Thanks casdave, that's exactly what's happened. The tants (especially on the preamp boards) have long since turned into, er, dynamic resistors. I changed the two in the preamps, and the hum and crosstalk disappeared from the preamps.
I've just got around to getting a big bag of good quality, low esr, high temp electros, and that's when I noticed none of the polys or ceramics were mentioned in any of the BOMs. And I agree, they're probably going to outlast me, unless they get overheated or crack. So it's not a big deal.
This latest saga came about because I've been getting huge amounts (tens of volts!) of hum on the outputs, and the main amp heatsink has been getting extremely hot to the touch.
So I did some troubleshooting, and I found the main rectifier was also getting hot, especially on one side. And when I verified the steady state voltage, instead of 30V DC, I was seeing 36VDC with 200mV RMS ripple (which is really wierd for this power supply), and worse, instead of -11.4V on the downstream side of the main pass transistor, I was getting -1.42V plus ripple.
I've just spent two hours checking every component on the psu, and despite seeing all the wrong voltages, all the components checked out perfectly (except the caps, but they don't explain the fault). It was driving me nuts (more nuts than usual).
I pulled off the front panel (I replaced the headphone socket a while ago, and accidentally shorted the outputs agains the front control panel, but I figured that out quick and no damage was done), and the turntable plate, and hey presto! the voltages were all back to normal, and nothing got hot.
It turns out that when I replaced the EOT kicker transistor (TR30) a while ago, I reused the TO5 heatsink and carefully bent it so the TO192 replacement could cool down... and I warped one corner just enough so it contacted the turntable plate, but only when it was pushed fully home.
That was why I was getting 11V sparking across the playback cover, and why I kept blowing the kicker transistor and cooking the kicker coil.
So there was the explanation for a whole bunch of seemingly completely unrelated faults over a 12-month period. And the snr for the playback is stupendously good now. It just goes to show.
Cheers,
"Leaky" Pete
Data is not Information; Information is not Knowledge; Knowledge is not Wisdom.