The next television you buy won’t contain a cathode-ray tube. And thanks to the imminent worldwide transition from analog to digital television, you probably will be buying a new TV soon--you and more than a billion other consumers around the world.
With the United States going fully digital in 2009 and other developed countries following closely on its heels, the entire television landscape will change dramatically in the next few years. What technology will dominate this new TV world? It will be flat, that’s for sure. All the new television technologies vying to dethrone the clunky, 100-year-old CRT are sleek and thin.
But will it be plasma, liquid crystal, or one of several new technologies not yet on retail shelves? Will they all survive, or do some of them have intrinsic technical weaknesses that will doom them in the next five or six years? I’m going to tell you where I’m placing my bets, but first, I’ll look at all the horses in this race.
Today, plasma televisions are what many consumers would like to have hanging on their walls; liquid-crystal-display televisions are what they can afford. Both are flat-screen systems with picture quality that, when displaying high-definition video, rivals film’s. Prices for both have dropped dramatically in the past few years; a 42-inch-diagonal plasma set that includes a tuner sells for about US $2000 today, compared with $5000 just four years ago. A 37-inch LCD set, a more popular size for that technology, runs about $1200.
This year, nearly 8 million plasma TVs will be manufactured worldwide, according to Gartner Dataquest--a unit of Gartner, headquartered in Stamford, Conn.--nearly double 2005’s production. Gartner estimates that LCD manufacturing plants will have rolled out nearly 42 million of those popular television displays by year-end. The vast majority of these sets, plasma and LCD alike, are produced in Japan, Korea, and Taiwan.
Both plasma and LCD TVs can have considerably bigger screens than CRTs can, but plasma screens can be absolutely huge: the largest screens available today measure 102 inches diagonally. Plasma images are also fairly bright and can be viewed clearly from virtually any angle. Today’s LCDs have a wider viewing angle than earlier models, but the field of view is still not quite as wide as that of other technologies.
Nevertheless, a plasma TV won’t be the last TV you buy. Here’s why: it’s got limited longevity, it’s power hungry, and it’s heavy. Like CRTs, plasma displays use red, green, and blue phosphors, but instead of hitting the phosphors directly with a beam of electrons, as in a CRT, a plasma display charges pockets of xenon and neon gas trapped between two glass panels with a honeycomb of cells.
In essence, every plasma display contains about a million miniature fluorescent tubes, evenly divided among red, green, and blue. When the charged, or ionized, gas releases ultraviolet photons, these photons strike the phosphors, which, in turn, emit the colored light that produces the television picture.
The longevity problem comes from the fact that the light-emitting efficiency of the phosphor coating decreases over time--that is, when a phosphor is stimulated by a photon, it releases less and less light. The problem is much worse in a plasma set than in a CRT because the plasma’s phosphors exist in a hostile environment; the electron beam in a CRT is much kinder to phosphors than are a plasma’s hot gases. In a plasma display, the contrast ratio--the difference between lit and unlit picture elements--drops quickly under normal use, as much as 50 percent in four to five years. At that point, the television image appears noticeably washed-out.
Manufacturers today claim 60 000 hours of use before the brightness falls by half (based on a few hundred hours of testing). Contrast, however, is more important than brightness. Recent tests by market research firm IDC, in Framingham, Mass., measured a 13 percent decline in the darkness of the black of a typical plasma television after four weeks of use; after five years of use, such a rapid decline could lead to blacks displaying as light grays.
Plasma displays also consume more power. Even though manufacturers have reduced the power consumption of typical plasma technology by 30 percent over the past five years, these units continue to need more power than comparable LCD TVs, particularly when displaying a white or light screen. This power consumption generates heat; if the sets are not cooled properly, heat build-up can damage components. Before you buy a plasma TV, consider this fact: Philips sent repair technicians to 12 000 U.S. homes this past spring to replace components in plasma TVs that could potentially overheat.
Plasma displays are also heavier than their flat-panel competitors. Because the glass panels that surround the gas are much thicker in plasma displays, a 40-inch plasma set weighs 43 kilograms, while the same-size flat-panel set weighs just 25 kg. Plasma technology requires such thick glass because the gas is very hot; thin glass would simply melt.
There are other problems, too, such as burn-in. It’s a particular problem nowadays because with 1000-channel cable and satellite services, TV networks feel an acute need to identify themselves all the time, usually with a static channel logo in the lower part of the screen. Again, because plasma technology is harder on phosphors than CRT technology is, burn-in happens faster and is more noticeable on plasma televisions. Manufacturers have done a lot to deal with this problem, and on new plasma sets, after approximately 12 hours of use without the static image, the burn-in will fade away. But it is still a drawback.
And as if those problems weren’t enough, plasma sets also don’t work well at high altitudes or, indeed, in any place where the ambient pressure is different from that of their internal gases. When such a differential exists, the TV’s power supply has to work harder to keep the gases ionized.
To be sure, plasma manufacturers have worked hard to address the technology’s deficiencies. They have developed longer-life phosphors, and they have made great strides in controlling light leakage between cells, successfully displaying darker blacks. But plasma sets with these improvements cost significantly more than competing LCD sets. And in any case, ordinary consumers are mostly buying the cheaper sets, in which the problems remain.
So what about the LCD , today’s most obvious alternative to plasma? A liquid-crystal television is, in effect, a sandwich with many ingredients. Its layers include a bright white backlight, a layer of liquid-crystal molecules, a matrix of thin-film transistors, two pieces of polarized glass, and colored filters. The transistors control the voltage applied to the three groups of liquid-crystal molecules that make up each picture element.
When the voltage is on, it twists the molecules, allowing light through the layers of glass and color filters; the molecules untwist when the voltage is off, blocking light. Each picture element consists of liquid-crystal molecules above a red, a green, and a blue filter. Switching the appropriate molecules on and off gives myriad combinations of red, green, and blue light, and therefore the palette of human vision.
Because of this many-layered structure, liquid-crystal TVs start off with a relatively poor contrast ratio--the difference between the brightest white and the darkest black that the screen can display. The white backlight that illuminates the displays is usually a cold cathode fluorescent tube, which operates on the same basic principles as the ubiquitous and venerable neon sign. The light has a rough path to travel through the many layers before it reaches the viewer’s eyes. Each layer absorbs some of the light, leading to reduced contrast and brightness.
LCDs are basically reliable. The component that limits their life is the backlight. The fluorescent tubes age over time; after about five years of normal home use, the tubes start to dim and their color temperature starts to drift--that is, the hue of the emitted light shifts from clean white toward the red end of the spectrum. The shift is gradual; viewers usually don’t notice the change until it is extreme. Because of this aging of the fluorescent tubes, the average usable life of a liquid-crystal TV is about seven to 10 years--close to that of ordinary CRT-based TVs.
To improve the longevity of liquid-crystal TVs, some manufacturers have recently started to use high-intensity light-emitting diodes (LEDs) as backlights. They aren’t cheap, though prices will come down as manufacturing volumes go up. Samsung sells a 46-inch model for about $9000; Sony is selling a similar set for about $12 000. In these models, instead of a fluorescent lamp, an array of red, green, and blue LEDs creates what appears to be white light.
Besides increasing the usable lifetime of an LCD, this LED-based lighting increases the color saturation of the display. Saturation is basically the purity of a color, or, more precisely, the relative bandwidth of the light. Light emitted at tighter bandwidths is more saturated; light occupying wider bandwidths looks washed-out.
When the color filters on an LED-illuminated display remove blue and green to display a red pixel, for example, the resulting red is at the single red frequency originally generated by the red LEDs. Do the same filtering to display red on a fluorescent-illuminated display, and a wide range of red frequencies creates a less-saturated color.
Saturation is particularly important for large high-definition liquid-crystal panels bigger than 37 inches, because picture-quality problems are more pronounced in larger panels. Increased saturation allows finer gradations of colors, enabling pictures to seem startlingly vivid, an effect particularly striking in scenery. A surfer on a red board pops out in a vast blue ocean; it’s the kind of imagery that might even get you addicted to the Travel Channel.
But even LEDs don’t last forever. Degradation starts becoming noticeable after about 60 000 hours of use--for most people, that means roughly 15 years. Although these LED/LCD televisions incorporate sensors to measure and adjust their hues as the diodes age, they, too, after about a decade of moderate use, will fade and the panel will dim. And they consume about twice as much power as conventional fluorescent-lit LCDs. A 42-inch LCD backlit with LEDs runs at around 250 to 300 watts, only a little less than a plasma panel.
Pretty soon--in about two years-- there will be a third horse in this race. The surface-conduction electron-emitter display, or SED, is just now starting to emerge as a serious contender in the race to replace the CRT. SED is an alternative flat-display technology emerging from Canon and Toshiba.
In an SED [see diagram, ” SED Science”], every single pixel of the display is, effectively, a cathode-ray tube. The cathode is a thin film of palladium oxide, chosen because it is electrically conductive and also extremely durable, resisting oxidation and corrosion even at high temperatures. As in a CRT, electrons emitted from the cathode hit phosphors--tiny dots of metals or rare-earth compounds that glow red, green, or blue when energized.
The result is a flat-panel display that uses less energy than a plasma screen does and yet has image quality close to that of the CRT, still the benchmark of all displays. Power consumption is low, relative to that of plasma, for the same reason as it is for the CRT: it takes a lot less energy to create an electron beam than it does to excite photons in a gas.
SED is a variation of field-emission display technology [see ”Watching the Nanotube,” IEEE Spectrum, September 2003]. The main difference is that SED, for its cathodes, uses palladium-oxide film rather than the cone-shaped bundles of carbon nanotubes employed in field-emission displays. (Nanotube-based field-emission displays have, so far, proven difficult to manufacture, mainly because of the difficulty of producing the nanotubes.) SEDs theoretically have a manufacturing advantage because they can be printed using an industrial inkjet printer not that much different from the inkjet printer in your home or office.
Last year, Toshiba and Canon began trial production of surface-conduction displays in the 40- to 50-inch range. Despite the theoretical manufacturing advantage, the companies appear to be having issues in bringing yields up to a commercially viable level. They say they’ll start mass production next July and ship SED televisions to Japanese retailers later in 2007.
The first SED televisions are likely to cost about 50 percent more at retail than comparable plasma sets. At the moment, it is not clear whether they will suffer from any long-term reliability or performance issues. Next year we should know if surface-conduction technology has a real chance of carving out a significant niche for itself; if Canon and Toshiba can’t make a reliable product in high enough volumes at a realistic retail price, the market just won’t buy into the technology.
So which technology will dominate over the next four to five years? Two winners will emerge, one for screen sizes smaller than 50 inches, one for screen sizes larger than 50 inches. Fifty is the magic number in the TV business, because it is, at least today, the upper limit for economical production of reliable panels that lay electronics on a glass substrate. It is also, not coincidentally, the smallest projection television screen size in mass production today.
For screens smaller than 50 inches, which account for the vast majority of sets sold throughout the world, the two most attractive technologies will be LCD and, conceivably, SED. But the prices of the SEDs are not likely to come down significantly before 2010. So the near-term winner will be the LCD.
Liquid-crystal technology will dominate not only the under-37-inch, or small-TV, category but also the midsize segment of 40- to 50-inch TVs. The lower cost and long-term reliability of LCDs will make them a better value than SEDs or plasma displays.
For consumers looking for the largest screens, those bigger than 50 inches, projection TVs are going to be the best bet for the near future. As plasma, liquid-crystal, and surface-conduction display sizes increase, yield goes down and therefore cost goes way up. A manufacturing line that produces four 42-inch displays at a time can manufacture only one 100-inch display, and that large display is more likely to have faulty areas, reducing yield. Nevertheless, 103â''inch plasma screens will soon hit the market, as will 65â''inch LCDs. But the prices are going to be ridiculous--about $70 000 and $15 000, respectively.
Projection TVs have gotten better lately. If you wince at the memory of the shadowy, washed-out images of the projection TVs of 20 years ago, you’re in for a surprise. Today’s projection technologies include Digital Light Processing (DLP) from Texas Instruments and micro-LCD and liquid crystal on silicon (LCOS) used in TVs from Hewlett-Packard, JVC, Mitsubishi, RCA, Sony, Samsung, and others.
Of the 10 million projection televisions that Gartner estimates will be manufactured this year, 25 percent are expected to be based on DLP, 9 percent on LCOS, and 66 percent on micro-LCD. All three technologies offer pictures both brighter and sharper than those viewed by moviegoers at traditional cinemas today. Many models also display deeper blacks and correspondingly higher contrast than most LCD or plasma displays.
A high-definition DLP system contains an array of just over 2 million hinge-mounted micromirrors, each about 20 square millimeters. A bright white light shines on the array. The mirrors change orientation to either reflect light to the screen or not--that is, to make an individual pixel light or dark on the screen. The mirrors, controlled by microscopic electrodes, switch up to several thousand times per second.
The system coordinates the switching with the rotation of a single color wheel, typically 7 centimeters in diameter, enabling the mirrors to create the red, green, and blue components of every one of the millions of pixels in a television image. To keep up with the 30-frame-per-second refresh rate of National Television System Committee (NTSC) video, the wheel has to move precisely and fast--first-generation color wheels, with three color segments, rotated at around 3600 revolutions per minute. Today’s color wheels, with seven color segments (two each of red, green, and blue, plus one of white) rotate at about 7200 rpm [see illustration, ” Mirror, Mirror”].
LCOS displays also redirect reflected light to create the television image, but they use individual liquid crystals to do so instead of micromechanical mirrors. The liquid crystals coat a reflective surface, typically on a 15-mm2 silicon chip, and change their orientation to block or allow the light to reach that reflective surface. In single-chip LCOS systems, either a color wheel or an array of LEDs illuminates the LCOS chip. In multichip LCOS technology, three separate chips, one for each primary color, combine optically to produce the visible image.
The third competing projection technology, micro-LCD, uses three transparent LCDs, one each for the red, green, and blue components of a full-color image. Each LCD measures from 18 to 33 mm diagonally, depending on the particular manufacturer and model. Mirrors split light from a metal-halide lamp into red, green, and blue beams, sending each beam of light through the appropriate LCD. The three beams pass through the LCDs into a prism, which combines the light back into a single beam to form a full-color image.
Each of these technologies has minor drawbacks. The heat from the high-intensity metal-halide projector lamp, over time, degrades the liquid-crystal coating in the micro-LCD panels, discoloring the TV picture. The spinning color wheel of single-chip DLP and LCOS systems can create a rainbow effect for some viewers, because it depends on the human vision system to retain images instantaneously after they are actually no longer visible and thus merge the red, green, and blue images into one. Some people’s eyes adjust better than others’. The rainbow is most noticeable when the picture has a lot of contrast, like a candle on a black background. In a football match with lots of motion and detail, the rainbow effect is hardly noticeable.
And all projection systems share one major problem: the lamp.
Projection systems typically use metal-halide projector lamps, because they are bright and give a consistent color level and brightness over their lifetimes. These lamps produce light by passing an electric arc through a high-pressure mixture of argon, mercury, and a variety of metal-halide gases. The precise mixture of halides affects the nature of the light produced, influencing the correlated color temperature and spectral intensity (making the light bluer or redder, for example).
The argon gas in the lamp is easily ionized, creating the arc across the two electrodes. Heat generated by the arc vaporizes the mercury and metal halides, which produce light as the temperature and pressure increase. About 24 percent of the energy used by metal-halide lamps produces light, making them generally more efficient than fluorescent lamps and substantially more efficient than incandescent bulbs such as halogen.
But these lamps last only 1000 to 2000 hours, and they are not cheap to replace, at $300 to $400 each. Longer-life lamps are available, such as the ultrahigh performance (UHP) lamps invented by Philips. These lamps generate an arc in a nearly pure mercury vapor under high pressure. The arc gap can be much smaller than that of alternative lamp technologies, as small as 1.3 to 1.0 mm across.
The smaller gap is more efficient; a 100-W UHP lamp in a projector can deliver more light to the screen than a 250-W metal-halide lamp. UHP lamps range from 100 to 200 W, with useful life spans ranging from 3000 to 10 000 hours. They are now available in video projectors and rear-projection TVs from all the major manufacturers.
But like LCD manufacturers, projection television manufacturers are moving toward replacing lamps with high-intensity LEDs, likely to be pervasive within the next three or four years. These LEDs will not be cheap either, but they should offer lifetimes measured in tens of thousands of hours. Such lifetimes will make the maintenance and operating costs of projection systems comparable to those of other available television displays.
Projection TVs are also smaller than they used to be: the boxes containing the projection optics and electronics are a lot shallower than their predecessors of 20 years ago, thanks to microdisplays’ replacing the earlier large-tube technology. As a result, the average depth of a projection television with a 50-inch screen today is only 0.43 meter.
By 2010, LCD TVs will dominate in sheer numbers, though mostly at the smaller screen sizes. Gartner projects that nearly 90 million LCD televisions will be manufactured worldwide in that year alone and will be worth about $30 billion in retail sales. Projection TV production will grow steadily, with 14 million manufactured in 2010. DLP technology will take the largest share of the projection television market, at 47 percent, compared with 35 percent for LCOS and just 18 percent for micro-LCD.
Meanwhile, plasma technology will gradually die, a victim of economics rather than of its shortcomings, for manufacturers today are investing far more in LCD production than they are in plasma, and at the same time, other technologies are emerging from the laboratories. Manufacturers like Panasonic and Pioneer, which have invested billions in plasma manufacturing facilities over the last five years, will either become niche players or convert to another technology, most likely LCD. Although if Canon and Toshiba are really successful with their SED technology, it is possible that Panasonic and Pioneer will jump onto that bandwagon.
And you’ll be happy with the LCD TVs in your bedroom and kitchen and the projection TV in your family room until 2015 or so. That’s when the next wave of display breakthroughs (organic LEDs, anyone?) [see sidebar, ”Now Showing on the Small Screen”] will bring paper-thin televisions out of the laboratory and into the marketplace and a new race for television technology dominance sends you TV shopping again.
About the Author
PAUL O’DONOVAN is a principal research analyst for Gartner Dataquest, based in Egham, England, and covers semiconductors and consumer electronics. Before joining Gartner 10 years ago, O’Donovan spent 12 years as a marketing engineer for National Semiconductor Corp., in Santa Clara, Calif.