More entries in the quickcam odyssey. The audio was back to being crap without anything having changed, but I decided to not do anything about it as it had quite frustrated me working on it previously.
Unfortunately, the quickcam fell on my desk earlier this week, which exposed a weak solder joint on a cap. Took it back apart, fixed that, but it was quirky again - sometimes starting, sometimes not, til it eventually quit working entirely. Generic troubleshooting for broken components, weak solder joints, removing and resoldering the ICs, checking voltages, etc, turned up nothing. Same static screen as in the original post - which seems to indicate an issue on the digital side of things. PIC not talking, or something similar. Checked all the caps (ceramic and electrolytic), resistors, etc - no progress. How frustrating.
Eventually I was about to quit and call it $20 down the drain. Which really pained me. Of course, this would be when I would
finally get a flicker of life. Mr quickcam likes playing hard-to-get, after all. Further troubleshooting found if I death-squeezed a certain part of the board near the PIC, the camera would start and work normally. I verified this was not a mechanical issue (bad joint, broken trace, etc). So this only leaves the really weird option:
my flesh was acting as a (poor) resistor and whatever was being bridged, it was enough to get the cam "booted" - after which it was fine.
Now, please don't get the impression that this is SOP for my troubleshooting processes, but in the name of science, in order to confirm the hypothesis: I moistened my finger (licked it) and touched the suspect area lightly.... it came right up. Repeatably. Hypothesis confirmed.
Some more probing after that found a 10k resistor between a certain capacitor beneath the PIC and +3v available on a diode nearby would bring the camera back to life. I have no idea why. This pin (RB6) seems to only be connected to pin 12 on the PIC, which is a GPIO without special function... but pulling it up to +3v available on the diode fixed my issue.
YMMV
I'm deeming this the "sphere of hatred". I've devoted more energy into this ball than I did SE/30 logic boards.
I have also attached a zip file with two logic analyzer captures of the quickcam being started by quickpict, previewing at full res for a few seconds, then capturing an image.
DSview software required, available here.
Further pictures underneath chips in case someone else gets deep into one of these. I left the crap chinese caps I fitted in the audio section as I have no interest in further troubleshooting to determine if the tantulums actually do or do not work there (seems I was battling multiple issues).
The potentiometer opposite from the crystal is used as a variable resistor and adjusts contrast. When adjusted properly (mostly a matter of trial and error, and personal preference) I saw about 14.3kohms on it. Can used any potentimeter 20k or greater to replace it. The adjustment range is fairly short. Too far and all goes gray (and never truly dark, despite the exposure values), too much the other way has no detail/is pure black and white.
QuickMovie supports live audio (Audio settings, on the audio menu) which is useful for testing.
Acetone works just fine to repair the clips. I recommend a little hotglue behind any repaired clips to reinforce them, see picture.
