Effects loop coupling capacitor blown again

[Timbo]

avt100-61-00-iss1.pdf

I had a Marshall AVT100X in the workshop this week with a dead effects loop. I found that a coupling capacitor in the return circuit (c253) was open circuit. I replaced it with another one of the same specs and all was well. For about 30 minutes! Now it’s gone again. Should I assume that the replacement was bad, or do you think there’s a problem with the rating of this component?

[pdf64]

That seems weird! It looks like neither side of the cap should be able to deviate outside the +/-15V supplies.

Front pcb?

[Timbo]

Yes, front PCB.

[Timbo]

Sorry, no - back PCB. Goes directly to pin 9 of cnb8 and then off to the front PCB via ribbon cable.

[Timbo]

Here’s the front PCB. The signal from c253 on the back PCB goes to the fx mix pot on the front PCB.

avt100-60-02-iss1.pdf

[Mars Amp Repair]

Hello Timbo,
Is it possible the customer is plugging the output of another amp into the return or send jacks? Seems like the only way enough voltage could be presented to that small coupling cap.
Just a thought, Glen

[Timbo]

No, just stomp boxes according to the customer.

[g1]

What do you mean by the cap being 'open circuit', and how did you check it?

[Timbo]

Hi, thanks for your reply. The cap actually tested ok with a Fluke dvm, however I read that that doesn’t actually prove that it’s ok. As the signal wasn’t making it through the cap (checking it through with a scope) I went ahead and replaced it. Sure enough the effects loop then worked fine. I checked it with a chorus pedal for around half an hour and it worked fine, so it went back to the customer. He called me up next day to say that it lasted around half a hour before breaking again. I pushed him for more details and he told me he was using a delay, chorus and reverb in series through the loop, and that the chorus pedal had also stopped working……. So, this is just speculation at the moment but it seems that either the chorus pedal stuffed up the amp, or vice versa. I’m not getting it back for a post mortem until mid week, so I won’t know for sure if it’s the same fault until then.

[g1]

I think the cap might just be a coincidence and you have an intermittent fault in the loop circuit. Replacing the cap may have jostled a bad connection to work again for awhile.

[Timbo]

Ok, I’ll check it over thoroughly, however how would that explain consistently seeing a signal in one side of the cap but not the other?

[J M Fahey]

Ok, I’ll check it over thoroughly, however how would that explain consistently seeing a signal in one side of the cap but not the other?

HOW did you get signal there?

Did you inject signal in the jack?

What signal are we talking about? Voltage? Frequency? Source?

Besides, not too sure about : "The cap actually tested ok with a Fluke dvm, however I read that that doesn’t actually prove that it’s ok. As the signal wasn’t making it through the cap (checking it through with a scope) I went ahead and replaced it."
I see a couple red flags there.
Please describe test and setup in your own words.

[Timbo]

Injected a 1khz sine wave into the input jack. Put a univibe stomp box into the effects loop and then traced the signal with an oscilloscope. I could see the modulated return signal at the output of IC203b, but nothing on the other side of C253, at pin 9 of CNB8. After replacing C253, the signal could be seen at pin 9 and at the amp output.

[g1]

I had wondered if you were measuring at the C253 cap directly so that answers my question. I take it that you are scoping at the component side of the board.
You could have a bad solder connection at the output pin of IC203B, or at pin 9 of CNB8. Or a dirty pin connection at that connector, if it is removable. As you have already replaced the cap C253, it's solder connections should be fine.
Another possibility is a broken pad solder pad right at the component or connector, or a cracked trace.