Capacitor in disguise

[PJC58-Hythe]

I recently repaired a Pioneer QX646 tuner amp and noticed the output pair midpoint point voltages on all four amplifiers were all over the place, one of them was at 3.9V where it should have been 20V. After a lot of probing around and head scratching I diagnosed the fault to faulty 1uF 25V input capacitors on all four amps so I replaced these an all was as expected again.

I noticed on closer inspection these didn't seem to be the same as a standard electrolytic, despite them being in a metal can and sleeved with plastic as one would expect. I peeled off the metal can and took a knife to what I found and discovered they are solid tantalum capacitors!

[kalee20]

Wow!

Well, replacing them fixed the fault, so they obviously were the culprits!

What do they 'look like' now - near short-circuit, or a resistance of a few kΩ? I'm always interested in failure modes.

For such an application, my inclination would be to replace with small 1μF 50V radial polyester capacitors - they're almost above suspicion and last for decades.

[DrStrangelove]

Always remember the Les Lawry-John's "Curse of the Blue Tants".

Other colours are equally evil.

The latest being the brown 10uF tant in a PC XT: dead short.

[PJC58-Hythe]

kalee20 they look more like a resistors, with the worst one of the four measuring a couple of K ohms. The annoying thing is if I'd known they were tants I'd of replaced them on site!

I've attached a snip of the circuit - C1 is the offending item.

[kalee20]

Thanks!

And that looks to be a good circuit, C3 as HF rolloff and C5 as anti-RF breakthrough smacks of care taken.

But... the indicated voltage 21.4V on the input base is getting rather close to the 25V input capacitor's voltage rating! That's not so impressive, given that tants have no self-healing mechanism.

[PJC58-Hythe]

kalee20 - needless to say they were replaced with 1uF 63V, and I agree, nice circuit design.