JCM900 Model 4100 100W Head - High Pitch Squeal

[nickb]

Sounds to me that there are 2 separate noises? The first is the "farting" around :02, and the other is the ringing sound about halfway through?

Justin

Do you think the ringing is just ordinary feedback - it is cranked right up? The farty noise almost sounds like and a bad lead ground where you getting massive hum overloading the amp. Have you tried a different guitar and lead? Anyway, do the stuff with the controls and we will have a more complete picture.

[Justin Thomas]

Whoops, didn't mean to confuse... not my amp, just an observation. The ringing sounds like something I've heard in Fender amps with reverb, usually if you end up with a microphonic preamp tube in the reverb circuits... If no tubes were changed, and no-one was poking at the wiring, I can only guess that some part has outlived itself... agree, do tests in post #11.

Justin

[mikeboone]

OK. It was working. No changes were made and it started oscillating one day. This kind of problem can be tricky to fix as you have one part of the circuit feeding back to an earlier part and there are many different ways for that to happen. There are more diagnostics you can do to find out what parts are involved and that will help narrow it down.

You've identified that the pentode/triode affects the problem. In pentode mode the gain is higher so this no real surprise. So I would not be at all surprised to learn that if in triode mode and you turn the gain up further it starts oscillating again. Anyway, what it does tell you is that the feedback is starting after the power tube plates somewhere. It doesn't tell us where it's getting back in. So how about you play with the controls and see what the effects are.

Some things to look for:
Are one or both channels affected?
Does the reverb level affect it?
Do the tone controls affect it?
Can you make it happen by having the gains at zero and raising just the masters or do you have to have to gains up?
Does your guitar have to be plugged in to make it happen?

Answers to these will tell us how much of the system is involved in the problem and hopefully allow us to narrow it down.

It does affect both channels
The tone stack affects it but only the way it sounds.
The reverb control has some affect but not drastic
With the gains at zero and the master volumes cranked, it doesn't take a lot of gain increase to make it happen.
It only happens with some input. It won't do it with no input even if all controls are max

It does happen when injecting a 440hz tone into the Effects Loop. I probed it with a scope and while the amp is pushed to the point of making the noise, the grid of V2a looks fairly normal. The output of the plate (past the coupling cap) shows a very distorted signal. I'll try to attach the pics

I have increased the grid stoppers from 1.5k to 5.6k, no change. I changed the bias split (leaks) from 150k to 220k, no change.

[nickb]

[ATTACH=CONFIG]37042[/ATTACH][ATTACH=CONFIG]37043[/ATTACH]

It does affect both channels
The tone stack affects it but only the way it sounds.
The reverb control has some affect but not drastic
With the gains at zero and the master volumes cranked, it doesn't take a lot of gain increase to make it happen.
It only happens with some input. It won't do it with no input even if all controls are max

Thanks - that helps. The amp's response varies with signal level so it needs to be driven to hit the bad spot.

It does happen when injecting a 440hz tone into the Effects Loop. I probed it with a scope and while the amp is pushed to the point of making the noise, the grid of V2a looks fairly normal. The output of the plate (past the coupling cap) shows a very distorted signal. I'll try to attach the pics

I have increased the grid stoppers from 1.5k to 5.6k, no change. I changed the bias split (leaks) from 150k to 220k, no change.

That was a really useful test. It clearly shows oscillation. It has also eliminated everything prior to the send jack. Also, you have a sig gen and 'scope and that will help.

Let's try to narrow it down further. Can you disconnect the wire that goes to the phase inverter C7 and drive that from your signal generator and see if your still have the problem if you drive it as hard.

EDIT: I later realized that you were talking about the plate of V2a meaning that it is involved making the above test a waste of time. The things I would be looking at now would be if there is some means for signal coupling into the grid of V2a- perhaps the shield on the cable to the return jack is open. Drewl above mentioned that he'd seen solder joint issues on that board too. Other things could be the one or both of 100pF caps in the phase splitter (C9 & C10 in the 5881 version) having gone bad.

[mikeboone]

Guys - thanks for all the help. got hit working. I relocated the grid stoppers to the sockets and it fixed the problem.

I saw that post but didn't quite understand it. I need to read up on he subject.


You guys are great. Happy New Year.

[nickb]

Guys - thanks for all the help. got hit working. I relocated the grid stoppers to the sockets and it fixed the problem.

I saw that post but didn't quite understand it. I need to read up on he subject.


You guys are great. Happy New Year.

I'm really glad to hear you fixed it. I have one of these on my bench right now - EL34s and no mods to the grid stoppers and no oscillation problems. If new tubes had been added then that would be a change, for example. The grid res mod is a good thing and I guess in this case it made just enough difference to fix the problem. The reason you saw the signal on the earlier plate was because it is involved- it's part of a rather big feedback loop. But what worries me, but not too much, is that it was working and then it broke. Therefore something changed and presumably is still broken, just iced over in a sense. Just sayin' cos it could came back to haunt you.

I would shove a square wave in effects return and look on the output for any sign of excessive ringing, just to be sure.

[pdf64]

It may be related to idiosyncracies of the particular power tubes used; I think that JJ EL84 have been noted as being more prone to oscillation than those from other manufacturers, eg http://www.ac30-guide.com/?page_id=410 so perhaps the same may be the case for their EL34?