Angles Of View

Arlie Conner is Vice President, Technology at Lightware, a manufacturer of LCD projection devices in Beaverton, Oregon. Mr. Conner, who has been Research Director for In Focus Systems, has worked for the last ten years on the technical design and development of Liquid Crystal Display systems. He can be reached at arliec@lightware.com and is interviewed here on the subject of

Monitoring the Future - the LCD

Da-Lite: Liquid Crystal Displays have become an extremely well established technology. But they weren't always so common. What originally prompted their wide acceptance?

Conner: Do you remember the earliest laptop computers? Initially many of them were using things like plasma or electroluminescent displays for their screens. And, until someone thought of LCDs, there was no simple solution to portable computer screens. However, principally because their power requirements are an order of magnitude lower than the alternative systems, LCDs rapidly became dominant.

Da-Lite: And how did they then migrate into projection display devices?

Conner: The really kind of fun breakthrough, the aha!, happened when someone first stuck a liquid crystal display onto an overhead projector. Once that had happened, it was suddenly group sharing of a big screen version of a computer monitor. And pretty instantaneously five or six companies sprang up to build these LCD panels or platens, as they were sometimes called.

Of course overhead projector manufacturers were delighted and it didn't take them very long to develop projectors with the significantly increased brightness which the LCDs demand. How bright, for instance, is your company's new overhead?

Da-Lite: Our Model 5000, as its name suggests, is conservatively rated at five-thousand lumens.

Conner: In addition to the panels for overhead projectors, those of us in the LCD world also recognized that there were certain benefits to be had if one could make an integrated system that wrapped the OHP and the LCD into one small package that could be fitted, as it were, into the overhead compartment on an airplane.

Da-Lite: Were the manufacturers of the LCDs cooperative?

Conner: They did take a little bit of convincing because they knew and still know today that the main market for their products is portable computers. But once aspects of that market began to call for smaller format LCDs (to go into screens for sub-notebooks, for instance), it wasn't so difficult to get smaller LCDs for inclusion in integrated projectors. A 10-inch LCD is what you may need for an OHP panel, but what you need for an all-in-one projector can be much, much smaller.

Da-Lite: OK; but regardless of its size, just how does a layer of liquid crystal become a functional projection device?

Conner: The LCD with which we are probably most familiar is called TN, for Twisted Nematic. That refers to the crystalline phase of the material. The term nematic means that the molecules of the crystal are long and rodlike and generally at room temperature are loosely organized such that, like a school of fish, they're all parallel and heading in the same direction.

Da-Lite: What about the "twisted" part?

Conner: If you sandwich a cell of TN liquid crystal material between cross polarizers, top and bottom, the alignment of the molecules at the bottom might be, say, North-South, but as you move up the column towards the top polarizer the intervening layers of molecules twist smoothly, layer by layer, so that, by the time you get to the top their alignment is East-West.

Da-Lite: So the twisting transition takes place in a plane perpendicular to the orientation of the molecules?

Conner: Think about it as a column of liquid crystal. At the top, everything points East-West. At the bottom everything points North-South. And in between there's a 90º gentle, spiral staircase. When polarized light is passed by the first polarizer, its orientation may be said to be North-South. And as it passes though the liquid crystal layers that polarization is rotated such that, like the molecules themselves, it has become East-West by the time it reaches the second polarizer which is more accurately termed the "analyzer."

Notice that even though twisting the molecules are still lying flat to the glass above and below them. But when you apply an electric field across the gap between the two glass polarizers, the molecules of the liquid crystal now want to tilt up which means that all of them are suddenly more or less perpendicular to where they had previously been pointing.

Da-Lite: What does that accomplish?

Conner: What it means is that the spiral staircase vanishes when the molecules tilt up and therefore you have a polarizing "valve" proportional to the applied electric field.

So if our model says that the first polarizer is going to pass light of only a North-South orientation and that light no longer encounters a spiral staircase to twist its polarization by 90º, when it reaches the analyzer (the second polarizer) which passes light of only an East-West orientation it can't get through. And you get black.

Da-Lite: If you vary the intensity of the electric field, then will the degree of tilt assumed by the molecules also vary?

Conner: Exactly. And that's how LCDs can produce grey scales of such surprisingly good gradation and uniformity. To produce grey scale requires good control of the cell gap (the spacing between the glass plates which normally is six to ten microns) and a proper understanding of the driver design.

Da-Lite: The drivers are what establish and control the pixels, aren't they?

Conner: Yes; and I'd like to point out that the fundamental benefit of an X/Y matrix like that which drives an LCD is that if you have 640 times 480 pixels, you have an awful lot of pixels. But we only have 640 plus 480 column and row drivers. So a flat panel LCD is really quite an economical design.

Da-Lite: Is that true whether or not the display is an active or a passive matrix?

Conner: Yes; although the performance characteristics of the active matrix are considerably better.

Da-Lite: In what way?

Conner: The contrast ratio of an active matrix is typically 100:1, whereas passive matrices might top out at 20:1.

Da-Lite: Why is that difference so large?

Conner: It results chiefly from the driving method. Passive matrices have a limitation on the voltage that you can have between an on and off pixel. Generally it's about six per cent. Active matrices are not similarly limited.

Da-Lite: Do most LCD projection devices now contain an active matrix display?

Conner: Yes; three of them in fact - one for each color, red, green, and blue. All high performance LCD projectors will use dichroic mirrors to separate the white light from the lamp into its constituent RGB wavelengths which are then passed through three identical black-and-white light valves before being recombined through the lensing system. The quality and saturation of the resultant colors are much superior to a single, filtered LCD.

Da-Lite: How efficient is an LCD projector? How much of their light source's initial brightness gets lost?

Conner: About 95%. Actually, it's worse than that. Almost none of the projectors is better than about 2% efficient.

Da-Lite: And what about resolution?

Conner: Well, perhaps I'm oversimplifying, but I think resolution depends on the state of the technology at any given time. Of course everyone would like more pixels but to get them you pay a cost in the complexity of their manufacture. Getting more dots onto the same size panel is always a challenge and never easy.

Da-Lite: Why can't you just use larger panels to increase your pixel count? That would increase their number without decreasing their size.

Conner: In fact that's what is going to be done. Although clever designers have been able to increase the number of dots from VGA to SVGA without enlarging the 1.3-inch diagonal panel size, to get to the larger number of dots implied by XGA or Workstation resolution they've had generally to go to the 1.8 or 2.4-inch diagonals.

However, because the manufacturing lines for these panels are set up to use 6-inch diagonal quartz substrates, the number of images you can get on a single substrate decreases sharply as their individual size is increased. Hence the unit cost goes up quickly and badly.

Da-Lite: And what do you see as the future for LCD projectors?

Conner: I'd say that LCDs are something whose time has not even fully come yet. Even as we speak the developmental improvements are staggering. Maybe it's not the same as the semi-conductor paradigm (where theoretically processing speed and memory capacity double while costs halve every six months), but LCDs have their own growth curve and their performance and cost have both been respectively improved and decreased radically over the past five years. It's hard to imagine where it's going next. But there's no clear end in sight.