I believe we recently had a discussion about an unsoldered ribbon cable that had managed to make "acceptable" contact for many years, until...
I just had two "bad soldering" experiences that might be of interest.
Case 1: Both of my JVC XP-A1000 hall synthesizers had leaking power-supply caps. About two years ago I replaced the caps in one of them. Then, six months back, I replaced them in the other.
The second started having problems with oddball noises in the rear channels, including some that sounded like idle noise. They came and went. I finally ripped into the unit, and unsoldered my suspicious-looking joint on the negative side of one of the caps. The pad came loose! I used a piece of heavy solid wire to restore the connection. Et viola, the noise stopped.
Case 2: I hadn't used my Fosgate Tate II 101A SQ decoder in some years, and tested it for a project I'd planned. The decoding was all shook up. This looked bad, because the custom Exar chips used for logic control are no longer made. In fact, they went out of production before the initial product run of the Fosgate unit was completed, 30 years ago!
One of the designers told me how to confirm that the logic chips were okay. They were (big sigh of relief). This left the phase-shift networks, which have
1% caps that can (supposedly) drift. Not only are the caps expensive and hard to obtain, but unsoldering them runs the risk of destroying the foil.The designer urged me to test the circuit's behavior (with a 'scope) before unsoldering. This really required a schematic -- but harman\kardon had destroyed all the schematics when it bought Fosgate! So I had no choice but to trace it out, starting with a close-up photo of the foil side.
Some of the solder joints didn't look so hot. I resoldered them. Need I go further? The unit is now working correctly.
"We already know the answers -- we just haven't asked the right questions."
-- Edwin Land