An automotive technician uses wiring schematics to help locate an electrical short. The drawings, however, aren’t printed on paper. Instead, he views digital renderings with a portable, head-worn display. The display screen is less than 0.5-in. square but the reflected virtual images appear to come from a 15-in., SVGA color monitor held at arm’s length. Resolution is a crisp 600 3 800 lines/in., the same quality as most desktop CRTs.
Tiny displays like this one may be the next development that extends computer technology further into consumer appliances and numerous kinds of portable equipment. Similar microdisplays may become standard issue in next-generation smart cell phones and other personal electronic devices such as DVD players and GPS receivers. What makes these tiny displays possible is the marriage of liquid crystals and microelectronics.
Although more complex, microdisplays operate much the same as simple, numerical LCDs in watches or calculators. The most common of these are called twisted-nematic (TN) displays. Nematics are rodlike molecules approximately 20 to 30 angstroms long and about 5 angstroms in diameter.
TN-microdisplays consist of nematic liquid-crystal material sandwiched between two flat panels spaced a few microns apart. Panel surfaces that contact the crystal material are treated with a thin polyimide coating. This coating is mechanically scrubbed to form tiny grooves. The crystal molecules at this interface tend to align parallel with the grooves. When the front surface has grooves pointing north/south, the backplane’s grooves may run east/west. Intermolecular bonds tend to keep the crystal molecules stacked flat on top of one another. However, the crossed grooves anchor the ends and help twist the stacked molecules into a helix shape, hence the display’s name.
Microdisplays come in two basic varieties, transmissive and reflective. Transmissive types contain two transparent panels and are illuminated through the backside. These displays may have groove-aligned polarizing filters applied to panel outside surfaces. Light polarized parallel with a filter enters the cell following the helical path set by the crystal molecules. When the light emerges, its polarization has rotated 90° from when it entered.
Reflective models, as the name implies, don’t transmit light but instead reflect it. One such microdisplay from Three-Five Systems, Tempe, Ariz., has a circular-polarized-mirror backplane and a clear top glass. A separate polarized beam splitter breaks incident light into two components, one perpendicular to the optical axis and the other parallel to the optical axis. These are denoted the s-component and p-component, respectively. The p-component passes through the splitter while the s-component reflects into the display top glass. This s-component of light becomes elliptically polarized as it traverses the crystal material. While the details are proprietary, the net result is that the reflected light exits with a p-component.
This reflected light passes back through the beam splitter and is made viewable with a magnifying eyepiece. Colorado Microdisplays Inc., Boulder, builds a similar display but places an analyzer in the viewer optical path. The analyzer permits only p-component light to pass which cleans up the signal. Such displays reflect light when no power is applied which gives them the familiar silver-gray color.
However, the displays would be of little use if it weren’t for a key property of liquid crystals. Nematic molecules, for example, have permanent electric dipoles so their long axis can be made to orient parallel to an applied electric or magnetic field. This molecular-level deforming, or so-called Freederickz transition, generally takes just a few milliseconds. Nematic crystal stacks begin untwisting at approximately a 1-Vdc potential with full-90° distortion occurring at about 3 Vdc.
To switch the display from light to dark, or somewhere in between (grayscale), such a voltage potential is applied between the two panels. In both transmissive and reflective types, a film of transparent, conductive indium tin oxide (ITO) acts as a common top panel electrode. With transmissive displays, the other surface may be patterned with thin-film transistors to form individual pixels (electrodes) of an active-matrix display. Larger versions of these displays are standard equipment in laptop computers.
Reflective microdisplays have a silicon backplane and contain FET-switched pixels and CMOS support circuitry. These backplanes are generally much less than an inch square yet can hold over 1 million pixels in some cases. Activating pixels cause the crystal molecules between the panels locally to untwist and align with the electric field direction. Now, s-component light entering this area doesn’t switch its polarization state and simply goes in and comes out unchanged. Since the beamsplitter only passes the parallel-polarized (p-component) light, active sections of the display appear dark.
The individual pixels are made to turn off and on from video signal control inputs to row and column registers. Depending on the type of control system pixel voltages can either be single-valued or analog. The choice of system depends on the type of liquid crystal used. Nematics, because of their voltage-dependent twisting, require analog controls. Levels of gray can be had by varying pixel voltage. So-called ferroelectric fluids, in contrast, have just two states, on or off. To get shades of gray requires what is called time-division multiplying. Here regular clock cycles used to time pixel signals are split into smaller pieces. The eye integrates the relatively shorter black and white flashes and perceives the color as gray.
The eye perceives light flashing on and off at frequencies above 30 Hz as a continuous source. If different colors of light flash in sequence, the eye blends the colors to perceive one color. For example, red and blue appear purple. By adding a third color, green, millions of color combinations are possible, including white. These three colors form the familiar RGB scheme used in CRTs. However, CRTs don’t use strobed light. They instead have pixels segmented into the three separate colors. The electron streams are directed at the different areas sequentially. Some transmissive microdisplays use this spatial sequencing scheme as well. Here, pixels are split into three addressable segments, each with its own color filter.
The future looks bright
Today there are over 30 microdisplay vendors, according to Chris Chinnock, editor of Microdisplay Report. And there seems to be at least as many microdisplay varieties as well. Most are built around the liquid-crystal model, while others, such as the virtual retinal display from Microvision Inc., Seattle, eliminate the screen altogether. Here, RGB laser light guided by a micromachined (MEMS) mirror paints full-color images on the retina itself.
Still, microdisplays alone are just one piece of the puzzle. They have to be integrated into products that consumers will buy. In that department there appears no shortage of ideas. At the recent Consumer Electronics Show in Los Vegas, for example, one company unveiled a 50-in. rear-projection HDTV (1,920 3 1,080 resolution) based on LCOS technology. Its developers say the set could retail for $2,000, about $8,000 less than conventional models.
|Some LCOS microdisplay makers
Colorado Microdisplays Inc. www.comicro.com
Three-Five Systems Inc. www.threefive.com
Displaytech Inc. www.displaytec.com
Kopin Corp. www.kopin.com
What are microdisplays?
There are two basic image-viewing methods, virtual and projection. Virtual microdisplays use optics to create magnified images that appear recessed inside a viewfinder. Camcorder viewfinders, for example, use such an optical scheme. Projection microdisplays also use magnifying optics but instead project an image on a screen. The screen may be external, as in the case of multimedia projector or contained within a housing such as in computer monitors or rear-projection TVs.