There's a blizzard of information on flat-panel displays. That's not so surprising, given the 30 varieties of displays available and the 85 or so manufacturers that make them. The technologies they employ run the gamut from fluorescent and plasma types, both almost century-old technologies, to the more youthful entries in the marketplace like organic light-emitting diodes (OLEDs), liquid crystal on silicon (LCoS), and inorganic electroluminescent types. These more recent arrivals are just beginning to capture the market's attention.
In such a climate, it would seem that steering a course to sound display decisions would be difficult. Fortunately for the user, display technology is evolutionary rather than revolutionary. Therefore, it's unlikely that some highly appealing display technology will suddenly spring out of the woodwork overnight, leaving the designer who already made a decision lamenting, "If I had only known about . . ."
Lawrence E. Tannas Jr. of Tannas Electronics, Orange, Calif., says, "It is not uncommon for display technology to take a generation to evolve. And even after 20 years, many display innovations that may eventually make it have yet to reach production. But their evolution continues on, at various rates of activity, depending on material advances and market demands." Tannas should know. He's a past president of the Society for Information Display (SID), author of several books on the subject, and a lecturer on display technology at the University of California at Los Angeles.
Still, despite the 30 varieties, the good news for the designer faced with selecting a display comes from the well-established display families. An example is active thin-film-transistor (TFT) displays, which are proven and will meet most needs in the 2- to 10-in. diagonal range. Furthermore, TFTs will most likely be the only choice in the 14- to 28-in. range.
As for the larger displays, when you start moving into screen sizes in the high teens and above, TFT prices begin accelerating and go ballistic, moving rapidly into four and five figures. Luckily, a proliferation of projection techniques are arriving that employ LCoS technology. They may prove to be an effective means of sidestepping the high costs of TFT. Also, not far off in the future lies the inorganic electroluminescent display.
If one is designing a small handheld unit where weight and low-power consumption are of paramount importance, he or she can find microdisplays beginning to pop up with diagonal dimensions of less than 1 in. and exhibiting astonishingly low power consumption. Can you find a display that consumes less than 6 mW? You bet.
Whatever the display, the eye-to-screen viewing distance is crucial. Just how far will the eye be from the display? It turns out that if you have 20/20 vision, you are endowed with a visual acuity that equates to one minute of arc. A valuable number, known as the "optimum viewing distance," is derived from this fact. It can assist a designer in establishing the pixel resolution that's necessary for a comfortable viewing of the display in an intended product (Fig. 1). This reduces to the angle subtended by adjacent pixels in the display of choice. The relationship is governed by:
α ≈ (d/L) 57.3 × 60
α = visual angle in minutes
d = the center-to-center distance between pixels
L = the distance from the display to the viewer's eye.
"Unlike photographic images," Tannas says, "you pay for every pixel." So, you want to see all the pixels you have paid for. If you get too close, the angle subtended by two adjacent pixels exceeds one minute. You then see the details of the pixels, which is sometimes called pixelation or raster noise. In the same vein, if you get too far away, so one minute of arc spans more than one pixel, "then you don't see all the pixels you have paid for," he concludes.
The term "artifact" identifies anything you don't want the viewer to see. By making the distance between adjacent pixels equal to one minute of arc, the viewer won't see the lines, corners, electrodes, or the pixels, but will see the image. By choosing the optimal distance, the viewer is assured that he or she will see all the information, but none of the artifacts.
Just like every other technology, there's no shortage of terminology in the world of displays. A decent familiarity with some of the key terms can be useful (see the table).
Keep a few things in mind with regard to display selection for portable applications where both power consumption and weight issues are so critical. When totaling the power requirements, include all of the electronics needed to produce a final image. This includes analog drivers and analog-to-digital converters (ADCs), as well as DRAM structures—and the inverter, if required. If the displayed contents change frequently, the higher refresh rates will raise power consumption. Incidentally, where a TFT LCD may consume 4 to 5 W, a microdisplay will consume between 10 mW and 1 W.
As far as weight goes—as a portion of the total product weight—a color TFT LCD, the backlighting, and a battery on a portable computer or portable DVD device can amount to 25% to 60% of the overall product's weight. Obviously, a microdisplay that might weigh in at a gram gives the designer a good head start in realizing a truly lightweight design.
There's a great deal of interest in the OLEDs, and there are good reasons for this. The OLED provides a high degree of brightness and a wide viewing angle. Unlike many LCDs in today's market, OLEDs are self-luminous and, therefore, require no external light source. As a result, they do away with bulky, less than desirable, cold-cathode fluorescent backlights (CCFLs). (They are less than desirable because they contain mercury.) In addition, plastic substrates become possible.
Active OLEDs, however, haven't quite arrived. They are expected to enter the marketplace by 2001, according to David J. Williams, general manager of OLED Display Technology, Eastman Kodak Co. The material and device architecture of the OLEDs were developed at Kodak, which has continued to improve the performance of the device over a period of years.
There are a number of layers in an OLED (Fig. 2). These are made up of thin coatings that form the active part of the device. Positive charges are injected at the indium-tin-oxide (ITO) anode, and negative charges are injected at the metal cathode. The charges travel through the organic materials, and when they recombine, they produce excited molecules that emit light. The color of light emitted—red, green, or blue—depends on the chemical composition of the layer.
Power consumption of the OLED is expected to be lower than a backlit-LCD. Also, the OLED weighs about half as much. This May, OLEDs exhibited at the SID Conference, Long Beach, Calif., were just 1.8 mm thick. This compares favorably with the LCD, where the thickness is typically 7 to 8 mm. As for the visual qualities, the brightness is approximately 150 candelas/m2. The contrast is approximately 1000:1 under dim laboratory conditions. There have been reports that the perceived brightness of an OLED at 20 candelas/m2 is equal to the perceived brightness of 50 candelas/m2 on an active TFT display.
The relative lifetimes of the three different colored materials degrade at slightly different rates. At present, Williams reports, lifetimes are approximately 5000 hours to half efficiency. Kodak hopes to extend them to 10,000 hours.
The voltage required by the OLED is in the 2- to 10-V range, which is particularly attractive in portable battery-powered applications. The viewing angle of 160° is quite satisfactory for most applications.
The passive versions, in which frame rates will be low, will be best suited for low-cost applications. An active-matrix OLED, though, enables a lower operating voltage. Of course, higher speed that's inherent in an active switch is enabled, too. And where faster refresh rates of 80 MHz and above are required, active versions are warranted. The scanning circuitry can be embedded in an electronic backplane, thereby eliminating any additional requirement for high-density interconnects and drivers.
It turns out that maximum efficiency in an OLED occurs below the maximum brightness point. If you were to operate an OLED at the highest brightness—moving beyond the peak in the maximum efficiency in lumens/W—it would, therefore, tend to speed degradation. "The lumens/W figures do diminish at the higher voltages," acknowledges Williams. "We will, however, be pushing the voltage down, as we move towards commercial products."
How do the critics size up the OLED? "It is pretty clear to me that active-matrix OLEDs—5 in. and below in diagonal—will become even cheaper than passive-matrix LCDs," says Barry Young, vice president of DisplaySearch, a market research firm that specializes in displays.
But another observer isn't so sure. "It is a beautiful technology, but OLEDs are going to have a hard time finding a market that liquid crystals can't fulfill cheaper and better," Tannas says. "OLEDs can't beat LCDs yet, and there is no promise that they will in the future," he speculates.
Coming on strong are microdisplays. Expected applications of these devices fall into two segments. The first segment encompasses virtual displays that employ optics to magnify an image—so as to wind up with a larger image that appears recessed inside a viewfinder.
A virtual image differs from a real image in that the viewer sees only the magnification of the original image. In this application, the appeal of the microdisplay is that it provides the viewer of portable devices with no-compromise image quality. In fact, with diagonals as small as 0.5 in., a microdisplay is said to provide the user with color and video capability comparable to or better than a television set or a PC monitor.
The second application segment is projection. Here, microdisplays are teamed up with optical systems that magnify and project an image onto a screen (Fig. 3). This screen can be distant, as is the case with a multimedia projector. Or, the screen can be contained within a housing, such as a computer monitor. When implemented in a front-projection device, a distant wall or screen is used to view the magnified image. In the case of rear projection, magnification optics are enclosed in the back or front of a monitor or TV housing and an integral screen is employed to produce the image within the self-contained system.
To obtain the image, microdisplays are used in either a transmissive or a reflective mode. In the transmissive mode, the illumination source is located in the back of the microdisplay. The light is modulated at each pixel to produce an image. In a reflective display, however, the external light source is reflected off the front of the display.
The advantages of the reflective technique is that approximately 90% of each pixel is used, yielding a light efficiency of 40%. This compares with a transmissive arrangement where the light can be thought of encountering the display in what is sometimes termed a "screen door effect." In this case, only 3% of the available light passes through the display.
Often in full-color microdisplays, a field-sequential approach is employed. It relies upon sequential illumination of the microdisplay turning on, successively, the red, green, and blue light sources from a side- or rear-mounted source. Therefore, only a third of the pixels are active at any given instant. If the sequence is fast enough, the human eye no longer discerns the separate colors and perceives a full-color image.
When moving images are portrayed, some controversy persists with regard to the refresh rate. Is the old standard of 60 Hz sufficient? Colorado MicroDisplay Inc., among others, believes that 85 Hz should be the lower bound for refresh rates. The company points out that rates of less than 85 Hz may cause many viewers to perceive motion artifacts or flicker. This causes discomfort in the form of eye fatigue. At 60 Hz, in sequential displays, viewers may discern color breakup. This describes the phenomenon in which the viewer's eye detects the red, green, and blue subframes.
To support the higher refresh rates and consuming just 15 mW, Colorado MicroDisplay has recently introduced a low-power, high-resolution, 320- by 240-pixel (QVGA) microdisplay that delivers up to 24-bit color depth (or 16.7 million colors). The display can handle refresh rates up to 120 Hz.
A companion CMD3XLB illumination controller is available to manage the LED-RGB illumination synchronization. The chip set that comprises the CMD3X2A microdisplay and CMD3XLB illumination controller costs approximately $25 in production volumes. It's especially well suited for mobile communications devices, digital still cameras, and smart phones. And the chip set has the capability to handle full-motion video, as well as text and still images. This means that whatever the media source, the viewer perceives a rock-solid image on the screen.
Based on LCoS technology and with the employment of a single-panel, field-sequential color method, the CMD3X2A's large reflective pixels overcome the pixelization, low-optical efficiency, and high-power consumption that are usually encountered in transmissive microdisplays. The display's power consumption, independent of the illumination and drivers, is 15 mW. This is the case even when operating at refresh rates of 120 Hz. Power-supply voltage is 5 V, the display diagonal is 4.8 mm, and the image size is 3.84 mm (horizontal) by 2.88 mm (vertical). Pixel pitch is 0.012 by 0.012 mm. Together, the two-chip set enables a display contrast ratio of 100:1.
Both chips are presently available. An evaluation kit also is available.
Another microdisplay man-ufacturer, Displaytech Inc., employs a ferroelectric liquid-crystal (FLC) material in its displays. The company says that this material is particularly well suited for camcorders and digital still cameras. The manufacturers apply a very thin layer, less than 1 µm, of FLC material on top of the silicon die. Because pixel size is governed by the liquid-crystal thickness, the company claims to have the smallest pixels in the industry, with a pixel pitch of 5 to 6 µm. The advantage provided by this technology is a thin, liquid-crystal bandgap that enhances the switching speed and tightens the interpixel gaps, enabling very small pixels.
Displaytech views its advantage in the sense that for a given pitch, the company is able to utilize the smallest dies. Or for a given die size, it can place more pixels on the die than anyone else—at present. For a given resolution, the company boasts that it can produce the smallest die. Or for a given die size, Displaytech claims to be able to achieve a higher-resolution display than any other company.
Because of their excellent image quality and low power consumption, Display-tech's ferroelectric LCDs are suitable for viewfinder applications, head-mounted displays, PDAs, communicating wristwatches, and wireless Internet devices.
Their QVGA (320- by 240-pixel) Display Module, with a 0.19-in. array diagonal, supports a frame rate that's adjustable between 50 and 100 Hz. Total module power consumption is below 135 mW. One of its features is the adjustable brightness to deal with changing ambient-light conditions.
A second product is the LightCaster WXGA (1280- by 768-pixel) display panel with a 0.78-in. array diagonal (Fig. 4). Color depth is 24 bits, and the frame rate is 60 Hz. This display is intended for rear-projection systems, HDTVs, rear-projection monitors, and front-projection systems. It employs a field-sequential color technique. With its two display modes it enables a designer to select either XGA (1024- by 768-pixel) or HDTV (1024- by 720- pixel) resolution with a wide-screen format (a 16:9 aspect ratio).
Three-Five Systems Inc. has its sights on four major display markets. According to Rob Harrison, senior director of marketing, these markets are "near-to-the-eye applications, including head-mounted displays, mobile web browsers, multimedia projectors for conference rooms, and rear-projection monitors for computers and rear-projection television." Robert L. Melcher, chief technical officer at Three-Five Systems, says, "We view LCoS as the one technology that is capable of addressing all four of those markets."
As Melcher describes the manufacturing process, to fabricate LCoS, liquid crystal is applied to a crystalline-silicon, CMOS, active-matrix array on a silicon chip. The LCoS operates in a reflection mode, illuminated by a strong arc lamp. The light passes through the liquid crystal, which is modulated by the voltage via the silicon backplane. The reflected light then carries the image to the viewer.
Recently, Three-Five has begun shipping its MD800D display. This is a standardized product that measures 0.42 in. diagonally and delivers SVGA (800- by 600-pixel) resolution, or 480,000 pixels. When used in a near-to-the-eye application, the display provides a virtual image equivalent to a 19-in. SVGA desktop monitor viewed at 2.5 ft. It allows the user to view an entire web page at one time without scrolling or translation software. This display consumes a total of only 100 mW, including all the electronics and illumination. Weighing in at 28.8 grams, it also is well suited for the handset market.
As for larger, direct-view units (diagonals over 40 in.), "the plasma display is the only game in town," says DisplaySearch's Young. Currently affordable only by the affluent, these displays contribute approximately $5000 to the selling cost of an HDTV.
A 60-in. plasma display was exhibited at the SID Conference by Plasmaco. The brightness was 550 candelas/m2 and the contrast ratio was greater than 500:1. Its resolution was 1366 by 768 pixels (WXGA). The display offered 256 gray levels and a viewing angle of 160°.
Much of the current interest in these leviathans is with the military. They need displays for command center and shipboard applications where a shallow panel depth of 3.5 to 4 in. is particularly appealing and, therefore, a decided advantage over the much deeper CRT.
Another industry looking at these large plasma displays is the film industry—for post-production viewing and for small screenings, as well as for displays hanging in lobbies. Additional applications include theme parks and airports.
Plasmaco is a research and development subsidiary of Panasonic (Matsushita Electric Industrial Co. Ltd.). Its parent is currently manufacturing 37- and 42-in. plasma displays in its Osaka, Japan, plant and will be coming out with a 50-in. version in October.
A breakthrough in achieving a low-cost flat-panel display may occur in the next two years if iFire Technology Inc. continues on track with its steady progression in improving its inorganic electroluminescent flat panel. At the SID Symposium, members of the company unveiled a 17-in. prototype electroluminescent flat panel with a brightness of 100 candelas/m2.
Still, the company isn't there yet. Sizes need to reach above 30 in. and brightness must be pushed up to 300 to 400 candelas/m2.
But iFire does have its eye on the HDTV market and expects to lower display costs from $5000 to $1000, enabling the selling price of the HDTV to plummet from approximately $9000 to $3000. If the display cost can drop, so that HDTVs will be marketable in far higher volumes, it makes sense that everything else—components costs, manufacturing costs, and the like—will tumble as well. The company's approach, based on inorganic electroluminescent display technology, relies on low-cost, thick-film processes. Its manufacturing steps employ a simple screen-printing technique that's well known in the ceramic-capacitor and pc-board industry.
For low-volume applications and/or where the objective is to cut nonrecurring design time and costs to the bone, a "plug and play" approach to displays may fill the bill (Fig. 5). Recently, Applied Data Systems Inc. announced a two-headed, single-board computer. It enables the running of two QVGA panels working off a 206-MHz, 32-bit StrongARM RISC processor. The processor delivers 235 MIPS at just 400 mW of power, enabling it to support two displays simultaneously and independently. Applications include point-of-sale terminals and gasoline pumps.
This single-board computer works with a variety of LCDs. The company has interfaced with over a half-dozen displays—from quarter VGA (320 by 240 pixels) up to XGA (1024 by 768 pixels). It's available immediately, with pricing in the low $300s—exclusive of the displays.
Active TFTs command the lion's share of the marketplace in midsize displays from 12 in. in diagonal and up to approximately 22 in. Typical of the recently introduced notebook computer displays is a 13.3-in. TFT-LCD module from LG Philips LCD America Inc. Designated the LP133X8, it's intended for PC makers who are designing new 13.3-in. notebooks and Internet appliances. At 390 grams and 5.2-mm thick, this XGA display is unusually light and thin. Philips says that these 13.3-in. displays can be as slim and light as traditional 12-in. displays, often for nearly the same price and with the same form factor.
The LP133X8 employs an unusual side-mounting technology. It offers XGA resolution (1024 by 768 pixels), which previously was confined to desktops. The average brightness is 150 candelas/m2 and the typical contrast ratio is 200:1. Pixel pitch is 0.264 mm, both horizontally and vertically. Power consumption is less than 4.5 W.
Another newcomer in the slightly larger TFT sizes is from NEC Electronics Inc. The company has introduced a 20.1-in., high-resolution, color LCD module with a digital interface. With its unusually wide, maximum viewing angle of 170° (top/down, left/right), the NL128102AC31-02 display module is equivalent to a 22- to 23-in. CRT monitor. It targets graphic design, desktop publishing, and medical applications.
The NL128102AC31-02 extends the company's line of SXGA displays. It builds on the company's recent introduction of an 18.1-in. version. Its interface is of the low-voltage differential signaling (LVDS) variety, delivering 8-bit RGB signals with 16.77 million colors and a resolution of up to 1280 by 1240 pixels.
|Companies And Organizations That Contributed To This Report|
Colorado MicroDisplay Inc.
Eastman Kodak Co.
iFire Technology Inc.
LG Philips LCD
NEC Electronics Inc.
Three-Five Systems Inc.