« Electric Gongs update | Home | Photocapacitors »

Interview: Circuit-bent video projector

A while back I picked up a bulky old Philips LCD video projector from MC Howards surplus, price tag $20. I wondered what would happen if I tried circuit-bending it. I hooked up with Austin Dorkbot #13 during SXSW 2008 as a presenter, and the pressure was on to produce something interesting.

I opened up the projector and did some research on the chips inside… I was able to find datasheets for the A-D and D-A converter chips and targeted them in the hunt for interesting visual effects. By inspecting the densely-populated SMT main board I gained some insight into the flow of video data in the projector. All of the inputs are analog video of different formats. The analog RGB signals are routed through A-D converters, and once digitzed, the data is held in registers for processing by a VLSI circuit that presumably re-formats the resolution and scan rate to accomodate the projector’s optics. From there the digitized video is converted back to analog via A-D converters (one for each color red, green, and blue), and hits the analog LCD drive circuitry.

To safely probe the various (mis)-connections I wanted to make, I wired up a 1k resistor and a thin wire. The resistor limits current flow and prevented damage as I made various connections. In the end I came up with six bends that were unique, and installed three toggles and a rotary switch on the front panel. These bends created color-shifting effects, lo-fi pixelization, and ghosting… not too interesting on their own, but when combined with a live camera feedback loop, I saw some rather interesting imagery. Focusing the camera on the projected image created the classic feedback tunnel effect, and when augmented by the bends I had wired, the tunnel was transformed into a beautiful fractal-like effect.

Due to a mishap, I accidentally blew up something in the blue channel’s drive circuitry so my projector only does red and green now. This occurred in a cavalier moment when I probed the circuit without a 1k safety resistor in the jumper wire… should have known better.