Frankenstein electronics!

[anagram]

I was resurrecting an old Philips EL3302 cassette player today, which I got for a birthday present about 50 years ago and has not been on for ~40 years.  Audio was heavily distorted and highly attenuated and I was poking around with an external audio amp to trace the signal path.  While probing the final stage, suddenly the audio returned!  No evidence of  loose connections -- more of a component popping back into life as a result of the probing.  I've no experience of this, but maybe very old (germanium in this case) transistors or capacitors can fail in a way that's recoverable with a bit of juice? Any ideas?

[Ian.M]

Some germanium transistors can suffer from tin whiskers.  e.g. https://nepp.nasa.gov/whisker/anecdote/af114-transistor/index.html

[anagram]

thanks Ian -- yes that looks the likely culprit. I guess a benefit of plastic transistor casings is that this can't happen.  btw the drive belts had turned into something very close to black construction adhesive but otherwise it seems to work fine.  The external psu is "a bit how ya doin'"  by modern standards though.

[RoGeorge]

If you plan to use it further, before any testing I'll replace all the electrolytic capacitors.

Don't spend time trying to measure each one, just replace them all anyways.  With time, electrolytic caps dry out, or they short, or they interrupt internally.  If you don't want to recap, at least let the player stay powered for a while (hours).  The oxides layers inside capacitors get dissolved in the electrolyte when kept discharged for very long time, and the oxide is expected to grow again (partially) while voltage is applied, so the capacitance will grow.  Not like new, but might be good enough to make it sing again.

Nothing wrong is expected to happen, but while keeping it powered, it's a good idea to do this while you are supervising, for just in case something else fails and it all starts smoking.   

Also press record a few times, in the hope that any internal oxides insides the play/rec mechanical switches will scratch away.

[anagram]

Thanks RoGeorge --yes,  the electrolytics are probably a bit shot, but audio seems ok.  I'll try leaving it on for a bit though as you suggest and might replace the one in the psu as that one looks a bit crusty.  I gave the multipole play/record switch a good scrub as initially I though it was the problem.  That switch is an interesting design, with a fully removable pcb finger containing the contact pads making it very easy to clean. This recorder gave me a lot of fun as a kid, so it's nice to see it working again!

[james_s]

If you plan to use it further, before any testing I'll replace all the electrolytic capacitors.

I usually recommend against this. I've seen lots of 40+ year old electrolytic capacitors that are still in good shape, and I've seen a lot of new problems introduced by novices re-capping gear, and often replacing good quality original parts with cheap no-name Chinese junk. Unless the specific are known for having a high failure rate test them and if they're still good, leave them alone.