Thursday, April 16, 2009

Adventures in Backlighting Part 2

The glowing blob you see above is a piece of a cut-to-size EL sheet, glowing on my workbench. I finally got it to work last night. Here's a pic of it unpowered (and with the lights on):

And here is what it looks like in low light -- kind of spooky:

If you read the previous post, you know that I was having trouble with it: every time I connected a piece of the EL sheet to the inverter, the +5V power supply would shut down. I was getting the power from one block of the Discombobulator, which had a Condor triple-voltage linear supply, with the +5 rated at 3A. There was only one module using the +5 in that block, but I figured maybe it was drawing a lot of current, so I disconnected it. With this, there was nothing but the EL inverter drawing from the +5. Still no joy; it shut down as soon as I connected the EL sheet. The second thing I tried was to go get another Discombobulator block. One of the three has a switching +5 power supply rate at 5A. Trying it with this, the power supply didn't shut down -- the inverter itself did.

Well, this left me completely puzzled. Was the sheet I was trying to light too big for the inverter? I tried cutting off a small piece (the piece you see above) and just using that. No difference. What was going on? I checked the piece with an ohmmeter, and was surprised to get readings ranging from about 1300 ohms to 20K. An EL light, being a capacitive device, should read open on a DC ohmmeter.

Tried it again, under low light. When I did this, I noticed that I could see small arcs when I pressed the EL sheet's tabs against the connector, just before the inverter shut down. The arcs evidently weren't in the EL sheet itself, since they weren't leaving burned or melted spots. They had to be in the connector. I looked over the connector and didn't see any obvious shorts. But, just to see, I removed the connector and rigged up a pair of jumper wires with some copper foil tape to connect the EL sheet to the inverter.

Voila! That worked. Something in the connector is creating a low-impedence short circuit path. Perhaps it's the adhesieve that was used to glue the metal contact surfaces to the plastic case. But in any event, that connector doesn't work. That's okay; I will rig up something to install it in the K5m when I get ready to do that.

Here's a larger piece, illuminated. This is the piece I'm going to cut to size to fit in the K5m's display. The crennelations are where the connecting tabs are.

By the way, if you play around with these things long enough, and you aren't very careful about the conductive surfaces, you will get zapped by the inverter eventually. And it will hurt. 90V at 400 Hz really stings. Lesson learned from recent experience.

No comments: