Plasma displays are closely related to the simple neon lamp. It has long been known that certain gas mixtures will, if subjected to a sufficiently strong electric field, break down into a “plasma” which both conducts an electric current and converts a part of the electrical energy into visible light. This effect produces the familiar orange glow of the neon lamp or neon sign, and can readily be used as the basis of a matrix display simply by placing this same gas between the familiar array of row and column electrodes carried by a glass substrate (Figure 4-12a). A dot of light can be produced at any desired location in the array simply by placing a sufficiently high voltage across the appropriate row-column electrode pair. The plasma display panel, or PDP, is clearly an emissive display type, in that it generates its own light. However, unlike the CRT, there is no easy means of controlling the intensity of the light produced at each cell or pixel. In order to produce a range of intensities, or a gray scale, plasma displays generally rely on temporal modulation techniques, varying the duration of the “on” time of each pixel (generally across multiple successive frames) in order to provide the appearance of different intensities.
Plasma displays may use either direct current (DC) or alternating current (AC) drive; each has certain advantages and disadvantages. The DC type has the advantage of simplicity, both in the basic structure and its drive, but can have certain unique reliability problems owing to the direct exposure of the electrodes to the plasma. In the AC type, the electrodes may be covered by an insulating protective layer, and coupled to the plasma itself capacitively. This results in an interesting side-effect; residual charge in the “capacitor” structure thus formed in a given cell of the AC display in the “on” state pre-biases that cell toward that state. Even after the power is removed, then, the panel retains a “memory” in those cells which were on, and the image can then be restored at the next application of power to the panel.
Color is achieved in plasma panels in the same way as in CRTs; through the use of phosphors which emit different colors of light when excited. In color plasma panels, the gas mixture is modified to optimize it for ultraviolet (UV) emission rather than visible light; it is the UV light that excites the phosphors in this type of display, rather than an electron beam. A typical AC color-plasma structure is shown in Figure 4-12b; note that in this type of display, barriers are built on or into the substrate glass, in order to prevent adjacent sub-pixels from exciting each others’ phosphors. Again, due to the difficulty of directly modulating the light output of each cell, temporal modulation techniques are used to provide a “gray scale” capability in these displays.
Figure 4-12 Plasma displays. In the typical monochrome plasma display panel (PDP), light is produced as in a neon bulb – a glowing plasma appears between the electrodes when a gas mixture is subjected to a sufficiently high voltage across them. In a color plasma panel, shown here as an AC type, the gas mixture is optimized for ultraviolet emission, which then excites color phosphors similar to those used in CRTs.
The fundamental mechanism behind the plasma display panel generally requires much higher voltages and currents than most other “flat-panel” technologies; the drive circuitry required is therefore relatively large and robust, and the structures of the display itself are larger than in other technologies. Owing to these factors, plasma displays have in practice been restricted to larger sizes – from perhaps 50-125 cm (20-60 inches) diagonal – and relatively low pixel counts. For this reason, plasma technology has not enjoyed the high unit volumes of other types, such as the LCD, but has seen significant success in many larger-screen applications such as television and “presentation” displays. The plasma display, especially in its color form, competes well against the CRT in those applications where, for reasons of space restrictions or environmental concerns, the much higher cost can be justified.
Electroluminescent (EL) Displays
Probably the simplest display, at least conceptually, is the electroluminescent or “EL” panel. Phosphor materials, in some cases identical to those used in the more common CRT, will glow not only when struck with an electron beam but also when subjected to a sufficiently strong electric field. Therefore, placing these materials between electrodes in the now-common row and column arrangement can produce an emissive display with an attractively wide viewing angle. In order to increase the light output, the rear substrate can be made reflective, although this can reduce the display contrast as it also reflects incoming ambient light. Another common design is to place a black (light-absorbing) layer at the rear of the structure, and/or place a circular polarizing layer on the front surface, both done in order to increase the display contrast . The circular polarizer will pass light produced by the display, but ambient light entering the panel and reflecting off the rear surface will not exit the panel due to the reversal of polarization occurring upon reflection.
Like the plasma displays, EL panels have been produced in both DC- and AC-drive versions, and are further classified by the nature of the electroluminescent layer (thin-film or powder); they can also employ either a passive-matrix or active-matrix drive scheme. Thin-film AC EL panels are currently the most popular commercial type. The technology does provide luminance control, although in this case it is by either varying the refresh frequency, or through the use of pulse-width modulation or other temporal techniques.
To date, EL technology has not seen the commercial success of the LCD or plasma types, and the use of this type of display has been for the most part restricted to certain industrial or military applications where the inherent ruggedness of the panels make them attractive. EL displays suffer from the need for high drive voltages and until recently relatively low luminance and contrast. The most common monochrome EL material (zinc sulfide, with manganese as an activator) produces a yellowish-orange light. Full-color EL panels have been very slow in coming. There are two options for producing color displays in this technology: first, a panel can be constructed using a white-emitting phosphor, and color filters applied over it as in the LCD case. The other option is to pattern individual sub-pixels of red-, green-, and blue-emitting phosphors, to form full-color triads as in the case of the CRT or color plasma types. Both of these have had their problems; the color-filter approach suffers from not having a sufficiently bright white-emitting phosphor available to tolerate the luminance reduction which comes from the filter layer. Using three separate phosphors, one for each primary, has resulted in workable full-color EL panels, but to date the blue phosphors used have not provided sufficient light output.
At the present time, EL technology appears to be in danger of being relegated to certain niche markets and applications, and potentially bypassed altogether due to advancements in other technologies. However, the history of EL development has been one of periods of rapid progress separated by times of relative stagnation, and it would be premature to count EL out just yet.
Organic Light-Emitting Devices (OLEDs)
A relatively recent development, just now coming to the market in commercial products, is the organic light-emitting device, or OLED. As a matrix display, the OLED panel functionally most resembles the EL types – a layer of light-emitting material placed between the electrodes which define the pixel array. However, unlike EL, the OLED materials operate at a much lower voltage, approximately in the same range as is used in the liquid-crystal types. Further, high-brightness OLED materials have already been demonstrated in all colors, including white, and so full-color operation is relatively easy to achieve. OLED displays promise a combination of the best of both the EL and LCD technologies: an emissive display with a wide viewing angle, good brightness and contrast, and yet with relatively modest power requirements and low operating voltages. Electrically, the OLED’s drive requirements are so similar to those of the LCD that it is expected that many LCD production lines could be converted to OLED production in a reasonably straightforward and economical manner.
While the OLED display structure superficially resembles EL, in that an emissive layer is located between the row and column electrodes, the actual OLED structure is somewhat more complex. The basis for the OLED is a layered structure of organic polymer semiconductors, arranged so as to produce light through a mechanism similar to that of the ubiquitous light-emitting diode. A typical OLED structure is shown in Figure 4-13.
Active-matrix OLED panels have already been shown to be practical alternatives to the TFT-LCD, with the potential to compare favorably with that technology in terms of both cost and performance. OLEDs are expected to compete with the LCD in practically all current LCD applications, from small calculator, PDA, cell phone, and similar displays, through notebook computer displays and panels intended for desktop monitors. The materials also appear to be a reasonable choice for use with polymer substrates, raising the possibility of low-cost, flexible color displays. Further, the OLED technology may scale to larger panel sizes than has been possible with the LCD, making it a potential alternative to plasma in at least the lower end of the size range covered by that technology. OLEDs are still in their infancy in terms of commercial development, but definitely seem poised to take a significant share of the worldwide display market in the near future.
Figure 4-13 The structure of an OLED. Not shown are the glass substrates between which these layers would be built. Note that the OLED, unlike the LED, is a current-driven, rather than voltage-driven, device.
Field-Emission Displays (FEDs)
For almost as long as there have been CRT displays, there have been those who have tried to construct a version of this technology – a display which uses electron beams to stimulate phosphors – in flat-panel form. Probably the closest commercial technology until recently, in terms of being analogous in operation to the CRT, has been the color plasma panels, which, as noted above, use UV light to excite the phosphors. The main problem with translating the CRT to a true “flat panel” display has always been the source of the electron beam; the conventional CRT uses a heated cathode along with extremely high accelerating voltages. Besides using a considerable amount of power, this approach does not readily lend itself to incorporation in a thin display. It should be noted that several manufacturers have attempted to make flat, thin (or at least thinner than normal) CRTs using more-or-less conventional heated-cathode electron sources, with varying degrees of success, but a true flat-panel equivalent to the CRT display required the development of a “cold” source of electrons which would operate at lower voltages.
This essentially defines the distinguishing feature of a class known as field-emission displays, or FEDs. These are display devices which produce light via phosphors, again excited by streams of electrons, but unlike the CRT the electrons originate from emitters which do not require heating above ambient levels. The term “cold-cathode CRT” has also been used to refer to this class of display.
There have been several different approaches to the problem of designing a practical electron emitter for these devices. The quantitative measure of the ease with which electrons may be driven off (or, from a different perspective, extracted from) a given surface, material, or structure is the work function, which may be expressed as either the potential required to cause electron emission (the work function potential) or the equivalent energy requirement in joules or electron-volts. In these terms, then, what is needed is a practical emitter design with a sufficiently low work function so as to permit adequate electron emission at ambient temperatures.
In a conventional, heated CRT cathode, achieving the required work function level is generally done through the use of certain materials to form the actual emitting surface; a common example is barium oxide, a layer of which is applied to the “top” surface of the metal CRT cathode. This is in general not practical in a flat-panel device; not only so such materials fail to provide a low enough work function on their own, but there are increased requirements for emission uniformity over a relatively small area and the requirement that the emitter be capable of fabrication using available FPD processes. Therefore, other approaches are used for FEDs. Most rely on the fact that sharp edges, points, or similar structures are relatively easy places from which to extract or inject charge (owing to the concentration of charge in such regions, and the resulting concentration of electric fields there; this is similar to the principle behind the common lightning rod). Such structures are fairly easy to produce using conventional silicon-IC processing techniques, which themselves are readily adaptable to flat-panel display production.
Several emitter designs have been used in the development of FEDs. Each of these involve structures which are sufficiently small so as to permit multiple emission sites per pixel (or sub-pixel, in the case of a color display), in order to address the need for overall uniformity. The Spindt cathode (named for its inventor, Charles “Capp” Spindt, then of the Stanford Research Institute) use a conical emitter, formed through standard IC fabrication processes, as the source of the electrons, which pass through a hole in a surrounding conductive layer which acts as a control grid. A similar approach uses a long sharp edge as the emitting structure, again with structure acting as a control grid placed above and to either side of the emission site. Recently, carbon nanotubes have shown great promise as the electron emitters for field-emission displays. These are microscopic hollow filaments of carbon, which can be deposited on the display substrate so that many are oriented orthogonal (or nearly so) to the surface. The nanotubes are small enough, and packed densely enough on the substrate, so as to provide many emission sites per pixel or sub-pixel, and thus fulfill the requirement for uniform electron emission across the area to be illuminated.
Outside of the unique requirements for the emitters, the FED is very similar in basic structure to the other flat-panel types, most closely resembling the plasma display (especially in the color form). A cross-section of a typical color FED is shown in Figure 4-14. This display uses the familiar row-and-column addressing scheme, with drivers located on the periphery of the panel. Conventional CRT phosphors are placed on the inner surface of the front glass, and barriers are constructed between pixels and sub-pixels to isolate them from each other, and also often form the spacers between the front glass and rear substrate. FEDs have been designed in both low-voltage (up to several hundreds of volts potential difference between cathode and anode) and high-voltage (thousands of volts) forms; each is promoted as having certain advantages. The tradeoff to be made is basically one of acceptable luminance at lower cathode currents, versus the complexities and costs involved with generating and controlling higher voltages.
Figure 4-14 The structure of the field-emission display, or FED. This drawing is not to scale, especially with respect to the electron emitters. These are typically microscopic, and of sufficient quantity that thousands of individual emission sites may comprise a single color sub-pixel.
As in the case of the OLED display, FEDs are just now entering the market commercially. These displays also promise high brightness and contrast at power levels and costs competitive with the TFT-LCD, and again offer the viewing angle and potential size advantages of an emissive display, requiring no backlight. FEDs do, of course, have some unique challenges, including the requirement for higher drive voltages and processes which to date have not been commonly used in smaller-size FPDs. Time will soon tell how successful these new FPD technologies are in the various display markets.
Perhaps the ultimate marriage of flat-panel display and silicon IC technologies, microdisplays have recently opened numerous new opportunities for electronic displays. Essentially a display constructed on (or even as) an integrated circuit, this class covers multiple technologies sharing two main distinguishing features: they may be considered “flat-panel” displays, but they are of such a small size (generally under 5 cm (2 inch) diagonal, and often less than 2.5 cm (1 inch)) that they are not used in a conventional direct-view manner. Instead, microdisplays are either used to generate the appearance (a “virtual image”) of a much larger display via magnifying optics, or the image of the display is projected onto a screen for viewing. Products using the former mode are often classed as “near-eye” applications, since the display device itself is physically located near the viewer’s eye; example are camera or camcorder viewfinders, or so-called “eyeglass” or “head-mounted” display systems. As the basis for projection displays, microdisplays become the hearts of products competing with direct-view monitors, televisions, and much larger presentation display systems.
Microdisplays may be categorized into two broad groups: those which are essentially miniaturized versions of any of several of the conventional FP technologies, such as LCD or OLED displays, and those which employ micro-electro-mechanical (MEM) structures to control light. The former category is currently dominated by the liquid-crystal types, often referred to as liquid crystal on silicon (LCoS) microdisplays. These are exactly what the name implies: a liquid-crystal display constructed on top of a silicon IC. The IC is basically a slightly modified memory array, in which the individual memory cells form the storage and drive elements for the LC pixels. The most obvious modification involves a slight change to the LC process – each element in the array must be topped by a large pad of reflective metal, formed as the last metallization step in the IC processing, which acts as both the driven electrical contact for the LC cell and the light-reflecting “back” of the display. A typical LCoS microdisplay is shown in Figure 4-15. Outside of the silicon IC substrate, this display is virtually identical to its larger, direct-view cousin. An alignment layer is deposited on top of the IC’s metal pads, a glass panel carrying the transparent upper electrode (ITO) is placed on top of the IC, and the cavity between glass and silicon filled with liquid-crystal material. (Note that only a single common electrode need be supplied by the glass, since the addressing of individual pixels is handled completely by the IC.) This example, as is the case with almost all LC microdisplays, is obviously a reflective display; polarized light enters through the glass, and is reflected from the metal pad at the “bottom” of each cell. The polarization of the light may either be altered by the LC or not, depending on its state, which provides the basis for the device serving as a display. (In most applications for displays of this type, the polarization of the light source and the polarizer through which the image are observed are generally physically separate from the microdisplay component itself, and in practice both functions are commonly provided by a single component.)
Figure 4-15 A typical liquid-crystal-on-silicon (LCoS) microdisplay. These devices are essentially LC displays built on top of a silicon IC, which provides both the lower electrodes (doubling as reflective surfaces) and the drive and interface electronics. Reflective microdisplays of this type may be used in both direct view (through magnifying optics, and then generally referred to as “near-eye” types) and projection applications.
It should be noted at this point that at least one company has successfully produced a liquid-crystal microdisplay which does not operate in the reflective mode. Kopin Corporation bases their displays on specially designed silicon circuits, as in the above types, but through a proprietary process transfers the circuitry to a glass substrate, and the microdisplay then constructed on that substrate is basically a miniaturized transmissive active-matrix LCD.
The other major class of microdisplay are the electromechanical types, which are characterized by their use of physically deforming or altering the position of structures within the device in order to control light. The most successful example of this class to date has been the “digital micromirror device”, or DMD, introduced by Texas Instruments in 1987, and which forms the heart of a technology which T.I. refers to as “digital light processing”, or DLP. In these devices, each pixel is actually a movable metal mirror, which tilts back and forth under the control of electrostatic forces driven by the integrated circuit below (as shown in Figure 4-16a). The mirrors’ tilt determines whether incoming light will be directed either through an optical system to the viewer (the “white” state for the pixel) or off to a “light trap” and so not seen by the viewer (the “black” state). The DMD has seen considerable commercial success in the conference-room and larger-screen projection markets, and is now one of the most serious challengers to conventional film projection in cinematic entertainment applications. Other examples of electromechanical microdisplays include Silicon Light Machines “grating light valve” device, in which strips are deformed electrostatically to control light via diffraction.
In any of these, however, full-color operation presents a unique challenge for the microdisplay, at least in the case of near-eye applications. These devices are in most cases too small to achieve color through individual color filters for each pixel, as is normally done with direct-view LCDs. (And in some cases, such as the electromechanical types, the color-filter method is simply not possible.) Instead, a field-sequential color drive scheme is more commonly employed. In this method, the color image is separated into three fields, one for each primary, and displayed in rapid succession. The light source is similarly switched between the three primary colors, in synchronization with the displayed fields. This results in the appearance of full-color image, and each pixel appears as the proper color over its full area; there are no separate color sub-pixels. In near-eye applications, the light source is most often implemented as a set of light-emitting diodes (LEDs), one in each of the three primaries. Figure 4-17 shows a typical near-eye display employing a reflective LCoS microdisplay with LED illumination.
Figure 4-16 The Texas Instruments Digital Micromirror Device, or DMD. In (a), two mirrors are shown in schematic form, illustrating how they may be tilted to direct light in different directions (this occurs due to electrostatic forces from electrodes on the underlying IC. (b) shows a series of photographs of actual mirrors and their underlying support structures. (c) shows the complete device, in its packaging.
Microdisplays may also be used as the basis for display systems providing normal “desk-top”-sized images, and even beyond to large, group-presentation displays, by projecting the image of the display on a screen of the desired size. Projection displays in general are described in more detail in the next section. The basic mode of operation of the microdisplay, however, is unchanged; it is simply a case of providing a considerably higher level of illumination, and then employing projection optics to image the display at the desired location.
Figure 4-17 An LCoS microdisplay product with an integral light source, for near-eye use. The LCoS device itself is mounted on the flexible substrate, underneath the black plastic structure that carries LEDs (on the upper left of the housing, as seen here) and a curved film which acts as a polarizing beamsplitter.
Simple in theory, at least; as we will see, projection displays have their own unique set of challenges.