How to figure out the pin out of this undocumented LCD?

Question

I have a very tiny LCD ripped from a piece of electronic scrap. The LCD is about 1 inch by .25 inches. It does not include a backlight(it had an external LED functioning as the backlight) and is basically just a flat piece of glass. There are 16 wires coming from the board it's on. (it is just a plain "breakout" board for the small SMD connector)

The LCD board says "LCD:91-42959-B01" which does not pull up any relevant results with google.

I've tried my best to test it with a multimeter, but nothing seems to even be connected. I'm not sure how most simple LCDs work, so what can I do to try and get this to work (hopefully) without ruining it in the process?

Also, because this is a flat piece of glass, I do not believe there will be a command type of interface, because as far as I can tell, there is no actual logic connected to it. It's just 16 wires going into clear glass. Also on the breakout board they are bundled into 4 bundles with 4 wires each(with brown, orange, orange, yellow being the colors in each bundle). I have no idea if this matters, but I figured maybe it could help

Answer 1

Driving raw LCDs is a complex business. Microchip have a good appnote on how to create an LCD driver with a PIC microcontroller, the ideas are all transferrable to other devices.

When reverse engineering devices, it's best to gain as much information as possible from a still working unit with a logic analzyer. If you have two units, you can then keep one working while you clone the driver hardware/software for your new system.

If you are lucky, your LCD may be a chip-on-glass module with a driver chip built in. By sniffing the data transactions with this, it might be possible to guess at the device.

Answer 2

If the display was a segmented display, and it has less than 64 segments total, you could try connecting pairs of pins through a 1K resistor to a function generator outputting +/- 5 volts. If you hit a row and a column, you should see one segment become somewhat dark, and possibly some other segments become less dark. If you hit a row and a row or a column and a column you may or may not see two or more segments become somewhat dark.

If the display is a dot matrix, it probably has a driver or controller. Without a working device to examine it in, I'd regard it as junk. If you had a working device, you could easily determine which pins are for power, and probing the other pins would let you see whether the waveforms look like those of a driver or controller, and what pins seem to be doing what. Without that, it might be possible to experiment with the device and figure things out, but if you guess wrong about things you could unknowingly damage the device, rendering further experimentation useless, and consequently waste a lot of time.

Answer 3

Very likely the display is driven directly from an MCU with an LCD-drive output. The resistance of an LCD segment is typically in the meg-ohms range. As a quick test, you can try applying about 3V to various pin pairs until you get a segment to change. If that fails, try 12V. You will have to then experiment with the voltage until you get the correct excitation voltage. Then, you'll need to modulate the drive waveform so that the LCD will get a 0-DC average voltage across it - otherwise, it will eventually die.