According to analysts at the Consumer Electronics Association, US citizens bought $113.5 billion in electronics last year, and those analysts forecast sales of $127.5 billion for this year (Reference 1). DTV (digital-TV) purchases, specifically of thin-and-flat televisions, represent a significant percentage of those sales figures; the total sales of DTVs last year reached $10.7 billion, or 7.3 million units, representing an increase of 78% in dollar sales and 63% in unit sales over 2003's numbers. Sales of LCD TVs topped $2 billion in 2004 and will surpass $3 billion this year. Unit sales of plasma TVs reached 853,000 in 2004 and are forecasted to grow to more than 1.4 million units this year. And a significant proportion of this year's DTV sales likely have already happened in preparation for last month's Super Bowl.

Particularly encouraging to the television vendors is the fact that consumers that already own at least one television are purchasing these new TVs. For example, a 2001 study by the Energy Information Administration, part of the Department of Energy, concluded that 98.9% of all US homes, regardless of their income level, contain at least one television (Reference 2). Additional good news to the television suppliers comes from the fact that LCD and plasma TVs are significantly more expensive than the CRTs they're obsoleting and more expensive than the large-screen CRT-alternative replacement. This trend mirrors one that's already well under way in the office cubicle: According to analyst company DisplaySearch, sales of LCD computer monitors definitively surpassed those of CRTs in the third quarter of last year (Reference 3). Fueling this trend is the allure of the all-important "digital" moniker in its multiple forms, including image sources, destination-display characteristics, and source-to-destination interface capabilities, as well as by aesthetic factors, such as thickness, weight, and screen size, particularly in the United States, with its comparatively large living rooms versus the rest of the world.

The outlook for CRTs seems grim. Don't count them out, though, say pundits such as Ray Soneira, PhD, the founder, president, and chief executive officer of DisplayMate Technologies, who notes, "The CRT is much maligned and has lots of terrific advantages in addition to the well-known disadvantages." What CRT shortcomings do emerging direct-view display alternative technologies address? What shortcomings do these alternatives have relative to the CRT and to each other? How is the CRT responding to the competitive threat, and how big a role will it play in the future? This article, along with EDN's Part 2 coverage in an upcoming issue, will address these and other related questions (see sidebar "Stay tuned").

Anyone who has used CRTs is well-aware of their deficiencies; they're bulky, power-hungry, and heavy. Form-factors issues have to date limited CRTs' screen sizes to 36 in. diagonal (wide screen) or 40 in. (4×3 aspect ratio). The color phosphors in their tubes inevitably fade. Manufacturers often quote 20,000-hour lifetimes to the half-brightness threshold, although your experience may vary, depending on your display's settings. The CRTs are also prone to image burn-in, although they're less sensitive to this effect nowadays than in the past. CRTs have a number of strengths, too, however; the biggest one, low cost, bodes well for the approach's future fortunes in applications that have no space, power, or weight constraints; that don't require a large screen; and that, conversely, are primarily fiscally motivated.

CRTs are notable for their high-quality images; color gamuts are comparatively wide. Thanks to an electron beam that the display's control circuitry can finely modulate and shut off when unnecessary, CRTs deliver deep black levels and wide contrast ratios, or ranges, as in "dynamic range." Within practical limits of design and cost, CRTs are also able to adapt themselves to multiple image resolutions and frame-sync rates and still maintain their top-notch pictures. Set the refresh rate too low, however, and observers tend to get headaches and suffer other fatigue symptoms after long periods of viewing; in contrast, fixed-pixel displays update only the portions of the screen that change during each refresh interval and are therefore not subject to flicker side effects.

The first response of the CRT camp to the looming plasma and LCD threat was low-distortion flat-CRT displays that lacked the curved front faces of their predecessors. More recently, companies such as LG Philips and Samsung have poured significant amounts of R&D dollars into developing thinner and lighter CRTs. Samsung's 30-in. thin CRT, for example, is only 16 in. thick and represents a 20% depth reduction and 10% weight reduction compared with a conventional CRT (Figure 1a). Innovations in all areas of TV design—electron-beam gun, tube, and aperture grill or shadow mask—were necessary to increase the maximum electron deflection angle from 110 to 125°, commensurate with the shorter tube and cabinet.

Canon and Toshiba, with their SED (surface-conduction electron-emitter display), take an even more enthusiastic stab at revamping the CRT. SEDs operate in a conceptually similar fashion to their predecessor, with electrons exciting phosphors to emit wavelength-specific photons. However, unlike with the CRT and its single-electron-beam approach, SEDs provide an electron gun for every pixel, thereby allowing the screen to be flat and shallow (Figure 1b). Toshiba will introduce its first SED sets this year, in limited quantities, and will primarily position them against plasma displays (Figure 1c).

According to Sean Collins, Toshiba America Electronic Components' director of business development, "SED will deliver to the industry a superior quality image that contains high resolution, fast response time, high contrast ratio, and dynamic color expression. These qualities, will combine with low power consumption and a thinner and lighter form factor, to distinguish SED as the ideal high-definition content display. Toshiba believes that the market for HDTV flat-panel displays will continue to grow and that SED will penetrate a significant share of this market, particularly in the space above 40 inches." As an aside to Canon's and Toshiba's work, another thin-CRT pioneer, Candescent Technologies, filed for bankruptcy last year. The purchaser of its intellectual property? Canon.

If you want to enjoy a big, slender display without paying more than $2000, it's likely that a plasma monitor will catch your eye. Plasma technology delivers in a svelte package many of the same image-quality strengths as CRT. But carefully examine the fine print before committing your credit card, or you might end up with buyer's remorse down the road.

Plasmas, like CRTs, are emissive displays. Minuscule cells between two panes of glass, which also house address and display electrodes, contain noble gases, such as xenon and neon. By charging the electrodes that intersect at the given cell to illuminate, the display's electronics excite the cell's gases via a voltage-potential difference, stimulating them to release ultraviolet photons that consequently excite color-specific phosphors. By varying the on/off rate and pattern, the display creates a full black-to-white intensity scale (Figure 2a).

You might think that a plasma screen, like a CRT, can produce pure black levels. After all, having no potential difference means that no current and, therefore, no photons flow through the noble gas, right? Unfortunately, as DisplayMate's Soneira explains, "All plasma panels have a special internal maintenance function that primes each pixel for pulse-width-modulation gas-discharge cycles ... The priming voltage ... produces the minimum luminance glow for the 'off' pixels." Even with this limitation, most of today's plasma displays deliver deeper blacks and, consequently, wider contrast ranges than their LCD peers. They also tend to have wider viewing angles—the degree to which an observer can be off-axis from the front of the display and still obtain a high-quality viewing experience—than LCDs.

However, plasma displays are as power-hungry as their CRT brethren of comparable screen sizes (Figure 2b). They're also heavier than their LCD counterparts, and operation at high elevations can sometimes be a challenge. Because of the increased pressure on the compressed gases within the dual glass substrate at high altitudes, the set's cooling system must work harder, resulting in higher fan or convection "buzzing" noise, and the plasma display's operating life proportionally decreases.

Plasma displays' phosphors also gradually degrade, although more slowly than they once did; manufacturers specify the half-life of latest generation plasma sets at 60,000 hours, although it depends on the display's APL (average picture level), a measure of image brightness. This degradation dims the intensity of the colors the displays output, and the displays are subject to burn-in when they project stationary images for long periods. Experiments by plasma manufacturers to vary the types and amounts of noble gases, in attempting to reduce burn-in effects, have generally been unsuccessful, according to Sharp Labs' Gary Feather, director of digital-audio/video systems. Instead, modern plasma displays take a more brute-force approach to preventing burn-in, moving the displayed image around on the screen slowly enough that a viewer won't notice.

Best Buy's recent newspaper advertisement of a $1500, 42-in. display exemplifies plasma's current market position. For that price, you'd be able to buy only a 30-in. LCD television—from a second-tier vendor. However, that LCD set would feature an HDTV display with 1280×768-pixel resolution. The plasma display, in contrast, features an only 852×480-pixel EDTV (enhanced-definition-TV) resolution, and it is also from an second-tier vendor. Upgrade your expectations to an HDTV plasma set from a tier-one supplier, and the price significantly increases, too, to near parity with the LCD alternative. In that same advertisement, a 43-in. Pioneer plasma set with "stretched" 1024×768-pixel resolution costs $4500; a 42-in.Panasonic plasma display with 1280×768-pixel resolution sells for $5500. Prices vary with the supply-versus-demand conditions at each point in time, of course; they also depend on whether the display includes a NTSC (and, in some cases, ATSC) tuner.

Dallas Mavericks owner and HDNet founder Mark Cuban claims to have talked to several people who think it's likely that prices for plasma displays will fall to less than $1000 by the end of next year if not sooner (Reference 4). Cuban's enthusiastic promotion of this prediction is ironic, because, although the article is about high-definition flat screens, a $1000 plasma display will likely not be a high-definition variant and will, therefore, not be the best way for consumers to view his HDNet network's 1080i broadcasts. Near-term fortunes aside, the long-term prognosis for plasma is unclear at best. The consensus at last fall's iSuppli Flat Information Displays conference, for example, was that, by decade's end, LCDs from below and DLPs (digital light processing) from above would squeeze plasma into insignificance. Terry Shea, general manager of corporate communications for JVC, comments, "Our focus ... will be our HD-ILA [liquid-crystal-on-silicon] rear-projection sets and direct view LCD. Our 2005 line includes two plasmas [42 and 50 in.], but plasma will not be a priority for JVC."

Plasma has historically shown itself able to more quickly ramp up to large screens than LCD (for manufacturing and bill-of-materials cost reasons that echo those in the long-standing hard-disk-drive-versus-flash-memory-density tug of war), and the trend continued at this year's CES. Samsung showed a prototype 102-in. plasma display behind closed doors, and the largest LCD was a 65-in. unit from Sharp (figures 2c and d). At these extreme dimensions, EDTV resolutions will obviously be inferior to HDTV line counts; by the time sets this large enter high-volume production, high-definition video content will also be commonplace, not only from over-the-air, cable, and satellite TV, but also from sources such as D-VHS and HDV tape and next-generation Blu-Ray and HD-DVD optical discs. Connecting displays to PCs, which relatively easily output 768 and higher line counts, also significantly increases the resolution expectation of those displays.

One key factor in the widely forecasted future-LCD-growth explosion at the expense of plasma is LCDs' diversity of applications, which lead to significant volume-cost efficiencies . The fact that, at equivalent screen sizes, LCDs are thinner, lighter, more power-thrifty, and, possibly, brighter than their plasma counterparts doesn't hurt LCDs' appeal, either. Unlike plasma displays, LCDs are not subject to burn-in issues, and, when LCDs' backlights fade after 75,000 hours or so of intensity-dependent use, users can easily and cheaply replace them; the LCD itself has an "infinite" lifetime. Conversely, LCDs have more restrictive viewing angles, less refined black levels, and, therefore, narrower contrast ranges than plasma displays. Historically, LCDs have also delivered a response on the order of 35 to 50 msec to on-screen image changes. Fast-motion 3-D game players and HDTV viewers, especially of sports programming, feel this slow response more acutely than users of traditional computing applications.

According to the free online encyclopedia, Wikipedia (www.wikipedia.com), "LCDs [have] transmissive or reflective modes. A transmissive LCD is illuminated from one side ... via a backlight and viewed from the opposite side. Activated cells therefore appear dark while inactive cells appear bright. This type of LCD is used in high-brightness applications such as pocket television receivers. The lamp used to illuminate the LCD in such a product usually consumes more battery power than the LCD itself. A reflective LCD, as used in pocket calculators and digital watches, is viewed by ambient light reflected in a diffuse (light-scattering) reflector behind the display. This type of LCD has lower contrast than the transmissive type, because the ambient light passes twice through the display before reaching the viewer. The advantage of this type is that there is no lamp to consume power." Combination transmissive-plus-reflective LCDs that, for example, extinguish the backlight in high-ambient-lighting environments, are also available.

Wikipedia further explains, "LCDs with a small number of segments are supplied with one electrical contact for each segment. This passive display structure becomes unwieldy when the number of elements increases. Medium-sized displays have a passive matrix structure. This type has one set of contacts for each row and column of the display. However, only one pixel can be addressed at any instant. The other pixels have to remember their last state until the control circuit has time to revisit them. This results in reduced contrast and a poor response to fast-moving images. The technology used is typically STN (supertwist nematic) or a double-layer version DSTN that corrects the color-shifting problem of STN. For high-resolution color displays such as large LCD monitors, an active-matrix system is used. The LCD panel contains, besides the polarizing sheets and cells of liquid crystal, a matrix of TFTs (thin-film transistors). These devices store the electrical state of each pixel on the display while all the other pixels are being updated. This method provides a much brighter, sharper display than a passive matrix of the same size."

The LCD-black-level issue fundamentally derives from the fact that the display employs an imperfect subtractive-light-transmission process (Figure 3a). These devices' polarizing sheets and intermediary crystalline material cannot block all the photons the backlight generates from passing through them and, eventually, to a viewer's eyes. Areas of the image that should be pure black, therefore, end up dark gray instead. In the near term, developers of large LCDs have improved black-level and contrast range performance by switching from a single-zone to a multizone array of CCFTs (cold-cathode fluorescent tubes) as the backlight, with a diffuser in front of the tubes to provide even illumination. When displaying dark images, the display automatically turns off some of the CCFT zones. In the longer term, a shift to LEDs as backlights is under way; small-format displays in which LEDs are the only feasible backlight option are leading this shift (Reference 5). In this case, the display can employ a large number of not only white LEDs, but also red, green, blue, and other colors, enabling adjustment of not only backlight luminance, but also its color characteristics (Figure 3b).

To mitigate motion artifacts that slow-switching LCD crystals cause, two techniques find common use. The first approach again focuses on the backlight because it can quickly switch on and off. In this approach, the display extinguishes the backlight between frames to make the resultant frame-to-frame transitions more distinct. Second, the LCD-control circuitry "overdrives" each subpixel crystal cell whose state is to change to stimulate it to more quickly switch. As a result, the latest generation LCDs have less-than-10-msec response times but only for abrupt white-to-black and black-to-white toggles; more gradual transitions form one shade of gray to another invariably take longer. Sharp's Feather flatly states of modern LCD TVs, "Motion blur is no longer an issue for our technology." Options for improving an LCD's horizontal- and vertical-viewing angles include altering the characteristics of the liquid-crystal material within the display and employing wide-viewing film in front of the LCD. Some LCD and notebook-PC vendors have also switched from traditional matte screen coatings to glossy approaches, such as Dell's UltraSharp, Hewlett-Packard's BrightView, and Sony's Xbrite. This technique makes the appearance of the display brighter and crisper, but the trade-offs include more noticeable scratches and fingerprints, a higher probability of eyestrain after long periods of use, and higher reflectivity. (You can potentially see yourself in the screen, and lights and other objects in the background can also reflect; many plasma displays also exhibit this shortcoming.)

One other lingering issue that LCD detractors doggedly bring up again and again involves bad pixels and subpixels—those stuck "on," those stuck "off," or both (Reference 6). Like both image sensors and semiconductor memories, the "yield" of LCDs can dramatically increase, and the cost per LCD can consequently decrease if manufacturers allow panels with defective pixels to pass testing. However, also like image sensors but unlike semiconductor memories, redundant extra pixel circuits cannot "patch" these defects. Manufacturers have different criteria for what threshold of defective pixels constitutes a "good" versus a "bad" panel, and this criteria may even depend on the model and the customer within a company's product line. Samsung, for example, recently announced that it would henceforth ship only 100%-functional LCDs—but only to Korea. A fully or even mostly functional panel becomes exponentially more challenging to manufacture as the pixel pitch decreases and the number of pixels vendors squeeze into a given-sized LCD consequently increases. Compare, for example, the prices of 1280×1024- and 1600×1280-pixel-native-resolution, 19-in. LCD computer monitors.

LCD companies are bringing a mind-boggling amount of manufacturing muscle to bear on the worldwide market opportunity, and, when this muscle doesn't exactly align with the time-varying customer demand, it produces DRAM-reminiscent boom and bust cycles of undersupply and oversupply. As Sharp's Feather explains, through the sixth generation of LCD factories, each generation roughly doubled the size of the LCD glass sheet that its predecessor delivered. Beginning with the seventh generation, the generation-to-generation doubling rate will slow, but it will still deliver incremental improvements in screen-size growth and cost reduction. Sharp's sixth-generation process, for example, creates sheets that measure 1800×1500 mm (5.9×4.9 ft), each of which can transform into eight 32-in. LCDs with 95% sheet-usage efficiency or two 65-in. LCDs. Sharp will fabricate the seventh-generation successor with 2160×2400-mm glass, which can make eight 40-in. or six 60-in. LCDs per sheet.

OLEDs promise to address many of LCDs' shortcomings, but the relevant vendors have been slow to transform marketing pitches into shipping products, and they're fighting the established and rapidly accelerating momentum of the LCD juggernaut. Because OLEDs, like plasma, are self-emissive, they're brighter than transmissive LCDs and dispense with the cost, bulk, and power draw of a backlight. At last year's SID (Society for Information Display) Conference, Philips showed its LifePix technology, which allows you to adjust the color characteristics of an image to match the limited gamut of an STN LCD or a TFT LCD. (Figure 4a and see sidebar "Continuing education").  Alternatively, you could choose a potentially easier design path, dispensing with the incremental postprocessing steps which LifePix and alternative algorithms represent and instead just employing a wide-gamut OLED display.

Perhaps OLED's biggest Achilles' heel is its limited lifetime, especially for blue-light-spectrum variants. To date, most blue-OLED elements lose half their brightness within 1000 hours of use, although manufacturers claim that recent developments extend this threshold to 10,000 to 30,000 hours, and Cambridge Display Technologies claims that its latest laboratory experiments have delivered 80,000-hour blue-LED life. The impact of this shortcoming is application-dependent; if the display is operating 24 hours a day, 365 days a year, 10,000 hours translates to only a bit more than a year of life. Conversely, if the display is on for only an average of one hour per day, it will provide more than 27 years' worth of operation before it reaches the half-illumination point. Keep in mind, too, that the purchaser of a cellular phone, who'll likely replace it in a few years' time, will have less stringent display-life expectations than someone who buys a $10,000 television.

Extrapolate a single pixel's worth of characteristics to an array of pixels, and the manufacturer's job becomes exponentially more difficult. The vendor must ensure not only consistent initial operation across the OLED array, but also consistent degradation over time. To date, OLEDs have achieved limited penetration in display-inclusive systems. Eastman Kodak, an early pioneer in OLED and the holder of a high percentage of the published technology patents, decided two years ago not to produce its OLED-inclusive, publicly announced LS633 digital still camera when it couldn't manufacture sufficient OLEDs. OLED technology has appeared in monochromatic form within several Pioneer car-audio head ends, though, as well as in a portable media player unveiled at January's CES. Also at that show, Samsung unveiled a 21-in. OLED-prototype display, and Seiko Epson last May previewed its 40-in. OLED (figures 4b to 4d).

For more information on the products and manufacturers this article covers, check the sidebar "For more information" on the Web version of this article at www.edn.com.

Author Information

Technical Editor Brian Dipert wonders when the oft-touted futuristic vision of wall-sized displays in homes will become commonplace and whether it will be in the form of direct-view technology, projection technology, or both.