The Dunkirk, Md., company, which holds an exclusive license on technology developed at Brookhaven National Laboratory, will give its first public demo of a 50-inch-diagonal prototype display at the DisplaySearch FPD Conference & High Resolution Symposium in Austin, Texas. Samsung (Seoul, South Korea) will demonstrate a 50-inch display based on the scheme at the CeBIT trade show, which convenes March 22 in Hannover, Germany.

"The contrast, color saturation and viewing angles were better than any other TV image I've seen on a large screen," said DisplaySearch analyst Mark Fihn, who viewed an early prototype late last year. "The SCRAM screen has the potential to be a breakthrough innovation."

The Austin event also promises to showcase the headway developers of electronic projectors have made in recent years in reducing size, weight and price while increasing brightness and pixel count. Among the exhibitors looking to further that trend are Analog Devices Inc. (Norwood, Mass.), which will take the wraps off a chip that it says will lead to smaller and cheaper projectors.

Product manager Ed Spence said ADI worked closely for about four years with a polysilicon-LCD and projector maker, which he declined to name, to develop a specialized driver chip that incorporates a digital-to-analog converter to address one difficulty posed by what is perhaps polysilicon's biggest virtue: integrated drivers.

Designated the AD8380, it "dramatically reduces the power, area and cost of driving small LCD displays," claimed Spence.

Rear-projection screens are the focus of 1998 startup SCRAM Technologies, which spent the past year and a half "refining the technology" and establishing a patent portfolio, according to chief executive officer Ray Kwong.

"We've been reluctant to fully disclose our lens and screen design until the intellectual-property situation was more solid," he said.

Kwong said the company now has "eight patents issued, probably 12 pending and 10 to 12 more being drafted." While it was still in stealth mode, he said, SCRAM demonstrated the technology to a number of OEMs and managed to quietly raise more than $7 million in two rounds of financing. "We'll very shortly finish the Series C round, which will be the last," said Kwong.

The SCRAM technology combines a lens that compresses a projected image and a screen that decompresses it, achieving what Kwong called "the holy grail of displays: a fairly thin form factor with extraordinary contrast."

"The screen consists of many layers of optical waveguides," Kwong explained. "The core material of the layers has a particular index of refraction, and the coating has a slightly lower index, so we achieve total internal reflection. That's significantly different from other screens, which have to bulldoze an image through with more light. We capture as much light at the output as we have at the input and achieve very high contrast."

Conventional projection systems expand the image from a small display "like a pyramid, and they need a lot of space inside the box to be able to do that," said Kwong. "But we compress the image in one plane, so what we're throwing around in our box is a ribbon of light."

That ribbon isn't expanded within the projection engine but passively by the screen itself. "Because we're passive in nature," Kwong said, "we can accept any resolution imager an OEM has and whatever his performance criteria are, and there won't be any significant degradation as we go from imager to screen."

The result is a sunlight-readable display within a relatively thin form factor. "We don't get reflections, because ambient light is absorbed between our layers," said Kwong.

SCRAM's 50-inch prototype is only about 9 inches deep, according to Kwong. "It's not paper-thin, but it's thin enough for people who want to be able to mount a display on a wall. Our 36-inch design is probably 6 inches deep."

The company's scheme will also contribute to reducing the costs of projection systems, according to Kwong. The screen is made of multilayer stacked plastic, he said, and the lens is injection-molded plastic. "The cost of light engines is coming down dramatically, and when they get down to several hundred dollars, we'll be talking about a bill of materials of below $1,000 for a 50-inch, high-definition, high-performance display."

Meanwhile, David Slobodin, chief technology officer at projector market leader InFocus Corp. (Wilsonville, Ore.), cited six "key technology enablers for further projector miniaturization and brightness enhancement." They include new thermal-management techniques; smaller microdisplays; shorter arc lamps, lamp assemblies and lamp drivers; smaller, lighter and more efficient optical architectures; smaller power supplies; and higher levels of electronics integration, he said.

The Analog Devices chip to be introduced at the DisplaySearch conference addresses several of those areas, specifically for projection systems based on polysilicon active-matrix (AM) LCDs and liquid-crystal-on-silicon (LCOS) displays. Polysilicon's claim to fame over conventional amorphous-silicon AM LCDs is higher electron mobility, which enables drivers and other circuitry to be integrated into the display panel itself, along with the usual active matrix of thin-film transistors and capacitors. Integrating the drivers sharply cuts down on the interface and interconnect needed to integrate displays.

"Instead of one data drive per [display] column, you can mux or switch in several higher-speed columns and bring them out to a reduced fanout," said Spence. "A state-of-the-art poly LCD for a projector, for example, uses six analog channels for SVGA input, vs. 800 or so for amorphous" LCDs.

But reducing the number of inputs to the display also has its challenges. "Not only do you have to provide a high-voltage dynamic range but you must operate at a considerably higher speed, which means driving a very low settling time into capacitive loads," Spence said. ADI and its anonymous customer worked together "to redefine the partitioning of the system to come up with a sensible solution," he said.

But Spence cited another problem: "When you're only driving so many fewer columns, you have to accurately match the columns. A mismatch in the signal accuracy driven in shows up as column color-saturation mismatch; you can see streaked lines, which show up pretty readily on a projector."

Previously, Spence explained, the customer had been using a sample-and-hold architecture to manage its polysilicon panels, "but this has several problems. First of all, it's inherently inaccurate. They were using CMOS D/A converters to feed the sample-and-holds — sort of an analog serial-to-parallel conversion, if you will — sampling things in sequence at higher speeds and presenting the panel with a parallel set of imaging signals.

"What we did was remix the integration. We absorbed the D/As at high speed, and because they're on the same silicon as the amplifier, we're able to laser-trim," Spence said. "So not only can we present a fast, high-dynamic-range output, but it's laser-trimmed-accurate, so matching is absolute.

"You can stack up a couple of 8380s in a 12-channel XGA system, for example, and get perfect color uniformity. That's the real breakthrough."

The architecture also suits LCOS projectors, Spence said. "Poly and LCOS both need the ability to integrate some switching functions on board the panel."

But two other consequences may be of just as much interest to projector makers: "The footprint on our solution went down dramatically, as [did] the power dissipation," he said.

Slobodin of InFocus noted that even the smallest projectors today have "fairly large circuit boards," and the main barrier to higher brightness and lower size and weight for projectors is thermal management.

"To date, we've been using conventional means, like fans and trying not to make too much noise," he said. "As we enter the next shrink [in ultraportable projectors], it will take some extra creativity and thought to get the heat out of the smaller volume."

'Dramatic reduction'

Spence predicted "a dramatic reduction in size" for coming polysilicon LCOS projectors — a shrink that will let LCOS play catch-up with projectors based on Texas Instruments Inc.'s digital micromirror devices.

"Drive electronics are a big part [of size and weight] because the engine and panel sizes are harder to change," he said.

There are, of course, a number of other system components beyond the electronics that affect the size, weight, price and imaging quality of a projector or rear-projection system.