Read an Excerpt
Polarization Engineering for LCD Projection
By Michael Robinson John Wiley & Sons
Copyright © 2005 John Wiley & Sons, Ltd
All right reserved. ISBN: 0-470-87105-9
Chapter One
Introduction
1.1 The Case for Projection
Daily life increasingly relies on electronic displays. Indeed, the information age is unimaginable without them. An electronic display is a device or system that converts an electronic signal representing video, graphic, or text information to a viewable image of this information. A display can be virtual, direct view, or projection. With a virtual display, there is no real image in space and the image information is brought to a focus only on the retina. Such displays are limited to one observer only. Direct-view displays are most familiar to the average person. The most common direct-view displays are cathode-ray tubes (CRTs) in televisions (TVs) and computer monitors. Other direct-view technologies, such as plasma displays, organic light emitting diode displays (OLEDs), and liquid crystal displays (LCDs), are starting to challenge the dominant position of the CRT in display applications. Active matrix LCD (AMLCD) computer monitors outshipped CRTs for the first time in 2003. These displays are all capable of high resolution and satisfactory luminance. However, it is difficult and expensive to make a direct-view display large enough to accommodate several viewers simultaneously.
The human eye has an angular resolution of approximately 1 minute of arc. Assuming an image is displayed at a distance of 2 meters from the viewer, the size of the display must be as large as ~70" to fully resolve the high-definition television (HDTV) content, which is shown in 1920 x 1080, ~0.6mm, full-color pixels (see Figure 1.1). It is certainly challenging, and expensive, to make a direct-view display of this size at present.
Projection displays utilize an optical imaging system to magnify a small picture created either by conventional direct-view technologies, such as CRTs, or by modulating the light from an illumination system with a device called a light valve or panel. A projection display can be operated either in front-projection mode, where the viewer and projector are on the same side of the screen, or in rear-projection (RP) mode, where the viewer and projector are on opposite sides of the screen. At the present time, projection systems offer the only economical solution to large, high-resolution displays. Figure 1.2 shows where projection displays figure in the display market with regard to resolution and screen size [Stupp E. H., 1999, p. 4].
1.2 History and Projection Technology Overview
1.2.1 Cinema Film
The history of projection systems begins with cinema movie projectors, which are the earliest and most familiar projection systems to the public. This type of projector is able to deliver a large, high-resolution image viewable by a large audience. The first machine patented in the United States, that showed animated pictures or movies, was a device called the "wheel of life" patented by William Lincoln in 1867. However, the Frenchman Louis Lumiere and his brother Auguste are often credited with inventing the first motion picture camera and projector in 1895. They presented the first projected moving photographic pictures to a paying audience. The first commercially successful projector was invented by Thomas Edison in 1896. The advantage of the film projector is that it displays very high-resolution images, which no modern projection technologies have surpassed as yet. Other types of film projectors include slide projectors and overhead projectors commonly used in classrooms.
The system layout of a typical cinema projection system is shown in Figure 1.3. It consists of an illuminator (lamp), film rotation drums, a sync shutter, and a projection lens. The film frame rate is 24 frames/sec, but is illuminated through a sync shutter operated at double the frequency to avoid flicker. A 16mm diagonal format is the typical film size used for motion pictures. Although the projection system is relatively simple and cheap, the film is not in digital format and must be physically copied for individual media content. It is therefore expensive to distribute the media and is clearly incompatible with the modern digital information age.
1.2.2 CRT-based Projection Systems
The most common projection systems are CRT based, as they dominate the middle and low-end rear-projection system market [Wolf M., 1937]. Three monochrome tubes, each optimized for luminance and beam width of a specific primary color, are imaged onto the screen. Since the path of the electron beam is relatively short, the beam spot size can be better controlled, minimizing any smearing effects. These features are required in projection systems to produce good resolution and chromaticity with high brightness. There are two configurations for CRT projectors, using either three lenses or a single lens as shown in Figure 1.4 and Figure 1.5 respectively. The optical coupling between the tube and projection lens is enhanced by the cooling fluid placed between the tube face front and the first optical surface of the lens. Furthermore, the tube faceplate is usually curved to improve the light collection by the lens [Stupp E. H., 1999, p. 202; Malang A. W., 1989].
Convergence of CRT projection systems is a major challenge. For good image quality, it is desirable to converge the images from three tubes to within about a half pixel. Since red and blue channels are in an off-axis arrangement in the three-lens CRT projection system, the off-axis tubes will generate a trapezoidal image (keystone distortion) [Hockenbrock R., 1982]. The angular dependence of the Fresnel reflection coefficients can cause color nonuniformity, which can be reduced by tilting the red and blue lenses. Suitable deflection circuits must also be implemented to correct for these errors.
Even though single lens CRT projectors are free from convergence errors arising from trapezoidal distortion, there are many other sources degrading convergence of CRT projectors due to optical, electrical, and magnetic issues. [George J. G., 1995]. Issues specific to single lens systems include the long back focal length (bfl) due to the dichroic combiner, and the relatively high [f.sub./#] required to avoid color non-uniformity stemming from the angular sensitive dichroic filter. High [f.sub./#] systems are typically low in brightness.
CRT projection cabinets are usually bulky. There is a trade-off between the cabinet size and image quality. A shorter focal length lens decreases the optical throw distance and allows a thin cabinet. However, it increases offset angles between tubes, which results in poor image quality due to increased electron beam deflection.
1.2.3 Schlieren Optics-based Projector
Among the earliest optical configurations employed in electronic projection systems was the schlieren optics-based projector. It was originally developed for the study of defects in lenses using dark-field optics [Fischer, F., 1940; Glenn W. E., 1958; Glenn W. E., 1979; Johannes H., 1979]. Diffracted beams can be either stopped or projected onto a screen depending on whether dark-field or bright-field optics are used. The higher contrast dark-field system is shown in Figure 1.6.
The projection panels in this system are diffractive light valves specifically based on phase gratings, which produce angular separation between the modulated and unmodulated beams. Systems can operate in either reflective or transmissive mode. The phase profile for the light valve is shown in Figure 1.7. The phase profile is flat in its non-diffracting state while imparting a spatially varying phase profile in its diffracting state. The maximum diffraction efficiency of a typical square phase profile can be achieved with ([pi], 0) phase modulation.
Two high-output systems have been made using dark-field schlieren optics: the Ediophor(r) [Johannes H., 1979; William S. A., 1997] and the Talaria(r) [Glenn W. E., 1958] systems. Both are based on electron beam-written diffractive gratings in oil films. The Eidophor(r) is operated in reflective mode, while the Talaria(r) is operated in transmission. Eidophor(r) projectors use three separate modulators for the three primary colors. They are among the highest luminance projectors ever made (~9500 lumens). Talaria(r) is no longer in production and can be operated with one, two, or three light valves.
A Schlieren projection system based on LC diffractive light valves was proposed by Bos et al. [Bos P. J., 1995]. It consists of a periodic structure with alternating left- and right-handed twisted nematic (TN) LC strips (Figure 1.8). Such a structure is realized by patterning LC alignment. When the retardance ([DELTA]nd) of the LC cell satisfies the first minimum condition (see Chapter 5), the output beams from adjacent strips have the desired [pi] phase difference. Furthermore, an advantage of this light valve is polarization insensitivity, as no polarizer is required. The gray scale can be well controlled by the applied voltage. Other diffraction structures based on LC light valves have been subsequently proposed [Yang K. H., 1998; Wang B., 2002], many of which can be operated in reflective mode.
In principle, dark-field schlieren systems can deliver high-contrast images. However, the demanding requirement of defect-free optical components results in expensive optical systems that are difficult to manufacture. Disclination lines at the boundary between two adjacent strips in LC diffractive light valves also degrade system contrast and reduce light throughput.
1.2.4 Microdisplay-based Projection Systems
Microdisplay-based projection is quickly overtaking CRT-based projection in the largescreen projection TV market. In the near future, microdisplay projection will displace CRT projection due to its superb image resolution, and brightness. There are three major microdisplay technologies, based on digital micromirror devices (DMDs), high-temperature polysilicon (HTPS), and liquid crystal on silicon (LCOS) technologies. Each technology has unique properties that influence the quality of the image.
1.2.4.1 Digital Micromirror Device (DMD)
The DMD was developed by Texas Instruments Inc. (TI) [Hornbeck L. J., 1983; Sampsell J. B., 1994; Hornbeck L. J. 1996] and is based on micro-electromechanical systems (MEMS) technology. Its fabrication is compatible with integrated circuit (IC) manufacturing. It consists of an array of aluminum mirrors (one per pixel), which are suspended above individual electrically addressed SRAM (Static Random Access Memory) cells by two thin metal torsion hinges attached to posts. A small tilting yoke, address electrodes, torsion hinges, and landing electrodes are created by successive photolithographic mask steps. A square mirror is fabricated that is integral to the post formed by each via. The sacrificial layers are then removed simultaneously. Figure 1.9(a) shows a photomicrograph of a DMD mirror array and its detailed structure is illustrated in Figure 1.9(b).
The working principle of the DMD is shown in Figure 1.10. Electrostatic forces are created between the mirrors and address electrodes connected to the SRAM nodes, at which positive and negative voltages (representing 1 and 0) are applied. These forces twist the mirrors one way or the other about an axis through the torsion hinges until the yoke hits a mechanical stop. The mirror rotation angle is typically ~10_, which determines the system [f.sub./#] and ultimately system brightness. TI recently developed DMD chips that operate with a tilt angle of ~12°.
The idea of a metal mirror suspended by deformable metal hinges was initially proposed by Van Raalte in 1970 [Van Raalte J. R., 1970]. However, the original device was operated in an analog mode, where the deflection was controlled by the voltage levels on the address electrodes. In practice, it is very difficult to use analog driving schemes to produce a uniform gray scale for an entire mirror array. The DMD developed by TI is bistable and its associated circuitry is entirely digital.
The layout of a projection system based on a DMD, also called digital light processing (DLP), is shown in Figure 1.11. A total internal reflection prism is used. When the DMD is in the +[theta] state, light incident onto the mirror will be reflected into the projection lens, producing a bright state. Conversely, when the DMD is in the - [theta] state, the light is reflected away and is internally absorbed. The projection system is operated in a binary mode. In very expensive DLP projectors, there are three separate DMD chips, one each for each primary color. However, in DLP projectors under $10 000, there is only one chip where full color is created by sequential R, G, and B illumination by a color wheel. Typical color wheels consist of red, green, and blue segments, although white segments can be introduced to boost brightness.
The gray scale in DLP projection systems is generated by time multiplexing enabled by the fast (~15[mu]s) switching between the mirror ON and OFF positions. For example, a single panel system operated at a 60 Hz field rate (shown in Figure 1.12) has a color field duration of 5.56 ms. For 256 gray levels, the shortest address interval required is about 22[mu]s, which is comparable to the DMD's switching speed. Signal correction is needed to avoid errors in the low gray-scale levels due to the finite switching speed. In a real DMD, the driving scheme uses so-called bit plane weighting, which dramatically improves manufacturability [Sampsell J. B. 1994; Tew C. 1994].
There are several unique advantages of DLP projection systems. They include:
Small package size, a feature most important in the mobile presentation market. Since DLP light engines consist of a single chip rather than three LCD panels, DLP projectors tend to be compact. All of the current 3 lb (1.4 kg) mini-projectors on the market are DLP based.
High contrast ratio. TI has developed a new generation of DMD, which increases the mirror tilt angle from 10° to 12° and features an absorbing coating to the substrate under the mirrors. These improvements significantly improve the DLP system contrast. Over 1000:1 system contrast is quite common for DLP systems.
High aperture ratio. DMD operates by deflecting suspended mirrors allowing the driving circuits to reside underneath. The gap between adjacent mirrors is usually less than 1[mu]m. Aperture ratios can therefore often exceed 90%. Visible pixel boundaries leading to the so-called 'screen door effect' are barely seen in DLP systems.
Good reliability. Tests indicate that current DMD performance is not degraded after thousands of hours of operation under harsh environmental conditions [Gouglass M. R., 1995; 1996].
Polarization independence. No loss is associated with polarizing the source.
Each technology has its own weaknesses, and DLP systems are no exception. These include:
Manufacturing complexity. The CMOS electronics in the underlying silicon substrate consists of a six-transistor SRAM circuit per pixel, and additional auxiliary addressing electronics. This drives the price of high-resolution DMD chips.
Color break-up-rainbow effect. DLP systems based on a single panel DMD require a spinning color wheel to achieve full color, resulting in a visible artifact known as color break-up, or the "rainbow effect." At any given instant in time, the image on the screen is red, green, or blue, and the technology relies upon the relatively slow response of the human visual system for the perception of full color. Unfortunately, a proportion of people in the population are sensitive to color break-up, resulting in eyestrain or even headaches. TI and DLP vendors have made progress to address this issue, such as increasing field rates from 180 Hz (1×) to 360 Hz (2×). Today, many DLP projectors being built for the home theater market incorporate a six-segment color wheel, which has two sequences of red, green, and blue, and spins at 120 Hz. Since R, G, and B are refreshed twice in every rotation, the industry refers to this as a 4× rotation speed. This further doubling of the refresh rate has again reduced the number of people who can detect color artifacts, but nevertheless it remains a problem for a number of viewers even today.
Temporal artifacts. Even in static images the binary nature of the DMD creates the sensation of temporal modulation. In fast-moving video images that cause the eye to move rapidly, object edges can become temporally unstable and appear fuzzy. Improvements to the addressing algorithms have reduced this effect, but it can still be the addressing algorithms have reduced this effect, but it can still be perceived under certain viewing conditions. Low gray-scale contouring also results from the binary addressing.
(Continues...)
Excerpted from Polarization Engineering for LCD Projection by Michael Robinson Copyright © 2005 by John Wiley & Sons, Ltd. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
