Bad Sozo caps???

[bluefireamps]

Hmmmm. What you mentioned about the standby switch has me wondering about the switch itself. A poor connection inside the switch dropping voltage could cause the problem you're experiencing. If that is the only way you are able to make the problem correct itself that just might be very telling.
Dave

[dehughes]

Hmmmm. What you mentioned about the standby switch has me wondering about the switch itself. A poor connection inside the switch dropping voltage could cause the problem you're experiencing. If that is the only way you are able to make the problem correct itself that just might be very telling.
Dave

Wouldn't that be an overly-simple solution....how could I test the switch?

[Reeltarded]

1. Hardwire the amp without a switch.

or

2. Installed, probe both sides and watch for the hot to drop on the amp side and not the in.

I hope this is your problem, because it's easy, and wouldn't that be grand!?

[dehughes]

1. I love your username...

and

2. Heck yes it'd be wonderful if that was the problem. I think I have a spare switch somewhere...but if not, that's about as affordable a fix as I can conceive of.

The standby switch does usually pop when I "re-set" the amp, and I do get some "voltage drop fizz/crackle" in the process, so who knows?

This is what I hate about issues like this....could be a number of problems, all of which exhibit the same symptoms. Kinda makes me wish it was just on fire, as then at least I'd know that the amp being on fire IS my problem.

[Cliff Schecht]

The switch is easy to test, just hardwire the amp on test (you could place a jumper on the switch to save hassle). Standby switches aren't really necessary in tube amps, we've discussed this to death in the past. They are actually unwanted if the switch is feeding into a choke because of the flyback voltages that can develop when you flip the switch (remember with an inductor, voltage can change instantaneously but current can't). The switches we use are rated for AC operation and typically aren't very good at handling DC, especially with any kind of power. That's why I won't use anything under 120V, 5A rating for a stand-by switch (I prefer a 10A rating), otherwise the switch is prone to arcing and eventually building up enough carbon residue to go intermittent.

[dehughes]

What would a faulty standby switch look like? I've never had one go bad before...

[Cliff Schecht]

It'll look fine, the issues are internal and (usually) hidden from sight.

[dehughes]

It'll look fine, the issues are internal and (usually) hidden from sight.

Right, I meant how would it look on a DMM or a scope? What am I looking for...or can it be tested outside of it being in the circuit and the problem being presently manifest?

[Cliff Schecht]

Yeah when the switch is off the DMM should read open (i.e. infinite resistance) and when closed it should read darn close to 0 Ohms. You can test it in circuit with the amp off and unplugged (and drained!).

[Reeltarded]

Jumper is brilliant! Just took half the work out of troubleshooting, and turned it into troublejamming!

When it was mentioned, I couldn't help but think back to the couple bad switches I have encountered.

[paulster]

It's hugely unlikely to be a faulty standby switch since they'll typically pass current or not, or will pass current with some resistance if the contacts have arced over, but not change behaviour in the 'on' position.

Whilst it would be a lovely solution it really looks like a cap charging slowly to a point that's shifting bias points to where the amp won't work properly and then removing the DC voltage temporarily resets it.

If that's the case you should be able to get the amp to exhibit the symptoms by leaving an iPod or similar plugged into the front end for a period with the power tubes pulled if you have a scope to see whether the problem is occurring, or by driving a dummy load.

[Cliff Schecht]

The hard part with leaky capacitors is sometimes they are placed in a spot in the circuit where you can't tell if they are leaking. Caps connected to ground at one end won't measure DC at the ground side, they'll measure ground. Only way to test that is in circuit is to lift the ground end, place a 1M (1/2W min) resistor from the now floating end to ground and measure the new resistor/cap connection. Do this with no signal and a grounded input so no AC noise throws off your DC reading (what you're interested in testing is DC blocking ability). If you measure any voltage across the 1M resistor then you have a leaky cap. If you're checking the capacitors right after a preamp plate then you can just pull the next stage tube and again measure from the 1M resistor to ground.

Be careful poking around in there. Lot of places to get shocked when probing certain caps..

[paulster]

True. It's not necessarily easy to find, but I've always found that working backwards helps. Look at the grid bias on a tube and if it looks suspect then the caps before are the ones to take a close look at if the bias at the stage before is good.

A JTM45 is a pretty simple circuit and it shouldn't take a whole load of troubleshooting to find a leaky cap in there without having to pull them out and reconfigure.

[dehughes]

So, interesting update:

For the past few months I've been running EL34B-STR tubes in the amp, and I've experienced none of the issues. Oddly enough, I don't remember a single instance of the amp dropping low end and gain.

Just last week I put the older TAD KT-66s back in, and every time that I've played the amplifier since, I've experienced the cut-out. It usually happens around the 30-40min mark. It seems the amp has to be played for a while before it cuts out. This has me thinking it may have been the KT-66s the whole time, but I swear I swapped them out when I was testing this out back at the start of the problem months ago.

That said, I've never had a set of tubes "go bad" in this way: Fine and dandy until the 40-ish minute mark where the amp looses a bunch of lows and mids, and the gain is reduced significantly...but it's all fixed with switching the standby switch off and then on again, albeit with an abnormal "pop" when the amp is put on standby.

Does that sound like a power tube issue? I mean, usually when power tubes have been bad in the past it's been pretty obvious....not like they just stop working like clockwork at the 40 min mark, then are fine from there on out. Perhaps there's something with this amp and this particular set of KT-66s? I'd go for that, but the fact that the genesis of the issue coincided with the Sozo cap installation has me wondering what's up.

[selloutrr]

current draw and heat.

easy way to answer your question about the tubes being bad. run them hot until they fail. use and oven mit remove them and put them on a test and test them for faults. let them warm up for a while and test in 5 min timed intervals for 90min or until failure. if they are good after 90 mins. running the same voltage as the amp. 90% sure it's not tube related.

is the heat causing the amp to produce cold solder joints when using the KT-66?

you wont beable to test this out of the cabinet as the heat build up plays a big roll in the heat transfer.