DVD PRO Conference
EMedia Expo
Current Issue
Article Archive
News Archive
Buyer's Guide
nbsp; Home Magazine eNewsletters Events Contact Navigation

EMedia Buyer's Guide

Current Issue Current Issue
DVDs in the Works


News Indices

CD TrackerCD/DVD-ROM IndexFact, Figures & FindingsConference Calendar



EMedia Buyer's Guide


eNewsletters
DVD TodayDigital EarfulTechVideo NEWSSubscript
Ad Links


It's a Gas, Gas, Gas: Plasma Display Technology Today

WK Bohannon

June 2001 | I was as happy as Ralphie when he got his BB gun in A Christmas Story the day I received my first plasma display. I had seen lots of plasmas at trade shows all around the world, and, finally, I was receiving an OEM unit from Fujitsu to evaluate all on my own. I was also as naive as a child, actually believing all the marketing hype about plasmas being wall-hanging TVs. However, when I saw the big truck backing into my driveway• emanating a loud, ominous, beeping noise warning all to stand clear•I began to wonder. Why should a slender, elegant wall-hanging TV need such a big escort? Was it some kind of wild elephant; a sort of modern-techno King Kong or Godzilla? I had envisioned a simple flat box containing a plasma display that would hang on my wall like an exquisite work of art.

Wrong. The big, noisy truck held a huge wooden box. With lots of whirling hydraulics and straining muscles, the delivery man and I hauled the box off the back of the truck and onto my porch. I signed the release, and was left to fend for myself. Certainly, this beastly crate could not contain the graceful, flat picture device I'd seen at Comdex. After about 30 minutes of work with a hammer and crow-bar, I unearthed another box; the wooden crate turned out to have been just protection, something to keep the cardboard box intact during its long sea voyage from Japan. Luckily, I was able to slide the cardboard box out of the end of the wooden crate, because there was no way that I could have lifted the whole thing on my own. Now, at last, I was ready to feast my eyes on the first plasma display to enter my domain.

But nestled in the cardboard box lay yet another enclosure, packed in a good deal of Styrofoam, along with enough parts, boxes, and instructions to intimidate the most stalwart father on Christmas eve. Receiving and assembling my first plasma was a lot more work than I ever imagined. I've unpacked and set up so many LCD projectors that I've lost count; then there's numerous TVs, monitors, projection CRTs, and rear-screen systems, and seldom have these products required much, if any, assembly.

Not so with my inaugural plasma installation. First, I had to assemble the plasma's stand, a heavy metal contraption capable of supporting the plasma's fragile, but hefty, weight. After that, I had to find a soft, flat surface and an assistant capable of lifting half of almost 90 pounds onto its easily cracked and scratched glass face. Then, I had to bolt the stand and frame to the back of the plasma, and stand it upright. Yes, the plasma looked thin, but weighed plenty, nonetheless. It required a wide, heavy stand to keep it upright and to prevent it from falling on its face and shattering into a million very expensive pieces. But this was all years ago, with my first plasma, and I figure they must have slimmed down over the years.

Wrong again. Today's plasmas are only a few pounds lighter than the first ones delivered two or three years ago; still weighing in between 70 and 80 pounds for the 42-inch (diagonal) models, and closer to 90 pounds for the 50-inch versions. The latest 60-inch units are heavier still, as would be expected, although I haven't weighed one. Those prototype products are so rare that I'll bet they travel by armored car.

That first plasma took at least two people to move around my home office, and I sweated every time I touched it out of fear of flexing it too much and cracking it. That hasn't changed either; one manufacturer told me that when they first started shipping plasmas, they lost about half of them due to breakage during shipment.

Despite the hype about "wall-hanging TVs," there was no way that I was going to perform any wall hanging (or banging) experiments with the first plasma I got to evaluate. I was able to measure it carefully, and then I put it back in its cocoon. However, I arranged for another plasma to be delivered to a friend's house for a wall hanging experience. He paid for it, and arranged for a team of construction workers to help mount the beast and attach it to the curved wall in his new "screening" room. Mounting the plasma to its heavy metal frame and wall bracket with sturdy 3/8-inch bolts was not the problem that actually hanging it proved to be. Neither of us did the heavy lifting required to bolt the unit to the wall with huge lag-bolts, but, afterwards, I did have my own struggle with the video cables. I needed a few more inches of cable to get the right angle on the plasma's BNC connectors, which were mounted on the bottom of the unit. But there was no way that I could pull the wires, as the plasma on its frame had been mounted close to the wall, and the angle was too acute to pull the wires through.

"Say guys, can't we just take this thing off for a minute to let me pull some more wire out?" I asked. Drenched in sweat from the effort expended to get the plasma into its current position, the crew replied, "Forget about that."

The Plasma Program, Players, and Performance

William Coggshall of Pacific Media Associates in Mountain View, Calif., estimates that about 35,000 plasma units (including 37-, 40-, 42-, and 50-inch sets) were sold in the U.S. for the "professional market" in 2000. Coggshall has been measuring the market for projection and display products for more years than I can remember, and he has recently begun to follow the emerging plasma market. These 35,000 units equal less than 10% of the number of projectors sold in the U.S., and an even smaller fraction of the number of rear-screen TV sets. Coggshall wouldn't release any numbers for U.S. market share for the various plasma players, but he did reveal that, within his overall estimate for market, most of the units were 42-inch sets. My own investigations tell me that most of the 42-inch sets were made by Fujitsu and its partners, Hitachi and Sony. Most, if not all, of the 50-inch sets were sold by Pioneer, who claims to have an 80% to 90% share of the 50-inch plasma market. Sources at Hitachi estimate that the total "world" market for plasma was closer to 200,000 units in 2000, which puts the U.S. in a relatively small segment in relation to the rest of the world.

So, the plasma market is still a small one, at least in the U.S., which may be one reason why California's power grid continues to hold on. Of course, everyone hopes that the plasma market will continue to grow, so much so that the major players have made enormous investments. Hitachi, Fujitsu, and Sony are all going in together to create a mega plasma factory in Miyazaki, on Japan's southern island of Kyushu. This factory•named "FHP" for Fujitsu, Hitachi Plasma (Sony seems to have been left out for only holding a 25% share)•will eventually be cranking out 70,000 units per month. The plant will produce a variety of 32-, 36-, 37-, 42-, and 50-inch products (and eventually 60-inchers) as well as a few 25-inch pieces. The resulting plasma products from the three companies will all use the same basic plasma components, and will all look pretty much the same except for slightly different colored anti-reflection, EMI coatings and other cosmetic touches and differentiation features.

Aside from the FHP powerhouse, the other major players•companies who make their own products in the plasma market•are NEC, Pioneer, and Panasonic. Some people might leave Panasonic out of the major player list, because they allegedly haven't shipped that many units, but I have no hard data to support that claim either way. A few years back, Panasonic acquired one of the "pioneering" U.S. plasma companies•Plasmaco. The gorgeous 60-inch prototype plasmas that Panasonic has been showing off were all supposedly produced in Plasmaco's U.S. plant. Plasmaco alums still hope to be in the running to actually produce 60-inch sets in the U.S. Other companies have shown large plasmas, which attract a lot of consumer and professional interest, but the ultimate product costs are high, as is the power consumption. What's more, the size and weight of the 60-inch units must be exponentially more inconvenient to ship than the 42-inch units. But none of that has stopped the development. As I reported earlier this year, Samsung has one-upped Panasonic•and every other plasma manufacturer•with its own home-grown, 63-inch, 1366x768- pixel plasma screen (model number SPD63P1H). Samsung still won't commit to its production plans, which are supposed to be in gear by the end of 2002. The price of the 63-inch set is supposed to be somewhere around $30,000 or so, depending upon the state of the market. The MSRP prices for other sizes range from around $9000, for the smallest 40- or 42-inch plasmas, to about $18,000, for the 50-inch models. Paying $30,000 for another step up isn't that outrageous, but it sure makes me think twice. If they do go into production, I'll keep an eye on the Web prices, which are generally at least 20% below the MSRPs.

Another Korean company, LG Electronics (formerly know as Lucky Goldstar) is also dipping into plasmas. LG, in case you haven't peeked inside your monitor lately, makes a huge quantity of the CRTs used in monitors and TVs. The company is also Apple's partner in LCD monitors, and now LG is going to get with the plasma program. It has been showing a 60-inch, 1280x720-pixel plasma prototype with the Zenith brand name (part number FD60X3R). LG's plasma prototype is perhaps closer to production than Samsung's, with a planned launch around June 2001, priced at about $28,000. But who can really tell when any of the companies making noise about 60-inch units will really go into full production and what the pricing will be. However, $28,000 or less for the first run of 60-inch models is not that crazy, since $25,000 is about where the 50-inch products started a couple of years ago. LG-Zenith was also showing a 42-inch, 852x480 RGB pixel, 16:9 aspect plasma, as well as a 36-inch in the same resolution and aspect, plus an unexciting 40-inch VGA plasma in 4:3 aspect.

Fujitsu, Hitachi, and Sony

As described earlier, it looks like Fujitsu, Hitachi, and Sony are the biggest guys in the market, with the most power in terms of "planned" production capacity. All of their units will be coming out of the same factory, with slight changes in features and marketing approaches. Sony, for example, has been showing a 42-inch, 16:9 aspect panel in the traditional 852x480 RGB pixel resolution (the PFM-500A3 monitor) along with a 1024x1024 resolution, 42-inch unit (PFM-42B1). Hitachi also sells a 25-inch SXGA plasma, and has plans for 32-inch and 36-inch HDTV plasma monitors, as well as a 60-incher for 2002 or so. Fujitsu's 42-inch, 852x480 plasma (the PDS-4209U) lists for $7999; and its version of the 1024x1024 resolution unit (the PDS-4221, without speakers, and PDS-4222, with speakers) has an MSRP of $15,999. Hitachi has recently announced a price drop to $9999, so I expect both Sony and Fujitsu to follow closely behind with their similar products. Competition is a good thing in this, as in any market.

The Hitachi 42-inch unit that I evaluated (the CMP402HDU, which now sells for $9999, down from $15,999) had a significant increase in resolution over competing 16:9 aspect ratio, 42-inch units. Hitachi (along with Sony and Fujitsu) has managed to squeeze in 1024x1024 RGB pixels versus the other companies' 852 or 853x480 RGB pixels, to achieve one of the highest resolutions seen to date in a 42-inch, 16:9 aspect unit. The Hitachi units are solid, middle-of-the-road performers in terms of both contrast and brightness. However, like all plasmas, the colors are different.

Hitachi doesn't use any color filters to "purify" the already-colored plasma emission colors. Nor does Hitachi vary its pixel size with color. Instead, the manufacturer claims to build a little bit of overall color correction into the anti-reflection, anti-electro-magnetic interference (EMI) filter that covers the front of its plasma screen (and blocks half the light). This filter changes the white point, but hardly affects the color saturation at all. According to my investigations, one of the other reasons behind the varying colors in all the plasma screens is the differing gas mixtures and tweaks to the colored phosphors. Every manufacturer uses its own secret sauce, and then, because of that secret, have to make certain adjustments in an attempt to get better colors. NEC uses what it does for gas and phosphor, and may need color filters to make that work. Whatever Panasonic uses causes its plasmas to require asymmetric pixel sizes to correct for it. In Hitachi's case, the resulting colors are again different from the others. This slight difference will extend to the other members of FHP, since Fujitsu and Sony supposedly use slightly different color corrections in their respective EMI filters.

In any case, even without additional color filters on each pixel, I thought that the Hitachi unit's color performance overall was a little better than the NEC's on some images. Maybe this was because the Hitachi unit had a slight edge over NEC's in contrast and white point. But don't get me wrong, both displays looked fine under optimum conditions. I also liked the performance of the Hitachi plasma with either NTSC or HD video or computer data input in true XGA resolution. Yes, Hitachi has true XGA resolution. Even its 37-inch, 4:3 aspect ratio CMP307XJ, which sells for $19,995, has true 1024x728 line resolution, while the CMP402HDU has 1024-line resolution, and either one is a great improvement over the VGA resolution in the other 42-inch plasma displays. Plasma really doesn't appear to resize very well, and you need more resolution than VGA to show the entire desktop in most modern computers.

When you ask a plasma (or, for that matter, any digital, pixilated display device) to resize beyond its native resolution, the pixel data starts to modulate or turn off and on quite rapidly, in an attempt to show more data with few pixels. This is okay in an LCD display, which turns on and off quite smoothly. In fact, if you "wiggle" an LCD's pixels hard enough, they just find some average value where everyone's happy. But that doesn't happen in a plasma, since the plasma's pixels like to be either on or off. So, when you flip a plasma's pixels too rapidly, they flicker, shimmy, or sparkle, and that looks a little weird. Because resized plasmas look out of focus, I tend to keep them as close as possible to their native resolutions, where they look their best. Plus, at this point, they're at their most stable and brightest, with the best contrast ratio.

Pioneer

Pioneer is the current heavyweight champ in the 50-inch market, but its status could change as soon as another company (like NEC, perhaps) enters a competitive 50-inch or larger plasma, and really starts to ship mass quantities. In addition to Pioneer's 50-inch product•the 1280x768 RGB pixel PDP 502 that now sells for $17,995•Pioneer also makes a 40-inch, 4:3 aspect model called the PDP N402, which currently sells for $8995.

Pioneer's 50-inch unit was designed as an XGA resolution computer monitor, and its 1280x768 RGB pixels can easily accept a computer's native 1024x768 ouput in either 4:3 aspect or in a "stretched" 16:9 full screen, wide XGA mode. As you can see on the graph of brightness versus APL, the big Pioneer makes the least amount of light for the most amount of power•because it is so big. However, the Pioneer's big size also makes up a little for its dimness; it doesn't look that dim when compared side by side, because it is so much larger•over 43%•than the others.

As shown in the contrast ratio graph, the Pioneer 50-inch plasma had decent, but relatively low, contrast ratio. This, however, was a huge improvement over earlier versions of Pioneer's 50-inch product. Pioneer has worked hard to improve both the brightness and the contrast in its 50-inch plasma over earlier units, while it has also brought down the cost to "reasonable" levels•all of which makes its 50-inch model one of the best sellers in the U.S. market.

Despite having slightly less brightness overall, the Pioneer's performance was very good. The big plasma produced realistic-looking images without appearing to be missing too much. When compared side by side with the same DVD- Video input, the Pioneer produced the cleanest, most noise-free, natural-looking image, by far. I thought that the Pioneer's images had a little too much red in them, and that was due to a slightly offset color balance that could not be adjusted. If Pioneer would just improve its user interface and allow people to make a few color adjustments (besides the hated tint slider), its product would be even better. In comparison, the Panasonic units produced colors that were maybe a little more true to life than the Pioneer's colors, which had no additional filtering, but the Pioneer's bigger, cleaner image gets the best-for-video nod from me.

Panasonic

Panasonic makes 42-inch plasmas, as well as a 50-inch, 1366x768 RGB pixel that is still probably not yet shipping. The company's semi-prototype 50-inch unit is called the 50-PHD3. Panasonic has also made several of its 60-inch prototypes that are used at shows to make us techno-geeks squirm with desire. But alas, I've heard no responsible person mention production. Panasonic's latest, actually-shipping 42-inch unit, the TH-42PWD3, sells for $9995. I have had hands-on time with several of Panasonic's 42-inch products. The latest one, the D3, demonstrates a big improvement contrast ratio, although the previous version, the D2, was no slouch. Some of this improvement came from the researchers at Plasmaco, who devised new methods to turn down the "dark current" or off-state energy in the Panasonic plasma's pixels. The result is quite obvious, but not quite up to the over-hyped, 3000-to-1 contrast ratio. However you want to measure or describe them, the black levels on the Panasonic plasma are really black, unlike some units (the Pioneer, in particular) where the black levels never get darker than a light gray.

Panasonic doesn't use any color filters to "purify" its already-colored plasma emission colors. And where the other manufacturers have designed RGB pixels with the same size for every color, Panasonic has developed pixels with different sizes and areas. The blue pixels in a Panasonic plasma are significantly wider than the red (which are the smallest) and green pixels; these adjustments reflect the manufacturer's attempt to achieve a more color-balanced brightness. This work results in the best-looking colors, which are actually a lot closer to the SMPTE-C color standard used in production video than any other plasma I've measured.

However, the Panasonic's colors were still not perfect. The reds and blues looked a little weak, which probably had more to do with the algorithm used to map and adjust color and gamma correction, than to the overall color output. On top of that, the Panasonic's white point, while a little better than the NEC's, lost out to the big Pioneer, which was like its video circuits, the best of the bunch.

NEC

NEC makes both 42-inch and 50-inch plasmas. In the past, it also made some smaller units (a 33-inch 4:3 aspect unit), but those never seemed to sell much in the U.S. NEC's 42-inch model, the PlasmaSync 42MP2, has a $9995 MSRP, and, as with all the other units, I'll bet that a little searching on the Internet will turn up a lower street price. Its low-end 50-inch model, the PlasmaSync 50PD1, sports the same VGA (853x480 resolution) as the smaller model, but with slightly bigger pixels. The 50-inch model retails for $12,995. I'm not sure what benefit there is to the 50-inch VGA model. The 42-inch version seemed to do a better job with image quality, while the 50-inch model seemed to be down on brightness and contrast. However, some of my complaints have supposedly been addressed in NEC's latest XGA, 1365x768 RGB pixel resolution 50-incher, the PlasmaSync 50MP1, which lists for $18,995. But I've only had hands-off demos with the 50MP1, so I can't vouch for its image quality.

Like that of other plasmas, NEC's unique design produces light by exciting a gas to produce UV radiation that further excites a colored light-producing phosphor. However, instead of relying completely upon the colored light produced by excited phosphors, NEC has placed tiny color filters in the cover glass over each pixel in its plasma. These color filters are similar to those used in large LCD monitors, but they operate a little differently. Where the color filters in an LCD monitor completely change the backlight's white light into colored light, the color filters in the NEC plasma just "purify" the already-colored light emissions.

Like the Hitachi 42-inch, the NEC plasma had a large amount of color saturation, but the NEC's colors were not that well balanced. Because of the color filters, the NEC's red colors were about right, but its blues were a little weak, and the greens were way, way off. I've measured error in the green color versus the SMPTE-C color standard for all of the plasma screens I've played with, and the NEC plasma is clearly the worst. I found that the NEC's greens were over-saturated, while the white point was a little farther away from true white than it was in most of the other units. This year's color saturation is a little lower than what I found in last year's plasma. Perhaps that change was done in an attempt to produce more balanced colors, but they're still not there. Some of the mismatched colors can be adjusted away using the tint control, but that takes a lot of time and patience•things I have in short supply.

Along with too much green, when compared side by side to a good CRT like the 36-inch Sony XBR Wega, the NEC's reds looked orange and the blues took on a grayish cast. Perhaps if NEC would let its power supplies crank up like Hitachi's and then cut back on the green some more, while allowing more red and blue through…well, I suppose they've already had that discussion back in the lab, and opted not to burn up more watts. But with better color balance, I think that the NEC plasma could compete quite well with any large-screen system•at least in evenness, if not in color performance. I also liked the performance of the NEC plasma with a computer data input in VGA resolution. But like the other VGA units, when the computer input was changed to SVGA resolution, the clear VGA image changed into a slightly noisier, chirpier-looking image. And speaking of noisy images, when the NEC plasma was compared side by side with the Panasonic and Pioneer plasmas•using the same DVD video signal•the NEC's images appeared to be noisier and less clear than the others.

ViewSonic

ViewSonic, a PC monitor powerhouse, recently introduced a 50-inch plasma screen, the VPW500. Its 1280x768 pixel resolution is the same as Pioneer's PDP 502, but ViewSonic doesn't really make its own products from scratch. Instead, it specifies, procures, and then puts its label on products OEM'd from other manufacturers. When I questioned ViewSonic about the source of its 50-inch product, company sources pointed to Pioneer, whose product I've evaluated at least twice before and liked. At any rate, ViewSonic should do well, offering people a good deal, since officials are telling me that the "estimated street price" of its version of 50 inches of plasma delight will be only $12,999.

Flat Screens of the Future

I'd like to see better, cheaper plasma displays in the future. And as I've said before, now more than ever, plasma needs to go on an energy consumption diet. I wish that they'd get down to around 35 watts per 100 square inches of screen size, and do that while still making 200 nits with a lot of picture content. That result would be a significant improvement over the current state of the art•and probably a ways off as yet•but I can wish, can't I? And while I'm wishing, I want better colors too. Forget about those weird green images, please. We went through that stage with LCD projectors, and we've come out the other side with mature products. Most LCD projectors today can make nice, clean, realistic-looking colors, and I expect that the one-chip DLP units will not be far behind. No one will want to use a plasma screen as a computer monitor if it can't make decent-looking WYSIWYG colors, especially with the latest move to Microsoft's new sRGB color standard.

The latest term to throw into the mix: sRGB is a new color standard based upon the latest HDTV ITU709 and SMPTE-240 color standards. It will allow all devices, monitors, scanners, printers, and projectors to produce the same colors. It's high time we had standards, because for too long there has been no color standard for computer peripherals. What this means is that if a device has a hard time doing the correct SMPTE-C video color (like plasmas), than it will now have a hard time doing the SMPTE-240 colors with computer inputs. Major players like Mitsubishi and NEC (for LCD and CRT monitors and projectors) and Epson (for scanners, printers, and projectors) have joined with Microsoft in developing sRGB-compliant devices. About time, indeed.

And, before I forget, there is more competition for plasma in the hang-on-the-wall, giant display screen category. The Rainbow company from New York•along with a lot of help from Philips and its Japanese partner HAPD (Hoshiden and Philips Displays)•has developed a method of carefully and seamlessly assembling three 21.4-inch LCDs into a 37.5-inch, 16:9 aspect ratio display with a resolution of 852x480. I was quite impressed with the image quality, even though, at first look, I mistook its giant LCD for a plasma! The Sharp Corporation showed a similar prototype 40-inch LCD structure several years back; however, the Rainbow assembly looked a little better with its great video response. It is even slated for production in the spring of 2001 at a $10,000 price point•allegedly competitive with plasma, though I think that Rainbow had better figure out how to get more cost out of it if they really want to be competitive. Company officials recently told me that Rainbow intends to get a premium over plasma pricing since its units have all the benefits of a LCD monitor (no dimming with increased picture content) and twice the brightness of a typical plasma screen. What's more, they consume less power: about 300 watts, which is much closer to my desired 35 watts per 100 square inches.

If that isn't enough to instill some healthy fear in plasma vendors' hearts, then watch out for a Canadian company called iFire. It has been quietly working away up in the frozen north of Toronto, Ontario to produce an entirely new type of large-screen, hang-on-the-wall display for HDTV and computer monitor applications. Even though iFire doesn't intend to dominate plasma in terms of lower power consumption, it does want to squash prices, while producing much more realistic-looking colors.

The key to iFire's technology is something called "thick-film" inorganic, electro-illuminesent (EL) display technology. Multi-color EL displays are kind of the holy grail for us techno-types. Those who have taken their laptops apart know about the EL backlights used to light up the early LCDs, but those were white lights, not multi-colored. Everyone has wanted to make sturdy, long-lasting, low-power, full-color EL displays for some time, but still no luck.

However, iFire has been working hard in conjunction with some major Japanese companies, such as TDK, to develop a "patterned" full-color, fast-switching, passive, matrix-driven EL display suitable for use as an HDTV, 42-inch display. The iFire technology consists of a two-color phosphor pixel pattern overlaid with a three-color RGB color filter like that used on LCD monitors and on NEC's plasmas. But since iFire's displays use cheaper substrates (one piece of glass instead of two) and simpler construction methods that don't require long bake cycles like those needed to form the plasma's pixel's ribs, an iFire display is supposed to be 40% cheaper than a correspondingly-sized plasma. Watch out, plasma.

All of this good stuff really starts in a couple of years, around 2003. By that time, whatever the cost or price is for a CRT-equipped display (like a rear-screen HDTV system), plasma is supposed to be two and a half to three times more expensive. Assume $3000 to $4000 for a decent HDTV around that time, which means a 42-inch plasma should cost about $7500 to $9000•just a little cheaper than what they are today, in terms of MSRP. Some people think that plasma will be even cheaper then, but, regardless of where plasma is, according to the optimistic folks at iFire and their Japanese partners, iFire displays will be even cheaper at no more than double the cost of a CRT-based set. That puts a decent-sized iFire at around $6000 MSRP initially, with the hope of reducing that cost to less than one and a half times the cost of a CRT-based system. If all that comes to pass, then by 2005 or so, we should be able to choose between large LCDs with great color and good brightness, large plasmas with more brightness and less energy consumption, and large EL from iFire•with the EL devices costing close to $100 per inch. That should drive all prices down to a level where more people can afford these exciting and glamorous displays. I can hardly wait.


Plasma Behind The Screen

So what's at the heart of these display devices that are amassing considerable cachet at the video-centric high end of the computer monitor and presentation universe? What justifies the often significant price differential between plasma displays and their comparably inexpensive CRT and LCD cousins? And where does all that light and color-production acuity come from?

A plasma display is made of a glass sandwich that traps a gas (usually a mixture of inert, rare gases like xenon and its cousins) in a partial vacuum surrounded by phosphor coatings and transparent electrodes. This glass sandwich also traps that gas in little pockets formed by etched ribs at each tiny pixel. The plasma display's thin glass panels are bonded around the edges and at the boundary of each pixel by the thin ribs that define each pixel's area. All those ribs and bonds allow the glass plates to be much thinner and lighter than the glass used in a CRT. But, since the plasma display is so wide and large (most popular sizes begin at 42 inches diagonal), it must be very well supported to prevent any blending, flexing, and subsequent cracking of that fragile glass sandwich. So, besides the weight of the glass, the much needed support frame adds a lot to the total weight.

CRTs are heavier than plasmas, because they support a lot more of the earth's atmosphere. For example, the Sony Wega's front faceplate is much thicker than a plasma's glass plate. This is because the CRT's faceplate, held only at the edges, remains unsupported across the rest of its area. Plasma, by contrast, uses the edges of each little pixel for support. And since a CRT contains a good vacuum (unlike a plasma's gas), the normal atmospheric pressure of 14.7 pounds per square inch adds up to several tons on a 36-inch display's faceplate with an area of about 612 square inches. All of that atmospheric pressure demands that the CRT's glass be several inches thick and, therefore, heavy. In spite of a 42-inch plasma's larger screen area (about 739 square inches), any atmospheric pressure is only applied to each individual pixel's 0.0016 square inch area.

Plasma versus lcd

Like a Liquid Crystal Display (LCD), a plasma display consists of an array of pixels, but that and their glass sandwich construction are about the only similarities between the two. An LCD does not emit light. Instead, LCD displays rely upon an array of fluorescent lamps arranged behind the LCD to shine through the pixels and to make a glowing image. A plasma display does not have a backlight•it emits it own light. In fact each pixel in a plasma display is like a tiny, separately controlled, fluorescent lamp.

In essence, a plasma display is nothing but a big desk lamp, since they both (plasma and fluorescent lamps) produce light by electrically exciting a gas surrounded by light-emitting phosphors. The excited gas emits UV radiation that then excites the phosphors to produce a visible image. A fluorescent lamp uses a variety of phosphors to produce a balanced white light containing all colors, whereas a plasma has split its colored phosphors into separately controlled red, green, and blue pixels, thus allowing a full-color image. The excited gas emits UV radiation that then excites the phosphors to produce a visible image. A fluorescent lamp uses a variety of phosphors to produce a balanced white light containing all colors, whereas a plasma splits its colored phosphors as explained above.

Plasma versus cRt

A TV set's cathode ray tube (CRT) uses phosphors too, but different kinds that use a different physical phenomenal. The image on a CRT-TV's screen is created by an electron beam that bombards and directly excites the color-striped phosphor "face-plate" on the CRT, causing it to glow with an image. The "electrons" in the plasma excite a mixture of rare gases to emit light in the far ultraviolet (UV) 150nm to 170nm range, which in turn excites low energy, colored phosphors with very different visible color characteristics as compared to a CRT's phosphors.

In addition, some plasmas•products from NEC, in particular•use colored filters on top of the glass sandwich to further "purify" the colored light being emitted by the plasma's phosphor.


Power Hungry Plasma

Experiencing plasma is not unlike the story of the blind men and the elephant. One blind man tries to pick up the plasma and discovers that it's very big and heavy. Another blind man feels the plasma's edges and finds out that it is quite thin and glassy. A third blind man touches the plasma while it's operating and discovers that it's hot. Which seems to blow the frequent comparison of a plasma screen's light production to a flourescent lamps•fluorescent lamps are supposed to be cool, right? Maybe they are cool if you only have one, but if you tie a whole bunch of them together it's going to eventually get hot. That's because plasmas use a lot of power to make their relatively puny amount of light.

The low-energy, UV-conversion phosphors used in plasma displays are a lot less efficient in converting electrical energy to light than the high-energy, electron-pumped phosphors used in CRTs. For example, a large direct-view CRT, like Hitachi's 36-inch HDTV set, requires about 185 watts of power to make an image that can exceed 300 nits in small areas. (A nit, by the way, is a measure of luminance, the physical measure of brightness.) However, the typical 42-inch plasma makes only about 200 nits in the same size picture area, yet requires closer to 400 watts of electrical power to do that. Plasmas make less light with more energy and, as you can imagine, it doesn't get any better with size; larger plasmas need even more power, at least 10 to 20 watts per diagonal inch in my estimate. This means that some 50-inch plasmas can gobble up over 500 watts of power. A big 60-inch plasma could exceed the 10 amp power limit for a standard 120 volt wall plug.

LCD monitors also need power, and they essentially use the same fluorescent light principal that makes a plasma glow. For example, a 21-inch LCD monitor has about 218 square inches of active screen area, and uses about a maximum of about 79 watts of fluorescent power to make 189 nits through the LCD. A 42-inch plasma with a 16:9 aspect ratio has about 3.4 times the screen area of a 21-inch, 5:4 aspect ratio SXGA LCD monitor•so if you multiply 79 watts by 3.4 you get about 269 watts for approximately the same level of brightness. But 269 watts is less than the 400 or so watts that some of the brightest plasmas consume, and there are some LCD monitors that use even less power. So why are plasmas so inefficient? There are several possible reasons•one is that the colored phosphors in a plasma are optimized for bright colors, rather than for an LCD backlight's white light, and that takes more power. But another reason is that a lot of the light in a plasma is lost. A plasma's pixel actually glows all around in all directions, but only the light coming out through the top layer of glass can be seen and utilized. The fluorescent lamps used in an LCD monitor's backlight are surrounded by efficient reflectors that attempt to use as much of the light as possible.

the truth and the light

I haven't been completely honest, but I'm going to come clean. An LCD monitor has essentially the same brightness no matter how much of its area is being "lit" up. If you make a tiny square in the middle of the screen white and leave the rest black, that little square can be 200 nits•however, that screen brightness doesn't change as you increase the size of the white square and decrease the size of the black surround. In fact, you can crank up the size of the white square all the way to the maximum screen size and still have the same 200 nits•ignoring any across-screen irregularities. However, with a plasma, if you run the same scenario and start with a small white square (approximately 10% of the available surface) surrounded by "turned-off" black screen, you can get about the same 200 nits as found in an LCD monitor. The first graph shows the relationship between average picture level (how much of the screen is being driven) and ANSI-across the screen, nine-point averaged nits. But once you increase the size of the "turned-on" white area (sometimes called the picture level), the plasma's brightness quickly drops. If you light up the entire 42-inch diagonal area of a 42-inch plasma, then the maximum surface brightness is only about 60 nits, and even less on a 50-inch plasma.

The second graph shows the relationship between the consumed power in watts and the average picture level in percent for the same recently-tested plasma displays. This graph uses the same x-axis order as the first graph for the same plasmas, and it is clear that the consumed power increases dramatically from the black, off-state to about the 33% average picture level (APL) condition. After that, further increases in APL do not result in an increase in power consumption, which indicates where the power is limited•while the brightness keeps on decreasing due to the increased APL. This graph shows that those 60 nits at the full screen 100% APL take the maximum power, close to 400 watts in some cases•a big increase from that required to make the maximum 10% APL brightness. That's a lot of power for only a little light. I once asked a designer if a plasma, like a CRT, needed some kind of power limitation to reduce the wear on its electron-generating parts.

"No," he said. "We could run a plasma at its maximum brightness level at 100% APL•if we wanted to break the glass." Turns out there is no good way to keep the plasma glass cool enough if the plasma is generating a lot of watts•more than 10 or so per diagonal inch. Therefore, the companies that make plasmas limit the power when the APL goes up to keep from overheating the glass and causing permanent damage, but this reduces their plasmas' brightness. In addition, plasma displays suffer from something called phosphor burn-in. Everyone knows about phosphor burn-in•that's why people use those cute screen savers. But plasmas are a lot more susceptible to this problem than CRTs. I've heard of plasmas being burned after only being left on overnight.

Plasma developers have created several automatic algorithms to reduce the chance of burn-in. Some plasmas, like the early Fujitsu units, automatically reduced their brightness levels after only a few seconds of max power, while others (such as NEC's current models) start to move the image slightly to reduce the phosphor burn. NEC also offers its customers the option of running the image in the "inverse" mode. For example, if you have a plasma art gallery, during the day you can show the true image and, at night, while no one is watching, you show the inverse image, which results in an "even" burn without any perceptible permanent patterns.

Another bit of truth to bring to light: plasma manufacturers have been guilty in the past of pretty gross exaggerations about their products' brightness. I've seen claims for 400 or 500 peak or maximum nits (sometimes they use the term candelas per square meter, which is the same as nits). I've never been able to measure more than 200 or so nits even with 10% APL, and so I discussed this little topic with the folks who wrote those wild specs. The guilty plasma manufacturers claimed that their brightness specs came from the basic max capability of the plasma component itself when a certain anti-reflection, anti-glare, and anti-EMI (Electro-Magnetic Interference) film is not present. That film layer on every single plasma screen that leaves the factory cuts the light output in half•thus accounting for the discrepancy. Bottom line, if somebody tells you a big number for brightness, believe only about 10% of that number with a 100% white image across the whole screen. Plasma manufacturers do readily admit to their products' excessive use of energy, however, they feel that the convenience of having thin, light, hang-on-the-wall displays offsets that waste. Plasma manufacturers are not going to combat their products' greedy power consumption at first, either. They want to concentrate on making cheaper display units instead, hoping to reach the point where $100 can buy 1 inch of diagonal plasma display size. They think that if 42-inch, 400-watt plasmas can be priced at something around $4000, instead of their current MSRP of close to $10,000, then a lot more plasmas can be sold.

Researchers hope to be able to develop more energy efficient phosphors•they want to develop substances that will emit twice as much light for a given UV energy input. Once that happens, they want to turn down the plasma power consumption to reach a level closer to a standard CRT's need to feed, and still be able to provide cheap displays with the same level of brightness as today's power-greedy systems. I sincerely hope they can accomplish that goal. I like the thin size of a plasma display, but I don't like its heavy weight, high cost, and certainly not its excessive power consumption.


Companies Mentioned in This Article

Fujitsu Computer Products of America
2904 Orchard Parkway, San Jose, CA 95134; 800/626-4686; Fax 408/594-3606; http://www.fcpa.com

Hitachi America, Ltd.
2000 Sierra Point Parkway, Brisbane, CA 94005-1835; 650/589-8300; Fax 650/583-4207; http://www.hitachi.com

iFire Technology, Inc.
15 City View Drive, Toronto, Ontario, Canada M9W 5A5; 416/246-1030; Fax 416/246-0458

LG Electronics
1000 Sylvan Avenue; Englewood Cliffs, NJ 07632; 201/816-2000; Fax 201/816-0636; http://www.lgeus.com

NEC Technologies, Inc.
1250 N. Arlington Heights Road, Itasca, IL 60143; 800/632-4636, 630/467-4567; Fax 630/467-4558; http://www.nectech.com

Panasonic Consumer Electronics Co.
One Panasonic Way, Secaucus, NJ 07094; 201/348-7000; Fax 201/348-7016; http://www.panasonic.com

Philips Consumer Electronics
64 Perimeter Center East, Atlanta, GA 30346; 770/821-3402; Fax 770/821-3876; http://www.philips.com

Pioneer New Media Technologies, Industrial Display Division
2265 E. 220th Street, Long Beach, CA 90810; 800/926-4329, 310/952- 2111; Fax 310/952-2990; http://www.pioneerusa.com

Plasmaco
180 South Street, Highland, NY 12528; 845/883-6800; http://www.plasmaco.com

Samsung America, Inc.
14251 East Firestone Boulevard, Suite 201, La Mirada, CA 90638; 562/483-7255; Fax 562-802-3011; http://www.samsung.com

Sony Electronics, Inc.
3300 Zanker Road, San Jose, CA 95134; 408/955-5068; Fax 408/955-5340; http://www.sel.sony.com

ViewSonic Corp.
381 Brea Canyon Road, Walnut, CA 91789; 800/888-8583, 909/869-7676; Fax 909/468-3756; http://www.viewsonic.com



WK Bohannon (manxrsrch@aol.com), founder of Manx Research, an independent evaluator of projection and display systems, has more than 25 years of experience in high-tech industries in areas from nuclear spectroscopy to high-energy laser systems and artificial intelligence. He worked as chief scientist for Display Products at Proxima Corporation from 1989 to 1994, and lived through the birth of today's small electronic presentations systems.

Comments? Email us at letters@onlineinc.com.


Copyright 2000-2001 Online, Inc.
213 Danbury Road, Wilton, Connecticut 06897-4007
203/761-1466, 800/248-8466
Fax 203/761-1444
info@onlineinc.com