Compiled by Steve Linke
Last updated 6/8/2010

Sony rear-projection liquid crystal microdisplay technology overview

Light produced by a high-intensity projection lamp enters the optical block and is split into three paths (one each for red, green, and blue) by a series of lenses, condensers, mirrors, filters, and/or beam splitters. Each color has an LCD assembly with polarizing filters and a liquid crystal microdisplay panel (approximately 1" across). On Sony's "3LCD" models, the light is transmitted through each LCD panel, and, on their "SXRD" (Silicon X-tal Reflective Display) models, the light is reflected off each SXRD panel. Finally, the three separately colored images are combined with a special prism into the final full-color image, which is then projected/enlarged onto the back of the screen by a projection lens.

For a more detailed explanation of 3LCD-based optical block technology, design, and failure, see below. For additional technical details about SXRD optical block technology from Sony, see this SXRD white paper, which compares conventional transmissive LCD and digital light processing (DLP) with SXRD technology. Also, see this HowStuffWorks article.

The following annotated photo shows various parts on the optical block of a 2004 3LCD model KDF-55WF655 (note that the LCD panel cover has been removed):

The light source for the images projected onto the screen is a high-intensity mercury-vapor arc lamp (100, 120, 132, 180, or 200 watts, depending on the model). The arc lamp is enclosed in a larger, clear glass outer bulb which is within a black plastic housing. The outer bulb includes a parabolic reflector that directs semi-collimated light (parallel rays) into the optical block. Mercury-vapor arc lamps produce a great deal of blue-violet and ultraviolet (UV) light as they warm up, but then they shift to a wider-spectrum with red, green, and blue-violet (white) visible light, as well as UV, peaks as they reach their normal operating temperature.

Pictured below is the 132-watt XL-2200 lamp used in several 2004 and 2005 3LCD models, including the KDF-55WF655 and KDF-60WF655, but other lamps are similar:

The electromagnetic light spectrum is shown below:

The photo below shows the internal parts of a 2004 3LCD model KDF-60WF655 optical block (structurally identical to the KDF-55WF655) with the LCD panel/prism block assembly removed from the center (courtesy of AVS Forum member "tdma"):

Below is a version of the above photo annotated to show all of the parts (cyan text and arrows) and the light path (large color-coded arrows). See the text following the photo for an explanation of optical block function:

The mercury-vapor arc lamp assembly (not shown) would be in the lower-left corner. It shines light (all wavelengths, including infrared, visible, and ultraviolet) into the optical block. The outer glass bulb on the lamp assembly likely filters some of the ultraviolet energy prior to the light entering the optical block. After entering the optical block, the "white" light is directed through two "fly's eye arrays" (objective and field), which create a more uniform brightness of light across the beam. After passing through a condenser lens, which concentrates the light, and bouncing off a mirror, the blue light path is split off by the blue dichroic mirror.

 

The "blue" light, which includes the ultraviolet (UV) portion of the light spectrum, then passes through a UV absorption filter that removes some of the UV light. Finally, the blue light bounces off another mirror and through the blue channel condenser lens, an orange-colored blue input polarizer, and the blue LCD panel. The orange-colored blue input polarizer is the final step to produce pure blue polarized light. The non-blue and improperly polarized light, including high-energy UV, is absorbed by this filter, turning it into heat. This leads to degradation in the absence of sufficient cooling.

Meanwhile, the "yellow" light that passes straight through the blue dichroic mirror next hits the green dichroic mirror, which splits off the "green" light and redirects it through the green channel condenser lens, the yellow-colored "green input polarizing filter", and the green LCD panel. The remaining "red" light travels through a couple of relay lenses and bounces off a couple of additional mirrors, redirecting it through the red channel condenser lens, the blue-colored "red input polarizing filter", and the red LCD panel.

Below are two photos with the input polarizers in place, but with the LCD panel/prism block removed. This photo shows the orange-colored input polarizer for the blue light path on the left, and the yellow-colored input polarizer for the green light path to the right. Note the open slots beneath each of the input polarizers, and the additional slots closer to the center of the cavity that line up with the LCD panels, used for air cooling:

This photo shows the yellow-colored input polarizer for the green light path on the left, and the blue-colored input polarizer for the red light path to the right:

The following annotated photo shows the LCD panel/prism block, which inserts into the cavity bounded by the input polarizers shown in the above photos. The prism area is indicated by a dashed teal square. Note that the LCD panels and output polarizers are secured to the prism with a gap of several millimeters, and that there is also a gap between the input polarizers and the respective LCD panels:

The LCD panels block the appropriate amount of light in each of the three separate color channels, and the remaining light goes through the output polarizers (one for each color that is similar to the respective input polarizers) attached between the LCD panels and the prism block (not shown). Finally, the three colors are recombined with the prism and projected through the projection lens off a large mirror behind the display screen, and then onto the display screen. The LCD microdisplay panels measure 0.61", 0.73", 0.74", 0.87", or 1.35" across, depending on the model.

Note that, on Sony's 3LCD models, the light is transmitted through each LCD panel as shown here. In contrast, on their SXRD models, the light is reflected off the panels, so the interior structure of the optical block, as well as some of the parts are different than shown here.

Parts in the blue light path, particularly the orange-colored input polarizer and the LCD panel itself, appear to become degraded more rapidly than those in the other light paths for reasons described later. Below are some examples. When degradation of parts in the blue light path leads to insufficient/improper polarization of the light (e.g., breakdown of polarizing layers in the filters or panel), stray blue light makes it through the liquid crystal panel and gets projected on the screen, giving rise to blue discolorations. When degradation of parts in the blue light path leads to improper blocking of light (e.g., scorching of the filters or panel), light from the green and red channels predominates to form a yellowish color in areas that should be white or lightly colored.

Degraded blue input polarizer from my 2004 3LCD model KDF-55WF655 (Steve Linke). Note the change in appearance from orange to cloudy white in a rectangular shape (the shape of the incoming beam of light), as well as the squarish oval in the middle:

Degraded blue input polarizer (slightly out-of-focus) from a 2004 3LCD model KDF-60WF655 (courtesy of "tdma" of AVS Forum). Note the similarity in the degradation pattern to the closely related model above:

Degraded blue input polarizer from a 2004 3LCD model KDF-55WE655 (courtesy of Howard De Leon by personal email):

Degraded blue input polarizer from a 2005 3LCD model KDF-E42A10 (courtesy "FreeKerXLX" of AVS Forum):

Degraded blue input polarizer from a KF-E42A10 (Brazil) (courtesy Wagner Rodrigues). Mr. Rodrigues apparently successfully replaced the filter with a standard polarizing filter for a camera.


Degraded blue LCD panel from a 2004 3LCD model KDF-60WF655 (courtesy of "tdma" of AVS Forum):

Degraded blue LCD panel from a 2004 3LCD model KDF-55XS955 (courtesy of Bob Scott):

Ultraviolet and deep blue light from the lamp

High-pressure mercury vapor lamps produce visible light with peaks in the red, green, and blue-violet ranges. However, they also produce large amounts of high-energy ultraviolet (UV) light, which gets filtered into the blue light path along with the blue-violet part of the spectrum. While some UV light likely is filtered out by the outer lamp bulb and UV filter(s) in the blue light path, a significant amount apparently remains unfiltered, as more effective UV filters remove too much of the visible blue-violet light. The polarizers and LCD panels are subject to degradation during prolonged exposure to UV and deep blue-violet visible light, at least in part due to the presence of organic dyes. Thus, the orange-colored input polarizer and the blue LCD panel tend to be the first to exhibit degradation. The blue output polarizer may also be subject to this damage, but to a far lesser degree.

From "Polarization Engineering for LCD Projection" by M. G. Robinson, J. Chen and G. D. Sharp (© 2005 John Wiley & Sons, Ltd):

 

The disadvantages of [3LCD] projectors include...Potential lifetime issues. [Liquid crystal] alignment layers used in [3LCD] light valves are organic polyimides (PIs). PI is susceptible to UV and deep blue light photochemical damage, which reduces operation lifetimes. UV filters with a long-wavelength cut-off are helpful, but tend to reduce the blue content of the final color imagery. The industry trend is toward UV filters with 50% transmission points between 430 and 435 nm...

 

Demerits [of SXRD] include...Lifetime. Long-term reliability of [SXRD] systems, as with [3LCD], is still a concern...Complexity. An [SXRD] optical system is more complex than either DLP or [3LCD] systems. Accurate control of the state of polarization is key to making [SXRD] products commercially competitive...

Visible light from the lamp converted to heat

As the intense visible light is filtered by the polarizers and LCD panels, the blocked light can be converted to heat. This conversion is likely most intense in the high-energy blue light path.

Direct heat from the lamp

The high-intensity projection lamp is directly adjacent to the optical block, and it creates a lot of direct heat. The filters and LCD panels may also be subject to degradation from this direct heat.

The following photo shows the inside case of my 2004 3LCD model KDF-55WF655 under the display area. The optical block inserts to the left, and the lamp portion sits on the right. The metal area by "lamp access" is where the lamp assembly can be accessed from the front of the TV for lamp replacement. Note the scorch marks on the plastic case where the lamp shines through its grid-shaped housing, which likely arises from intense heat and/or exposure to UV light escaping from the lamp.

As further evidence of the heat level generated by the mercury-vapor arc lamps, the WE610, WE620, WE655, and XBR950 models (2003-2004 3LCDs) have a documented issue in which parts adjacent to the lamps (100 or 120 watts, depending on the model) become warped, cracked, and/or scorched. Sony has extended the warranty on these models for this type of damage. The following photo is an example of the interior of a KF-42WE610 adjacent to the lamp (photo courtesy of AVS Forum member "Dayclone"):

Dust

The optical blocks are not sealed--at least in the 2003-2004 3LCD models. There is an air intake filter leading to the fan on the bottom of the optical block that cools the parts by directing air up through slots, but the filter is constructed of a material that can only exclude very large particles. In many cases, this fan becomes clogged with dust, which it then also circulates throughout the optical block. In addition, the top of the optical block is exposed with only an angled piece of plastic positioned over the LCD panels with open sides. Thus, dust can accumulate on the internal parts, which can further reduce cooling efficiency, as well as directly cause visual anomalies.

On the 3LCD models, there is a small centrifugal "lamp cooling fan" mounted to the top-left of the optical block (as you face the TV). This fan draws unfiltered air from inside the TV and blows it directly across the face of the lamp from top to bottom through perforations in the lamp housing. As the air comes out the bottom, it is directed out the back by another larger fan near the back of the TV. Interestingly, the air input to the fan is not far from the path of the outgoing air heated by the lamp, which could affect cooling efficiency.

In addition, there is a somewhat larger centrifugal "optical block cooling fan" mounted inside the optical block on the bottom-right (as you face the TV). This fan draws air from the void behind the projection screen through a crude filter and blows it up through multiple small ports beneath various parts in the light path--the main condenser lens/fly's eye array near the beginning and the polarizers and LCD panels for each of the three color paths. The heated air presumably exits through the top of the optical block and out the sides of the open angled cover. There is an extra port above the condenser lens/fly's eye array that directs air from these parts into the area of the angled cover.

Picture quality longevity studies commissioned by DLP competitor Texas Instruments have been conducted on liquid crystal projectors and TVs at the Rochester Institute of Technology's Munsell Color Science Laboratory and an independent testing organization called Intertek. Studies on 3LCD models were conducted on front projectors, and the manufacturer(s) of the tested models were not revealed, but they are likely from Epson and/or Sony, the two main members of the 3LCD Group. Studies on SXRD^TM models were conducted directly on Sony TVs.

The Munsell Color Science Laboratory conducted a preliminary study on five 3LCD projectors. The study commenced in May of 2002 on models that were popular at the time, and the results were reported in March of 2003. All five projectors showed early degradation of the blue channel (within 2,500 hours of usage), manifesting as blue or yellow discolorations, with eventual signs of degradation in the red and green channels. A theory was presented that organic compounds in the blue polarizers and LCD panels degraded during exposure to high frequency blue and UV light, causing the projectors to be unsuitable for use long before the end of their expected life spans.

The following slide from the preliminary 3LCD projector study shows examples of the blue and yellow discolorations developing over time (left to right):

This article on Projector Central (June, 2003) presents additional details of this preliminary study, including the following statement:

Manufacturers recognize that the organic compounds in LCD panels and polarizers are susceptible to high heat and light energy stress, and will eventually break down if deployed in high stress environments...

This article on ExtremeTech (April, 2003) also presents additional details of the preliminary study. They interviewed a representative named Bob Getner from NEC, another 3LCD projector manufacturer, and described his comments as follows:

[Getner] claimed LCD projectors can be made for long life use, especially when designers focus on proper cooling. NEC designs its projectors to keep the insides as cool as possible to avoid detrimental effects. Getner was aware of the blue LCD problem and said NEC's blue panels have a U/V filter to help protect against harmful light energy. Getner claims LCD projector design and lifespan is a function of engineering for expected usages.

A Texas Instruments "white paper" published in SPIE Proceedings Vol. 4980 (2003) also discusses the preliminary study (see Section 6.2 - Picture Reliability on pages 8-9).

Although the study was done on front projectors, the results were very consistent with observations of LCD panel and polarizer degradation in Sony's 3LCD TVs correlating with blue and yellow discolorations, except that the longevity of the TVs tends to be modestly higher.

Follow-up 3LCD projector study (Intertek)

The preliminary 3LCD study described above was criticized for using an unusually harsh duty cycle (continuous operation),  testing too few units (five), and not properly controlling environmental temperatures (77-85 degrees Fahrenheit ambient). So, Intertek conducted a follow-up study in 2005 to address these criticisms by testing three different duty cycles (continuous, 5.5 hr on/2.5 hr off, and 1.5 hr on/2.5 hr off) in triplicate on three LCD models (27 total units) at a constant ambient temperature of 75 degrees Fahrenheit.

The results of this study through December 2, 2005 are published in another Texas Instruments "white paper". In particular, see the Summary on page 6 and the Conclusion on page 61. Similar to the preliminary 3LCD projector study (above), initial signs of degradation (blue and/or yellow discolorations) arose between about 1,700 and 3,000 hours, with significant deterioration between about 2,000 and 4,000 hours in all units that reached those usage levels. The following statement appears in the Conclusion:


The following images are extracted from the white paper on the follow-up 3LCD study, showing examples of blue and yellow discolorations developing over time (left to right), as well as degradation of the LCD panels and polarizers over time (click on the images for larger versions or view the original white paper):

Projector Central also has an article that included references to this follow-up study. In particular, see the "Weaknesses and Limitations of LCD" section, which includes the following:


Again, although this follow-up study was done on front projectors, the results were very consistent with observations of LCD panel and polarizer degradation in Sony's 3LCD TVs correlating with blue and yellow discolorations, except that the longevity of the TVs tends to be modestly higher.

Sony SXRD^TM rear-projection TV study (Intertek)

Intertek also conducted a longevity study between March and October of 2006 on seven Sony KDS-R50XBR1 SXRD rear-projection models purchased on the open market. They used three different duty cycles: one was left on continuously, three were 9.5 hr on/2.5 hr off, and three were 3.5 hr on/2.5 hr off. The "Results and General Summary" contained the following statement:


Similar to the 3LCD models, Intertek determined that the yellowing of the displayed image arose due to a problem with the blue channel. These observations were all consistent with problems customers were experiencing during this same time period, right down to the location on the right side of the screen.

Interestingly, these observations also were later confirmed by Sony during discovery for the XBR1 class action lawsuit. In the judge's Opinion and Order approving the settlement of that case, it was revealed that Sony acknowledged the yellow stain issue and claimed that it was caused by a "microscopic material" in the liquid crystal panels, disrupting their uniformity over time during prolonged exposure to UV light produced by the projection lamp. This was consistent with the problem in the blue channel observed by Intertek.

Sony also acknowledged that the issue typically started at the edge of the screen and enlarged over time, exactly as observed by Intertek. Sony claimed that the extent of the discoloration depended on the amount of microscopic material present in the panel, which varied from TV to TV, and the frequency of usage by the consumer. They also claimed that service records indicated that the issue always appeared within the first 3,000 hours of usage, if it was going to happen. They went on to claim that, between 1/2006 and 10/2006, they worked to reduce both the amount of the microscopic material and the amount of UV light exposure, and that, by 10/2006, they were producing non-defective optical blocks for the 2005 SXRD TVs.

Sony threat of legal action

Mr. Ronald Wasinger, a Sony Vice-President in their Law Department, wrote a letter to Texas Instruments at the conclusion of the Intertek study (in 10/2006) threatening them with legal action if they distributed the Intertek SXRD^TM study to the press or for marketing. Ironically, Mr. Wasinger suggested that the Study contained false and misleading claims in violation of federal and state deceptive trade practices and consumer protection laws. Here is an excerpt from that letter:


In his letter, Mr. Wasinger also made reference to the earlier 3LCD projector studies:


In reality, though, it appears that all of the conclusions reached by Intertek in their Texas Instruments-sponsored studies have turned out to be true. And it is particularly disturbing that Mr. Wasinger would write a letter denying problems with yellow discolorations and threatening legal action at a time (10/2006) when Sony, by their own later admission during the XBR1 class action litigation, had allegedly just completed re-engineering of the XBR1 optical blocks in hopes that it would solve the yellow stain issue. In fact, Mr. Wasinger himself was involved in the XBR1 litigation. Of course, it is also disappointing that whatever re-engineering was done did not effectively resolve the discoloration issue, as it continues to plague XBR1 owners, as well as subsequent XBR2, A2000, A2020, and A3000 owners.

It has been noted elsewhere that Sony is a Silver Sponsor of the San Diego and Imperial Counties chapter of the Better Business Bureau (BBB), and that Mr. Wasinger is on the Board of Directors that governs it. However, Mr. Wasinger is only 1 of 25 Directors on that Board and would presumably recuse himself from any cases involving Sony. In the absence of any hard evidence to the contrary, I assume the BBB can provide fair and impartial mediation on matters involving Sony. However, they are not necessarily set up to apply complex merchantaiblity laws to individual cases, and they tend to accept any discount offer from Sony as fair compensation and consider the cases closed, so BBB mediation is not necessarily the best option.

Sony patents for enhanced optical block cooling methods

Sony has a number of patents from the late 1990s and early 2000s that contain language making it clear that Sony inventors were well aware of the problems with intense heat, light, and dust in optical blocks leading to deterioration of the parts within. The inventions in these patents focus on methods to properly cool, prevent dust accumulation, and inhibit ultraviolet-mediated damage in optical blocks. For example, the patents propose the use of additional heat-dissipating glass plates, metal radiation absorbers, heat sinks (e.g., heat tubes), liquid-based cooling systems, and electronic cooling devices (e.g., Peltier devices), as well as the use of contiguous connections  between parts in the light path with heat-conducting glass to both reduce the effects of dust and improve cooling. However, these cooling methods were not implemented in Sony's production optical blocks. Below are some sample excerpts from some of the patents.

US Patent 6,132,049 (priority date 9/16/1997): "Picture display apparatus and cooling apparatus for optical apparatus" (the liquid crystal panel and polarizer in the blue light path are particularly susceptible to damage when the high-energy blue light is converted to heat).


US Patent 5,757,443 (priority date 10/13/1995): "Transmission-type display device with a heat-dissipating glass plate external to at least one liquid crystal substrate" (heat radiating from the light source and intense light cause the liquid crystal in the liquid crystal panels to overheat and deteriorate).


US Patent 6,057,894 (priority date 10/15/1997): "Projection type liquid crystal display having a dichroic prism and polarizing and phase shifting properties" (dust and dirt cause a deterioration of image quality).


US Patent 7,123,334 (priority date 12/4/2001): "Liquid crystal display device and liquid crystal projector device" (organic compounds in optical block parts, such as the liquid crystal panels and polarizing plates exhibit deterioration of reliability due to exposure to intense light and heat).


US Patent 7,535,543 (priority date 12/15/2004): "Liquid crystal display apparatus and cooling device" (light absorbed by optical block parts, such as the liquid crystal panels, heats them up, leading to deteriorated performance and shortened lifetime).


US Patent 7,666,325 (priority date 9/29/2005): "Liquid crystal orientation layer and liquid crystal display element" (ultraviolet light deteriorates the orientation layers in liquid crystal panels).

Additional pictures of blue visual anomalies

(pictures of blue anomalies from my KDF-55WF655 in July of 2007--original optical block)

The following photographs were taken with a Sony CyberShot camera of my KDF-55WF655 TV on the DVD input using a wide-screen DVD with a solid 75% black/25% white background. At the time, my optical block had been in use for ~28 months, and the service menu indicated that it had been operating for ~6,000 hours.

A zoomed version of this first photo was used as an example in Sony's warranty extension announcement (see the upper-right photo in the collage of four above). Note that the blue haze is largely restricted to the 4:3 viewing area, even though the current source is wide-screen, suggesting some sort of burn-in.

In the next picture, a 4:3 program that was letter-boxed was being viewed for quite awhile. The blue line appeared at the lower border of the letterbox. It remained there when the aspect was changed to the zoom setting and when the wide-screen DVD was played. Over time, the blue line diffused away, but the general blue haze remained.

In the following photo, the blue color has turned into a number of random squiggly lines that resemble areas of high contrast in a paused frame of a TV program recorded on a digital video recorder. It remained there when the aspect was changed to the zoom setting and when the wide-screen DVD was played. Over time, the intense blue lines diffused away, but the general blue haze remained.

Many other posted pictures of the problem show an uneven hazy border of blue at the bottom or other edges of the screen. Others show bright blue blobs that tend to be in the corners. See the eCoustics Sony Projection LCD TV Problems message board thread for many more examples.

Optical block pictures

(pictures of the optical block from my KDF-55WF655, or a KDF-55XS955 courtesy of Bob Scott when noted)

Picture 1. "Front" view of the entire optical block. The optical block is in the center bottom of the TV case near the front. If you could see through the front of your TV, this is what you would see (the lamp housing is accessible through a door on the front of the TV). There is a centrifugal fan under the "fan housing" on the upper left that has a port that leads to the lamp area in the front-left. There is another centrifugal fan on the lower right (behind and below the cylindrical plastic protrusion with the white air filter around its perimeter that is just to the right of the projection lens). This fan appears to blow air up through LCD panel and lens/mirror areas. There is a black metal "LCD panel cover" that covers the innards, but it is open on the sides, so dust can move freely into it.

Picture 2. Left view of the optical block. The open side of the black "LCD panel cover" can be seen here, which opens near the "LCD fan port". The other side of the LCD panel cover is also open.

Picture 3. Top view of optical block with the LCD panel cover removed. The ribbon cables leading to the three LCD panels (blue, green, and red) are now visible.

Picture 4. Close-up view of the LCD innards.

Picture 5. Extreme close-up of LCD innards. Note the lettering on the ribbon cables leading to the LCDs. The third one is labeled "RL".

Picture 6 (KDF-55XS955). Optical block with the lamp housing (on the left), the LCD panel cover and housing (on top), and the LCD panel/prism/projection lens/air filter module (on the right) removed. Note that the centrifugal fan on the lower-right is now visible.

Picture 7 (KDF-55XS955). LCD panel/prism/projection lens/air filter module that was removed from the optical block (see Picture 6). Note that the blue LCD panel has been removed from its holder.

Picture 9. Fan and lamp housing removed from the optical block (top view of left side). The centrifugal fan pulls air from beneath the circular opening and directs it onto the lamp through the port built into its housing. The lamp would be on the right in the metal clips.

Picture 10. Fan removed from optical block. Note the dust build-up in the fins. There were also dust balls accumulated under the lamp housing. Presumably, the fan blows out the "lamp fan port" built into the housing directly onto the lamp to cool it. There are perforations at the top and bottom of the lamp itself that are aligned with the port, presumably to enhance cooling. The other centrifugal fan on the other side of the optical block presumably cools the LCD panel and lens areas by blowing air up through them. Also note that there are two other, larger fans in at least some models. On the KDF-55WF655, one is oriented horizontally straight back from the lamp, directing air out the back of the TV. This fan presumably clears hot air coming off the lamp. The other fan is on the far right as you face the TV, oriented vertically over the chassis unit that contains all of the audio and video connections (left side when facing the back). The air seems to be deflected off the case and out the back of the TV by this fan.

Picture 11. LCD fan port from the LCD side. This shows the "LCD fan port" opening on the side of the housing that contains the LCD innards. This opening leads to a lens area adjacent to the lamp. This may serve as an outlet for cooling air from the optical block fan that blows up from underneath the lenses. It empties into the same area (under the LCD panel cover) as the other cooling air that blows up across the LCD panels and filters. The effectiveness of this cooling system is questionable. It is also a concern that the whole thing is open to dust.

---

© Copyright 2009-2011 by Steven P. Linke. All rights reserved.