Zenith CCII 25GC45 Oil Cap overheating

[wb2mep]

Thanks guys. That confirmed what I was thinking, not much else other than the cap and the transformer to go wrong and these caps are reaching their "use-by" dates. But when a new replacement part fails shortly after installation, it usually means you fixed the symptom, not the real fault.
I read that the value of that capacitor on these constant-voltage transformers has to be correct as it forms a 60 Hz tuned circuit with the inductance of the transformer.
I still had the cap I removed from the set. It's a 3.5 uF, 6%, 440 V, but measured 3.99 uF on my DMM. I heated it with my heat gun and it rose up to 4.25 uF. So I think what was happening is that the resonant frequency didn't match the 60 Hz drive frequency, caused higher current thru the cap and transformer, heating it up, going further off resonance resulting in thermal runaway.
I do have one more N.O.S. capacitor to try when I have time to get back in to this set, this one measures 3.3 uF, increasing when heated.

[Tim Tress]

The issue with those capacitors is electrical leakage. That was a common failure in the old days. A leaky capacitor will give incorrect readings on a digital tester; the best instrument for testing capacitors is one which will test for leakage at the rated voltage.

[JohnCT]

Originally Posted by wb2mep

It's a 3.5 uF, 6%, 440 V

If you're not concerned with originality, you can pick up something like this:

https://www.digikey.com/en/products/...50AA4J/6556394

John