Arcing or sparking between traces?

[kvandekrol]

I'm building a DeadRinger (layout/schematic), no mods, and am having an issue that I don't know how to address. When I plug it in, after about 10-15 seconds it starts crackling audibly - and I can see the sparks flashing through the PCB substrate from the top. The first time I plugged it in, I didn't catch it until it had burned the substrate between the pads of C13 (the 100uF power filter capacitor), unsoldering C13 in the process. I replaced C13, scratched out the charring, and did a number of continuity tests to make sure there weren't any bridges anywhere in that area. Now C13 appears to be okay, but I'm getting sparks on the outer two pins of Q1. I replaced Q1, but it's still happening.

I haven't thought it wise to leave it on for more than a few seconds after the sparks start flying, so I can't provide full voltages on all of the transistors and the IC, but I can verify that the collector of Q1 and the positive pin of C13 both read 9.6V as they should. (I'm using 25v capacitors, so they're not blowing due to overvoltage - in fact, the original C13 still looks perfectly fine).

What could possibly be causing this? I'm not sure what to do next because I can't leave it on long enough to probe around! I'm using a 9V wall wart with the correct (center negative) polarity. I did etch the board myself, but this is about the 15th board I've etched and I've never yet had any issues like this with other projects...

[darron]

i'm not familiar with this circuit. it's only 9V and causing arcing between traces? crazy!

so.... i suppose the first question is what sort of power supply are you using?

since you've built other project i would't ask to make sure your soundings etc. are all right.


i want to watch to see what people suggest. learning is fun.

[deadastronaut]

sparking?...wow...


could it be diodes/fets wrong way round?......i'm curious too..

[Galego]

Weird. Are you sure you let the board etch long enough? I usually leave it a few minutes after I can no longer see any copper between traces.

[R.G.]

PCB cancer.

No, not being funny. When a PCB starts arcing, it carbonizes a path under the arc. The carbon traces are conductive, so they keep conducting. If there is enough current available, it keeps charring and eating more PCB.

Whatever started it - perhaps a microscopic thread of copper left on the board, is probably gone by now. Could have been a defect in the original PCB material.

To stop this you have to interrupt the current path by cutting or scraping away any trace of charring to restore it to an open circuit.

I first ran into this when working on a 200A 5V power supply. Only there, the current had charred away about 2 square inches of PCB area.

[kvandekrol]

Yikes. Diodes, caps and transistors were all oriented correctly, so my guess is that it's malignant. Sounds like I'm going to be etching a new board and giving it another shot. I don't doubt there could have been some residual copper that started the whole thing off... that sounds like the most likely culprit ( maybe my etchant was a little weak).

Am I safe to reuse parts from the old board, or would it be better not to? I will probably ditch everything that was connected to the +9V just in case, as it's less than a dollar in parts and I have spares, but hopefully it'll be okay to reuse the other caps and resistors?