Intermittant fault, the volume swells and fades.

[wiseowl]

I have an intermittant fault which is driving me nuts, I seem to have been through everything and it all checks out OK. Sods law plays its part inasmuch as whenever I try to check the amp live it won't misbehave.

The fault is basically that the volume fades out and then comes back over a second or two, either it settles down or continues to do this. Any suggestions welcomed.

Thanks

Martin

[guitarmike2107]

What circuit is it? Has it being like this since new or has it developed?

When does it happen, i.e. only when you play with the power amp cranked up or at any time regardless of the amp settings?

Is the recovery time consistent or if you keep playing does it stay muted?

Have you tried different valves

have you tried reflowing any solder joints yet?

Make sure your grid leaks are all grounded properly and that there values are within the data sheet specs for the valves. if these are over sized or have a bad connection the bias will drift into cut off, and it takes a bit of time to recover.

[rockinrob86]

this sounds like an intermittent connection. Check solder joints.

If you are flipping your amp over or something like that to check it live, you may be fixing the problem by moving the amp.

[zaphod_phil]

Also warming up various parts under the chassis with a hair dryer can help expose problems in a less intermittent way.

[wiseowl]

It's a 36 watt lite, inasmuch as it uses the front end of the 18watt lite and the power section of the 36 watt, from the schematics on this site.

This is a problem that has developed recently, I have replaced the valves, apart from the rectifier, and I've been through all solder joints and re flowed them. The fault only appears when reasonanbly loud, ie level suitable for a small hall. Once there the sound cycles down and up about once a second.

Interesting idea about the flipping it upside down fixing the problem, I'll try that next time.

It could be temperature related Phill, I'll try the hairdrier idea too.

Thanks for your input.

[jaysg]

If you can make it happen at will, or at least predict it, monitor the DC level in a few places to see if a supply is pumping.

[guitarmike2107]

Like Jaysg said if you can get it to cycle it should be measurable somewhere.

I take it you don’t have a spare rectifier to try?

My guess is that the cyclic nature of it (1hz) means its a cap problem.

[wiseowl]

Unfortunately I can't predict it or make it happen, which is why it's so frustrating.

No I don't have a spare rectifier, I'm going to swap it out for a few diodes to take it out of the equation, as soon as I get chance to pick some up.

I have a spare set of smoothing caps, I'll swap them out and see if that helps.

Thanks again for the advice.

Martin