed, better support for force feedback (which shakes the game controller realistically to match the action onscreen), and tilting (tilt the gamepad and the


onscreen action responds). As with USB mice, your USB-connected gaming controllers are only as good as their software drivers. Be sure to install the latest software available to keep up with the latest games.

Compatibility Concerns If you play a lot of older games designed in the heyday of the 15-pin gameport, consider keeping a gameport-type controller. Even though the vendors of USB game controllers strive to make the USB port emulate a game port for use with older games, some older games can't be fooled. If you have problems using a USB game controller with a specific game, check the game's Web site for patches, as well as your game controller's Web site for tips and workarounds.

Programmable Game Controllers Some of the latest controllers are programmable, enabling you to develop profiles that provide you with features such as one-button operation of keyboard shortcuts or other game commands and adjustments to range of motion. To use these profiles, be sure you install the driver software supplied with your game controller and change to the profile you have customized for a particular program whenever you start that program.

Choosing the Best Game Controller If you're a serious gamer who likes various types of games, you might need about as many game controllers as a golfer needs clubs. An increasing number of specialized controllers are on the market, and even familiar favorites like joysticks and steering wheels are available in new variations. Table 18.5 provides a quick reference to the many types of game controllers available today for PCs. Table 18.5. Game Controller Overview Controller Suitable For Type

Joystick

Desirable Features

Action games, Force feedback; programmable actions; including flight rudder panels and shooting

Sample Products

Microsoft Sidewinder Precision 2; Logitech WingMan Extreme Digital 3D; Thrustmaster Top Gun AfterBurner II Joystick


Microsoft Precision Racing Wheel; Logitech MOMO Force and Speed Force; Thrustmaster Force Feedback Racing WheelPC

Steering wheel

Driving games

Force feedback; programmable actions; foot pedals

Game pad

Driving and sports games

Force feedback; motion sensing; Microsoft Sidewinder Freestyle Pro; programmable buttons; mouse and keyboard Thrustmaster FireStorm Dual Power emulation Gamepad; Logitech WingMan Rumblepad

Voice chat Most and multiplayer control games

Voice communication with other players in a game (such as conferences with members of Microsoft Game Voice your team or taunting an opponent)

Strategy controller

Programmable buttons; shortcuts; record actions as you play; can control camera movement in many 3D strategy titles

Strategy games

Microsoft Sidewinder Strategic Commander


Wireless Input Devices For several years, many manufacturers have offered cordless versions of mice and keyboards. In most cases, these devices have used either infrared or short-range radio transceivers to attach to standard serial or PS/2 ports, with matching transceivers located in the mouse or keyboard. Wireless input devices are designed to be easier to use in cramped home-office environments and where a large-screen TV/monitor device is used for home entertainment and computing. Many manufacturers, including Microsoft, Logitech, and second-tier vendors, offer bundled kits that include a wireless keyboard and mouse which share a transceiver. Because many of these keyboards and mice have the latest features, including programmable keys, multimedia and Internet-access keys, and optical sensors, these wireless combos are often the top-of-the-line products from a given vendor and are often less expensive than buying the keyboard and mouse separately.

How Wireless Input Devices Work The three major technologies used by wireless input devices are as follows: Infrared (IR) Proprietary radio frequency Bluetooth All three technologies use a transceiver connected to the PS/2 or USB ports on the computer. Because many wireless transceivers are designed for use with a mouse and keyboard, PS/2-compatible versions have two cablesâ&#x20AC;&#x2039;one for the mouse port and one for the keyboard port. A USB-compatible transceiver needs only one USB port to handle both devices if the system supports USB Legacy (keyboard) functions. The transceiver attached to the computer draws its power from the port. The transceiver receives signals from the transceiver built in to the mouse or keyboard. These devices require batteries to function; therefore, a common cause of wireless device failure is battery run-down. Early generations of wireless devices used unusual battery types, but most recent products use off-the-shelf alkaline AA or AAA batteries. Rechargeable batteries are usually not supported, although Logitech's MX 700 has built-in rechargeable batteries and a transceiver


that doubles as a charger. Although all three technologies rely on battery power, the similarities end there. IR devices have a relatively short range (12 ft. maximum) and must have a clear line-of-sight between the input device and transceiver. Anything from a Mountain Dew can to a sheet of paper can block the IR signal from reaching the transceiver, assuming you're aiming the transmitter built in to your input device correctly in the first place. Some late-model IR devices have transceivers that can receive signals through a relatively wide 120° range, but this technology is much more temperamental than the others and has been abandoned by most vendors. Figure 18.15 shows how range and line-of-sight issues can prevent IR input devices from working correctly.

Figure 18.15. A wireless mouse using IR technology must be within range of the transceiver, at the correct angle to the transceiver, and not blocked by any objects. Otherwise, it cannot work.


Because of the problems with IR devices shown in Figure 18.15, almost all vendors of wireless input devices now use radio waves (RF) for transmission between the device and transceiver. RF-based wireless devices have no line-ofsight problems, but most have a limited range of about 6 ft. from the transmitter (see Figure 18.16).

Figure 18.16. A wireless mouse using RF must be within range of the transceiver, but unlike IR-based wireless mice, the angle of the mouse to the transceiver doesn't matter and radio signals can't be blocked by books, paper, or other obstacles.

Although RF overcomes line-of-sight issues that can cripple an IR mouse, early versions of RF products had a high potential for interference from other devices and from other devices in use in the same room because of a limited range of channels. For example, early Logitech wireless MouseMan products required the user to manually select the channel used by the transceiver and mouse. If more than six users in a small room had wireless devices, interference was practically inevitable and user error could lead to a user's mouse movements showing up on the wrong computer screen.


Fortunately, improvements in frequency bands used and automatic tuning have enabled all users of a particular type of device to avoid interference with other electronic devices or with each other. For example, Logitech's current line of wireless products uses its patented Palomar technology. Although the 27MHz frequency used by Palomar has become a de facto standard for most recent wireless input devices (it's also used by Microsoft and IBM for their wireless products), Logitech allows users to enable a digital security feature that uses one of more than 4,000 unique codes to prevent accidentally activating another computer with a wireless device or signal snooping by another user. Most vendors use similar technology but might use a much smaller number of codes. The range of 27MHz RF devices is shortâ&#x20AC;&#x2039;about 6 ft.â&#x20AC;&#x2039;but the transmitter can be located behind the computer or under the desk without loss of signal. Although most wireless products use proprietary radio transceivers, the Bluetooth wireless standard can now be used by some keyboards and mice, such as Microsoft's pioneering Wireless Optical Desktop mouse and keyboard bundle or the Wireless IntelliMouse Explorer for Bluetooth. Logitech also offers a Bluetoothenabled input device: its Cordless Presenter handheld pointing device. Bluetooth-enabled devices have an effective range of up to 30 ft. and might be compatible with other brands of devices that are also Bluetooth enabled. For more information about Bluetooth, see Chapter 20, "Local Area Networking," p. 1073.

Having used both IR and RF types of wireless devices, I can tell you that a radiofrequency input device beats an infrared input device hands down for use at home or in a small one- or two-person office. Because infrared requires an unobstructed direct line between the transceivers, when I used an infrared keyboard/pointing stick combination at a client site, I was constantly re-aiming the keyboard at the transceiver to avoid losing my signal. When I used a radio mouse, on the other hand, there were no line-of-sight issues to worry about. Historically, the only advantage to infrared was cost, but the problems of reliability in my mind outweigh any cost savings. Additionally, a wide range of prices for RF wireless products, including attractive keyboard and mouse combinations, make RF input devices affordable for almost everyone. If you're planning to use a computer to drive a big-screen TV or as a presentation unit, consider Bluetooth-enabled devices available from Microsoft, Logitech, and others because of their longer range (up to 10 meters, or 33 ft.).

Power Management Features of Wireless Input Devices


A wireless mouse is useless if its batteries fail, so several vendors of wireless products have developed sophisticated power-management features to help preserve battery lifeâ&#x20AC;&#x2039;especially with optical mice, which use power-eating LEDs to illuminate the mousing surface. For example, the Logitech Cordless MouseMan Optical's LED sensor has four operating modes, as shown in Table 18.6. Table 18.6. Logitech Cordless MouseMan Optical Power Management Mode

LED Flashing Rate

Notes

Normal

1,500 per second

Used only when mouse is being moved across a surface

Glow

1,000 per second

Used when mouse stops moving

Strobe

10 per second

Mouse not moved for more than 2 minutes

Flash

2 per second

Mouse not moved for more than 10 minutes

Wireless keyboards are activated only when you press a key or use the scroll wheel available on some models, so they tend to have longer battery lives than mice. Conventional ball-type mice also have longer battery lives than optical mice, but the convenience and accuracy of optical mice outweigh battery-life issues for most users.

Wireless Pointing Device Issues Before you invest in wireless pointing devices for multiple computers, you should be aware of the following issues: Line-of-site issues. Infrared devices won't work if the IR beam between the pointing device and the transceiver attached to the system is blocked. These units are not as suitable for casual in-the-lap use as radio-frequency units are. Radio-frequency interference. Although early wireless mice used analog tuners that were hard to synchronize, today's wireless input devices typically use digital selectors. However, if several similar devices are used in close quarters, a transceiver might actually receive data from the wrong mouse or keyboard. Also, metal desks and furniture can reduce range and cause erratic cursor movement. Most wireless devices operate around 27MHz, minimizing interference from devices such as cordless phones. If you plan to install several different computers using wireless input devices in the same room, set up one at a time and allow about half an hour between installations if possible


to let each unit synchronize with its transceiver. Check with the vendor for other tips on overcoming interference issues. Battery life and availability. Early wireless devices sometimes used unusual, expensive batteries. Today's units run on common battery types, such as AAA or AA. Battery life is usually rated at about 6 months for keyboards or balltype mice and about 2â&#x20AC;&#x2039;3 months for optical mice. Be sure you have spare batteries for the input device to avoid failures due to running out of battery power. Some vendors provide software that gives users an onscreen warning when batteries run low. Furthermore, when using an optical wireless mouse, you should try working on brighter or whiter surfaces. Many optical mice adjust their sensors based on the illumination of the surface, which is why you might see the light in the mouse change intensity. The less intense the internal LED operates, the less battery power being used. Location. The range of wireless devices can vary from 6 ft. with conventional RF devices to as much as 30 ft. with Bluetooth-based devices. Consider where the device will be used before making your purchase. For instance, in an office where multiple devices might be used at the same time, a close-range device might be more desirable to avoid cross talk among devices. On the other hand, the home user who wants to sit away from the screen while maintaining control might want an extended range, making Bluetooth-enabled devices a better choice. User experience. Different users will have different expectations of wireless input devices, but in general, the more a wireless input device acts like its wired siblings, the better. The fact that a device is wireless should not compromise its functionality. If things such as reliability, connection, or driver problems hinder proper usage, the device isn't worth using. Hardcore gamers who need the fastest response time possible generally favor the responsiveness of a wired optical mouse over any wireless mouse. Although minimal, some lag time does exist. Some mice can require up to 0.25 centimeter of movement before responding. This lag time can also affect users doing graphical work requiring the superior consistency and accuracy of a wired optical mouse, although the latest dual-sensor wireless optical mice have accuracy on par with wired optical mice. Pointer speed. Conventional wired optical mice transmit their positions about 120 times per second, whereas wireless mice that use a USB receiver transmit their positions about 40â&#x20AC;&#x2039;50 times per second. If you use a mouse to play fast-action games, you might find a corded mouse a better choice because of the more frequent position updates it provides.


Troubleshooting Wireless Input Devices If your wireless input device does not work, check the following: Battery failure. The transceivers attached to the computer are powered by the computer, but the input devices themselves are battery-powered. Check the battery life suggestions published by the vendor; if your unit isn't running as long as it should, try using a better brand of battery or turning off the device if possible. Lost synchronization between device and transceiver. Both the device and the transceiver must be using the same frequency to communicate. Depending on the device, you might be able to resynchronize the device and transceiver by pressing a button, or you might need to remove the battery, reinsert the battery, and wait for several minutes to reestablish contact. Interference between units. Check the transmission range of the transceivers in your wireless units and visit the manufacturer's Web site for details on how to reduce interference. Typically, you should use different frequencies for wireless devices on adjacent computers. Blocked line of sight. If you are using infrared wireless devices, check the line of sight carefully at the computer, the space between your device and the computer, and the device itself. You might be dangling a finger or two over the infrared eye and cutting off the signalâ&#x20AC;&#x2039;the equivalent of putting your finger over the lens on a camera. Serial port IRQ conflicts. If the wireless mouse is connected to a serial port and it stops working after you install another add-on card, check for conflicts using the Windows Device Manager. Disconnected transceiver. If you have moved the computer around, you might have disconnected the transceiver from its keyboard, PS/2 mouse, serial, or USB port. You can plug a USB device in without shutting down the system, but the other types require you to shut down, reattach the cable, and restart to work correctly. USB Legacy support not enabled. If your wireless keyboard uses a transceiver connected to the USB port and the device works in Windows, but not at a command prompt, make sure you have enabled USB Legacy support in the BIOS or use the PS/2 connector from the transceiver to connect to the PS/2 keyboard port.


Chapter 19. Internet Connectivity Relating Internet and LAN Connectivity Cable Modems and CATV Networks Digital Subscriber Line Fixed-Base Wireless Broadband Internet Connectivity via Satellite with DirecWAY or StarBand Integrated Services Digital Network Comparing High-Speed Internet Access Leased Lines Securing Your Internet Connection Asynchronous (Analog) Modems Modem Standards Sharing Your Internet Connection Internet Troubleshooting


Relating Internet and LAN Connectivity Communication between computers is a major part of the PC computing industry. Thanks to the World Wide Web (WWW), no computer user is an island. Whether using a modem or broadband technology, virtually all PCs can be connected to other computers, enabling them to share files, send and receive email, and access the Internet. This chapter explores the various technologies you can use to expand the reach of your PC around the block and around the world. It might surprise you to see discussions of protocols and networking setup in both this chapter and the LAN chapter of this book, but a modem connection is really just another form of networking. In fact, 32-bit versions of Windows from Windows NT and Windows 9x all the way through Windows XP have all but blended the two services into a single entity. The reason for this combination is that the typical target for a modem connection has changed over the years. Computer users a decade ago dialed in to bulletin board systems (BBSs), which are proprietary services that provide terminal access to other computers. However, BBSs are practically extinct today. Similarly, proprietary online services such as America Online and CompuServe (now owned by AOL but maintained as a separate service), which have also been around for many years, have dropped their proprietary client software and protocols and have been reborn as gateways to the Internet. With the explosive growth of the Internet, modem and network technologies were joined because both could use the same client software and protocols. Today, the most popular suite of networking protocolsâ&#x20AC;&#x2039;TCP/IPâ&#x20AC;&#x2039;is used on both LANs and the Internet. When you dial in to an Internet service provider (ISP), you are actually connecting to a network using a modem instead of a network interface card, and when you use most broadband services, your path to the Internet typically starts with a network interface card, built-in network port, or network-to-USB adapter. Although most new PCs still include an analog modem, an increasing number of PC users are abandoning analog modem Internet access for the faster world of broadband access. As of the end of the third quarter of 2002, for example, there were an estimated 18 million subscribers to residential broadband Internet services such as DSL and cable modems in the United States and Canada. About 11.9 million of these broadband subscribers had selected cable modem service, compared to just 2 million as of February 2000â&#x20AC;&#x2039;and about 6.1 million subscribers used DSL connections. Some DSL and cable modem service providers have had a rough time financially in 2001 and 2002, but broadband Internet is an increasingly popular choice for both experienced computer users and novices. Broadband subscribers have been increasing by at least 14% per quarter every quarter since early 2000. Plus, a recent survey suggests that as many as one-


third of current dial-up users plan to switch to broadband Internet access within a year. Because broadband Internet is the fastest-growing part of Internet access and because it is completely different from the familiar analog modem environment many computer users have grown up with, broadband Internet devices are covered first in this chapter. Analog modems, however, still have their uses and a large user base. Travelers, those without broadband options, or those on a budget still need an analog modem. Also, additional analog modems are required for some types of broadband service, and sometimes as a backup Internet access source. You'll find analog modem coverage in the second portion of the chapter.

Comparing Broadband and Analog Modem Internet Access Even though most new PCs purchased at retail include some type of analog modem you can use for Internet and email access, you are likely to find that Internet and email access with an analog modem aren't sufficient for your needs if you use these services for more than a few minutes each day. Here are some reasons you should consider switching to a broadband service: Speed. The fastest analog modems can download data at a maximum rate of 56Kbps (limited in the United States to just 53Kbps by the FCC), whereas broadband services start at 128Kbps for ISDN. Newer forms of broadband, such as DSL and cable modems, start at 384Kbps and typically exceed 500Kbps. Similarly, broadband services can upload data at several times the speed of an analog modem. Convenience. Cable modems and some types of DSL and satellite broadband Internet service are always on, providing you with an immediate connection as soon as you open your Web browser or email client. Analog modems require you to dial up the server and wait up to a minute before you can check your email or surf the Web. Similarly, always-on broadband services can provide you with immediate notification of incoming email, whereas analog systems can check for incoming email only if you stay online and tie up your phone line. Telephone line usage. In the wake of the September 11th terrorist attacks on the United States, keeping telephone communications lines open for emergencies has been a major consideration for many people. Most analog modems do not support call waiting, making it difficult for callers to reach you with important messages while you're online unless you use call-forwarding or call-notification software. Although some dial-up ISPs provide software that can alert you to incoming calls, in most cases you must find and install such


software yourself. By contrast, most broadband services keep your telephone line free so you can check email or surf the Web and use the telephone at the same time. Price. One disadvantage of broadband communications is apparent when you see your bill: It costs two to three times as much per month as dial-up access with an analog modem. However, millions of U.S. users believe the additional speed and convenience of broadband make the extra cost of the service per month a worthwhile investment. If you use the Internet enough to justify installing a second phone line just for Internet use, the price gap narrows considerably because you can use most broadband services without tying up your single existing phone line. If you have cable TV, most cable TV providers now also offer cable broadband service and provide a discount off the normal price for customers who have both cable TV and cable Internet service. Ease of reconnection after an operating system upgrade. Because broadband Internet is usually based on automatically configured TCP/IP network settings, you should be able to keep your broadband connection running during a Windows upgrade with little difficulty. Just verify that you have the correct drivers for your Ethernet adapter (used for most broadband connections) before you perform the upgrade, note your computer and workgroup name, and you should be able to go online as soon as the upgrade is completed. A dial-up connection is often much tougher to keep working, especially because of the different methods used by various Windows versions for handling dialup networking.

Broadband Internet Access Types Thanks to the combination of huge multimegabyte downloads needed to update software and support hardware, dynamic Web sites with music and full-motion video, and increased demand for online services, even the fastest analog modem (which can download at just 53Kbps) isn't sufficient for heavy Internet use. More and more users are taking advantage of various types of broadband Internet access solutions, including: Cable modem DSL Fixed-base wireless Satellite-based services


ISDN Leased lines At least one of these services might be available to you, and if you live in a large to medium-size city, you might be able to choose from two or more of these broadband solutions. The first portion of this chapter focuses on these solutions.

High Speed = Less Freedom Although high-speed services such as cable modems, DSL, and others all represent major improvements in speed over existing dial-up analog or 56Kbps connections, one big drawback you should consider is the loss of freedom in choosing an ISP. With an analog or a 56Kbps modem, you can choose from a wide variety of services, including Local ISPs (personalized service) National ISPs with dial-up access across the country (great for travelers) Online services with customized content plus Web access (AOL and CompuServe) Family-friendly filtered Internet access (Mayberry USA and Lightdog) Business-oriented Web hosting plus Internet access plans from many vendors At present, if you want faster speed, you must use the ISP provided with your high-speed service. Whether it's your local telephone company, a third-party vendor, or your friendly cable TV operator, their ISP is your ISP. When you evaluate a high-speed service, remember to look at the special features and services provided by the ISP and its track record for reliability and keeping customers happy. After all, the quality of the work your ISP does is reflected in the quality of your broadband connection.


Tip If you want the extra speed and convenience of broadband Internet but want to shield your family from some types of Internet content, several filtering services are available that work with your preferred broadband service, including these: Cleanweb's Filter Your Internet (FYI) at www.cleanweb.net S4F FamilyConnect FilterPak at www.familyconnect.com Bsafe Online at http://www.bsafehome.com


Cable Modems and CATV Networks For many users, cable modem serviceâ&#x20AC;&#x2039;which piggybacks on the same cable TV (CATV) service lines that bring your TV many channelsâ&#x20AC;&#x2039;represents both a big boost in speed from that available with ISDN and a major savings in initial costs and monthly charges. Unlike ISDN, cable modem service normally is sold as an "all you can eat" unlimited-access plan with a modest installation charge (often waived) and a small monthly fee for the rental of the cable modem. Because more and more cable networks support a single standard, you can also choose to buy your cable modem from any of several vendors in some cases, saving even more money over the long term.

Connecting to the Internet with a "Cable Modem" As with ISDN, the device used to connect a PC to a CATV network is somewhat inaccurately called a modem. In fact, the so-called "cable modem" (a name I will continue to use, for the sake of convenience) is actually a great deal more. The device does indeed modulate and demodulate, but it also functions as a tuner, a network bridge, an encryptor, and an SNMP agent. To connect your PC to a CATV network, you do not use a serial port as with analog modem technologies or ISDN terminal adapters. Instead, the most typical connection today uses 10/100 Ethernet or USB (10/100 Ethernet is faster). If your computer doesn't include an Ethernet adapter or 10/100 Ethernet card, you can install a 10/100 Ethernet adapter into a PCI slot or connect most recent cable modem models to a USB port on your computer. Some older cable modem services utilize an internal adapter for one-way service and use the conventional modem for uploads; this type of service is called telco return. Even though you get fast downloading, telco return systems tie up your phone line and are not recommended.


Tip For maximum speed and ease of sharing, I recommend that you connect your cable modem to a 10/100 Ethernet card or port or use a USB 2.0 port (built into many new machines or available as an add-on card for older systems). USB 1.1 ports limit the throughput of your cable modem.

The Cable Modem and the CATV Network The cable modem connects to the CATV network using the same coaxial cable connection as your cable TV service (see Figure 19.1). Thus, the cable modem functions as a bridge between the tiny twisted-pair network in your home and the hybrid fiber/coax (HFC) network that connects all the cable customers in your neighborhood.

Figure 19.1. A typical hybrid fiber-coax cable TV network that also includes two-way cable modem service.

A few cable modem CATV systems have been built using the older one-way (download-only) coax cable, but this type of cable is much slower for cable modem use and is obsolete for both CATV and data communications. The industry has largely replaced coax with HFC. Before you sign up for CATV Internet service, find out which type of service is being offered. Only the two-way, HFC-based systems allow you to use the Internet independently of the telephone system; one-way cable modem service requires an analog modem for uploading page requests, files, and email. The modem can be built into your one-way cable modem (these are called bundled cable modems) or be a separate external analog modem. In either case, going online with a one-way cable modem ties up your phone line.


Digital CATV service, which brings your TV many more channels and a clearer picture, requires the cable TV provider to upgrade to an HFC physical plant. Thus, digital CATV service is a precursor to two-way cable modem service; pure-coax CATV systems can't be used to transmit digital service or handle two-way cable modem traffic. CATV systems that have been upgraded to digital service are capable of providing two-way cable modem service, after suitable head-end equipment is installed at the CATV central office. A good rule of thumb, therefore, is that CATV systems that don't offer digital cable TV might offer only one-way cable modem service or no cable modem service at all. A typical two-way cable modem connection is shown in Figure 19.1. Originally, cable modems were not sold to users of CATV Internet access but were leased by the CATV companies offering Internet access to their cable modem customers. This is because each cable modem on a particular CATV network had to match the proprietary technology used by the network. In late 1998, DOCSIScompliant cable modems began to be used by some CATV companies. DOCSIS refers to devices that meet the Data Over Cable Service Interface Specification standards established by Cable Television Laboratories, Inc. (CableLabs). Modems that meet DOCSIS standards are now referred to as CableLabs Certified cable modems. Visit the CableLabs Web site at www.cablelabs.org for a complete list of cable modems that are CableLabs Certified. Many vendors of traditional modems and other types of communications products, such as Zoom Telephonics, 3Com, GVC, General Instruments, Philips, Motorola, Cabletron, Toshiba, Cisco, D-Link, and many others, now make CableLabs Certified Cable Modemâ&#x20AC;&#x2039;compliant hardware. The models supported by your CATV Internet provider might vary according to the DOCSIS standard it supports. Table 19.1 provides a brief overview of the differences in these standards. Table 19.1. DOCSIS Standards Overview DOCSIS Standard

Benefits

Notes

1.0

Basic broadband CATV (cable modem) service

Original DOCSIS version

1.1

Supports tiered service (different speeds at different costs), faster uploading, home networking, and packet telephony while reducing costs

Backward-compatible with DOCSIS 1.0

2.0

Faster performance for downloading and uploading compared to DOCSIS 1.0 and Backward-compatible 1.1; supports high-speed two-way business services with DOCSIS 1.1, 1.0

Although most cable modems are now available for about $100, compared to $30â&#x20AC;&#x2039;$70 for typical analog modems, you should check with your CATV Internet provider before purchasing one to determine which models are supported by your provider and whether your CATV Internet provider still requires you to lease the


cable modem. If you plan to keep cable modem service for more than a year, I recommend purchasing a CableLabs Certified cable modem, but if you're unsure of your longterm plans, leasing isn't all that expensive. Typical lease costs for the device add only about $10/month to the monthly rate of $30â&#x20AC;&#x2039;$40 for cable modem service.

Types of Cable Modems Cable modems come in several forms: External cable modem "box." It requires a 10BASE-T or 10/100 Ethernet NIC or a USB port. Some vendors provide "bundles" that combine the external cable modem with the NIC at a lower price than leasing or purchasing both separately. This type of cable modem is designed for fast uploads as well as fast downloads and works only on newer "two-way" CATV Internet connections. Almost all cable modems sold by retailers fall into this category. Internal cable modem with a 56Kbps analog modem built in. This type of cable modem provides fast downloads but uploads at analog speeds only; this type might use either an ISA or a PCI slot. Internal cable modem for use with a separate 56Kbps analog modem. This type might use either an ISA or a PCI slot. As with the previous model, uploads are at analog speeds only. Internal cable modem for use with two-way service. This type of cable modem is rare, but because some internal cable modems are now CableLabs Certified, you might see some cable ISPs offer these. Some also might be designed for one-way service, enabling the cable ISP to install a single modem for use with both types of service. External cable modem with a 56Kbps analog modem built in. This type of cable modem is sometimes designed to work with both one-way and two-way services. Cable modem models that must be used with (or include) an analog modem are designed for older "one-way" CATV Internet connections; these will tie up your phone line when transmitting page requests as well as when sending uploads.


Note For more comprehensive coverage of cable modem service types, installation, sharing, security, and use, see the Absolute Beginner's Guide to Cable Internet Connections by Mark Edward Soper (published by Que, 2002).

CATV Bandwidth Cable TV uses what is known as a broadband network, meaning the bandwidth of the connection is split to simultaneously carry many signals at different frequencies. These signals correspond to the channels you see on your TV. A typical HFC network provides approximately 750MHz of bandwidth, and each channel requires 6MHz. Therefore, because the television channels start at about 50MHz, you would find channel 2 in the 50MHz​56MHz range, channel 3 at 57MHz​63MHz, and so on up the frequency spectrum. At this rate, an HFC network can support about 110 channels. For data networking purposes, cable systems typically allocate one channel's worth of bandwidth in the 50MHz​750MHz range for downstream traffic​that is, traffic coming into the cable modem from the CATV network. In this way, the cable modem functions as a tuner, just like your cable TV box, ensuring that your PC receives signals from the correct frequency. Upstream traffic (data sent from your PC to the network) uses a different channel. Cable TV systems commonly reserve the bandwidth from 5MHz to 42MHz for upstream signals of various types (such as those generated by cable TV boxes that enable you to order pay-per-view programming). Depending on the bandwidth available, you might find that your CATV provider does not furnish the same high speed upstream as it does downstream. This is called an asymmetrical network.


Note Because the upstream speed often does not match the downstream speed (and to minimize noise, which tends to accumulate because of the tree-and-branch nature of the network), cable TV connections usually are not practical for hosting Web servers and other Internet services. This is largely deliberate because most CATV providers are currently targeting their traditional home user market. As the technology matures, however, this type of Internet connection is likely to spread to the business world as well. There are now specialized domain name services that can be used to "point" Web surfers to your cable modem or DSL connection. Some cable ISPs require you to switch to a more expensive business plan if you want to host a server (including a P2P server such as Gnutella) on your cable modem connection. In fact, hosting a server on a residential cable modem service could be a violation of your contract and could lead to cancellation of your service.

The amount of data throughput that the single 6MHz downstream channel can support depends on the type of modulation used at the head end (that is, the system to which your PC connects over the network). Using a technology called 64 QAM (quadrature amplitude modulation), the channel might be capable of carrying up to 27Mbps of downstream data. A variant called 256 QAM can boost this to 36Mbps. You must realize, however, that you will not achieve anything even approaching this throughput on your PC. First of all, if you are using a 10BASE-T Ethernet adapter to connect to the cable modem, you are limited to 10Mbps. A USB 1.1 port is limited to 12Mbps, but even this is well beyond the real-life results you will achieve. As with any LAN, you are sharing the available bandwidth with other users in your neighborhood. All your neighbors who also subscribe to the service use the same 6MHz channel. As more users are added, more systems are contending for the same bandwidth, and throughput goes down. In November 1999, ZDTV (now TechTV) tested five brands of cable modems in typical operations and found that the overhead of CATV proved to be a major slowdown factor. The cable modems were first connected directly to the server to provide a baseline for comparisons. Some of these tests showed speeds as high as 4Mbps. However, when the CATV cable was connected and the same tests were run, the best performer dropped to just 1.1Mbps, with others running even slower.


Tip You can improve the performance of your cable modem or other broadband Internet access device by making changes to your system Registry that affect the size of the TCP receive window and other settings. I recommend the system tweaks available at the SpeedGuide.net Web site (www.speedguide.net). You can find tweaks for all Windows versions from Windows 95 through Windows XP, along with instructions on how to make the changes manually and easy methods for setting your system's Registry back to its default values.

Widespread reports from cable modem users across the country indicate that rush hourâ&#x20AC;&#x2039;type conditions occur at certain times of the day on some systems, with big slowdowns. This rush hour is due to increasing use of cable modem systems in the late afternoon and early evening, as daytime workers get home and pull up the day's news, weather, stocks, and sports on their Internet connections. Because cable modems are shared access, this type of slowdown is inevitable and becomes exceptionally severe if the CATV Internet provider doesn't use a fast enough connection to the rest of the Internet. To minimize this problem, many CATV Internet providers use caching servers at their point of presence connections to the Internet. These servers store frequently accessed Web pages to enable users to view pages without the delays in retrieving them from the original Web sites. By adding multiple T1 or T3 connections to the Internet backbones and using caching servers, ISPs can minimize delays during peak usage hours.

CATV Performance The fact that you are sharing the CATV network with other users doesn't mean the performance of a cable modem isn't usually spectacular. Although the CATV network takes a big cut out of the maximum speeds, you'll still realize a throughput that hovers around 512Kbps, almost 10 times that of the fastest modem connection and four times that of ISDN. You will find the Web to be an entirely new experience at this speed. Those huge audio and video clips you avoided in the past now download in seconds, and you will soon fill your hard drives with all the free software available. Add to this the fact that the service is typically quite reasonably priced. Remember that the CATV provider is replacing both the telephone company (if you have two-way service) and your ISP in your Internet access solution. The price can be about $40â&#x20AC;&#x2039;$50 per month (including cable modem rental), which is twice that of a normal dial-up ISP account, but it is far less than ISDN, does not require a telephone line (if your provider has two-way service), and provides 24hour access to the Internet. The only drawback is that the service might not be available yet in your area. In my opinion, this technology exceeds all the other Internet access solutions available today in speed, economy, convenience, and widespread availability. Its nearest rival is DSL, which is still not as widely available geographically and is plagued with poor coordination between ISPs and


telephone companies. Because cable modem Internet service providers provide the physical plant, provide ISP services, and can provide equipment, you can get service installed in just days and avoid the finger-pointing common with other types of broadband Internet service.

CATV Internet Connection Security Because your PC is sharing a network with other users in your neighborhood and because the traffic is bidirectional on systems using two-way cable modems, the security of your PC and the network becomes an issue. In most cases, some form of encryption is involved to prevent unauthorized access to the network. CableLabs Certified (DOCSIS) cable modems have built-in encryption, but older one-way modems might not have this feature. If you use an operating system such as 32-bit Windows that has built-in peer networking capabilities and your provider doesn't use CableLabs Certified cable modems or some other form of encryption, you might be able to see your neighbors' computers on the network. The operating system has settings that enable you to specify whether other network users can access your drives. If these settings are configured improperly, your neighbors might be able to view, access, and even delete the files on your hard drives. Be sure the technician from the cable company installing the service addresses this problem if your cable modem hardware doesn't provide encryption. If you want to use a cable modem along with sharing access on your computer (for printing, file storage, and so on), I'd recommend that you use passwords for any shared drives, but you're even safer if you disable file and printer sharing on the system you connect to the cable modem. For more information on securing any type of Internet access, see "Securing Your Internet Connection," p. 1046.


Digital Subscriber Line The biggest rival to the cable modem in the broadband Internet business is the digital subscriber line (DSL). DSL, like its predecessor ISDN, appeals to the telephone companies who might be able to use the existing POTS analog wiring to provide high-speed Internet access. Not every type of DSL is suitable for existing wiring; however, all but the fastest, most expensive types can sometimes be used with the existing POTS plant. DSL is also appealing to businesses that don't have access to cable modems but are looking for a high-performance, lower-cost alternative to the expensive ISDN services that top out at 128Kbps.


Note Some technical discussions of DSL refer to xDSL. The x stands for the various versions of DSL being proposed and offered by local telephone companies and ISPs. DSL generally is used to refer to any type of digital subscriber line service.

One advantage of DSL compared to its most popular rival​cable modems​is that cable modems share common bandwidth, which means that a lot of simultaneous use by your neighbors can slow down your connection. If you use DSL you don't have this concern; whatever bandwidth speed you pay for is yours​period.

How DSL Works DSL takes advantage of the broadband nature of the telephone system, using the system's capability to carry signals at multiple frequencies to allow both highspeed Internet traffic and phone calls at the same time. Two methods for sending and receiving signals are used by the most common type of DSL, Asymmetric DSL (ADSL): Carrierless Amplitude/Phase (CAP) Discrete Multitone (DMT) Most early DSL installations used CAP, which splits the telephone line into three frequency bands. Exact frequency usage varies by system, but most typically, the divisions resemble the following: Voice calls use frequencies from 30Hz to 4KHz. This frequency is also used by answering machines, fax machines, and alarm systems. Upstream data such as Web page requests and sent email uses frequencies between 25Hz and 160Hz. Downstream data such as received Web pages and email uses frequencies between 240KHz and 1.5MHz. Some systems use the 300Hz​700Hz range for downstream data and frequencies of 1MHz and above for upstream data. Because voice, downstream, and upstream data use different frequencies, you can talk, surf, and send email at the same time.


DMT, the system used by most recent ADSL installations, divides the telephone line into 247 channels that are 4KHz wide. If a particular channel has problems, a different channel with better signal quality is used automatically. Unlike CAP, DMT uses some channels starting at around 8KHz to send and receive information. Both types of signaling can have problems with interference from telephones and similar devices, so devices called low-pass filters are used to prevent telephone signals from interfering with signals above the 4KHz range, where DSL signals begin. The location of these filters depends on the type of DSL you use and whether you are installing DSL service yourself. At the central switch, DSL data is transferred to a device called a DSL access multiplexer (DSLAM), which transfers outgoing signals to the Internet and sends incoming signals to the correct DSL transceiver (the correct name for the socalled "DSL modem" that connects to your computer).

Who Can Use DSLâ&#x20AC;&#x2039;and Who Can't DSL services are slowly rolling out across the country, first to major cities and then to smaller cities and towns. As with 56Kbps modems, rural and small-town users are probably out of luck and should consider satellite-based or fixed wireless Internet services where available for a faster-than-56Kbps experience. Just as distance to a telephone company's central switch (CS) is an important consideration for people purchasing an ISDN connection, distance also affects who can use DSL in the markets offering it. For example, most DSL service types require that you be within about 18,000 feet (about 3 miles) wire distance to a telco offering DSL; some won't offer it if you're beyond 15,000 feet wire distance because the speed drops significantly at longer distances. Repeaters or a local loop that has been extended by the telco with fiber-optic line might provide longer distances. The speed of your DSL connection varies with distance: The closer you are to the telco, the faster your DSL access is. Many telcos that offer some type of DSL service provide Web sites that help you determine whether, and what type of, DSL is available to you. If you want to locate DSL service providers in your area, compare rates, and see reviews from users of the hundreds of ISPs now providing DSL service, set your browser to http://www.dslreports.com. The site provides a verdict on many of the ISPs reviewed, summarizing users' experiences and ranking each ISP in five categories.


Note If you want to connect DSL to your SOHO or office LAN, check first to see what the provider's attitude is. Some users report good cooperation, whereas others indicate they were told "we can't help you" or were told that DSL "couldn't be connected to a LAN." Again, check around for the best policies. Lowcost switch/router combinations from companies such as Linksys and D-Link and Microsoft's Internet Connection Sharing provide relatively easy ways to share both DSL and other types of high-speed connections.

Even if your telco's central switch is well within wire distance range of your location, that's no guarantee that you qualify for DSL service. The design and condition of the wiring plant connecting your location with the central switch can prevent you from qualifying for DSL service. Because DSL service depends on successful sending and receiving of high-frequency data, a telephone wiring plant that blocks high-frequency signals can't be used for DSL service. Some of the typical issues with telephone lines that aren't DSL-friendly include: Loading coils. These amplifiers boost voice signals and are sometimes called voice coils. Unfortunately, these block the high-frequency signals needed by DSL service. Bridge taps. Used to extend service to new customers without running separate lines all the way back to the central switch. Bridge taps can create a circuit that's too long for DSL service. Fiber-optic cables. Used to carry a lot of signals in a small physical space, fiber-optic cables use analog-to-digital (A/D) and digital-to-analog (D/A) converters where they connect to copper telephone lines. A/D and D/A converters can't pass DSL signals through to their destinations.

Major Types of DSL Although the term DSL is used in advertising and popular discussions to refer to any form of DSL, many, many variations of DSL are used in different markets and for different situations. This section discusses the most common forms of DSL and provides a table that compares the various types of DSL service. Although many types of DSL service exist, you can choose only from the service types offered by your DSL provider: ADSL (Asymmetrical DSL). The type of DSL used most often, especially in residential installations. Asymmetrical means that downstream (download) speeds are much faster than upstream (upload) speeds. For most users, this is no problem because downloads of Web pages, graphics, and files are the


major use of Internet connections. Maximum downstream speeds are up to 1.6Mbps, with up to 640Kbps upstream. Most vendors who offer ADSL provide varying levels of service at lower speeds and prices, as well. Voice calls are routed over the same wire using a small amount of bandwidth, making a single-line service that does voice and data possible. ADSL is more expensive to set up than some other forms of DSL because a splitter must be installed at the customer site, meaning that you must pay for a service call (also called a truck roll) as part of the initial setup charge. CDSL (Consumer DSL). A slower (1Mbps upstream) form of DSL that was developed by modem chipset maker Rockwell. It doesn't require a service call because no splitter is required at the customer site. G.Lite (Universal DSL, and also called DSL Lite or Splitterless DSL). Another version that splits the line at the telco end rather than at the consumer end. Downstream speeds range from 1.544Mbps to 6.0Mbps, and upstream speeds can be from 128Kbps to 384Kbps. This is becoming one of the most popular forms of DSL because it enables consumers to use self-install kits. Note that the DSL vendor might cap the service at rates lower than those listed earlier in the chapter; check with the vendor for details. SDSL (Symmetrical DSL). This type of DSL service provides the same speed for upstream as for downstream service. Generally, SDSL is offered to business rather than residential customers because it requires new cabling (rather than reusing existing phone lines). A long-term contract frequently is required. Table 19.2 summarizes the various types of DSL. Table 19.2. DSL Type Comparison DSL Type

Description

Data Rate Downstream; Upstream

Distance Limit

Application

IDSL

ISDN Digital 128Kbps Subscriber Line

18,000 feet on 24-gauge wire

Similar to the ISDN BRI service but data only (no voice on the same line)

CDSL

Consumer DSL 1Mbps downstream; less from Rockwell upstream

18,000 feet on 24-gauge wire

Splitterless home and small business service; similar to DSL Lite

DSL Lite Splitterless (same DSL without as the truck roll G.Lite)

From 1.544 Mbps to 6Mbps downstream, depending on the 18,000 feet on 24-gauge wire subscribed service

The standard ADSL; sacrifices speed for not having to install a splitter at the user's home or business


HDSL

High bit-rate 1.544Mbps duplex on two Digital twisted-pair lines; 2.048Mbps SubscriberLine duplex on three twisted-pair lines

12,000 feet on 24-gauge wire

T-1/E1 service between server and phone company or within a company; WAN, LAN, server access

SDSL

1.544Mbps duplex (U.S. and Canada); 2.048Mbps (Europe) Symmetric DSL on a single-duplex line downstream and upstream

12,000 feet on 24-gauge wire

Same as for HDSL but requiring only one line of twisted pair

ADSL

Asymmetric 1.544Mbps to 6.1Mbps Digital downstream; 16Kbps to Subscriber Line 640Kbps upstream

1.544Mbps at 18,000 feet; 2.048Mbps at 16,000 feet; 6.312Mpbs at 12,000 feet; 8.448Mbps at 9,000 feet

Used for Internet and Web access, motion video, video on demand, and remote LAN access

Rate-Adaptive Adapted to the line, 640Kbps to RADSL DSL from 2.2Mbps downstream; 272Kbps Not provided Westell to 1.088Mbps upstream

UDSL

Unidirectional DSL proposed Not known by a company in Europe

VDSL

12.9Mbps 52.8Mbps Very High 4,500 feet at 12.96Mbps; downstream; 1.5Mbps to Digital 3,000 feet at 25.82Mbps; 2.3Mbps upstream; 1.6Mbps to Subscriber Line 1,000 feet at 51.84Mbps 2.3Mbps downstream

Not known

Similar to ADSL

Similar to HDSL

ATM networks; Fiber to the Neighborhood

Table copyrighted by and used with permission of whatis.com, Inc.

With any type of DSL, an external device called a DSL modem is attached to the computer through either of the following: A crossover cable running to a 10BASE-T or 10/100 Ethernet card or port in the computer A USB cable running to a USB port in the computer An RJ-11 (standard telephone) cable is attached between the DSL modem and the RJ-11 port that has been set up for DSL service. To prevent telephone signals from interfering with DSL frequencies, splitters or microfilters must be installed on a DSL line. If you choose a technician-installed form of DSL, a device called a splitter is used at your location to prevent interference. Splitter-based DSL allows faster speeds than splitterless DSL installations, but the wait for a technician to show up and add the splitter can add days or weeks to your installation process. If you self-install DSL, you will install small devices called microfilters to block


interference from telephones, answering machines, and similar devices. These devices might fit behind the faceplate of the wall outlet used for DSL service or inline between the phone, answering machine, or fax machine and the wall outlet (see Figure 19.2).

Figure 19.2. Two types of DSL self-installations; if a splitter is used to set up a separate DSL line, the microfilters shown here are not necessary.


Tip If you have a security system attached to your telephone line, watch out for problems if you select DSL as your preferred broadband access method. Security systems are often designed to seize the line, interrupting a phone call in progress to send an alarm to the security company. This feature won't work with normal microfilters, so you should purchase a special DSL Alarm filter to allow your alarm system to coexist with your DSL installation. Get more information about the alarm microfilter and alternative DSL installation options from www.hometech.com/learn/dsl.html.

DSL Pricing DSL pricing varies widely, with different telephone companies offering different speeds of DSL and different rates. One thing that's true about the most commonly used flavors of DSL is that they are usually an asymmetrical serviceâ&#x20AC;&#x2039;with download speeds faster than upload speeds. ADSL installations can typically be run over existing copper wires, whereas SDSL installations usually require that new highquality copper wires be installed between the CO and the subscriber's location. For unlimited use, typical residential DSL pricing ranges anywhere from $50 to $80 a month depending on the download speed, which ranges from 256Kbps to 1.5Mbps. Business DSL pricing ranges from $50 to as high as $500 per month. The wide variance is partly due to the upload speeds permitted. The lower-cost plans typically use a lower upload speed (some variation on ADSL or G.Lite); in contrast, the more expensive plans often use SDSL. Check carefully with your vendor because your traditional telephone company might not be the only DSL game in town. Some major cities might have as many as half a dozen vendors selling various flavors of DSL.

DSL Security Issues Unlike other types of broadband access, DSL is a direct one-to-one connection that isn't shared; you have no digital "neighbors" who could casually snoop on your activities. However, as with any broadband "always-on" connection, intrusion from the Internet to your computer is a very real possibility. For more information on securing any type of Internet access, see "Securing Your Internet Connection," p. 1046.


Technical Problems with DSL Telecommunications has always had its share of difficulties, starting with the incredibly slow and trouble-plagued 300bps modems used on early PCs, but as speed increases, so do problems. DSL connections are often very difficult to get working correctly because DSL, as you've seen, combines the problems of adding high-speed data access to the telephone line with network configuration using TCP/IP (the most powerful and most complex network protocol in widespread use; see Chapter 20, "Local Area Networking," for details). A review of comments from DSL users in various forums, such as DSLReports.com and others, shows that the most common problems include the following: Poor coordination between the DSL sales department of the telco or thirdparty provider and the installers. This can lead to broken or very late appointments for installation; if possible, contact the installer company to verify the appointment. If possible, opt for a self-install version of DSL to avoid problems with late or missing appointments. Installers who install the hardware and software and then leave without verifying it works properly. Ask whether the installer carries a notebook computer that can test the line; don't let the installer leave until the line is working. Poor technical support before and after installation. Record the IP address and other information used during the installation; read reviews and tips from sources listed earlier in the chapter to help you find better DSL providers and solutions you can apply yourself or ask your telco or provider to perform. Lower speeds than anticipated. This can be due to a poor-quality connection to the telco from your home or business or problems at the central switch; ask the installer to test the line for you during initial installation and tell you the top DSL speed the line can reach. On a healthy line, the problem is often traceable to a very low value for the Windows Registry key called RWIN (receive window), which should be adjusted from its default of 8192 (8KB) to a value as high as 32768 (32KB) or even 65535 (64KB). If your system previously was used with a dial-up modem, the value for RWIN can be as low as 2144; low values force your DSL connection to receive data at rates hardly faster than those for a dial-up analog modem connection. For interactive tests that will help you find the best value to use for RWIN or other Registry options, find line problems, and adjust your configuration, go to http://www.dslreports.com and follow the DSLR Tools link from the home page.


Because of the problems with trying to retrofit an aging voice-oriented telephone network with high-speed Internet service, many pure DSL companies are having financial problems. Some once-prominent DSL ISPs went out of business in 2000â&#x20AC;&#x2039;2001, leading to service cancellations in some cases. Before you sign a longterm contract for DSL service, you should determine what your options are if your telco, DSL line provider, or ISP drops DSL service.


Fixed-Base Wireless Broadband If you're beyond the reach of CATV-based Internet or DSL and aren't interested in trying to install a satellite dish, you still might be able to get broadband Internet service through a fixed wireless broadband Internet provider. These services use various frequencies of microwave signals to connect to the Internet. To learn more about fixed-base wireless broadband, see the Technical Reference section on the DVD packaged with this book.


Internet Connectivity via Satellite with DirecWAY or StarBand If you're in an area where cable TV doesn't exist, or you already have a DirecTV or Dish Network satellite dish, take a look at the southern sky from your home, condo, or apartment building. If you have a good, clear 45° window view to the sky toward the equator and you want fast downloads of big files, a satellite-based service such as DirecWAY (formerly DirecPC) or StarBand might be the high-speed choice for you.


Note Geosynchronous satellites used for satellite Internet/TV service are visible in the southern sky for users in the Northern Hemisphere (North America, Europe, and Asia); if you're in the Southern Hemisphere (South America, Australia, Africa), these satellites are located in the northern sky.

Depending on the product you choose for satellite Internet, you might be able to use a single dish for both satellite Internet and satellite TV.


Tip If you want both high-speed Internet access and satellite TV with a single dish, you can add DirecTV to the DirecWAY dish at any time. The StarBand dish can work with both Dish Network (TV) and StarBand (Internet) services in the continental United States and Canada. However, if you decide to add DirecWAY to an existing DirecTV setup, you will need to replace your existing DirecTV dish unless you installed the larger DirecDUO dish (which works with DirecTV and DirecPC or DirecWAY).

DirecWAY DirecWAY was originally called DirecPC, but Hughes Network Systems renamed it in mid-2001, shortly after rolling out a two-way version of the DirecPC service; this section discusses the two-way service. The original version of DirecPC/DirecWAY, and the only version available until early 2001, is strictly a hybrid system, meaning that incoming and outgoing data streams and other operations are actually routed two different ways: Downloading uses the 400Kbps (peak speed) satellite connection. At top speed, a DirecWAY customer can receive data at speeds about seven times faster than with a 56Kbps modem. Peak traffic loads can slow the satellitebased service download speed. Uploads and Web page requests require the use of a conventional analog modem. This version of DirecWAY is no longer on the market, having been replaced by the current two-way service. For more information on one-way (telco return) DirecPC/DirecWAY, see Upgrading and Repairing PCs, 13th Edition, available on the DVD packaged with this book.

DirecWAY Requirements The DirecWAY service requires you to purchase and install a small satellite dish as part of the necessary hardware. It's similar but not compatible to those used for DirecTV and USSB satellite services. You can add DirecTV to DirecWAY service at any time because the DirecWAY satellite dish is also compatible with DirecTV signals. The dish is connected to what's called a satellite modem, a USB device used to receive data. The current two-way DirecWAY service uses the 35'' wide DirecWAY satellite dish to send and receive data. DirecWAY works with Windows 98SE, Windows 2000, Windows Me, and Windows XP on systems with a 333MHz AMD K6 or Intel Pentium IIâ&#x20AC;&#x2039;class processor or faster. Windows 98SE and Windows Me systems need at least 64MB of RAM, whereas Windows 2000 and Windows XP


systems need at least 128MB of RAM. The software installs from a CD-ROM drive. Your analog modem is used only for initial account setup (a process called commissioning by DirecWAY) or troubleshooting with a two-way system.

Purchasing DirecWAY Service DirecWAY can be purchased from several partners in the United States, including these: Earthlink. Earthlink Satellite Powered by DirecWAY; its Web site is at www.earthlink.com. DirecTV. Its Web site is at Directv.direcway.com. AgriStar. Its Web site is at www.agristar.com. National Rural Telecommunications Cooperative (NRTC). Its Web site is at www.nrtc.org; call your rural electric/telephone cooperative or independent telephone company for details. Depending on the vendor you choose and the payment plan offered, you might be able to spread the cost of the satellite dish and satellite modem out over an extended time, instead of paying for all the equipment up front. You also must have the system professionally installed. The cost might be included in your system, or there might be an additional charge, although some vendors provide price breaks on equipment or installation. Monthly service charges are around $70/month when equipment is purchased up-front; charges can vary. Check the Get DirecWAY Service link at the DirecWAY Web site (www.direcway.com) for the latest vendor and pricing information.

DirecWAY's FAPâ&#x20AC;&#x2039;Brakes on High-Speed Downloading? A big concern for those wanting to exploit the high-speed download feature continues to be DirecWAY's Fair Access Policy (FAP), which was introduced long after the original DirecPC service was started. FAP uses unpublished algorithms to determine who is "abusing" the service with large downloads. Abusers have their download bandwidths reduced by about 50% or more until their behavior changes. A class-action lawsuit was filed in July 1998 by DirecPC users who objected to this policy. Users said that Hughes Network Systems, Inc., the developer of DirecPC, was simultaneously selling the system on the basis of very


fast download times and then punishing those who wanted to use it in the way DirecPC had sold it to the public. As a result of the class-action lawsuit, DirecWAY partners' Web sites now offer usage guidelines that are supposed to help you avoid being "FAPped." However, many DirecPC and DirecWAY users whose comments are available on the alt.satellite.direcpc newsgroup (also available via the groups.google.com Web site) complain that the guidelines are misleading. For example, the current guidelines for residential customers state that if you download more than 169MB during the period from 5 a.m. to 2 a.m. the following morning (21 hours), you will exceed the maximum usage threshold for FAP, and your download speed will be slowed down to approximately 47Kbps for about eight hours. The limit for off-peak hours (2 a.m. to 5 a.m.) is about 225MB; limits are higher for business customers. For a more detailed discussion of the real-world impact of FAP on both residential and business users and software you can use to track downloading, see the Fair Access Policy page at http://www.copperhead.cc/fap.html.


Note I'm concerned about FAP limitations when it's time to download the increasingly bulky service packs for operating systems, Web browsers, and office suites. Some users have reported their service speeds dropping by 50% or more after downloading just 40MB of files. At a time when a single service pack can be about 40MB by itself, this isn't good news. Because DirecWAY refuses to release details of its FAPcalculating algorithm and its public guidelines seem misleading, you should probably avoid downloading a lot of large files in sequence if you want to avoid being FAPped. Because different satellites are used by the different DirecWAY partners, it pays to research which DirecWAY versions perform best. In addition to the DirecPC/DirecWAY newsgroup, check out the DirecPC Uncensored! Web site at www.copperhead.cc for speed tests and tweaks you can make to your system. You can also use the freeware FAPBuster software available from http://home.mindspring.com/~testftptest/ to help track usage and determine when you are close to exceeding FAP.

DirecWAY's Real-World Performance Benchmark addicts will find that DirecWAY performs poorly on ping tests because the complex pathway your data must travel (ground to space and back again) results in pings taking at least 400ms​600ms. Interactive benchmarks are also disappointing. The delays caused by communicating with a geosynchronous satellite over 22,500 miles in space make two-way DirecWAY a poor choice for these applications, although download speeds can be much faster than dial-up modems. Speeds vary widely, but according to DirecPC Uncensored!, DirecWAY can reach download speeds over 2,000Kbps. To achieve results like this, use the tips available on the DirecPC Uncensored! Web site to adjust your system's configuration.

StarBand In April 2000, StarBand​the first consumer-oriented two-way satellite network​was introduced after being tested as Gilat-At-Home. StarBand uses an external USB modem and a satellite dish that supports both StarBand Internet and Dish Network satellite TV. In fact, the feature set of DirecWAY in its current two-way form is almost identical to StarBand. StarBand and DirecWAY services work as shown in Figure 19.3.

Figure 19.3. The StarBand service can receive both Dish Network TV programs (left) and StarBand Internet traffic (right) with its single 24''x36'' satellite dish. Two-way DirecWAY service also supports both TV and Internet on its single dish and works in a similar fashion.


StarBand provides download speeds ranging from 150Kbps to 1,000Mbps (1Gbps) and upload speeds ranging from 50Kbps to 150Kbps or higher, depending on the satellite modem used. StarBand's satellite modems include the new StarBand Model 480Pro, released in early 2003. The 480Pro contains a four-port router and can be used with non-Windows operating systems. The Model 360 is the standard model for residential users; it's a smaller, faster model than the original Model 180 and can be interfaced through either USB or Ethernet ports. StarBand supports Windows 98/98SE, Windows Me, Windows 2000, and Windows XP. StarBand has partnered with several other companies, including W.A.Y.S. Inc. LLC (www.waysinc.com), SIA (satellite-internet-access.net), and US Online (www.usonline.com). StarBand equipment pricing and monthly service fees are generally similar to DirecWAY two-way's price structure, although some vendors might offer special promotional packages and bundles.


Tip You can find excellent tips, tricks, utility software, and user-provided help at the StarBand Users Web site: www.starbandusers.com.


Integrated Services Digital Network The connection speed of analog modems is limited by Shannon's Law (see the section "56Kbps Modems," later in this chapter). To surpass the speed limitations of analog modems, you need to use digital signals. Integrated Services Digital Network (ISDN) was the first step in the move to digital telecommunications. With ISDN, you can connect to the Internet at speeds of up to 128Kbps. Because ISDN was developed by the telephone companies, you can purchase a variety of service plans. Depending on the ISDN service you choose, you can use it strictly for Internet service or use it to service multiple telephony applications such as voice, fax, and teleconferencing. Depending on where you live, you might find that ISDN service is available for Internet uses, or your local telco might offer faster DSL service as an alternative. Because ISDN was not originally designed for Internet use, its speed is much lower than other broadband options. Also, ISDN costs about twice what a typical ADSL or cable modem connection costs per month. ISDN doesn't require as high a line quality as DSL, so it can be offered in areas where DSL can't work without a major upgrade of the telephone system.

How Standard ISDN Works Because ISDN carries three channels, it allows integrated services that can include combinations such as voice+data, data+data, voice+fax, fax+data, and so on (see Figure 19.4).

Figure 19.4. An analog modem connection (top) connects only your PC to the Internet or other online services, whereas ISDN (bottom) can connect your computer, fax, and many other devices via a single ISDN terminal adapter.


On a standard ISDN connection, bandwidth is divided into bearer channels (B channels) that run at 64Kbps and a delta channel (D channel) that runs at either 16Kbps or 64Kbps, depending on the type of service. The B channels carry voice transmissions or user data, and the D channel carries control traffic. In other words, you talk, surf, or fax through the B-channel lines. Two types of ISDN service exist: basic rate interface (BRI) and primary rate interface (PRI). The BRI service is intended for private and home users and consists of two B channels and one 16Kbps D channel, for a total of 144Kbps. The typical BRI service enables you to use one B channel to talk at 64Kbps and one B channel to run your computer for Web surfing at 64Kbps. Hang up the phone, and both B channels become available. If your ISDN service is configured appropriately, your Web browsing becomes supercharged because you're now running at 128Kbps. The PRI service is oriented more toward business use, such as for PBX connections to the telephone company's central office. In North America and Japan, the PRI service consists of 23 B channels and 1 64Kbps D channel for a total of 1,536Kbps, running over a standard T-1 interface. In Europe, the PRI service is 30 B channels and 1 64Kbps D channel, totaling 1,984Kbps, which corresponds to the E1 telecommunications standard. For businesses that require more bandwidth than one PRI connection provides, 1 D channel can be used to support multiple PRI channels using non-facility associated signaling (NFAS). The BRI limit of two B channels might seem limiting to anyone other than a small office or home office user, but this is misleading. The BRI line can actually accommodate up to eight ISDN devices, each with a unique ISDN number. The D channel provides call routing and "on-hold" services, also called multiple call signaling, allowing all the devices to share the two B channels.


Note When speaking of ISDN connections, 1 kilobyte equals 1,000 bytes, not 1,024 bytes as in standard computer applications. As you saw earlier, this is also true of speed calculations for modems. Calculations that use 1,000 as a base are often referred to as decimal kilobytes, whereas the ones based on 1,024 are now called kibibytes or binary kilobytes.

If you need a more powerful, more flexible (and more expensive) version of ISDN, use the PRI version along with a switching device, such as a PBX or server. Although PRI allows only one device per B channel, it can dynamically allocate unused channels to support high-bandwidth uses, such as videoconferencing, when a switching device is in use along with PRI.

Acquiring ISDN Service To have an ISDN connection installed, you must be within 18,000 wire feet (about 3.4 miles or 5.5km) of the CO (telco central office or central switch) for the BRI service; wire feet refers to the distance traveled by the telephone wires serving your location, not straight-line distance. For greater distances, expensive repeater devices are needed, and some telephone companies might not offer the service at all. Prices for ISDN service vary widely depending on your location. In the United States, the initial installation fee can range from $35 to $150, depending on whether you are converting an existing line or installing a new one. The monthly charges typically range from $30 to $50, and sometimes you must pay a connecttime charge as well, ranging from 1 to 6 cents per minute or more, depending on the state. Keep in mind that you also must purchase an ISDN terminal adapter for your PC and possibly other hardware as well, and these charges are only for the telephone company's ISDN service. In addition, you must pay your ISP for access to the Internet at ISDN speeds. Typically, when all charges are included, you can pay up to $100 or more per month for ISDN service in a residential setting, and more for a small business connection. Residential plans are often dial-up, requiring you to make a connection to the ISP's server every time you want to go online, whereas business plans are usually always-on, with immediate connection.


Note Although ISDN Internet access provided by your local telephone company usually has a single price for the ISDN line and ISDN Internet access, most third-party ISDN ISPs provide pricing for only Internet access. These costs might appear to be much less expensive at startup and per month than what the telco's ISDN package costs, but this is misleading because the telco's charges for ISDN service aren't included. Add up the costs from both the ISP and the telephone company for a true picture of thirdparty ISDN Internet service pricing.

Because ISDN pricing plans offer many options depending on the channels you want and how you want to use them, be sure you carefully plan how you want to use ISDN. Check the telco's Web site for pricing and package information to get a jump on the decision-making process. Although ISDN is unique among broadband Internet services for its capability to handle both voice and data traffic, its relatively high cost and low speed make it a poor choice for most small-office and home-office users.

ISDN Hardware To connect a PC to an ISDN connection, you must have a hardware component called a terminal adapter (TA). The terminal adapter takes the form of an expansion board or an external device connected to a serial port, much like a modem. In fact, terminal adapters often are mistakenly referred to as ISDN modems. Actually, they are not modems at all because they do not perform analog/digital conversions. Because an ISDN connection originally was designed to service telephony devices, most ISDN terminal adapters have connections for telephones, fax machines, and similar devices, as well as for your computer. Some terminal adapters can also be used as routers to enable multiple PCs to be networked to the ISDN connection.


Caution To achieve the best possible performance, you should either purchase an external ISDN terminal adapter that connects to your computer's USB port or use an internal version. A terminal adapter with compression enabled easily can exceed a serial port's capability to reliably send and receive data. Consider that even a moderate 2:1 compression ratio exceeds the maximum rated speed of 232Kbps, which most high-speed COM ports support. USB 1.1 ports, on the other hand, can handle signals up to 12Mbps, easily supporting even the fastest ISDN connection. USB 2.0 ports available on the latest computers support speeds up to 480Mbps.


Note For more information about ISDN hardware and configuration, see Upgrading and Repairing PCs, 11th Edition, available in electronic form on the DVD-ROM packaged with this book.


Comparing High-Speed Internet Access One way of making sense out of the confusing morass of plans available from cable modem, DSL, fixed wireless Internet, and satellite vendors is to calculate the average cost per Kbps of data downloaded ($/Kbps). You can calculate this figure yourself by dividing the service cost ($SC) per month by the rated or average speed of the service ($SPD): $SC / $SPD = $/Kbps For example, a typical cable modem service costs $50 per month, including cable modem lease, and has an average (not peak) speed of 500Kbps. Divide $50 by 500Kbps, and the cost per Kbps equals 10 cents. Use this formula with any broadband or dial-up service to find the best values. Don't forget to calculate the cost of required equipment (as in the example). If you must pay for equipment or installation upfront​as you will need to do with satellite, fixed wireless, and ISDN Internet plans​divide the upfront cost by the number of months you plan to keep the service and add the result to the monthly service charge to get an accurate figure. How does a typical 56Kbps modem compare, assuming 50Kbps download speeds? Using Juno Web ($14.95 per month) and assuming no charge for an analog modem, the cost per Kbps is 29.9 cents per Kbps​almost three times as much for service that is at least 10 times slower than a typical cable modem. Generally, the services stack up as shown in Table 19.3, from slowest to fastest when download speeds are compared. Table 19.3. Comparing Typical Speeds for Various Types of Internet Connections Connection Type

Speed

V.34 annex analog modem

33.6Kbps

V.90/V.92 analog modem

53Kbps (due to FCC regulations; hardware capable of 56Kbps)

ISDN (1BRI)

64Kbps

ISDN (2BRI)

128Kbps

ADSL

384Kbps [1]

DirecWAY or StarBand (two-way)

500Kbps


Cable modem or fixed wireless

512Kbps [2]

ADSL

512Kbps [1]

ADSL

1GB[1]

Cable modem or fixed wireless

1.5Gbps [2]

The values in this table are estimated ratings from the vendor; substitute average actual values when available.

[1] DSL bandwidth depends on the package you select; higher bandwidth packages carry higher monthly

fees. [2] Cable modem/wireless bandwidth can depend on the package you select; higher bandwidth packages

carry higher monthly fees. Also, speeds can vary with network traffic; ask vendor for details.

Another way to compare Internet connection types is by feature, as in Table 19.4. Table 19.4. Comparing High-Speed Internet Access by Feature Always Shared with On? Other Users?

Ties Up Reliability Affected By? Phone Line?

Cable modem (two-way)

Yes

Yes

No

Cable outages

Cable modem (one-way)

No

Yes

Yes

Cable outages; phone line Ethernet or USB; might need outages external analog modem

Fixed wireless Yes (two-way)

Yes

No

Transmitter outages

Ethernet

Fixed wireless No (one-way)

Yes

Yes

Transmitter outages; phone line outages

Ethernet or PCI slot; might need external analog modem

DirecWAY (two-way)

Yes

Yes

No

Satellite outages

USB

StarBand (two-way)

Yes

Yes

No

Satellite outages

USB

DSL

Yes

No

No

Phone line outages; telco Ethernet or USB network changes

Service

Connect Type

Ethernet or USB

Having a Backup Plan in Case of Service Interruptions


Because no high-speed connection is immune to service interruptions, you should consider having some type of backup plan in place in case of a significant service outage. If your high-speed Internet access uses an ISP that can also accept 56Kbps connections, you might still be able to use your regular modem for emergencies. However, this might require an extra charge in some cases. You could also consider using a free trial subscription to an ISP that uses a conventional modem. If you temporarily switch to a different ISP​especially one that uses its own client, such as AOL​be sure to back up your current Internet configuration information before you install the client software. Your best bet is to use an Internet-only ISP whose dial-up connection can be configured manually with the Dial-Up Networking Connection Wizard or Network Setup Wizard in XP. Then, you can construct a new connection without destroying your existing configuration. If you don't want to spend $15​$25/month for backup service, or if you travel occasionally and want a low-cost way to work online when you're away from broadband, consider the following prepaid services, which let you purchase blocks of time as desired: Slingshot (www.slingshot.com) AOL PrePaid Card (www.aol.com/prepaid_aol/index.adp) MaGlobe Prepaid Global Internet (www.maglobe.com) Because prices and local access numbers vary, check with the vendors before you purchase a starter kit to ensure that coverage is available in the areas you prefer and to verify that a toll-free number option (which uses prepaid service time at a faster rate) is available as an alternative.


Note Each type of Internet connection uses a particular combination of TCP/IP settings. TCP/IP is the protocol (software rules) used by all computers on the Internet. TCP/IP is covered in Chapter 20, but for now keep in mind that different TCP/IP settings are required for modem access and access through a NIC or USB port device (cable modem, DSL, and DirecWAY or StarBand). Modems usually have an IP address provided dynamically by the ISP when the modem connects with the ISP. The other types of Internet access devices might have static IP addresses that don't change or have dynamically assigned IP addresses. IP addresses are just one of the network settings that, if changed, prevent you from connecting to the Internet.


Leased Lines For users with high bandwidth requirements (and deep pockets), dedicated leased lines provide digital service between two locations at speeds that can far exceed ISDN and are as fast or faster than DSL or cable modem. A leased line is a permanent 24-hour connection to a particular location that can be changed only by the telephone company. Businesses use leased lines to connect LANs in remote locations or to connect to the Internet through a service provider. Leased lines are available at various speeds, as described in the following sections.

T-1 and T-3 Connections To connect networks in distant locations, networks that must support a large number of Internet users, or especially organizations that will be hosting their own Internet services, a T-1 connection might be a wise investment. A T-1 is a digital connection running at about 1.5Mbps. This is more than 10 times faster than an ISDN link and is more than double the speed of most fast DSL connections. A T-1 can be split (or fractioned), depending on how it is to be used. It can be split into 24 individual 64Kbps lines or left as a single high-capacity pipeline. Some ISPs allow you to lease any portion of a T-1 connection that you want (in 64Kbps increments). Ameritech, for example, offers a flexible T-1 service it calls DS1; it's available at full bandwidth or in various fractional sizes. Figure 19.5 shows how a T-1 line is fractioned.

Figure 19.5. Full T-1 service uses all 24 lines (each one is 64Kbps) as a single pipeline; a fractional T-1 service of 256Kbps could use slots 1â&#x20AC;&#x2039;4 only, for example.

An individual user of the Internet interacts with a T-1 line only indirectly. No matter how you're accessing the Internet (dial-up modem, ISDN, DSL, cable modem, DirecWAY, StarBand, or fixed-base wireless), your ISP typically will have a connection to one or more T-1 or T-3 lines, which connect to the backbone of the Internet. This connection to the backbone is sometimes referred to as a point


of presence (PoP). When you make your connection to the Internet, your ISP shares a small chunk of that T-1 pipe with you. Depending on how many other users are accessing the Internet at your ISP or elsewhere, you might experience very fast to slow throughput, even if your modem connection speed remains constant. It's a bit like splitting up a pizza into smaller and smaller slices to accommodate more people at a party: The more users of a high-speed connection, the slower each individual part of it will be. To keep user connections fast while growing, ISPs add full or fractional T-1 lines to their points of presence. Or, they might switch from a T-1 connection to the even faster T-3 if available.


Note Equivalent in throughput to approximately 28 T-1 lines, a T-3 connection runs at 45Mbps and is suitable for use by very large networks and university campuses. Pricing information falls into the "if-you-haveto-ask-you-can't-afford-it" category.

If your Internet connection is on a corporate LAN or your office is located in a downtown building, your relationship to a T-1 line might be much closer. If your building or office is connected directly to a T-1, you're sharing the capacity of that line with just a relatively few other users rather than with the hundreds or thousands of dial-up users a normal ISP is hosting at one time. Full or fractional T-1 lines are being added to more and more apartments and office buildings in major cities to allow residents and workers faster Internet access. In these cases, a LAN connection to the T-1 is usually provided, so your Internet access device is a network card, rather than a modem or ISDN terminal adapter. With the rise of the Internet and the demand for high-speed data access for networks, the price of T-1 links in the United States has fallen drastically since the late 1990s, although you will still pay in the hundreds of dollars for typical service offerings. T-1 service can be acquired from either your local telco or thirdparty firms. Fractional T-1 or burstable T-1 (which allows you to have differing levels of bandwidth up to the entire T1 1.5Mbps depending on demand) costs less than full T-1 service. For a large organization that requires a lot of bandwidth, the lower cost of T-1 services today make installing a higher-capacity service and growing into itâ&#x20AC;&#x2039;rather than constantly upgrading the linkâ&#x20AC;&#x2039;more economical than ever. Although the speed of T-1 links resembles the maximum rates available with DSL or cable modem service, most types of T-1 service provide constant bandwidth (unlike cable modems) and bypass the potentially severe problems of trying to retrofit old phone lines with digital service (unlike DSL).

Comparing Conventional High-Speed Services Some telcos who formerly posted pricing for ISDN, T-1, or other high-end telecommunications services now have a "call us" button on their Web sites because pricing is complicated by many factors, including Location (state and locality because telephone companies are regulated public utilities) Fixed and variable costs Usage


Installation costs Your needs Be sure to consider hardware and usage costs when you price services, and (for items such as ISDN terminal adapters and network cards) compare the official offerings with products available elsewhere. If you decide to provide some of the equipment yourself, find out whose responsibility repairs become. Some companies provide lower-cost "value" pricing for services in which you agree to configure the hardware yourself and maintain it. If you have knowledgeable staffers who can handle routers and other network configuration, you can save money every month, but if not, go with the full-service option.


Securing Your Internet Connection Because any Internet connection must use TCP/IP, which uses built-in logical ports numbered 0â&#x20AC;&#x2039;65,535 to service different types of activity, any user of the Internet can be vulnerable to various types of Internet attacks, even if precautions such as turning off drive and folder sharing have been followed. Such vulnerabilities increase drastically with always-on broadband services such as DSL and cable modems. Steve Gibson of Gibson Research Corporation (makers of the classic SpinRite disk maintenance program) has established a free Web-based service called Shields Up that you should try with any Internet-connected PC to see how safe or vulnerable you are. The Shields Up portion of the Gibson Research Corporation Web site (http://www.grc.com) probes your system's Internet connection security and Internet service ports. After testing your system, Shields Up provides reviews and recommendations for proxy server and firewall software (such as ZoneAlarm, Norton Internet Security, and Sygate Personal Firewall) that you can use to help secure your system. With the increasing importance of the Internet and the multiple vulnerabilities we've seen since 2000 to Internet-borne viruses, Trojan horses, and denial-ofservice attacks, Shields Up provides a valuable service for any Internet user. Chapter 20 discusses how using a router to share your Internet connection can also help protect your system against intruders.


Note Proxy servers and firewalls are subjects that go far beyond the scope of this book. If you want to learn more, I suggest you pick up a copy of Upgrading and Repairing Networks also published by Que. Que's Absolute Beginner's Guide to Personal Firewalls by Jerry Lee Ford is another excellent resource for users who want to install a personal firewall, such as ZoneAlarm or BlackICE PC Protector, instead of a hardware firewall.


Asynchronous (Analog) Modems If you want to connect to the Internet without spending a lot of money, an analog modem can serve as your on-ramp to the rest of the computing world. Modems are standard equipment with most recent systems and continue to be popular upgrades for systems that do not have access to broadband solutions, such as two-way cable modem or DSL lines. Even with some types of broadband access (such as One-way DirecWAY and one-way cable modem), modems are still needed to send page requests and email. The word modem (from modulator/demodulator) basically describes a device that converts the digital data used by computers into analog signals suitable for transmission over a telephone line and converts the analog signals back to digital data at the destination. To distinguish modems that convert analog and digital signals from other types of access devices, modems are frequently referred to as analog modems; because you must dial a telephone number to reach a remote computer, they are also referred to as dial-up modems. The typical PC modem is an asynchronous device, meaning it transmits data in an intermittent stream of small packets. The receiving system takes the data in the packets and reassembles it into a form the computer can use.


Note Because it has become such a familiar term, even to inexperienced computer users, the word modem is frequently used to describe devices that are, strictly speaking, not modems at all. For example, earlier in this chapter you read about broadband solutions such as ISDN, cable modems, DirecWAY, DSL, and StarBand. Although all these services use devices commonly called "modems" to connect your PC to fast online services, none of them converts digital information to analog signals. However, because these devices look similar to a standard modem and are used to connect PCs to the Internet or to other networks, they are called modems.

Asynchronous modems transmit each byte of data individually as a separate packet. One byte equals 8 bits, which, using the standard ASCII codes, is enough data to transmit a single alphanumeric character. For a modem to transmit asynchronously, it must identify the beginning and end of each byte to the receiving modem. It does this by adding a start bit before and after every byte of data, thus using 10 bits to transmit each byte (see Figure 19.6). For this reason, asynchronous communications have sometimes been referred to as start-stop communications. This is in contrast to synchronous communications, in which a continuous stream of data is transmitted at a steady rate.

Figure 19.6. Asynchronous modems frame each byte of data with a start bit and a stop bit, whereas synchronous communications use an uninterrupted stream of data.

Synchronous modems generally are used in leased-line environments and in conjunction with multiplexers to communicate between terminals to Unix- or Linux-based servers and mainframe computers. Thus, this type of modem is outside the scope of this book. Whenever modems are referred to in this book, I will be discussing the asynchronous, analog variety. (Synchronous modems are not found in typical computer stores and aren't included in normal computer configurations, so you might not ever see one unless you go into the data center of a corporation that uses them.)


Note During high-speed modem communications, the start and stop bits are usually not transmitted over the telephone line. Instead, the modem's data compression algorithm eliminates them. However, these bits are part of the data packets generated by the communications software in the computer, and they exist until they reach the modem hardware. If both ends of an analog modem connection don't use the same value for start and stop bits, the connection transmits gibberish instead of usable data.

The use of a single start bit is required in all forms of asynchronous communication, but some protocols use more than one stop bit. To accommodate systems with different protocols, communications software products usually enable you to modify the format of the frame used to transmit each byte. The standard format used to describe an asynchronous communications format is parity/data bits/stop bits. Almost all asynchronous connections today are therefore abbreviated as N-8-1 (No parity/8 data bits/1 stop bit). The meanings for each of these parameters and their possible variations are as follows: Parity. Before error-correction protocols became standard modem features, a simple parity mechanism was used to provide basic error checking at the software level. Today, this is almost never used, and the value for this parameter is nearly always set to none. Other possible parity values you might see in a communications software package are odd, even, mark, and space. Data Bits. This parameter indicates how many bits are actually carried in the data portion of the packet (exclusive of the start and stop bits). PCs typically use 8 data bits, but some types of computers use a 7-bit byte, and others might call for other data lengths. Communications programs provide this option to prevent a system from confusing a stop bit with a data bit. Stop Bits. This parameter specifies how many stop bits are appended to each byte. PCs typically use 1 stop bit, but other types of protocols might call for the use of 1.5 or 2 stop bits. In most situations, you will never have to modify these parameters manually, but the controls are almost always provided. In Windows 9x/Me/2000/XP, for example, if you open the Modems control panel and look at the Connection page of your modem's Properties dialog box, you will see Data Bits, Parity, and Stop Bits selectors. Unless you use the Windows HyperTerminal program to establish a direct connection to another computer via phone lines, you might never need to modify these parameters. However, if you need to call a mainframe computer to perform terminal emulation for e-banking, checking a library's catalog, or working from home, you might need to adjust these parameters. (Terminal emulation means


using software to make your PC keyboard and screen act like a terminal, such as a DEC VT-100 and so on.) Many mainframe computers use even parity and a 7-bit word length. If your PC is set incorrectly, you'll see garbage text on your monitor instead of the other system's login or welcome screen.


Modem Standards For two modems to communicate, they must share the same protocol. A protocol is a specification that determines how two entities will communicate. Just as humans must share a common language and vocabulary to speak with each other, two computers or two modems must share a common protocol. In the case of modems, the protocol determines the nature of the analog signal the device creates from the computer's digital data. Bell Labs (which set standards for early 300bps modems) and the CCITT are two of the bodies that have set standards for modem protocols. CCITT is an acronym for Comité Consultatif International Téléphonique et Télégraphique, a French term that translates into English as the Consultative Committee on International Telephone and Telegraph. The organization was renamed the International Telecommunication Union (ITU) in the early 1990s, but the protocols developed under the old name are often referred to as such. Newly developed protocols are called ITU-T standards, which refers to the Telecommunication Standardization Sector of the ITU. Most modems built in recent years conform to the standards developed by the CCITT/ITU. The ITU, headquartered in Geneva, Switzerland, is an international body of technical experts responsible for developing data communications standards for the world. The group falls under the organizational umbrella of the United Nations, and its members include representatives from major modem manufacturers, common carriers (such as AT&T), and governmental bodies. The ITU establishes communications standards and protocols in many areas, so one modem often adheres to many different standards, depending on its various features and capabilities. All modems sold today support the following ITU protocols: ITU V.90 (modulation) ITU V.42 (error correction) ITU V.42bis (data compression) However, earlier modems supported many industry-standard and proprietary protocols for modulation, error correction, and data compression. Most modems today also support the proprietary Microcom Network Protocol (MNP) MNP10 and MNP10EC error-correction standards to provide better connection during conventional wired and wireless (cellular) communication


sessions. The latest modems also support the newest ITU standards: V.92 (modulation) and V.44 (data compression). All current protocols are discussed later in this chapter.


Note To learn more about earlier industry-standard and proprietary protocols, see Chapter 18 of Upgrading and Repairing PCs, 11th Edition, available in electronic form on the DVD-ROM packaged with this book.


Note The term protocol is also used to describe software standards that must be established between different computers to allow them to communicate, such as TCP/IP.

Modems are controlled through AT commands, which are text strings sent to the modem by software to activate the modem's features. For example, the ATDT command followed by a telephone number causes the modem to dial that number using tone dialing mode. Applications that use modems typically generate AT commands for you, but you can control a modem directly using a communications program with a terminal mode or even the DOS ECHO command. Because almost every modem uses the AT command set (originally developed by modem-maker Hayes), this compatibility is a given and should not really affect your purchasing decisions about modems. The basic modem commands might vary slightly from manufacturer to manufacturer, depending on a modem's special features, but the basic AT command set is all but universal.


Note A list of the basic AT commands can be found in the Technical Reference on the DVD included with this book. However, the best source for the commands used by your modem is the manual that came with the device. Although most modem users will never need to review these commands, if you use MS-DOSâ&#x20AC;&#x2039;based communication programs or some specialized Windows programs, you might be required to enter or edit an initialization string, which is a series of AT commands sent to the modem before dialing. If these commands are not correct, the modem will not work with these programs.

Bits and Baud Rates When discussing modem transmission speeds, the terms baud rate and bit rate are often confused. Baud rate (named after a Frenchman named Emile Baudot, the inventor of the asynchronous telegraph printer) is the rate at which a signal between two devices changes in 1 second. If a signal between two modems can change frequency or phase at a rate of 300 times per second, for example, that device is said to communicate at 300 baud. Thus, baud is a signaling rate, not a data-transmission rate. The number of bits transmitted by each baud is used to determine the actual data-transmission rate (properly expressed as bps or Kbps). Modern pure-analog (33.6Kbps and slower) modems transmit and receive more bits per baud than the original 300bps modems (which also ran at 300 baud).


Note To learn more about bits versus baud, see Chapter 18 of Upgrading and Repairing PCs, 11th Edition, available in electronic form on the DVD packaged with this book.

Modulation Standards Modems start with modulation, which is the electronic signaling method used by the modem. Modulation is a variance in some aspect of the transmitted signal. By modulating the signal using a predetermined pattern, the modem encodes the computer data and sends it to another modem that demodulates (or decodes) the signal. Modems must use the same modulation method to understand each other. Each data rate uses a different modulation method, and sometimes more than one method exists for a particular rate. Regardless of the modulation method, all modems must perform the same task: Change the digital data used inside the computer (ON-OFF, 1-0) into the analog (variable tone and volume) data used by the telephone company's circuits, which were built over a period of years and were never intended for computer use. That's the "mo(dulate)" in modem. When the analog signal is received by the other computer, the signal is changed back from the analog waveform into digital data (see Figure 19.7). That's the "dem(odulate)" in modem.

Figure 19.7. The modem at each computer changes digital (computer) signals to analog (telephone) signals when transmitting data or analog back to digital when receiving data.

The three most popular modulation methods are as follows: Frequency-shift keying (FSK). A form of frequency modulation, otherwise known as FM. By causing and monitoring frequency changes in a signal sent


over the phone line, two modems can send information. Phase-shift keying (PSK). A form of phase modulation in which the timing of the carrier signal wave is altered and the frequency stays the same. Quadrature amplitude modulation (QAM). A modulation technique that combines phase changes with signal-amplitude variations, resulting in a signal that can carry more information than the other methods. All modem protocols since ITU V.34 (33.6Kbps maximum speed) up through the current ITU V.90 and ITU V.92 standards (56Kbps maximum speed) are full-duplex protocols. A full-duplex protocol is one in which communications can travel in both directions at the same time and at the same speed. A telephone call, for example, is full duplex because both parties can speak at the same time. In half-duplex mode, communications can travel in both directions, but only one side can transmit at a time. A radio call in which only one party can speak at a time is an example of half-duplex communications. These protocols are automatically negotiated between your modem and the modem at the other end of the connection. Basically, the modems start with the fastest protocol common to both and work their way down to a speed/protocol combination that will work under the line conditions existing at the time of the call. The ITU V.90 and V.92 protocols are the industry-standard protocols most commonly used today; V.92 modems also support V.90.


Note The 56Kbps standards that represent the highest current increment in modem communication speed require a digital connection at one end and are therefore not purely analog. Other high-speed communication technologies such as ISDN and cable network connections don't perform digital-toanalog conversions, so they should not be called modems, strictly speaking.

V.90 V.90 is the ITU-T designation for a 56Kbps communication standard that reconciles the conflict between the proprietary U.S. Robotics (3Com) x2 and Rockwell K56flex modem specifications developed in 1996 and 1997. The last ISA modems manufactured by major vendors typically support V.90, as do many PC Card and PCI modems built from 1998 to 2001. See "56Kbps Modems," p. 1053.

V.92 V.92 is the ITU-T designation for an improved version of the V.90 standard that provides faster negotiation of the connection, call-waiting support, and faster uploading than is possible with V.90. Most PCI and PC Card modems sold by major vendors since mid-2001 to the present are V.92 compatible. See "56Kbps Modems," p. 1053.

V.90 and V.92 are the current communication protocols supported by ISPs; any modem you want to use today should support at least the V.90 protocol.

Error-Correction Protocols Error correction refers to the capability of some modems to identify errors during a transmission and to automatically resend data that appears to have been damaged in transit. Although you can implement error correction using software,


this places an additional burden on the computer's expansion bus and processor. By performing error correction using dedicated hardware in the modem, errors are detected and corrected before any data is passed to the computer's CPU. As with modulation, both modems must adhere to the same standard for error correction to work. Fortunately, most modem manufacturers use the same errorcorrection protocols.

V.42, MNP10, and MNP10EC Error-Correction Protocols The current error-correction protocols supported by modems include Microcom's proprietary MNP10 (developed to provide a better way to cope with changing line conditions) and MNP10EC (an enhanced version developed to enable modems to use constantly changing cellular telephone connections). V.90 and V.92 modems (as well as some older models) also support the ITU V.42 error-correction protocol, with fallback to the MNP 4 protocol (which also includes data-compression). Because the V.42 standard includes MNP compatibility through Class 4, all MNP 4â&#x20AC;&#x2039;compatible modems can establish error-controlled connections with V.42 modems. This standard uses a protocol called Link Access Procedure for Modems (LAPM). LAPM, similar to MNP, copes with phone-line impairments by automatically retransmitting data corrupted during transmission, ensuring that only error-free data passes between the modems. V.42 is considered to be better than MNP 4 because it offers approximately a 20% higher transfer rate due to its more intelligent algorithms.


Note For information about the MNP 1â&#x20AC;&#x2039;4 protocols, see Chapter 18 of Upgrading and Repairing PCs, 11th Edition, included in electronic form on the DVD packaged with this book.

Data-Compression Standards Data compression refers to a built-in capability in some modems to compress the data they're sending, thus saving time and money for modem users. Depending on the type of files the modem is sending, data can be compressed to nearly onefourth its original size, effectively quadrupling the speed of the modemâ&#x20AC;&#x2039;at least in theory. This assumes that the modem has V.42bis data compression built in (true since about 1990) and that the data hasn't already been compressed by software. Thus, in reality, the higher throughput caused by data compression applies only to HTML and plain-text files on the Web. Graphics and Zip or EXE archives have already been compressed, as have most PDF (Adobe Acrobat Reader) files. Another factor that influences the throughput of a modem is the type of UART chip used by the serial port included in an internal modem or connected to an external modem, or the use of a USB port instead of a serial port. To learn more, see "Can Non-56Kbps Modems Achieve Throughput Speeds Above 115,200bps?" in the Technical Reference section of the DVD-ROM packaged with this book. As with error correction, data compression can also be performed with software. Data can be compressed only once, so if you are transmitting files that are already in a compressed form, such as Zip archives, GIF or JPEG images, or Adobe Acrobat PDF files, there will be no palpable increase in speed from the modem's hardware compression. The transmission of plain-text files (such as HTML pages) and uncompressed bitmaps, however, is accelerated greatly by modem compression.

MNP5 and V.42bis Current data-compression standards in modems include Microcom's MNP 5 and the ITU V.42bis protocols. V.42bis is a CCITT data-compression standard similar to MNP Class 5, but it provides about 35% better compression. V.42bis is not actually compatible with MNP Class 5, but nearly all V.42bis modems include the MNP 5 data-compression capability as well. V.42bis is superior to MNP 5 because it analyzes the data first and then determines whether compression would be useful. V.42bis compresses only data that needs compression. MNP 5, on the other hand, always attempts to compress


the data, which slows down throughput on previously compressed files. To negotiate a standard connection using V.42bis, V.42 also must be present. Therefore, a modem with V.42bis data compression is assumed to include V.42 error correction. When combined, these two protocols result in an error-free connection that has the maximum data compression possible.

V.44 At the same time that the V.92 protocol was introduced by the ITU in mid-2000, a companion data-compression protocol called V.44 was also introduced by the ITU. V.44 uses a new lossless LZJH compression protocol designed by Hughes Network Systems (developers of the DirecWAY satellite broadband Internet service) to achieve performance more than 25% better than that of V.42. Data throughput with V.44 can reach rates of as much as 300Kbps, compared to 150Kbpsâ&#x20AC;&#x2039;200Kbps with V.42bis. V.42bis was developed in the late 1980s, long before the advent of the World Wide Web, so it is not optimized for Web surfing the way V.44 is. V.44 is especially designed to optimize compression of HTML text pages.


Note V.44 is the latest compression algorithm to be based in part on the work of mathematicians Abraham Lempel and Jakob Ziv in the late 1970s. Lempel and Ziv's work also has been used in the development of LZW (Lempel-Ziv-Welch) compression for TIFF image files, GIF compressed image files, PKZIP-compatible compression, and other data compression methods.

Proprietary Standards In addition to the industry-standard protocols for modulation, error correction, and data compression that are generally defined and approved by the ITU-T, several protocols in these areas were invented by various companies and included in their products without any official endorsement by any standards body. Some of these protocols have been quite popular at times and became pseudo-standards of their own. The only proprietary standards that continue to enjoy widespread support are the Microcom MNP standards for error correction and data compression. Others, such as 3Com's HST, CompuCom's DIS, and Hayes' V-series, are no longer popular. For details about the MNP classes, see "MNP Classes" in the Technical Reference section of the DVD-ROM packaged with this book.


Note For more information about older proprietary protocols, see Chapter 18 of Upgrading and Repairing PCs, 11th Edition, included in electronic form on this book's DVD-ROM.

56Kbps Modems At one time, the V.34 annex speed of 33,600bps (33.6Kbps) was regarded as the absolute speed limit for asynchronous modem usage. However, starting in 1996, modem manufacturers began to produce modems that support speeds of up to 56,000bps. These so-called "56K" or "56Kbps" modems are now universal, although the methods for breaking the 33.6Kbps barrier have changed several times. To understand how this additional speed was achieved, you must consider the basic principle of modem technology​that is, the digital-to-analog conversion. As you've learned, a traditional modem converts data from digital to analog form so it can travel over the Public Switched Telephone Network (PSTN). At the destination system, another modem converts the analog data back to its digital form. This conversion from digital to analog and back causes some speed loss. Even though the phone line is physically capable of carrying data at 56Kbps or more, the effective maximum speed because of the conversions is about 33.6Kbps. An AT&T engineer named Claude Shannon came up with a law (Shannon's Law) stating that the maximum possible error-free data communications rate over an all-analog PSTN is approximately 35Kbps, depending on the noise present. However, because many parts of the United States's urban telephone system is digital​being converted to analog only when signals reach the telephone company's central office (or central switch)​it's possible to "break" Shannon's Law and achieve faster download rates. You can, in some cases, omit the initial digital-to-analog conversion and send a purely digital signal over the PSTN to the recipient's CO (see Figure 19.8). Thus, only one digital-to-analog conversion is necessary, instead of two or more. The result is that you theoretically can increase the speed of the data transmission, in one direction only, beyond the 35Kbps specified by Shannon's Law​to nearly the 56Kbps speed supported by the telephone network. Prior to the new ITU V.92 standard, the transmission in the other direction was still limited to the V.34 annex maximum of 33.6Kbps. However, both the modem and the ISP must have support for the ITU V.92 standard to overcome this limitation for uploading speeds.

Figure 19.8. V.90-based 56Kbps connections enable you to send data at standard analog modem rates (33.6Kbps maximum) but enable you to receive data nearly twice as fast, depending on


line conditions.

See "ITU V.92 and V.44â&#x20AC;&#x2039;Breaking the Upload Barrier," p. 1055, for more information on how the V.92 standard enables faster uploading.

56Kbps Limitations Thus, 56Kbps modems can increase data transfer speeds beyond the limits of V.34 modems, but they are subject to certain limitations. Unlike standard modem technologies, you can't buy two 56Kbps modems, install them on two computers, and achieve 56Kbps speeds. One side of the connection must use a special digital modem that connects directly to the PSTN without a digital-to-analog conversion. Therefore 56Kbps modems can be used at maximum speeds only to connect to ISPs or other hosting services that have invested in the necessary infrastructure to support the connection. Because the ISP has the digital connection to the PSTN, its downstream transmissions to your computer are accelerated. If both sides of the connection support standards pre-dating V.92, your communications back to the ISP are not accelerated. On a practical level, this means you can surf the Web and download files more quickly, but if you host a Web server on your PC, your users will realize no speed gain because the upstream traffic is not accelerated unless you and your ISP both use V.92-compliant modems. If you connect to another regular modem, your connection is made at standard V.34 annex rates (33.6Kbps or less). Also, only one digital-to-analog conversion can be in the downstream connection from the ISP to your computer. This is dictated by the nature of the physical connection to your local telephone carrier. If additional conversions are involved in your connection, 56Kbps technology will not work for you; 33.6Kbps will be your maximum possible speed.


Note Although most advertising for 56Kbps modems refers to them as simply "56K" modems, this is inaccurate. "K" is most often used in the computer business to refer to kilobytes. If that were true, a "real" 56K modem would be downloading at 56,000 bytes per second (or 448,000 bits per second)!

With the way the telephone system has had to grow to accommodate new exchanges and devices, even neighbors down the street from each other might have different results when using a 56Kbps modem.


Caution 56Kbps modem communications are highly susceptible to slowdowns caused by line noise. Your telephone line might be perfectly adequate for voice communications and even lower-speed modem communications, but inaudible noise easily can degrade a 56Kbps connection to the point at which there is only a marginal increase over a 33.6Kbps modem, or even no increase at all. If you do have a problem with line noise, getting a surge suppressor with noise filtration might help. Hotel connections through telephones with data jacks typically provide very slow connections with any type of modem. Even if you have a V.90- or V.92-compliant 56Kbps modem, you will be lucky to achieve even a 24Kbps transmission rate. The analog-to-digital conversions that occur between your room's telephone and the hotel's digital PBX system eliminate the possibility of using any of the 56Kbps standards the modem supports because they depend on a direct digital connection to the central switch (CS). As an alternative, more and more hotels and motels provide Ethernet-based access to broadband Internet service; a growing number even provide wireless Ethernet access with Wi-Fi/IEEE-802.11b hardware. Depending on the location, you might be able to use your normal Ethernet card or connect to a USB adapter provided by the hotel. If you want high-speed Internet access as part of the package with your next hotel or motel stay, contact the specific lodging location or chain Web site for details and pricing.

Early 56Kbps Standards To achieve a high-speed connection, both modems and your ISP (or other hosting service to which you connect) must support the same 56Kbps technology. The first 56Kbps chipsets were introduced in late 1996: U.S. Robotics's x2 used Texas Instruments (TI) chipsets. Rockwell's K56flex was supported by Zoom and other modem makers. These rival methods for achieving performance up to 56Kbps were incompatible with each other and were replaced in 1998 by the ITU's V.90 standard.


Note For more information about K56flex and x2, see Upgrading and Repairing PCs, 11th Edition, available in electronic format on the DVD-ROM supplied with this book.

Unfortunately, the 56Kbps name is rather misleading, in regards to actual transmission speeds. Although all 56Kbps modems theoretically are capable of this performance on top-quality telephone lines, the power requirements for telephone lines specified in the FCC's Part 68 regulation limit the top speed of these modems to 53Kbps. The FCC has been considering lifting this speed limitation since the fall of 1998, but it still applies to modems as of early 2003.

V.90 V.90 was introduced on February 5, 1998, and was ratified by the ITU-T on September 15, 1998. Its ratification ended the K56flex/x2 standards "war": Shortly thereafter, most modem manufacturers announced upgrade options for users of x2 and K56flex modems to enable these products to become V.90 compliant. Some modem vendors offer upgrades for K56flex and x2 modems to the V.90 standard. If you purchased your modem before the V.90 standard became official, see your modem vendor's Web site for information about upgrading to V.90.

ITU V.92 and V.44â&#x20AC;&#x2039;Breaking the Upload Barrier 56Kbps protocols, such as the early proprietary x2 and K56flex and the ITU V.90 standard, increased the download speed from its previous maximum of 33.6Kbps to 56Kbps. However, upload speeds, which affect how quickly you can send email, page requests, and file transfers, were not affected by the development of 56Kbps technologies. Upload speeds with any of these 56Kbps technologies are limited to a maximum of 33.6Kbps. This causes severe speed lags for both pure dial-up users and those who depend on analog modems for upstream traffic, such as users of one-way broadband solutionsâ&#x20AC;&#x2039;for example, one-way (Telco Return) cable modems, One-way DirecWAY, and one-way (Telco Return) fixed-base wireless Internet services. Other shortcomings of existing 56Kbps technology include the amount of time it takes the user's modem to negotiate its connection with the remote modem and the lack of uniform support for call-waiting features. In mid-2000, the ITU unveiled a multifaceted solution to the problem of slow connections and uploads: the V.92 and V.44 protocols (V.92 was previously referred to as V.90 Plus).


V.92, as the name implies, is a successor to the V.90 protocol, and any modem that supports V.92 also supports V.90. V.92 doesn't increase the download speed beyond the 56Kbps barrier, but offers these major features: QuickConnect. QuickConnect cuts the amount of time needed to make a connection by storing telephone line characteristics and using the stored information whenever the same phone line is used again. For users who connect to the Internet more than once from the same location, the amount of time the modem beeps and buzzes to make the connection will drop from as much as 27 seconds to about half that time. Bear in mind, though, that this reduction in connection time does not come about until after the initial connection at that location is made and its characteristics are stored for future use. Modem-on-Hold. The Modem-on-Hold feature allows the user to pick up incoming calls and talk for a longer amount of time than the few seconds allowed by current proprietary call-waiting modems. Modem-on-Hold enables the ISP to control how long you can take a voice call while online without interrupting the modem connection; the minimum amount of time supported is 10 seconds, but longer amounts of time (up to unlimited!) are also supported by this feature. Modem-on-Hold also allows you to make an outgoing call without hanging up the modem connection. Modem-on-Hold, similar to previous proprietary solutions, requires that you have the callwaiting feature enabled on your telephone line and also requires that your ISP support this feature of V.92.


Note Although Modem-on-Hold is good for the Internet user with only one phone line (because it allows a single line to handle incoming as well as outgoing calls), it's not as good for ISPs because when you place your Internet connection on hold, the ISP's modem is not capable of taking other calls. ISPs that support Modem-on-Hold might need to add more modems to maintain their quality of service if this feature is enabled. More modems are necessary because the ISP won't be able to count on users dropping their Internet connections to make or receive voice calls when Modem-onHold is available.

PCM Upstream. PCM Upstream breaks the 33.6Kbps upload barrier, boosting upload speed to a maximum of 48Kbps. Unfortunately, because of power issues, enabling PCM Upstream can reduce your downstream (download) speed by 1.3Kbpsâ&#x20AC;&#x2039;2.7Kbps or more. PCM Upstream is an optional feature of V.92, and ISPs who support V.92 connections might not support this feature. Modems that support V.92 typically also support the V.44 data-compression standard. V.44, which replaces V.42bis, provides for compression of data at rates up to 6:1â&#x20AC;&#x2039;that's more than 25% better than V.42bis. This enables V.92/V.44 modems to download pages significantly faster than V.90/V.42bis modems can at the same connection speed. When will you be able to enjoy the benefits of V.92/V.44? Although most major modem vendors have been offering V.92/V.44-compliant modems since late 2000, ISP interest in this standard has been tepid. Only one national ISP, Navipath (which provided support for many local and regional ISPs), offered V.92 access in 2001, and Navipath went out of business in September 2001. Prodigy began to offer V.92 service early in 2002. According to the V.92 News & Updates page at Richard Gamberg's Modemsite (www.modemsite.com/56K/v92s.htm), many vendors of ISP equipment continue to drag their feet on V.92/V.44 support, in part because it often requires expensive upgrades to terminal equipment. Even after upgrading to support V.92/V.44, some existing terminal hardware is incapable of working with the desirable PCM Upstream feature. Additionally, many major modem vendors have produced so-called "V.92" modems that don't support major V.92 features. Check user reviews available at the Modemsite Web site and others before you buy a particular V.92 modem model.


Tip Wondering whether it's time to get a V.92/V.44 modem? Before you update your current model or buy a new modem, do the following: Contact your ISP to see whether (or when) it plans to support V.92/V.44 and to determine which features will be supported. Check the V.92 ISPs listing at (www.v92.com) and contact ISPs in your area for more details.

Although the change from x2/K56flex to V.90 was a no-cost one for many modem owners, the upgrade from V.90 to V.92/V.44 isn't as painless. In many cases, only the most recent V.90 modems will be eligible for a free upgrade to V.92/V.44 firmware. Contact your modem vendor for details. Can your existing V.90-compatible modem be upgraded to V.92/V.44? As with earlier 56Kbps modem standards, the answer will likely be, "It depends." Some Lucent LT Winmodem (Agere Systems) modem drivers for V.90 also might include V.92 commands; see Modemsite's Lucent modem section for the latest information (http://www.modemsite.com/56k/ltwin.asp). For modems based on other chipsets, check with your modem vendor. As with earlier 56Kbps standards, you shouldn't worry about V.92/V.44 support until your ISP announces that it is supporting these standards. Because the V.92 standard has several components, find out which features of V.92 your ISP is planning to support before you look into a modem firmware update or modem replacement.

Fax Modem Standards Even though the first experimental facsimile equipment was developed at the end of World War II, it took many years for faxing to become commonplace. Similarly, the first fax boards for computers were not introduced until the late 1980s as separate devices. Later, fax capabilities were incorporated into modems. Today, virtually all modems also meet the ITU-T Class 3 fax standards, enabling them to send data to and receive data from other ITU-T Class 3 fax machines and multifunction devices. Many recent multifunction devices also support the newer ITU-T.30E recommendation for color faxing. Fax modems don't meet this standard as shipped from the manufacturer, but you can download free color fax software developed by HP (Impact ColorFax) that works with most fax modems. Find Impact ColorFax at the BlackICE Software Web site (www.blackice.com). For more information about ITU-T fax protocols, see "ITU-T Fax Protocols" in the


Technical Reference on the DVD packaged with this book.

Analog Modem Recommendations An analog modem for a PC can take the form either of an external device with its own power supply that plugs into a PC's serial port or USB port or of an internal expansion card you insert into a PCI or PC Card bus slot inside the computer. Very few ISA-slot analog modems are still on the market because the majority of recent systems no longer have ISA slots. Most manufacturers of modems have both internal and external versions of the same models. External versions are slightly more expensive because they include a separate case and power supply and sometimes require you to buy a serial modem cable or USB cable. Both internal and external modems are equally functional, however, and the decision as to which type to use should typically depend on whether you have a free bus slot or serial port, whether you have USB ports and Windows 98/Me/2000/XP, how much room you have on your desk, the capabilities of your system's internal power supply, and how comfortable you are with opening up your computer. I often prefer external modems because of the visual feedback they provide through indicator lights. These lights let you easily see whether the modem is still connected and transmitting or receiving data. However, some communication programs today include onscreen simulations of the lights, providing the same information. In other situations, however, an internal modem is preferable. If you are using an older computer whose serial ports do not have buffered UART chips, such as the 16550, many internal modems include an onboard 16550 UART-equivalent chip. This onboard UART with the modem saves you the trouble of upgrading the UART serial port. Also, external 56Kbps modems can be hampered from achieving their full speeds by the limitations of the computer's serial port. An external USB model or an internal model using the PCI slot might be preferable instead. Use Table 19.5 to see how internal and external units compare. Table 19.5. External Versus Internal Modems Features Built-in 16550 UART or higher

External

No (uses computer's USB port or serial port UART).

Price Higher. comparison

Internal

Yes (if 14.4Kbps or faster).

Lower.


Extras to buy

RS-232 Modem Interface cable or USB cable in some cases.

Ease of moving to another computer

Easy​unplug the cables and go! USB modems require a functioning USB port on the computer and Windows 98 or above.[1] RS-232 serial modems require you to Difficult​must open case and remove card, open shut down the computer first before disconnecting or other PC's case and insert card. reconnecting the modem; USB modems can be hotswapped.

Power supply

Plugs into wall ("brick" type).

None​powered by host PC.

Reset if modem hangs

Turn modem off, and then on again.

Restart computer.

Monitoring Easy​external signal lights. operation

Interface type

RS-232 serial or USB port; some models support both types of connections. (Parallel-port modems were made a few years ago, but never proved popular and have been discontinued.)

Nothing.

Difficult​unless your communication software simulates the signal lights. PCI or ISA; PCI is preferred for its extra speed, capability to allow mapping of COM 3 and COM 4 to unique IRQs instead of COM 1/3 and COM 2/4 sharing IRQs, and capability to work in so-called "legacy-free" systems that no longer include any ISA slots.

[1]

Although late versions of Windows 95 OSR 2.x have USB support, many USB devices actually require Windows 98 or better. Use Windows 98/Me/2000/XP to achieve more reliable support for USB devices.

See "UARTs," p. 965.

Not all modems that function at the same speed have the same functionality. Many modem manufacturers produce modems running at the same speed but with different feature sets at different price points. The more expensive modems usually support advanced features, such as distinctive ring support, caller ID, voice and data, video teleconferencing, and call-waiting support. When purchasing a modem, be sure it supports all the features you need. You also should make sure the software you plan to use, including the operating system, has been certified for use with the modem you select. If you live in a rural area, or in an older city neighborhood, your telephone line quality might influence your decision. Look at comparison test results carefully and pay particular attention to how well various modems perform with noisy lines. If your phone line sounds crackly during a rainstorm, that poor-quality line makes


reliable modem communications difficult, too, and it can limit your ability to connect at speeds above 33.6Kbps. Another feature to consider is the modem's resistance to electrical damage. Some brands feature built-in power protection to shield against damage from digital telephone lines (higher powered and not compatible with modems) or power surges. However, every modem should be used with a surge protector that allows you to route the RJ-11 telephone cable through the unit for protection against high-voltage surges. All modems on the market today support V.90 or V.92, and even if your particular location can't support those speeds, your modem might still offer advanced features, such as voicemail or simultaneous voice and data. Keep in mind that V.90/V.92 connections seem to work better for many users if their modems also support x2. If you prefer a modem made by a vendor that also supports K56flex, try to buy a modem that contains both types of standards in its firmware (referred to by some vendors as "Dualmode" modems).

Choosing to Upgrade If you bought your modem in 1997 or later, or if it was included in your computer, chances are good that it came with 56Kbps support or that you've upgraded it to some form of 56Kbps support. However, even though today's V.90/V.92 modems still have the same maximum speed, other changes in modem design have occurred that might make a modem upgrade desirable for you. And if you are still using a 33.6Kbps modem or even slower model, you should get a 56Kbps modem if your line quality can support it. Table 19.6 summarizes some of the major changes in analog modem design and features in recent years, along with advice on who should consider these features. Analog modems introduced from 2001 to the present still run at the 53Kbps FCC maximum with the potential to run at up to 56Kbps if the FCC regulations change, but they might offer one or more of the following advanced features: Call-waiting support PCI expansion slot for internal modems USB connection for external modems Faster performance for gaming


MNP10 and MNP10EC V.92/V.44 support Table 19.6. Analog Modem Features and Types Modem Feature

Benefit

Improves performance on poor-quality phone lines by MNP10EC quickly adjusting line speed up support and down with changing conditions.

PCI bus

Who Should Buy

Cautions

Don't buy MNP10 support by mistake; Anyone with poor-quality lines, MNP10EC is much better and includes especially if modem at other end MNP10. Check with ISP to see whether also has MNP10EC support; users cellular connections are supported; who want to use modem with speed of such connections can be cellular phones. 14.4Kbps or slower.

Works in PCI slots that Verify one or more PCI slots are dominate current systems and Anyone without ISA slots or who available; look for other features listed have replaced ISA slots. PCI is planning to move the modem here to make change as useful as modems can share IRQs with to another computer in the future possible; see also "Modems Without a other devices and be redirected and prefers internal modems. UART (WinModems)" later in this to open IRQs. chapter.

Works with USB connectors on Although Windows 95 OSR 2.5 most current and all future Anyone who wants portability and (Win95"C") has USB support, many USB computers; high-speed has USB connectors along with devices require Win98 or better; make connection connection and capability to Windows 98 or above. sure USB ports have been activated in connect many devices via a BIOS. hub. Faster PING and response for Gaminggaming, which is more optimized important than data modem throughput.

Callwaiting support

Anyone who uses call-waiting and Allows you to answer the doesn't like voice callers to get phone and not lose the modem busy signals; check with connection, instead of disabling manufacturer to see whether call-waiting, which most your existing modem can be modems require. upgraded with a driver.

Faster uploads, negotiation, V.92/V.44 Modem-on-Hold, better support throughput on downloads.

Voice support

Game optimizations are not useful for Anyone who plays a lot of online ordinary Web surfing. These modems games or wants to get started. are more expensive than others.

Anyone whose ISP supports all V.92 and V.44 features.

Allows digitizing of received calls; computerized answering Anyone who wants to use PC as machine and faxback; inbound a telecommunications center. and outbound phone calls via computer.

Modems Without a UART (WinModems)

Maximum talk time might be only a few seconds; make sure you have the phone near the computer for fast "hello, I'll call you later" answers to avoid exceeding time limit.

Check with your ISP to see whether V.92/V.44 support will be introduced and which features will be supported.

Check quality of voice recording; can use up a lot of disk space.


Modems without a UART chip, sometimes referred to as WinModems after the pioneering U.S. Robotics version, can save you money at purchase time but can cause problems with speed and operating-system compatibility later. For users wanting an inexpensive internal modem, a modem that doesn't use a traditional UART instead of a UART-equipped internal or external modem looks like a great deal, often costing less than $40, compared to $80 or more for a UART-equipped "hardware" modem. But, there is no free lunch for modem users. What can you lose with a modem that lacks a UART? First, you need to realize that there are actually two types of UART-less modems: those that rely on Windows and the CPU for all operations (these modems are also called controllerless modems) and those that use a programmable digital signal processor (DSP) chip to replace the UART. Both types of modems use less power than traditional UART-based modems, making them better for use with notebook computers. Although both are "software modems," there's an enormous difference in what you're getting. A Windows-based modem must run under Windows because Windows provides the brains of the modem, a cost-cutting move similar to that used by some low-cost host-based printers. You should avoid this if you're planning to try Linux, move the modem to a Macintosh, or use an old MS-DOSâ&#x20AC;&#x2039;based communications program. If you have no drivers for your modem/operating system combination, you'll have no luck using the software modem. Software modems that lack a DSP have a second major strike against them: They make your CPU do all the work. Although today's computers have much faster CPUs than those required for typical software modems (Pentium 133 minimum), your modem can still slow down your computer if you multitask while downloading or surfing the Web. Most of the modems bundled with computer systems are software modems, and the major chipsets used include Lucent LT (now Agere Systems), Conexant (formerly Rockwell) HCF, U.S. Robotics WinModem, ESS Technology's HSPcompliant chipsets, Intel's Modem Silicon Operation (formerly Ambient), and PCTel. Except for U.S. Robotics, the other companies produce chipsets that can be found in the modems made by many manufacturers. For best results, do the following: Make sure your modem uses a DSP. Typically, these modems don't require a particular CPU or a particular speed of CPU.


Consider modems built around the Lucent/Agere LT chipset. These modems have a DSP, and Lucent/Agere's firmware is frequently revised to achieve the best results in a rapidly changing telephony environment. Use the modem manufacturer's own drivers first. But software modems can often use any manufacturer's drivers for the same chipset with excellent results; in particular, Lucent/Agere LT chipset modems typically can use any Lucent/Agere LT driver from any modem manufacturer. Don't delete the old software driver when you download and install new modem software. As with UART-equipped modems, the latest firmware isn't always the best. Look carefully at the CPU, RAM, and operating system requirements before you buy your modem.


Tip Many manufacturers sell both traditional and UART-free modems. If you have an older system or want the option to use MS-DOSâ&#x20AC;&#x2039; or Linux-based communications programs, the hardware modem with a traditional UART might cost more but be a better choice.

Finding Support for "Brand-X" Modems Many computer users today didn't install their modems, or even purchase them as a separate unit. Their modems came bundled inside the computer and often have a bare-bones manual that makes no mention of the modem's origin or where to get help. Getting V.90 firmware updates, drivers, or even jumper settings for OEM modems such as this can be difficult. One of the best Web sites for getting help when you don't know where to start is the "Who Made My Modem?" page (www.56K.com), which features Links to the FCC's equipment authorization database (enter the FCC ID to locate the vendor) Using ATI commands to query the modem chipset Lookup by chipset manufacturer Search engine tips Links to major modem and chipset manufacturers

Squeezing Performance from Your 56Kbps Modem Although many users of 56Kbps modems have seen significant improvements in their connect speeds and throughputs over their previous V.34-type modems, many have not or have seen only sporadic improvements. According to the research of Richard Gamberg, available online at his Modemsite Web site (http://www.modemsite.com), a combination of five factors comes into play to affect your ability to get reliable connections in the range of 45Kbpsâ&#x20AC;&#x2039;53Kbps (the current FCC maximum): The modem The modem firmware/driver


Your line conditions The ISP's modems The ISP's modem firmware It's up to you to ensure that you match your modem 56Kbps type to the 56Kbps standards your ISP supports, and that you use the best (not always the latest!) modem firmware and drivers, as discussed in the previous section. Other modem adjustments recommended by Modemsite include Modifying existing modem .INF files used by Windows 9x to accurately reflect connection speed Disabling 56Kbps connections (!) when playing games to minimize lag times


Note This last suggestion might seem odd, but "fast" modems are designed to push a large amount of data through for downloads, and gameplay over modems actually deals with small amounts of data instead. The lag time caused by 56Kbps data handling can make a regular "fast" modem seem slow. If you want a fast modem for both downloading and game playing, see "Choosing to Upgrade," earlier in this chapter.

The site also hosts a forum area for discussing modem configuration, reliability, and performance issues.

Telco "Upgrades" and Your Modem In addition to the well-known analog-to-digital conversion issue that prevents some phone lines from handling 56Kbps modems at anything beyond 33.6Kbps, other local telephone company (telco) practices and services can either prevent 56Kbps from ever working or take it away from you after you've enjoyed it for a while. If you were getting 45Kbps or faster connections with your 56Kbps modem but can no longer get past 33.6Kbps, what happened? Some local telephone companies have been performing network "upgrades" that improve capacity for voice calls but prevent 56Kbps modems from running faster than 28Kbps. The cause seems to be the telephone companies' change from a signaling type called RBS (Robbed Bit Signaling) to SS7 (Signaling System 7), which changes how data used by the modem for high-speed access is detected. Caller ID devices connected to your phone line use RBS or SS7 signals to obtain information from incoming phone calls. If you use a caller ID box on the same phone circuit as your modem (even if it's connected in another room), you might not be able to get fast connections, or you might experience frequent disconnects. If you notice a drop in connection speed or reliability after you install caller ID, disconnect the caller ID box from the wall jack while you're online and see whether the speed and reliability of your modem connections improve. What else can you do? You can install the latest firmware available for your modem model or chipset. You can also check with your local telephone company to see whether it can update its firmware to solve the problem. Even if your modem has different firmware, checking on an upgrade might still be useful because this problem is likely to become widespread as telephone numbers, exchanges, and area codes continue to multiply like weeds and telephone network upgrades must keep pace.


Sharing Your Internet Connection Whether you have a 56Kbps dial-up modem or a broadband connection, one connection is often not enough for a home or small-office setting. You can share your connection with other computer users with one of the following methods: Computer-based sharing solutions. These work by connecting the computer with Internet access to a network with the other computers with which you want to share the connection. The computer with the connection acts as a gateway to the Internet. Router-based sharing solutions. These work by connecting all the computers on a network with a router or gateway connected to the Internet. Most routers are designed to work with broadband devices that use a USB or 10BASE-T connection, but a few work with analog modems. Typical computer-based sharing solutions include Microsoft Internet Connection Sharing (ICS). Introduced in Windows 98 Second Edition (Win98SE) and also a part of Windows Me, Windows 2000, and Windows XP Third-party gateway or proxy-server programs such as Wingate, Winproxy, and others Both ICS and third-party programs can also work with non-Windows computers because the TCP/IP network protocol, the standard protocol of the Internet, is used for networking. See "Network Protocols Overview," p. 1078, for details of the TCP/IP protocol.

Router-based solutions are available for popular types of home and small-office networks, including: 10BASE-T and 10/100 Ethernet


IEEE 802.11a and 802.11b wireless Ethernet HomePNA (phone-line) networks

Comparing Gateways, Proxy Servers, and Routers To the typical user, it doesn't matter whether a gateway, proxy server, or router is used to provide shared Internet access. Traditional gateway programs such as Microsoft ICS, Sygate Home Network, and WinGate use a method of shared access called Network Address Translation (NAT), which enables sharing by converting network addresses into Internet-compatible addresses during the file request and download process. This process requires little client PC configuration but doesn't permit page caching, content filtration, firewalls, or other useful services that can be provided by a proxy server. Proxy servers traditionally have required tricky configurationâ&#x20AC;&#x2039;sometimes at an individual application level. However, products such as WinProxy combine the ease of configuration of a gateway with the extra features of a proxy server. Popular third-party sharing programs include WinProxy (www.winproxy.com), WinGate (www.wingate.com), and Sybergen SyGate (www.sybergen.com). Many home-oriented networks and modems are bundled with these or similar products, so if you're in the market for a new modem or are building a small network, ask whether a proxy server or gateway program for Internet sharing is included. If not, you can download free trial versions from the previously listed Web sites. If you don't like leaving a computer on at all times to provide Internet access to the network, a router is the only way to go. Routers also provide better firewall protection for all computers on the network, and some, such as certain Linksys models, can be configured to require networked PCs to be running specified firewall or antivirus software before Internet access is granted. The most common routers for broadband Internet access also contain a switch, so you won't need a separate connection device for your home network.

Microsoft Internet Connection Sharing Windows 98 Second Edition, Windows Me, Windows 2000, and Windows XP feature a built-in gateway program called Internet Connection Sharing (ICS), which allows users to share a single Internet connectionâ&#x20AC;&#x2039;either dial-up or broadband. Because ICS is a gateway and clients use TCP/IP networking to use the gateway, only the gateway computer needs to use Windows 98 SE, Windows Me, Windows 2000, or Windows XP. Any computer using TCP/IP with the option to set up a gateway can be used as a client, including computers using older versions


of Windows 9x and other operating systems.

Requirements for ICS ICS requires a network interface card (NIC) to be installed in the host computer and a network connection to each guest computer to share the host's Internet connection. If the Internet connection is made through a NIC (as is the case with DSL, twoway cable modems, or fixed wireless broadband connections), two NICs are required: one for the Internet connection and one for sharing the connection. If you want to use a computer with limited expansion slot space inside as the ICS gateway, you might be able to either connect to a broadband Internet device or your favorite network solution with a USB port or use a USB-to-Ethernet adapter. ICS requires special configurations or might not work with some one-way services that use an analog modem (such as some cable modem, fixed wireless, or DirecWAY versions) because these devices use a separate connection for downloading and uploading. Figure 19.9 shows a typical home network using ICS to share a cable modem.

Figure 19.9. When ICS is used to share a cable modem, you might need two network connections. This computer has a builtin network port (attached to the cable modem), so the user must add another network card to connect to the home network.


Overview of the ICS Configuration Process The configuration process has two parts: Installing ICS on the gateway computer Configuring the clients to use the ICS gateway to reach the Internet The process of installing ICS on the gateway computer varies with the Windows version: For Windows 98SE and Windows Me, install ICS through the Add/Remove Programs icon in the Control Panel as a Windows component. For Windows 2000 and Windows XP, ICS is built into the Network icon in the Control Panel. If you need to install a network interface card or device before you can share your Internet connection, you must do this before you can configure ICS on the gateway computer.

Configuring ICS on the Gateway Computer with Windows 98SE


and Windows Me If ICS was not installed when Windows was installed, install it by selecting Start, Settings, Add/Remove Programs, Windows Setup. Select ICS from the Internet Tools category; Windows Me includes ICS as part of its Home Networking Wizard. Next, specify whether you are using a dial-up connection (modem or ISDN) or a high-speed connection (LAN, including cable modem or DSL). If you select dial-up, choose the dial-up connection (which must be set up already) you'll be sharing, followed by the NIC that connects you with the client PCs that will share the connection. Windows creates a client configuration floppy and prompts you to reboot the computer. When you view the Network Configuration in the Control Panel after rebooting, you should see the following: Three adapters (your actual NIC, the dial-up adapter, and a new one called Internet Connection Sharing) Three Internet Connection Sharing protocol entries, listing the adapters mentioned earlier in the chapter Three TCP/IP protocol entries, listing the adapters mentioned earlier in the chapter The TCP/IP protocol entry for Home must point to the NIC that connects the clients to the host PC; the TCP/IP protocol entry called Shared must point to DialUp Networking; and the remaining TCP/IP protocol entry must point to Internet Connection Sharing. Also, check the TCP/IP configuration for Home (the NIC) and verify the IP address; it should be 192.168.0.1. This IP address might need to be provided to the computers that will share the Internet connection. If the settings aren't correct, remove ICS and start over. Start an Internet connection on the gateway (host) computer before continuing.


Tip For a more detailed description of setting up ICS and networking on a computer running Windows 98SE, see Special Edition Using Microsoft Windows 98 Second Edition by Ed Bott and Ron Person (Que, 2000). For a more detailed description of setting up ICS and networking on a computer running Windows Me, see Special Edition Using Microsoft Windows Me by Ed Bott (Que, 2000).

Configuring ICS on the Gateway Computer with Windows 2000 and Windows XP If you want to share a dial-up analog modem or ISDN connection with Windows 2000, look for the ICS tab on the properties sheet for the connection in the Network and Dial-up Connections folder. Enable ICS and enable On-Demand Dialing (which will allow anyone on the network to open a connection as necessary). You also must enable the Remote Access Auto Connection Manager service in the Computer Management Control Panel. Configure this service to run Automatically. If you want to share a broadband connection with Windows 2000, look for the Sharing tab in the properties sheet for the connection to your broadband Internet device in the Network and Dial-up Connections folder. Enable sharing.


Tip For a more detailed description of setting up ICS and networking on a computer running Windows 2000, see Special Edition Using Microsoft Windows 2000 Professional by Robert Cowart and Brian Knittel (Que, 2000).

Windows XP uses the Network Setup Wizard to configure ICS for both the ICS gateway and client computers. You can also configure ICS manually through the properties sheet for your connection in the My Network Places folder.


Tip For a more detailed description of setting up ICS and networking on a computer running Windows XP, see Special Edition Using Microsoft Windows XP Home Edition (Que, 2002) or Special Edition Using Microsoft Windows XP Professional (Que, 2002). Both books are coauthored by Robert Cowart and Brian Knittel.

Configuring ICS on the Client Computers Use the ICS configuration disk created during the ICS installation process or the ICS client setup program available on the Windows CD to configure ICS on other Windows versions. If you prefer to set up ICS manually on client computers, do the following: Install the TCP/IP protocol on each client. Ensure that the following values are set up in the TCP/IP protocol for each client PC: P address: automatically obtained WINS Resolution: disabled Gateway: None DNS: disabled Each client will be assigned an IP address by the ICS server. Restart each client after performing the previous steps. Use a Web browser on each guest to verify the connection is working; Internet Explorer should not have any dial-up settings configured for it and should have no LAN settings enabled. Netscape Navigator/Communicator should be set to Direct Connection to the Internet. Some versions of Netscape Navigator might not work unless you create a Dial-Up Networking adapter on the guest and set its gateway, as explained earlier in the chapter. If you are using Windows 9x or Me, reboot before you test the connection.


Note Useful Web sites that cover this process in more detail include http://www.practicallynetworked.com http://www.duxcw.com/digest/Howto/index.html The Practically Networked Web site also provides useful tips for sharing one-way broadband services.

Routers for Internet Sharing Just as an ICS gateway has two IP addressesâ&#x20AC;&#x2039;one for the network and one for the Internetâ&#x20AC;&#x2039;so does a router. Most routers are sold for use with two-way broadband Internet access devices such as two-way cable modems and fixed wireless broadband services or DSL lines. Most of these devices connect to the computer via a 10BASE-T Ethernet port, as seen in Figure 19.10.

Figure 19.10. Front (top) and rear (bottom) views of a typical broadband router with a built-in four-port switch, the Linksys EtherFast Cable/DSL Router, BEFSR41. Top photo courtesy Linksys.


When you use a router to share your Internet connection, the WAN port on the router replaces the network card connection originally used to connect your computer with the cable modem or DSL modem. All computers on the network connect to LAN ports and can share files and printers with each other as well as share Internet access. The router can be configured to provide either dynamically assigned or fixed IP addresses to each computer connected to it through the LAN ports and can be configured to use the same MAC address (a unique hardware address assigned to each network component) originally used by the network card first connected to the cable modem or DSL modem. This prevents the ISP from determining that you're sharing the connection. The WAN port on the router can be configured to obtain an IP address from the cable modem or DSL modem or to have a fixed IP address, depending on the configuration required by the ISP. As long as the router is running and properly connected to the cable modem or DSL modem, any computer connected to it can go online just by opening its email client or Web browser. Figure 19.11 shows a typical 10/100 Ethernet home network configuration that uses a router with a built-in switch to share a cable modem.

Figure 19.11. When you use a router to share a cable modem or


DSL modem, you need only one network card in each computer. This router incorporates a switch that can connect up to four computers to the cable modem.

If you have a wireless network, connect the wireless access point or gateway device designed for your network to the cable modem or other broadband Internet device (some also work with analog modems). The wireless access point or gateway device will transfer data to and from computers on the wireless network and the Internet. For more information about choosing and installing wired and wireless networks, see Chapter 20.


Internet Troubleshooting This section deals with hardware problems that can cause Internet problems. Software problems usually are caused by incorrect configuration of the TCP/IP protocol required by all types of Internet connections. For more information about TCP/IP or other software problems, see Chapter 20.

Diagnosing Problems with a Shared Internet Connection Although each Internet sharing product has individual configuration issues, the following tips provide general guidelines useful for solving problems with all of them.

Check Your Host Configuration If your host isn't set up correctly, it can't share its connection with clients. Check the bindings for TCP/IP or other protocols used to create the shared connection. If you are using Microsoft's ICS and two Ethernet cards, you will see entries in the Network configuration on the host computer for each Ethernet card and for the ICS software itself.

Check Your Client Configuration Make sure your clients have correct TCP/IP, DHCP, and other protocol settings for the host. The ping command can be used to check the Internet connection; try pinging a Web site by opening a Windows command prompt and typing a command, such as ping www.selectsystems.com. If you have a working Internet connection, you should see the IP address for the Web site you specify and the round-trip time (or ping rate) for four signals sent to the Web site. If you get no response or see an error message, you might have a configuration problem with your TCP/IP configuration.


Note Because pinging can also be used for denial-of-service (DoS) attacks by hackers on Web sites, some Web sites don't respond to pings. Use ping when your system is working properly to find a Web site that will respond to ping and use that site for troubleshooting as described previously.

Verify that the host has a working Internet connection that's active before you try to share it. Check the sharing program's documentation to see how guests can dial the host's modem to start a connection if necessary.

Speed Will Drop with Multiple Users It's normal for the speed of an Internet connection to drop with multiple users, but if you're concerned about the degree of decline, check with the sharing software provider for Registry tweaks and other options to improve performance.

Diagnosing Connection Problems with Signal Lights Signal lights are found on most external broadband devices, such as cable modems, wireless broadband routers, and DSL modems. The signal lights indicate whether the unit is receiving signals from the computer, sending data to the network, or receiving data from the network and whether the unit can "see" the networkâ&#x20AC;&#x2039;even if no data is currently flowing through the unit. On many units, the power light also is used to indicate problems. If the light is normally green, for example, a red light might indicate the unit has failed. Other lights flash when data is being sent or received. On cable modem or wireless broadband routers, look for a signal lock light; this light flashes if the unit is trying to lock onto a signal from the cable network or wireless transmitter. Learn the meaning of the lights on your broadband device to help you diagnose problems; the user manual or vendor's Web site will provide the troubleshooting information you need for the particular broadband device you use.

Analog Modem Fails to Dial 1. Check line and phone jacks on the modem. Use the line jack to attach the modem to the telephone line. The phone jack takes the same RJ11 silver cord cable, but it's designed to let you daisy-chain a telephone to your modem, so you need only a single line for modem and telephone use. If you have reversed these cables, you will not get


a dial tone. If the cables are attached properly, check the cable for cuts or breaks. The outer jacket used on RJ-11 telephone cables is minimal. If the cable looks bad, replace it. If the modem is external, make sure the RS-232 modem cable is running from the modem to a working serial port on your computer and that it is switched on. Signal lights on the front of the modem can be used to determine whether the modem is on and whether it is responding to dialing commands. If the modem is a PC Card (PCMCIA card), make sure it is fully plugged into the PCMCIA/PC slot. With Windows 9x/Me/2000/XP, you should see a small PCMCIA/PC Card icon on the toolbar. Double-click it to view the cards that are currently connected. If your modem is properly attached, it should be visible. Otherwise, remove it, reinsert it into the PCMCIA/PC Card slot, and see whether the computer detects it.


Connecting a PC Card Modem via a Dongle Some PC Card modems do not use a standard RJ-11 cable because the card is too thin. Instead, they use a connection called a dongle, which runs from the PC Card to the telephone. If this proprietary cable is damaged, your modem is useless. You should purchase a spare from the modem vendor and carry it with you. And if you find the dongle is too short to reach the data jacks in a hotel room, buy a coupler from your local RadioShack or telephone-parts department and attach a standard RJ-11 cable to your dongle via the coupler. Dongles are also used with some PC Card network cards for the same reason; the RJ-45 twisted-pair cable connector is too wide to attach to a standard PC Card. To avoid using dongles, look for a network or modem card that has a standard RJ-11 or RJ-45 connection built in to it.

Make sure your modem has been properly configured by your OS. With Windows 9x/Me/2000, use the Modems control panel to view and test your modem configuration (with Windows XP, you can use the Modem Troubleshooter). Select your modem and click the Diagnostics tab. This displays the COM (serial) ports in your computer. Select the COM port used by the modem, and click the More Info tab. This sends test signals to your modem. A properly working modem responds with information about the port and modem. If you get a Couldn't Open Port error message, your modem isn't connected properly. It might be in use already by a program running in the background, or there might be an IRQ or I/O port address conflict with another card in your computer. Whether you have a modem installed, every COM port that is working will display its IRQ, I/O port address, and UART chip type when you run Diagnostics. The UART type should be 16550 or above for use with any modern modem.


Note You can also test your modem response by setting up a HyperTerminal session (discussed earlier) to send the modem commands. If the modem fails to respond, this is another indication of a problem with the modem-PC connection.

Computer Locks Up After Installing or Using Internal Modem, Terminal Adapter, or Network Card The usual cause of lockups after you install an internal card is an IRQ conflict. Internal analog modems that use ISA slots typically cause the "curse of the shared IRQ," especially if a serial mouse is also used. PC Card and PCI modems can share IRQs safely, and USB mice use the same IRQ as the USB port itself. For more information about shared IRQs with ISA modems and serial mice, see the Technical Reference section on the DVD packaged with this book.

Computer Can't Detect External Modem 1. Make sure the modem has been connected to the computer with the correct type of cable. For external modems that use an RS-232 serial port, you might need a separate RS-232 modem cable, which has a 9-pin connector on one end (to connect to the computer) and a 25-pin connector on the other end (to connect to the modem). Some external modems have an integrated modem cable. Because RS-232 is a very flexible standard encompassing many pinouts, be sure the cable is constructed according to the following diagram: PC (with 9-pin COM portâ&#x20AC;&#x2039;male)

Modem (25-pin portâ&#x20AC;&#x2039;female)

3

TX data 2

2

RX data 3

7

RTS

4


8

CTS

5

6

DSR

6

5

SIG GND

7

1

CXR

8

4

DTR

20

9

RI

22

If you purchase an RS-232 modem cable prebuilt at a store, you'll have a cable that works with your PC and your modem. However, you can use the preceding chart to build your own cable or, by using a cable tester, determine whether an existing RS-232 cable in your office is actually made for modems or some other device. Make sure the COM (serial) port or USB port to which the modem is connected to is working. The Windows 9x/Me/2000/XP diagnostics test listed earlier can be useful in testing the serial port, but third-party testing programs such as AMIDIAG have more thorough methods for testing the system's COM ports. These programs can use loopback plugs to test the serial ports. The loopback plug loops signals that would be sent to the modem or other serial device back to the serial port. These programs normally work best when run from the MS-DOS prompt. Some diagnostics include a loopback plug to test serial ports. Loopback plugs may vary in design depending on the vendor. To ensure that the USB port is working, check the Device Manager in Windows; a working USB port is listed as a USB Root hub and a PCI to USB Universal Host Controller in the Universal Serial Bus device category. Any external USB hubs also are listed in the same category. If this category is not listed and the ports are physically present on the computer, make sure you are using Windows 98/Me/2000/XP (only a few late versions of Windows 95 have USB support). If you are, be sure the USB ports are enabled in the system BIOS.


Check the power cord and power switch.

Using Your Modem Sound to Diagnose Your Modem If you listen to your modem when it makes a connection, you might have realized that different types of modems make distinctive connection sounds and that different connection speeds also make distinctive sounds. The various types of 56Kbps modems have distinctly different handshakes of tones, buzzes, and warbles as they negotiate speeds with the ISP's modem. Learning what your modem sounds like when it makes a 56Kbps connection and when it settles for a V.34-speed connection can help you determine when you should hang up and try to connect at a faster speed. The Modemsite's troubleshooting section has a number of sound samples of various modems during the handshaking process. Use RealPlayer to play the samples, available at www.modemsite.com/56k/trouble.asp (click the Handshakes link). Compare these sound samples to your own modem; be sure you adjust the speaker volume for your modem so you can hear it during the call.


Chapter 20. Local Area Networking Focus of This Chapter Defining a Network Client/Server Versus Peer Networks Network Protocols Overview Hardware Elements of Your Network Network Cable Installations Wireless Network Standards Network Protocols Other Home Networking Solutions Putting Your Network Together Tips and Tricks Direct Cable Connections Troubleshooting a Network


Focus of This Chapter This chapter concentrates on how to build and use a peer-to-peer network, the lowest-cost network that is still highly useful to small business and home-office users. This type of network can be created by adding network hardware to any recent version of Windows, from Windows 9x and NT through Windows Me, 2000, and XP. As you'll see, most peer-to-peer networks can be "grown" into client/server networks at a later point by adding a dedicated server and the appropriate software to the server and client PCs. Thus, this chapter provides the hands-on and practical information you need to create a small-office, workgroup, or home-office network. If you are managing a corporate network using Linux, Unix, Windows NT Server, Windows 2000 Server, Windows Server 2003, or Novell NetWare, you will also be concerned with matters such as security, user profiles, SIDs, and other factors beyond the scope of this book.


Note Networking is an enormous topic. For more information about client/server networking, wide area networking, the Internet, and corporate networking, I recommend Upgrading and Repairing Networks, Fourth Edition, from Que.


Defining a Network A network is a group of two or more computers that intelligently share hardware or software devices with each other. A network can be as small and simple as two computers that share the printer and CD-ROM drive attached to one of them or as large as the world's largest network: the Internet. Intelligently sharing means that each computer that shares resources with another computer or computers maintains control of that resource. Thus, a switchbox for sharing a single printer between two computers doesn't qualify as a network device; because the switchboxâ&#x20AC;&#x2039;not the computersâ&#x20AC;&#x2039;handles the print jobs, neither computer knows when the other one needs to print, and print jobs can interfere with each other. A shared printer, on the other hand, can be controlled remotely and can store print jobs from different computers on the print server's hard disk. Users can change the sequence of print jobs, hold them, or cancel them. And, sharing of the device can be controlled through passwords, further differentiating it from a switchbox. Virtually any storage or output device can be shared over a network, but the most common devices include: Printers Disk drives CD-ROM and optical drives Modems Fax machines Tape backup units Scanners Entire drives, selected folders, or individual files can be shared with other users via the network. In addition to reducing hardware costs by sharing expensive printers and other peripherals among multiple users, networks provide additional benefits to users:


Multiple users can share access to software and data files. Electronic mail (email) can be sent and received. Multiple users can contribute to a single document using collaboration features. Remote-control programs can be used to troubleshoot problems or show new users how to perform a task. A single Internet connection can be shared among multiple computers.

Types of Networks Several types of networks exist, from small, two-station arrangements to networks that interconnect offices in many cities: Local area networks. The smallest office network is referred to as a local area network (LAN). A LAN is formed from computers and components in a single office or building. A LAN can also be built at home from the same components used in office networking, but, as you'll see later, special home-networking components now exist to allow the creation of what can be called a home area network (HAN). Home area networks. A home area network (HAN) often uses the same hardware components as a LAN, but it is mainly used to share Internet access. Powerline, low-speed wireless, and phoneline networks are used primarily in HAN rather than LAN environments. HANs are often referred to as smalloffice/home-office (SOHO) LANs. Wide area networks. LANs in different locations can be connected together by high-speed fiber-optic, satellite, or leased phone lines to form a wide area network (WAN). The Internet. The World Wide Web is the most visible part of the world's largest network, the Internet. Although many users of the Internet still use modems over a dialup connection rather than a LAN or WAN connection, any user of the Internet is a network user. The Internet is really a network of networks, all of which are connected to each other through the TCP/IP protocol. Programs such as Web browsers, File Transfer Protocol (FTP) clients, and newsreaders are some of the most common ways users work with the


Internet. Intranets. Intranets use the same Web browsers and other software and the same TCP/IP protocol as the public Internet, but intranets exist as a portion of a company's private network. Typically, intranets comprise one or more LANs that are connected to other company networks, but, unlike the Internet, the content is restricted to authorized company users only. Essentially, an intranet is a private Internet. Extranets. Intranets that share a portion of their content with customers, suppliers, or other businesses, but not with the general public, are called extranets. As with intranets, the same Web browsers and other software are used to access the content.


Note Both intranets and extranets rely on firewalls and other security tools and procedures to keep their private contents private. If you want to learn more about firewalls and security, I recommend picking up copies of the following: Absolute Beginner's Guide to Personal Firewalls, ISBN: 0-7897-2625-4 Upgrading and Repairing Networks, Fourth Edition, ISBN: 0-7897-2817-6

Requirements for a Network Unless the computers that are connected know they are connected and agree on a common means of communication and what resources are to be shared, they can't work together. Networking software is just as important as networking hardware because it establishes the logical connections that make the physical connections work. At a minimum, each network requires the following: Physical (cable) or wireless (infrared [IRDA] or radio-frequency) connections between computers A common set of communications rules, known as a network protocol Software that enables resources to be shared with other PCs and controls access to shared resources, known as a network operating system Resources that can be shared, such as printers, disk drives, and CD-ROMs Software that enables computers to access other computers with shared resources, known as a network client These rules apply to the simplest and most powerful networks, and all the ones in between, regardless of their nature. The details of the hardware and software you need are discussed more fully later in this chapter.


Client/Server Versus Peer Networks Although every computer on a LAN is connected to every other computer, they do not necessarily all communicate with each other. There are two basic types of LANs, based on the communication patterns between the machinesâ&#x20AC;&#x2039;client/server networks and peer-to-peer networks.

Client/Server Networks On a client/server network, every computer has a distinct role, that of either a client or a server. A server is designed to share its resources among the client computers on the network. Typically, servers are located in secured areas, such as locked closets or data centers, because they hold an organization's most valuable data and do not have to be accessed by operators on a continuous basis. The rest of the computers on the network function as clients (see Figure 20.1).

Figure 20.1. The components of a client/server LAN.

Servers A dedicated server computer typically has a faster processor, more memory, and more storage space than a client because it might have to service dozens or even hundreds of users at the same time. High-performance servers also might use two


or more processors, use the 64-bit version of the PCI expansion slot for serveroptimized network interface cards, and have redundant power supplies. The server runs a special network operating systemâ&#x20AC;&#x2039;such as Windows NT Server, Windows 2000 Server, Windows Server 2003, Linux, Unix, or Novell NetWareâ&#x20AC;&#x2039;that is designed solely to facilitate the sharing of its resources. These resources can reside on a single server or on a group of servers. When more than one server is used, each server can "specialize" in a particular task (file server, print server, fax server, email server, and so on) or provide redundancy (duplicate servers) in case of server failure. For very demanding computing tasks, several servers can act as a single unit through the use of parallel processing.

Clients A client computer communicates only with servers, not with other clients. A client system is a standard PC that is running an operating system such as Windows 9x, Windows Me, Windows 2000 Professional, or Windows XP. These versions of Windows contain the client software that enables the client computers to access the resources that servers share. Older operating systems, such as Windows 3.x and DOS, require add-on network client software.

Peer-to-Peer Network By contrast, on a peer-to-peer network, every computer is equal and can communicate with any other computer on the network to which it has been granted access rights (see Figure 20.2). Essentially, every computer on a peer-topeer network can function as both a server and a client; any computer on a peerto-peer network is considered a server if it shares a printer, a folder, a drive, or some other resource with the rest of the network. This is why you might hear about client and server activities, even when the discussion is about a peer-topeer network. Peer-to-peer networks can be as small as two computers or as large as hundreds of systems. Although there is no theoretical limit to the size of a peer-to-peer network, performance drops significantly and security becomes a major headache on peer-based networks with more than 10 computers. Also, Microsoft imposes a 10-station limit on computers running Windows 2000 Professional or XP Professional that are sharing resources with other systems. For these reasons, I recommend that you switch to a client/server network when your network climbs above about 10 stations.

Figure 20.2. The logical architecture of a typical peer-to-peer network.


Peer-to-peer networks are more common in small offices or within a single department of a larger organization. The advantage of a peer-to-peer network is that you don't have to dedicate a computer to function as a file server. Instead, every computer can share its resources with any other. The potential disadvantages to a peer-to-peer network are that typically less security and less control exist because users normally administer their own systems, whereas client/server networks have the advantage of centralized administration.

Comparing Client/Server and Peer-to-Peer Networks Client/server LANs offer enhanced security for shared resources, greater performance, increased backup efficiency for network-based data, and the potential for the use of redundant power supplies and RAID drive arrays. Client/server LANs also have a much greater cost to purchase and maintain. Table 20.1 compares client/server and peer-to-peer server networking. Table 20.1. Comparing Client/Server and Peer-to-Peer Networking Item

Client/Server

Peer-to-Peer

Access control

Via user/group lists of permissions; single password Via password lists by resource; each resource provides a user access to only the resources on his list; requires a separate password; all-or-nothing users can be given several different levels of access. access; no centralized user list.

Security

High because access is controlled by user or by group identity.

Performance High because the server doesn't waste time or resources handling workstation tasks.

Low because knowing the password gives anybody access to a shared resource.

Low because servers often act as workstations.


Hardware Cost

High, due to specialized design of server, highperformance nature of hardware, and redundancy features.

Low because any workstation can become a server by sharing resources.

Software Cost

License fees per workstation user are part of the cost of the Network Operating System server software (Windows NT and Windows 2000 Server, .NET Server, and Novell NetWare).

It's free; all client software is included with any release of Windows 9x, Windows NT Workstation, Windows 2000 Professional, Windows Me, and WindowsXP.

Backup

Centralized when data is stored on server; enables use Left to user decision; usually mixture of backup of high-speed, high-capacity tape backups with devices and practices at each workstation. advanced cataloging.

Duplicate power supplies, hot-swappable drive arrays, and even redundant servers are common; network OS Redundancy normally is capable of using redundant devices automatically.

No true redundancy among either peer "servers" or clients; failures require manual intervention to correct, with a high possibility of data loss.

Windows 9x, Windows Me, Windows NT, Windows 2000, and Windows XP have peer-to-peer networking capabilities built into them. Because Windows, from version 95 forward, uses Plug and Play technology, installing network interface cards in a collection of these systems is relatively easy. Simply connect them with the correct kind of cable, and build your own peer-to-peer network.


Network Protocols Overview The protocol you choose to run is the single most important decision you make when setting up a local area network. This protocol defines the speed of the network, the medium access control mechanism it uses, the types of cables you can use, the network interface adapters you must buy, and the adapter drivers you install in the network client software. The Institute of Electrical and Electronic Engineers (IEEE) has defined and documented a set of standards for the physical characteristics of both collisiondetection and token-passing networks. These standards are known as IEEE 802.3 (Ethernet) and IEEE 802.5 (Token-Ring). IEEE 802.11 defines wireless versions of Ethernet.


Note Be aware, however, that the colloquial names Ethernet and Token-Ring actually refer to earlier versions of these protocols, on which the IEEE standards were based. Minor differences exist between the frame definitions for true Ethernet and true IEEE 802.3. In terms of the standards, IBM's 16Mbps Token-Ring products are an extension of the IEEE 802.5 standard. There is also an older data-link protocol called ARCnet that is now rarely used.

The two major choices today for wired networks are Ethernet and Token-Ring, although Ethernet and its variations are by far the most popular. Other network data-link protocols you might also encounter are summarized in Table 20.2. The abbreviations used for the cable types are explained in the following sections. Table 20.2. Wired LAN Protocol Summary Network Type

ARCnet

Speed

2.5Mbps

Maximum Number of Stations 255 stations

Cable Types

RG-62 coax UTP/Type 1 STP

Notes

Obsolete for new installations; was used to replace IBM 3270 terminals (which used the same coax cable).

Per segment:

Ethernet 10Mbps

10BASET-2 UTP Cat 3 (10BASE-T), Thicknet (coax; 10BASE-5), 10BASEThinnet (RG-58 coax; 2-30 10BASE-2), fiber-optic 10BASE- (10BASE-F)

Largely replaced by Fast Ethernet; can be interconnected with Fast Ethernet by use of dual-speed hubs and switches; use switches and routers to overcome "5-4-3" rule in building very large networks.

5-100 10BASEFL-2

Per segment: Cat 5 UTP 2

Fast Ethernet can be interconnected with standard Ethernet through the use of dual-speed hubs, switches, and routers. The most common variety is 100BASE-TX; alternative 100BASE-T4 is not widely supported.

Per Gigabit 1,000Mbps segment: Cat 5 UTP Ethernet 2

Gigabit Ethernet can be interconnected with Fast or standard Ethernet through the use of multispeed hubs, switches, and routers.

Fast 100Mbps Ethernet

TokenRing

4Mbps or 16Mbps

72 on UTP, 250UTP, Type 1 STP, and fiber 260 on optic type 1 STP

High price for NICs and MSAUs to interconnect clients; primarily used with IBM mid-size and mainframe systems.

Wireless protocols, such as Wireless Ethernet IEEE 802.11 variants, are discussed later in this chapter. See "Wireless Network Standards,"


p. 1101.

A few years ago, the choice between Token-Ring or Ethernet wasn't easy. The original versions of standard Ethernet (10BASE-5 "Thick Ethernet" and 10BASE-2 "Thin Ethernet") used hard-to-install coaxial cable and were expensive to build beyond a certain point because of the technical limitations expressed by the "5-43" rule (see the following note).


Note The 5-4-3 rule gets its name from the fact that Ethernet signals can travel through no more than five segments, four repeaters or hubs, and three populated segments (cable segments with two or more stations) before they become unreliable. Because 10BASE-T hubs act as repeaters and a 10BASE-T segment has only one station per cable, creating large networks is easier with 10BASE-T Ethernet than with older forms of Ethernet.

Initially, Token-Ring's 16Mbps version was substantially faster than 10BASE versions of Ethernet and had larger limits on the numbers of workstations permitted per segment. Currently, however, the popularity and low cost of Fast Ethernet; the use of easy-to-install twisted-pair cabling for standard, 100Mbps Fast, and even 1,000Mbps Gigabit Ethernet; and the use of hubs and switches to overcome classic Ethernet station limitations have made Fast Ethernet the preferred choice for workgroup-size networks and a competitor to Token-Ring for larger networks. A properly designed Fast Ethernet network can be upgraded to Gigabit Ethernet in the future.

Ethernet With tens of millions of computers connected by Ethernet cards and cables, Ethernet is the most widely used data link layer protocol in the world. Ethernetbased LANs enable you to interconnect a wide variety of equipment, including Unix and Linux workstations, Apple computers, printers, and PCs. You can buy Ethernet adapters from dozens of competing manufacturers. Older adapters supported one, two, or all three of the cable types defined in the standard: Thinnet, Thicknet, and Unshielded Twisted Pair (UTP). Current adapters, on the other hand, almost always support UTP only. Traditional Ethernet operates at a speed of 10Mbps, but the more recent (and most popular of the Ethernet flavors) Fast Ethernet standards push this speed to 100Mbps. The latest version of Ethernet, Gigabit Ethernet, reaches speeds of 1,000Mbps, or 100 times the speed of original Ethernet.

Fast Ethernet Fast Ethernet requires adapters, hubs, and UTP or fiber-optic cables designed to support the higher speed. Some early Fast Ethernet products supported only 100Mbps, but almost all current Fast Ethernet products are combination devices that run at both 10Mbps and 100Mbps, enabling you to gradually upgrade an older 10Mbps Ethernet network by installing new NICs and hubs over an extended period of time. Both the most popular form of Fast Ethernet (100BASE-TX) and 10BASE-T


standard Ethernet use two of the four wire pairs found in UTP Category 5 cable. An alternative Fast Ethernet standard called 100BASE-T4 uses all four wire pairs in UTP Category 5 cable, but this Fast Ethernet standard was never popular and is seldom seen today.

Gigabit Ethernet Gigabit Ethernet also requires special adapters, hubs, and cables. Most users of Gigabit Ethernet use fiber-optic cables, but you can run Gigabit Ethernet over the same Category 5 UTP cabling that Fast Ethernet and newer installations of standard Ethernet use. Gigabit Ethernet for UTP is also referred to as 1000BASET. Unlike Fast Ethernet and standard Ethernet over UTP, Gigabit Ethernet uses all four wire pairs. Thus, Gigabit Ethernet requires dedicated Ethernet cabling; you can't "borrow" two wire pairs for telephone or other data signaling with Gigabit Ethernet as you can with the slower versions. Most Gigabit Ethernet adapters can also handle 10BASE-T and 100BASE-TX Fast Ethernet traffic, enabling you to interconnect all three UTP-based forms of Ethernet on a single network. Neither Fast Ethernet nor Gigabit Ethernet support the use of thin or thick coaxial cable originally used with traditional Ethernet, although you can interconnect coaxial-cable-based and UTP-based Ethernet networks by using media converters or specially designed hubs and switches.


Note For more information about Ethernet, Fast Ethernet, Token-Ring, and other network data-link standards, see the "Data Link Layer Protocols" and "High-Speed Networking Technologies" sections found in Chapter 19 of Upgrading and Repairing PCs, 11th Edition, available in electronic form on the DVD provided with this book.


Hardware Elements of Your Network The choice of a data-link protocol affects the network hardware you choose. Because Ethernet, Fast Ethernet, Token-Ring, and other data-link protocols use different hardware, you must select the protocol before you can select appropriate hardware, including network interface cards, cables, and hubs or switches.

Network Interface Cards On most computers, the network interface adapter takes the form of a network interface card (NIC) that fits into a PCI slot on a desktop computer or a PC Card (PCMCIA) slot on a notebook computer. Although network cards for older systems might use the ISA or EISA slot standards, these don't support high-speed network standards and are obsolete. Many recent systems incorporate the network interface adapter onto the motherboard, but this practice is more commonly found in workstations and portable computers and rarely in servers because most network administrators prefer to select their own NICs.


Note You can also adapt the USB port for Ethernet use with a USB-to-Ethernet adapter, but because USB 1.1 ports run at only 12Mbps (compared to Fast Ethernet running at 100Mbps), a performance bottleneck results. USB 2.0-to-Fast Ethernet adapters are now available from several vendors and are a suitable choice if your system has USB 2.0 ports but lacks a free PCI or PC Card slot.

Ethernet and Token-Ring adapters have unique hardware addresses coded into their firmware. The data link layer protocol uses these addresses to identify the other systems on the network. A packet gets to the correct destination because its data link layer protocol header contains the hardware addresses of both the sending and receiving systems. 10/100 Ethernet network adapters range in price from under $20 for client adapters to as much as $300 or more for server-optimized adapters. Token-Ring adapters are much more expensive, ranging in price from around $170 for client adapters to more than $500 for server-optimized adapters. Network adapters are built in all the popular interface-card types and are also optimized for either workstation or server use. For first-time network users, so-called network-in-abox kits are available that contain two 10/100 Fast Ethernet NICs, a small hub or switch, and prebuilt UTP cables for less than $90. When combined with the builtin networking software in Windows, these kits make networking very inexpensive. A number of 10/100/1000-BASE-TX Gigabit Ethernet adapters for use with UTP cable can be purchased for under $60, and some 10/100 switches now also feature Gigabit Ethernet ports. For client workstations (including peer servers on peer-to-peer networks), the following sections contain my recommendations on the features you need.

Speed Your NIC should run at the maximum speed you want your network to support. For a Fast Ethernet network, for example, you should purchase Ethernet cards that support 100BASE-TX's 100Mbps speed. Most Fast Ethernet cards also support standard Ethernet's 10Mbps speed, allowing the same card to be used on both older and newer portions of the network. To verify dual-speed operation, look for network cards identified as 10/100 Ethernet. Your NIC should support both half-duplex and full-duplex operation: Half-duplex means that the network card can only send or only receive data in a single operation. Full-duplex means that the network card can both receive and send


simultaneously. Full-duplex options boost network speed if switches are used in place of hubs. For example, 100Mbps Fast Ethernet cards running in fullduplex mode have a maximum true throughput of 200Mbps, with half going in each direction. Unlike hubs, which broadcast data packets to all computers connected to it, switches create a direct connection between the sending and receiving computers. Thus, switches provide faster performance than hubs; most switches also support full-duplex operation, doubling the rated speed of the network when full-duplex network cards are used. Although, at one time, switches were hard to justify on a workgroup peer-to-peer LAN because of their extra cost, switches are now available for little more than the price of hubs with a similar number of ports and are recommended for LANs of any size. If you plan to use your Ethernet-based home or office network to share an Internet connection, consider purchasing a router with an integrated switch. This combination makes networking simple and requires less space than separate routers and switches. Many router/switch combinations cost little more than a switch or router would separately. You can also purchase a wireless access point and router with an integrated switch so both wired and wireless workstations can share a single Internet connection. For other types of networks, you can use a gateway to connect your network with the Internet.

Bus Type If you are networking desktop computers built from 1995 to the present, you should consider only PCI-based NICs (these computers typically have three or more PCI slots). Although many older computers still have at least one ISA or combo ISA/PCI expansion slot, the superior data bus width and data transfer rate of PCI make it the only logical choice for networks of all types. The integrated NIC found on some recent PC motherboards is also a PCI device. Alternative interfaces include USB or PC Card/Cardbus adapters that are often used for portable systems. Table 20.3 summarizes the differences between all the types of interfaces used by network cards. Table 20.3. Bus Choices for Client PC NICs Bus Type

Bus Width (Bits)

Bus Speed (MHz)

Data Cycles per Clock

Bandwidth (MBps)

8-bit ISA (AT)

8

8.33

1/2

4.17

16-bit ISA (AT-Bus)

16

8.33

1/2

8.33


EISA Bus

32

8.33

1

33

MCA-16 Streaming

16

10

1

20

MCA-32 Streaming

32

10

1

40

MCA-64 Streaming

64

10

1

80

PC-Card (PCMCIA)

16

10

1

20

CardBus

32

33

1

133

PCI

32

33

1

133

PCI 66MHz

32

66

1

266

PCI 64-bit

64

33

1

266

PCI 66MHz/64-bit

64

66

1

533

USB 1.1

1

12

1

1.5

USB 2.0

1

480

1

60

Note: ISA, EISA, and MCA are no longer used in current motherboard designs. MBps = Megabytes per second ISA = Industry Standard Architecture, also known as the PC/XT (8-bit) or AT-Bus (16-bit) EISA = Extended Industry Standard Architecture (32-bit ISA) MCA = Microchannel Architecture (IBM PS/2 systems) PC-Card = 16-bit PCMCIA (Personal Computer Memory Card International Association) interface CardBus = 32-bit PC-Card PCI = Peripheral Component Interconnect USB = universal serial bus

Although a few ISA-based NICs are still on the market, their slow speeds and narrow bus widths severely restrict their performance. Most ISA-based Ethernet cards can't support speeds above 10Mbps and thus don't support Fast Ethernet or Gigabit Ethernet. A few vendors make 10/100 Ethernet cards for the ISA slot, but their performance is substantially lower than PCI cards. If you are shopping for a network card for a laptop or notebook system, look for Cardbus types, which are significantly faster than PC Cards and those using USB.


Network Adapter Connectors Ethernet adapters typically have a connector that looks like a large telephone jack called an RJ-45 (for 10BASE-T and Fast Ethernet twisted-pair cables), a single BNC connector (for Thinnet coaxial cables), or a D-shaped 15-pin connector called a DB15 (for Thicknet coaxial cables). A few older 10Mbps adapters have a combination of two or all three of these connector types; adapters with two or more connectors are referred to as combo adapters. Token-Ring adapters can have a 9-pin connector called a DB9 (for Type 1 STP cable) or sometimes an RJ45 jack (for Type 3 UTP cable). Figure 20.3 shows all three of the Ethernet connectors.

Figure 20.3. Three Ethernet connectors on two NICs: RJ-45 connector (top center), DB-15 connector (bottom right), and BNC connector (bottom left).

The following figures provide profile views of the most common types of NIC connections. Figure 20.4 shows a 10BASE-2 NIC configured to be at the end of a network; the T-adapter connected to the BNC connector has a Thinnet (RG-58) cable attached to one side and a 50-ohm terminator at the other end.

Figure 20.4. An Ethernet 10BASE-2 NIC configured as the last station in a Thin Ethernet network.


Figure 20.5 shows a 10BASE-T NIC with its UTP cable attached.

Figure 20.5. An Ethernet 10BASE-T NIC with a UTP cable attached.

Virtually all standard and 10/100 Ethernet NICs made for client-PC use on the


market today are designed to support unshielded twisted-pair (UTP) cable exclusively; Gigabit Ethernet cards made for wired (not fiber-optic) networks also use only UTP cable. If you are adding a client PC to an existing network that uses some form of coaxial cable, you have three options: Purchase a combo NIC that supports coaxial cable as well as RJ-45 twistedpair cabling. Purchase a media converter that can be attached to the coaxial cable to allow the newer UTP-based NICs to connect to the existing network. Use a switch or hub that has both coaxial cable and RJ-45 ports. A dual-speed (10/100) device is needed if you are adding one or more Fast Ethernet clients. For maximum economy, NICs and network cables must match, although media converters can be used to interconnect networks based on the same standard, but using different cable.

Network Cables Originally, all networks used some type of cable to connect the computers on the network to each other. Although various types of wireless networks are now on the market, most office and home networks are still based on one of the following wired topologies: Coaxial cable Twisted-pair cabling

Thick and Thin Ethernet Coaxial Cable The first versions of Ethernet were based on coaxial cable. The original form of Ethernet, 10BASE-5, used a thick coaxial cable (called Thicknet) that was not directly attached to the NIC. A device called an attachment unit interface (AUI) ran from a DB15 connector on the rear of the NIC to the cable. The cable had a hole drilled into it to allow the "vampire tap" to be connected to the cable. NICs designed for use with thick Ethernet cable are almost impossible to find as new hardware today.


10BASE-2 Ethernet cards use a BNC (Bayonet-Neill-Concilman) connector on the rear of the NIC. Although the thin coaxial cable (called Thinnet or RG-58) used with 10BASE-2 Ethernet has a bayonet connector that can physically attach to the BNC connector on the card, this configuration is incorrect and won't work. Instead, a BNC T-connector attaches to the rear of the card, allowing a thin Ethernet cable to be connected to either both ends of the T (for a computer in the middle of the network) or to one end only (for a computer at the end of the network). A 50-ohm terminator is connected to the other arm of the T to indicate the end of the network and prevent erroneous signals from being sent to other clients on the network. Some early Ethernet cards were designed to handle thick (AUI/DB15), thin (RG-58), and UTP (RJ-45) cables. Combo cards with both BNC and RJ-45 connectors are still available but can run at only standard Ethernet speeds. Figure 20.6 compares Ethernet DB-15 to AUI, BNC coaxial T-connector, and RJ-45 UTP connectors to each other, and Figure 20.7 illustrates the design of coaxial cable.

Figure 20.6. An Ethernet network card with thick Ethernet (DB15), thin Ethernet (RG-58 with T-connector), and UTP (RJ-45) connectors.

Figure 20.7. Coaxial cable.


Twisted-Pair Cable Twisted-pair cable is just what its name implies: insulated wires within a protective casing with a specified number of twists per foot. Twisting the wires reduces the effect of electromagnetic interference (that can be generated by nearby cables, electric motors, and fluorescent lighting) on the signals being transmitted. Shielded twisted pair (STP) refers to the amount of insulation around the cluster of wires and therefore its immunity to noise. You are probably familiar with unshielded twisted-pair (UTP) cable; it is often used for telephone wiring. Figure 20.8 shows unshielded twisted-pair cable; Figure 20.9 illustrates shielded twisted-pair cable.

Figure 20.8. An unshielded twisted-pair (UTP) cable.

Figure 20.9. A shielded twisted-pair (STP) cable.


Shielded Versus Unshielded Twisted Pair When cabling was being developed for use with computers, it was first thought that shielding the cable from external interference was the best way to reduce interference and provide for greater transmission speeds. However, it was discovered that twisting the pairs of wires is a more effective way to prevent interference from disrupting transmissions. As a result, earlier cabling scenarios relied on shielded cables rather than the unshielded cables more commonly in use today. Shielded cables also have some special grounding concerns because one, and only one, end of a shielded cable should be connected to an earth ground; issues arose when people inadvertently caused grounding loops to occur by connecting both ends or caused the shield to act as an antenna because it wasn't grounded. Grounding loops are situations in which two different grounds are tied together. This is a bad situation because each ground can have a slightly different potential, resulting in a circuit that has very low voltage but infinite amperage. This causes undue stress on electrical components and can be a fire hazard.

Most Ethernet and Fast Ethernet installations that use twisted-pair cabling use UTP because the physical flexibility and small size of the cable and connectors makes routing it very easy. However, its lack of electrical insulation can make interference from fluorescent lighting, elevators, and alarm systems (among other devices) a major problem. If you use UTP in installations where interference can be a problem, you need to route the cable away from the interference, use an external shield, or substitute STP for UTP near interference sources.

Network Topologies Each computer on the network is connected to the other computers with cable (or some other medium). The physical arrangement of the cables connecting computers on a network is called the network topology. The three types of basic topologies used in computer networks are as follows: Bus. Connects each computer on a network directly to the next computer in a linear fashion. The network connection starts at the server and ends at the last computer in the network. Star. Connects each computer on the network to a central access point. Ring. Connects each computer to the others in a loop or ring. These different topologies are often mixed, forming what is called a hybrid network. For example, you can link the hubs of several star networks together with a bus, forming a star-bus network. Rings can be connected in the same way.


Table 20.4 summarizes the relationships between network types and topologies. Table 20.4. Network Cable Types and Topologies Network Type

Standard

Cable Type

Topology

Ethernet

10BASE-2

Thick coaxial

Bus

10BASE-5

Thin (RG-58) coaxial

Bus

10BASE-T

Cat 3 or Cat 5 UTP

Star

Fast Ethernet

100BASE-TX

Cat 5 UTP

Star

Gigabit Ethernet

1000BASE-TX

Cat 5 UTP

Star

Token-Ring

(all)

STP or coaxial

Logical Ring

The bus, star, and ring topologies are discussed in the following sections.

Bus Topology The earliest type of network topology was the bus topology, which uses a single cable to connect all the computers in the network to each other, as shown in Figure 20.10. This network topology was adopted initially because running a single cable past all the computers in the network is easier and less wiring is used than with other topologies. Because early bus topology networks used bulky coaxial cables, these factors were important advantages. Both 10BASE-5 (thick) and 10BASE-2 (thin) Ethernet networks are based on the bus topology.

Figure 20.10. A 10BASE-2 network is an example of a linear bus topology, attaching all network devices to a common cable.


However, the advent of cheaper and more compact unshielded twisted-pair cabling, which also supports faster networks, has made the disadvantages of a bus topology apparent. If one computer or cable connection malfunctions, it can cause all the stations beyond it on the bus to lose their network connections. Thick Ethernet (10BASE-5) networks often fail because the vampire tap connecting the AUI device to the coaxial cable comes loose. In addition, the T-adapters and terminating resistors on a 10BASE-2 Thin Ethernet network can also come loose or be removed by the user, causing all or part of the network to fail. Another drawback of Thin Ethernet (10BASE-2) networks is that adding a new computer to the network between existing computers might require replacement of the existing network cable between the computers with shorter segments to connect to the new computer's network card and T-adapter.

Ring Topology Another topology often listed in discussions of this type is a ring, in which each workstation is connected to the next and the last workstation is connected to the first again (essentially a bus topology with the two ends connected). Two major network types use the ring topology: Fiber Distributed Data Interface (FDDI). A network topology used for large, high-speed networks using fiber-optic cables in a physical ring topology Token-Ring. Uses a logical ring topology


A Token-Ring network resembles a 10BASE-T or 10/100 Ethernet network at first glance because both networks use a central connecting device and a physical star topology. Where is the "Ring" in Token-Ring? The ring exists only within the device that connects the computers, which is called a multistation access unit (MSAU) on a Token-Ring network (see Figure 20.11).

Figure 20.11. A Token-Ring network during the sending of data from one computer to another.

Signals generated from one computer travel to the MSAU, are sent out to the next computer, and then go back to the MSAU again. The data is then passed to each system in turn until it arrives back at the computer that originated it, where it is removed from the network. Therefore, although the physical wiring topology is a star, the data path is theoretically a ring. This is called a logical ring. A logical ring that Token-Ring networks use is preferable to a physical ring network topology because it affords a greater degree of fault tolerance. As on a bus network, a cable break anywhere in a physical ring network topology, such as FDDI, affects the entire network. FDDI networks use two physical rings to provide a backup in case one ring fails. By contrast, on a Token-Ring network, the MSAU can effectively remove a malfunctioning computer from the logical ring, enabling the rest of the network to function normally.

Star Topology The most popular type of topology in use today has separate cables to connect


each computer to a central wiring nexus, often called a hub or concentrator; a switch can also be used in place of a hub. Figure 20.12 shows this arrangement, which is called a star topology.

Figure 20.12. The star topology, linking the LAN's computers and devices to one or more central hubs, or access units.

Because each computer uses a separate cable, the failure of a network connection affects only the single machine involved. The other computers can continue to function normally. Bus cabling schemes use less cable than the star but are harder to diagnose or bypass when problems occur. At this time, Fast Ethernet in a star topology is the most commonly implemented type of LAN; this is the type of network you build with most preconfigured network kits. 10BASE-T Ethernet and 1000BASE-TX Gigabit Ethernet also use the star topology. 10BASE-T can use either Category 3 or Category 5 UTP, whereas Fast Ethernet and Gigabit Ethernet require Category 5 UTP or greater.

Hubs and Switches for Ethernet Networks As you have seen, modern Ethernet workgroup networks are based on UTP cable with workstations arranged in a star topology. The center of the star uses a multiport connecting device that can be either a hub or a switch. Although hubs and switches can be used to connect the networkâ&#x20AC;&#x2039;and can have several features in commonâ&#x20AC;&#x2039;the differences between them are also significant and are covered in the following sections.


All Ethernet hubs and switches have the following features: Multiple RJ-45 UTP connectors Diagnostic and activity lights A power supply Ethernet hubs and switches are made in two forms: managed and unmanaged. Managed hubs and switches can be configured, enabled or disabled, or monitored by a network operator and are commonly used on corporate networks. Workgroup and home-office networks use less expensive unmanaged hubs, which simply connect computers on the netwo