Matchless Chieftain Oscillating

[rf7]

This amp was in before for new power tubes, but came back with a cracked/vented EL34. I put in new power tubes and it's got a 20KHz oscillation going on that it did not have when it was in before. It seems to be coming from V1 in the preamp. I can have only V1 and the PI and the power tubes in, an it still causes the oscillation. Pulling V1 always kills the oscillation. Any ideas?

[stokes]

Check the de-coupling/filter cap for that V1.

[rf7]

I replaced it already. I hoped that would do it, but it didn't.

[pdf64]

Is this the right schematic http://www.webphix.com/schematic%20h...s_chieftan.pdf
I would consider fitting grid stoppers to the tube bases of V1, V2 and V3, about 22k for each grid.
Pete

[Jazz P Bass]

Seeing that there is not any signal connection from V1 to V5 (PI), then maybe it comes down to a power supply issue.

Have you measured the ripple on the PI decoupling cap?

Maybe replace that one also.

Link: http://www.freeinfosociety.com/elect...sschieftan.pdf

Note: very poor quality pdf file.

[stokes]

Thats what I was thinking,Jazz,but if it were the PI filter it wouldnt stop when V1 is pulled.Maybe the replacement is bad?It happens.Check for AC voltage at the filter for V1.Also check that the dropping resistor feeding that cap hasnt drifted too low.

[rf7]

Is this the right schematic http://www.webphix.com/schematic%20h...s_chieftan.pdf
I would consider fitting grid stoppers to the tube bases of V1, V2 and V3, about 22k for each grid.
Pete

It looks like the right schematic. The main difference is that most of the 22uf filters are 33uf in the amp. Only the PI filter is 22uf.

I put a 22k grid stopper on tube V1, and with tubes V2 & V3 removed, it is still oscillating. Note that the scope triggers quite well with V2 & V3 out. With them in, the oscillation is more complicated and jittery on the scope.

Seeing that there is not any signal connection from V1 to V5 (PI), then maybe it comes down to a power supply issue.

Have you measured the ripple on the PI decoupling cap?

Maybe replace that one also.

Link: http://www.freeinfosociety.com/elect...sschieftan.pdf

Note: very poor quality pdf file.

Yeah, that’s why I replaced the local filter cap at first.

I jumper clipped a new cap across the PI filter and it did not help. The PI filter is a newer cap too, someone already had replaced it.

Thats what I was thinking,Jazz,but if it were the PI filter it wouldnt stop when V1 is pulled.Maybe the replacement is bad?It happens.Check for AC voltage at the filter for V1.Also check that the dropping resistor feeding that cap hasnt drifted too.

I jumper clipped a new cap across the V1 filter and there was no improvement.

As far as 120Hz ripple on the Filters, they look ok. The preamp and PI filters all exhibit a small version of the oscillation on their waveform. They also have about a once a second overall motion up and down on the scope.

Dropping resistor measures 23.5K ohms. Should be ok.

[stokes]

Jumping a good cap across a bad one doesnt always cure the issue.It seems apparent the osc.is in the power supply.As Jazz says,the fact that it is at the out put with those preamp tubes removed bears that out.The fact that it only disappears when you pull V1 indicates it is that stage causing the issue.You say the "120hz ripple" looks ok,but just for grins put an ac meter on them.Heres the thing I dont like about jumping a cap across a bad one.If the bad cap is shorted even puting a good one across it,you still have a short,and in effect you sre shorting the good cap.

[pdf64]

It could be some sort of grounding issue - check all chassis ground screws are tight etc.
Pete

[Chuck H]

It could be some sort of grounding issue - check all chassis ground screws are tight etc.
Pete

+1

[rf7]

Well, I called Matchless and tried a bunch of grounding things that they told me to try, none of which helped. So I put it back in the cabinet, and lo and behold, the oscillation was gone. Ugh!!! Banging of head against wall ensues.

[Chuck H]

AAaargh!!! This is actually something I've experienced. I've thought to warn others about in past threads but it's been a long time. Indeed an amp in a fully shielded environment may behave even though it won't when outside the chassis. Another consequence of the high gain nature of guitar amps. Even the small amount of crosstalk that makes the amp unstable out of the chassis, or when installed creates inductive relationships with other leads and circuits, can affect an amps overall character. Especially when overdriven. This effect is much to complicated and seemingly random for "me" to get a grasp on controlling it, but I do know it happens most often as a happy accident when the layout helps create a specific, desirable tone.

[nickb]

Glad to hear you fixed it.

They were asking for trouble when laying out the chassis and putting the input V1 next to the output.

One strategy the can help narrow down the parts involved is to carefully touch the grids of the tubes with an small insulated screwdriver while monitoring the output on a scope. If it changes frequency, it's part of the loop. The solutions are usually much the same:
(1) Change lead dress i.e separate wires with big signals from those with small signals
(2) Move sensitive wires closer to the chassis
(3) Replace sensitive wires with screened ones. Twisted pairs can work too. Ground at one end only.
(4) For stubborn cases, add grid stoppers.

[Chuck H]

Ah, but if the amp sounds especially good the way it is (when installed in the cabinet) then changing any crosstalk issues will alter the tone! And believe me, it's a real phenomenon. I mean, if one circuit is coupling to another in ANY way (even through radiant fields, small ground lead resistances, etc.) it still qualifies as another circuit. If you remove that circuit something could change. For better or worse.

[rf7]

One thing about this version of the amp: the V5 PI tube is in the middle of the chassis, right behind the V2 tube, which is different than the layout shown on the official looking Matchless schematic. Maybe they moved it in later versions? I'd like to see if there is any difference tonally between the 2 layouts.