Amp restoration and older ceramic capacitors

[Rhodesplyr]

Sometimes the job is just to repair a guitar amp, but sometimes I get restoration requests where the goal is to make it sound as much like new as possible, given what's available to work with.

With that in mind, how do you tend to deal with ceramic capacitors in older guitar amps? I'm not talking about the C0G/NP0 types since these don't suffer particularly from aging effects (usually only for picofarad values). I'm talking about the Z5U, Y5U Class II ceramics that do change somewhat over time. I know they rarely fail outright, though they can go microphonic.

But take this situation: you've got a 1960 Gibson with push-pull outputs, and, with your test equipment, with the caps still in circuit, you can tell that the two coupling caps 1) aren't well matched and 2) have developed higher dissipation factor than they should have. These are caps small enough that their high-pass filter function is going to be in the audible range, which makes a difference.

Do you leave them because the wonky sound of >50 year old ceramic caps is part of the vintage mojo, or do you use new capacitors that have the same dielectric?

In the past, I've usually just left them alone, but I like to reexamine my restoration practices as I work on more amps.

[bob p]

For me it would depend upon whether I'm working on my amp or somebody else's. If it's somebody else's, I'll explain the situation to them, and the possible outcomes, and once they understand the choices, I let them tell me what to do.

For my amp, when a part is out of spec I clip the leads short and throw it into the garbage so there's no going back.

[Bruce / Mission Amps]

Take no chance. Ask the customer what they want, what they are expecting and tell them what the possible consequences are.
I made that mistake once on an old tweed Super Amp. When done the thing sounded quiet and just like new... exactly what he hated.
I never got him satisfied because I didn't know what he wanted from the start... which was an unrealistic expectation on his part.

[Rhodesplyr]

In this case, the customer is someone I know well and who likes what I do when given a free rein. He bought the amp in barely-working condition, and it has a random speaker with a tiny magnet in it. What I may do is to do the other repairs first, which include the standard filter caps, etc...., plus reconing the blown original Jensen speaker. Then, I'll clip some other capacitors in parallel to get some idea of how the ceramics are coloring the sound--in good or bad ways.


Now, if you were to replace ceramics would you replace them with current-production ceramics like Cera-Mites? I have no particular aversion to ceramics in musical instrument amps. I have a customer with one of those Wards Airline Supro amps with lights over the controls that's full of them, and it sounds great.

[Enzo]

How can you tell the caps are not matched if they are still in circuit? All that other stuff is hanging on them electrically. And if he wants it to sound liek new, who knows how "matched" they were to begin with. They certainly were not using 1% parts. Those are notoriously sloppy tolerances.

Does he have any rememberance of how it ought to sound, or is this just a bit of personal philosopy on his part?

CEramic caps change with the weather, change capacitance with vaqrying voltage across them, are microphonic. I fail to see why someone would prefer them, other than for historic reasons. But if they work, hell, leave them alone. SOme vintage gotta be stock sort of amp, OK deal with it. But to just get an amp working well, I'd prefer film caps or silver micas for the small values. I leave the ceramics in teh trem oscillator though. WHy not?

[Steve Conner]

If I find any part that has drifted out of spec, I have to assume that it's unreliable and will keep on drifting further and eventually fail outright. In the garbage it goes. I get especially suspicious of coupling capacitors, since they can destroy valuable power tubes if they start to leak.

In the case of components in a fancy vintage amp, I would note the actual values that they drifted to, to leave myself the option of replacing them with new parts of those values if the customer thought it sounded too "new".

[Rhodesplyr]

How can you tell the caps are not matched if they are still in circuit? All that other stuff is hanging on them electrically.

It's a cathodyne PI, so the caps are fairly well isolated. The resistors are not particularly out of value. The difference is that the plate resistor is AC-grounded through only a 10uF cap, so I clipped in a 2200uF cap from it's + supply to ground, which makes it look grounded to AC. At that point, the impedance across the caps should look roughly similar, but it doesn't, checking both with an Anatek Blue ESR meter that tests at 100kHz and an impedance bridge that tests at 1kHz.

I normally wouldn't go to that kind of trouble, but I was curious :-)

[bob p]

If I find any part that has drifted out of spec, I have to assume that it's unreliable and will keep on drifting further and eventually fail outright. In the garbage it goes.

Great minds think alike.

In the case of components in a fancy vintage amp, I would note the actual values that they drifted to, to leave myself the option of replacing them with new parts of those values if the customer thought it sounded too "new".

I like Steve's idea for "blueprinting" a vintage amp. As a matter of practice, I print out a schematic and run through all of the parts, writing down the observed values before I start work. then I print out a second schematic that I'll write changes on to document the work that I've done. That way I've got a before/after snapshot that shows exactly what's been changed.