LCD Projector Teardowns
Sharp LCD Projector model XG-E1200U - 2006-06-08
One of my most favorite teardowns ever! I went into it, knowing nothing about what would be inside a typical LCD projector. This particular unit had previously adorned the ceiling of the university EE computer lab. It had an annoying defect where multiple columns of pixels (regularly spaced, probably every 8th or 16th column) were constantly blue regardless of video input. When I took over as Electrical Lab Supervisor, I was thrilled to replace it with a newer model so I could take apart this old clunker.
I only took a handful of photos, beginning with the spread of all the critical components shown on the work bench. To get the metal halide lamp working, I had to connect all major circuit boards together as they normally would be, then power on the whole system with the front power button or remote control. I wanted to get the lamp's power supply (blue colored board near top right of photo) to work independently of the video input/processing boards, but I was unable to isolate the control signals that were used to keep the lamp turned on. It was difficult because the power supply board powered the entire system, not just the lamp. No doubt, there were multiple power-on/shut-down signals integrated into the system, that would shut down the power supply if any one subsystem was malfunctioning or disconnected.
The best part of this teardown was the beautiful colors created by the optics. All the "magic" happens with the dichroic filters, with one filter splitting white light into red and cyan, and another filter splitting cyan light into green and blue. You can see that final mirrors in the red and blue light paths are also dichroic, passing any leftover cyan or green light, respectively, that manages to get through the first set of filters, and letting the dark plastic chassis absorb the unwanted colors.
A final bit of "magic" happens once more with the multi-chroic prism. Three tiny LCD screens (seen at bottom right corner of bench-top photo) would normally be mounted within the red/green/blue light paths, immediately surrounding three sides of the prism. The green image shines straight though the center of the prism, while red and blue images are reflected off the internal surfaces. All three images are recombined and exit the fourth side of the prism before entering the main lens assembly.
I never bothered to figure out what caused the vertical blue lines in the projected images. After the teardown, it was apparent that the most likely cause would have been a problem with the video processing board's blue-image LCD transmitter, or the LCD driver IC integrated into the flat-flex/LCD, or a connection between them.
Panasonic LCD Projector TV model PT-43LC14 - 2017, May-June
My grandmother wanted to get rid of this old internal projection TV that had seen better days. I was happy to take it off her hands so I could try to relive the fun that I experienced while taking apart the Sharp LCD projector 11 years prior. This time, I would be armed with a much better digital camera, and increased knowledge/experience that I could use to finally get a metal halide lamp power supply working independently of any pesky control signals from the video processing board.
Most of the TV's body is empty space. There's a triangular speaker box on either side, and a large trapezoidal mirror to reflect the image onto the projector screen. All the good stuff is crammed in the base of the TV.
After dismantling the bulk of the TV in the shack, I took the primary projection system to my apartment work space to see if it still worked. At first it did not work because while trying to get the projection system out of the TV body, I damaged the cable that carried video signals from the main processing board to the LCD driver board. Once repaired, I was successfully able to get a setup image projected onto the ceiling. However, I was unsuccessful in getting other images displayed via VGA and HDMI from my laptop. It might have still worked with the RF or Composite that my grandmother was using, but I couldn't be bothered to move the whole thing over to my apartment living room.
This photo below shows the arrangement of red, green, and blue image-component LCDs arranged around three sides of the multi-chroic prism.
I was amazed that the blue setup image was so clear and focused on the ceiling (or at any other distance, as I later discovered) without any focusing required. In hindsight, it makes sense for the optics to be designed this way so the TV could be manufactured without any mechanical adjustments needed for clear images on the screen.
With external lighting set up just right (in this case, a couple of Ikea LED lamps), it was remarkable how the fish-eye lens looked just like a blue-colored HAL 9000 from 2001: A Space Odyssey.
Given the highly modular nature of this projector's construction, the metal halide lamp had its own dedicated power supply. This is no ordinary SMPS. It has some microcontroller smarts built into it to ignite and drive the lamp at optimal voltages and currents.
There is one cable that carries power-on/shut-down signals between the main processing board and the lamp power supply. With some cable splicing, braeadboarding, and DMM probing, I was able get the lamp to stay on indefinitely. Previously, in normal operation, simply disconnecting the video transmission cable would cause the lamp to turn off. With those annoying safety systems bypassed (including a thermistor and thermal fuse that were mounted on the lamp housing), I was able to take some spectacular photos of the optics.
The optics arrangement is similar to that from the Sharp projector, dismantles 11 years ago. However, with this Panasonic projector, the red color seems to be more like orange. This may be by design, or more likely, the lamp's color spectrum shifted toward blue due to its accumulated usage-time. I'm not sure if such a thing happens with aging metal halide lamps in general, but it's a theory.
What would happen if I turned the optics around and injected light into the fish-eye lens? I started by removing the prism/lens assembly and shining the Ikea lamps directly into the lens. The prism takes in the white light and splits it into red, green, and blue images projected onto the tiny LCD modules. Even an image of my work bench setup could be projected onto the LCDs with remarkable clarity.
I removed the LCDs to see if an image could still be seen without them. Indeed, it was possible, and with an elaborate setup, I was able to take some green-hued self-portraits (since I was just using the green side of the prism). I enjoy the resulting image because it looks like a selfie with an accumulation of Instagram filters, when in fact, there was no editing whatsoever. What you see in the photo shown here is exactly what the camera recorded, with the room and my face illuminated by white light. Some slight distortion can be seen on my face (nose larger than usual) due to the fish-eye lens.
Something else was revealed from the LCDs being removed from the prism assembly: 2-dimensional diffraction gratings. By placing the LCD directly over the camera lens, I could get spectacular colors to come out the LED lamp, especially when I overlapped two LCDs simultaneously, or moved the camera on the tripod during a time exposure.
One final thing I wanted to do with the optics was to duplicate the classic physics demonstration of the three primary colors of light - red, green, and blue - all being combined and partially overlapping to create other colors - yellow, cyan, magenta, and white. I made some cardboard-and-electrical-tape filters with carefully located holes, and inserted them in place of the LCDs surrounding the prism.
The image was projected onto a piece of paper laid above the lens just as clearly as it was projected onto the ceiling. I was extremely pleased to create such a display because the last time I had done anything like this was for my 6th grade science fair project in 1993. In that case, I used red, green, and blue 100W flood lamps mounted inside a large box, and projected the colors onto a sheet of white cardboard a few feet away.
Here's two photos of the lamp and power supply setup I had been using. On the right side of the first image, there is the metal halide lamp power supply box stacked on top of the system power supply. The system power supply generates 3.3V, 5V, 12V, and audio amplifier voltages, but what I really needed was the 17V and 19V outputs that go to the microcontroller circuits in the lamp's power supply. The second image shows the multiple connections I had to make to fool the power supply into thinking it was connected to a fully functional television system. The blue LED is there to light up whenever there is a critical fault condition that would normally be signaled from the power supply to the video processing board.
In the end, I packed everything up for further experimenting another day. Some of the things I would like to do are:
Modify the power supply to reduce its size/complexity and make it battery-powered so I can have a bad-ass, blindingly bright flash light.
Use the multi-chroic prism to create a white laser by combining red, green, and blue laser light shining into its appropriate sides.
Create an artistic mobile of dichroic filters (from this and other LCD projectors) hanging on strings, and mounted in front a sunny window.
Create a replica HAL-9000 interface, with a red light shining through the fish-eye lens.
This page last updated: 2017-07-23
