What To Do With Old LCD Screens: Hack Your Own Electrochromatic Glass

There’s something decidedly science fiction-like about electrochromatic glass. A wave of a hand or a voice command and the window goes dark (or goes transparent). You can get glass like this today or you can add (pricey) film to existing glass, if you prefer.

[Artem Litvinovich] thought about using LCDs as window panes twenty years ago, but the cost was, of course, prohibitive. He recently realized that he had easy access to LCDs out of broken laptops and decided to see if it would be useful as a small window.

After tearing down the screen (and there are a lot of pictures of the tear down), [Artem] needed to find the bias voltage pin on the LCD connector. He grounded a 12V battery pack and put a 10K resistor on the positive lead. Then he touched the pins until he got a slight change in the panel’s opacity.

Once he identified the right pin, he removed the other conductors to prevent current draw from the remaining circuitry. It also prevents the onboard drivers from trying to control the panel and interfering with the externally-applied bias voltage.

The electronic curtains don’t draw much power, but they also don’t go completely transparent (you can see a demo on the video below). Still, an interesting hack and one maybe someone can improve upon. This looks like a fairly simple hack compared to, say, driving an LCD with an FPGA. Certainly, finding the bias line is easier than reverse engineering the entire display.

47 thoughts on “What To Do With Old LCD Screens: Hack Your Own Electrochromatic Glass

  1. First thought would be for a clear church window, place the LCD into it, show the signs of the cross,or stained glass images. But I guess if people see things moving or animated on a church window, its a message from above, no adverts here. Good demo.

        1. While they are basically 1 big pixel, they have a lot more than 2 states. My welding mask is adjustable from a shade #4 at light state, all the way to a #13 dark state depending on what you’re welding, or if you’re grinding.

  2. I learned recently that applying a DC voltage to a LCD panel will cause the transparent indium tin oxide (ITO) electrode material to migrate, eventually causing panel failure. To prevent this, LCD panels have to be driven from AC, with positive and negative cycles exactly equal in voltage and time. The ITO still migrates, but it does so equally in both directions, and no net change results.

    So if you’re going to use this long-term, better drive it with a transformer or other AC source.

      1. Note that the link provided by Elliot Williams seems to indicate the opposite, namely that the transparent ITO gets covered by the chemicals of the LCD.
        ITO seems to be sensitive to moisture, and somewhat to heat, but I don’t seem to be able to find any issue with it migrating as you suggest.

        The end results are the same though, you need to avoid too much DC as you said.

  3. “don’t go completely transparent” is an understatement. For a mono display you lose at least 50% through the polarisers,and for colour, another at least 60-70% through the RGB filters, so maybe 30% transmission if you’re lucky

  4. I’ll reiterate my comment here: When driving into a low on the horizon sun (below the visor, and blindingly in front of me) I want a system tracking my eyes and the Sun and then drawing a “shadow” over only the sun so I can see the road!

    1. This. Plus dimming oncoming “high” or misaligned headlights would be awesome too. I have often thought about how to do this but could never come up with an economical way to do it. Plus, anything blocking the driver’s view is generally considered illegal in the US.

      1. I had the same idea independently. If the LCD screen’s unpowered state is “opaque”, it would completely block the driver’s view if power to the screen is lost for any reason.

        Under the same condition (unpowered is “opaque”), an old TV screen would make an excellent sunshade! As soon as you turn off the car, sunlight won’t come through the windshield.

  5. I have an ongoing project using the rescued layers from a set of LCD screens with two possible applications, both replacing the LCDs (which were broken to begin with) with glass or plexi and likely redoing the array of LED’s:
    a) Lightbox for design work.
    b) Lighting fixture using white LED’s and IR ones (security application idea).

  6. @Al Williams – I wonder how fast the opaque/transparent transition rate can be. I envisioned modulating a light beam with this (VLC or Visual Light Communications). However, I wanted to do so with a a high-wattage spotlight like a 18-million Candlepower Q-Beam Spotlight. Of course you could simply put a high-power TRIAC in line to the Q-Beam’s bulb and just drive (gate 12vdc) it directly with audio, Morse Code, or PCM (Pulse code modulation) digital data. But that would fatigue the bulb with too many off/on transitions.This method could put in front of the beam by a few feet to reduce overheating the LCD glass. The low-voltage AC signal could be applied to the glass causing it to interrupt the beam of light without damaging the spotlight’s bulb. You could use a low-power TRIAC (TRIAC’s are meant for gating low–to-moderate-voltage AC. You can also gate DC with a TRIAC. SCR’s latch and need to be reset. TRIAC’s don’t. But both can handle higher wattage).

    Alexander Graham Bell (et al) did something like this with an improvement to his sunlight-based Photophone in the 19th century using a Kerr Cell to modulate artificial light (i.e. lanterns). However, Kerr Cells require very high voltages to operate. This idea only requires less than 12 volts AC to operate. But can it handle high-speeds like voice and digital? Or can it only do slow-speed Morse Code?

    By putting a solar cell and audio amplifier with a telescope or Fresnel lens miles away from this gadget you could actually communicate over light with it. With a Q-Beam spotlight you could do cloud bounce communications over the tops of buildings and mountains or maybe even over the horizon (maybe not). However. Morse Code would be visible to everybody as a flashing/blinking light. But a audio or digital one would not be visible to the naked eye. They would only see a constant spotlight and not know it contained modulated-light media content.



    1. triacs do latch, they turn off during zero crossing so you would not be able to modulate anything faster than whatever frequency your power source is. and modulating hid bulbs would be very bad for the bulb. once you extinguish the arc you need to re-ignite it, and the hot start ignition voltage is quite a few thousand volts higher than the cold start ignition. then you would also be thermal cycling it, and since they run so hot you would probably still be emitting light for just a small bit during the cooldown. as for modulating the light beam with an lcd, you could probably do it. panel heating is going to depend on how much power you’re shining on the panel and how effectively it blocks the light from leaking through. people on lumenlabs had a lot of success using mercury arc bulbs and old lcd panels to build projectors, so that might actually be practical. Though if you really wanted to make it simple, just get ahold of those cheap 100 watt ebay leds. then you can just modulate them with a mosfet and the whole thing would be much simpler.

      1. andres – Yes you are right but I meant LATCH in where a manual reset was required like with SCR’s. But what you said about the TRIAC’s AC power source intrigues me. That is a consideration I had not anticipated.

        My idea would not use ARC lamps and such. A 18 million CP Q-Beam uses a 12v HALOGEN lamp bulb. And directly driving it would ruin it too. Your 100W LED MOSFET idea is pretty cool. Thanks.

        BTW – VLC using arc lamps dates back to WW2. The NAZI’s used a IR VLC which modulated a glass prism wrapped in a voice coil. They called it the Light Talker. Even Bell used a VLC technique in 1880’s using a voice modulated Kerr Cell (Pockel’s Cell) after his famous sunlight driven Photophone, The Allies used an arc lamp VLC but they too used a method like I propose to modulate the signal (Kerr Cell?). However, Bell did modulated a gas flame somehow for the gas-lit version the Photophone – I think.

        Ues simpler is better…

        1. Triacs latch just the same way as SCRs, requiring the voltage that they’re switching to be cut off, in order to switch off. It’s just that, since triacs are usually used on AC mains, the voltage naturally turns off 50/60 times per second. Well, actually 100/120 times, twice each cycle. So using a Triac on DC would be the same as an SCR. They’re essentially a pair of SCRs connected in opposite polarity.

          Because of the massive amount of light absorbed by the polarisers and LCD screen, most of your light would be wasted. Probably not a good idea, certainly not efficient way of modulating light. Forrest M Mimms, among others, published in his famous notes, a system for modulating audio onto a torch’s lightbulb. I think it was pretty much an audio amp suited to driving the bulb. Was good enough for speech, though of course the bulb’s thermal mass prevented high-frequency modulation.

          Simply modulating LEDs would, I think, be the best way. Of course you could use IR if you wanted invisible. The response of LEDs is not linear to voltage like an incandescent bulb is, so maybe convert the signal to pulses first? Something like the ADC on a small microcontroller should handle audio OK. You could convert to PWM, then converting back to audio could be a simple matter of a capacitor in the receiver to smooth the pulses out.

          Either way, LED or bulb, I think modulating the light source would be best. Simpler is better!

          1. Thanks Greenaum… all very valid points re:TRIAC idea. I guess that idea is now moot. I think I’ll still try it with slow-Morse Code and a bright LED or something. I have a few busted laptops I could cannibalize the LCD. I just have this curiosity to see how it plays out. I like Al Williams HaD posting here. I must admit I did not think about it before. I was going to go with Pockel Cells but I was told by HaD SME’s that they are over hundreds of dollars (or more) and require HV to operate. This idea is free and requires only some batteries to make it work. I could probably drive the LCD flasher with a common transistor.

  7. I’ve always wanted to find a cheap source for polarizing sheets. Cut strips about and 1/2″ wide and alternate polarization. Make two pieces like this, place them on top of each other, and add a small lever to shift one sheet. Simple on/off window.

  8. how does the lcd panel stands intensive sun light? i use to have an ideea like this using normal lcd(black and white, like on old tetris games:) but they could not stand the intensive sun light.( foil for my car:) :)

  9. I’ve always wondered before why ‘this’ wasn’t more the aim of the original ‘Google Glass’ Project… I mean, no, they would be a lot more like ‘Google Shades’ and there is a reasonably good chance you might be completely blind at night… But you get the whole of the concept, augmentation, plus vision.

    Where the investment comes in is in spinning a ‘glass shaped’ and curved/flexible lcd and routing the connections in a much nicer way.

    Obviously the ‘right blend’ of technologies simply does not exist yet, but I was surprised something like this wasn’t used as a ‘first approach’.

  10. Artem Litvinovich] thought about using LCDs as window panes twenty years ago,

    As LCDs were being introduced to the public (early 1970’s) one of the “possible uses” they trumpeted would be walls that become windows or dark as desired.

    I did see in a Hospital ICU in Sao Paulo ( circa 15 years ago) small LCD windows (about 10 inches by 10 inches) on doors to the patient rooms that could be used for patient privacy.

    1. In the beginning of 2015 I spend 2 months in a ICU in the Netherlands. All ICUs had large (100x150cm, multiple planels) with electrochroamatic glass, also for privacy reasons. Almost killed me when life-support tubing disconnected and no-one noticed because the glass was opaque. Managed to hold the tubes together myself. Nurses came in about 5 minutes, despite of me pressing the help button multiple times. Good times :)

  11. I wonder if you could do something similar with e-paper displays, and therefore be able to leave the windows shaded indefinitely without using power (even if you had to shade them with a solid stipple rather than alpha). I’m not really clear on whether or not the opaque backing is an integral part of e-paper.

      1. Speaking of shades, I was unable to convince my wife that Venetian blinds should be reflective on one side and flat black on the other, so depending on the season we could reflect sunlight away from the window or absorb it.

        1. The black part isn’t necessary because any sunlight that enters the room will be absorbed. The IR might bounce off a surface or two, but it’ll get absorbed eventually somewhere in the room.

          You used to see awnings over windows – in the summertime, the sun is higher and the awning shields the window. In wintertime, the sun is lower, and sunlight comes into the windows.

    1. There’s already the electrochromatic windows Chris mentions. I remember in the 1980s, a TV show showed a pair of Japanese sunglasses with a little battery, a couple of buttons let you customise the darkness.

      In electrochromatic windows, electricity forces a chemical reaction, that, I think, forces a metal (iron?) between it’s elemental state, and being part of a clear compound. It’s reversible, takes a few seconds from light to dark and back again.

      What would be good, is hacking them to use a building as a giant display. Would be slow, of course, and low-res. You couldn’t play Pong or Tetris, but there must be something cool you could do. Give it a smiling, winking face.

  12. This article is quite old, but with some luck maybe somebody will read my comment and answer my questions anyway?
    I have a few more or less broken Samsung LCD TVs, and i would like to replicate this hack with their LCD panels.
    But first, when in the article’s video the panel seems to be transparent by default when unpowered, i’ve noticed that the TVs panels are completely black/opaque by default. Why is this?
    Also, the guy doesn’t clearly explain which signal he is abusing, nor why it works. He vaguely says to find 2 spots linked together by a trace on each side of each ribbon connected to the panel. I’ve not been able to visually find them on the multilayered PCB of my panels. And i’m reluctant to blindly probe various spots, since i could maybe fry something doing that.
    Also, the guy says that when probing right spot, the panel should become a little bit darker, but since my panels are already fully dark, that doesn’t make sense!
    I suppose the trick is to use some of the various voltages, but which one(s)? Vcom, VGH, VGL, AVDD, VGamma… There are a lot, and names are quite different from a manufacturer to another.
    And also, since some people says it’s bad to apply DC voltage, how to do it with AC instead?
    Thanks a lot for your help!

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.