JCM800 style amplifier troubleshooting

[Tallablokki]

Hi!

I built a hot-rodded Marshall JCM800-style amplifier from a kit. I have attached the schematic.

The amplifier works, but there is a fairly loud background hiss, and the amp also occasionally oscillates. I have narrowed the problem down to the first tube stage of the overdrive channel. Turning the Gain control all the way down removes all of the noise.

Grounding the grid of the first gain stage of the overdrive channel does not eliminate the hiss. However, grounding the signal after the anode, at capacitors C10 and C11, removes the noise completely, so the source of the problem appears to be somewhere before that point.

The plate resistor is quite large, 330k, but adding a snubber capacitor in parallel with it (47 pF or 100 pF) does not change either the amount or the character/frequency content of the hiss. This makes me wonder whether the noise is actually caused by the thermal noise of the plate resistor.

Channel switching and the boost function are controlled by relays. The channel selector switch in particular (overdrive channel – boost – clean channel) produces a very loud popping noise when switching. Relays K1 and K2 also appear to be microphonic.

I tried measuring the relay supply voltage with an oscilloscope, but my oscilloscope skills are not good enough to determine whether the relay supply is behaving correctly.

Activating the Boost function causes the amplifier to oscillate when both the Gain and Master Volume controls are turned all the way up. The Boost function adds an additional gain stage before the overdrive channel. However, switching the Boost on can still trigger the oscillation even if the tube used for that additional gain stage is physically removed from the amplifier.

Because of this, I have started wondering whether the hiss and oscillation could somehow be related to the relay switching circuit. Unfortunately, I do not know how to investigate or locate the problem any further.

There is another, possibly unrelated problem with the effects loop. When the effects loop is engaged, additional noise appears. The designer of the amplifier suggested that the noise could originate from the 100 V Zener diode in the effects-loop circuit. I added a resistor in parallel with the Zener, which reduced the voltage to below 100 V, but I did not notice any change in the noise level.

Yesterday I noticed that I had accidentally installed two Omron G5V-2 5 VDC relays instead of the specified G5V-2 12 VDC versions.
Could powering the 5 V relay coils from the 12 V relay supply have damaged any part of the relay switching circuit, for example the relay coils themselves, the driver transistors, diodes, or the power supply components?
I have now replaced both relays with the correct 12 VDC versions. However, this made no noticeable difference to the background hiss, so the incorrect relays do not appear to have been the direct cause of the noise.

I have been actively trying to find the source of the hiss for several months, but so far I have not been able to locate it. I would really appreciate any help or advice, because apart from these issues, the amplifier sounds very good.

Thank you very much in advance for any help!

[xtian]

Hi, and welcome to the board! First thing you need to do is disconnect the NFB loop (from output jack to NFB pad on your PCB) and see how this affects the noise and oscillation.

[Tallablokki]

Thanks Xtian!

Disconnecting nfb wire doesn't remove the hiss. Hard to say if there is more or less noise. Amp seems a bit brighter.

Amp no longer oscillates when gain and volume on full and activating boost switch. But it still oscillates when V1 (the boost tube) is pulled out.

[Fischerman]

Might be helpful to post pictures of the actual amp, focusing on the early preamp stages and switching.

[Helmholtz]

How did the oscillation manifest?

[xtian]

Amp no longer oscillates when gain and volume on full and activating boost switch. But it still oscillates when V1 (the boost tube) is pulled out.

This may indicate you have your wires reversed, and you're getting positive feedback instead of negative. Swap your OT primaries (typically blue and brown, not the center tap), reconnect NFB, and reevaluate.

[Tallablokki]

Might be helpful to post pictures of the actual amp, focusing on the early preamp stages and switching.

Here are some pictures. There are 0,6W resistors in V2 grids. And V2A grid resistors one terminal looks awful, because I heated it so many times. But it works and has good connection underside of pcb board.

[Fischerman]

Amp no longer oscillates when gain and volume on full and activating boost switch. But it still oscillates when V1 (the boost tube) is pulled out.

This may indicate you have your wires reversed, and you're getting positive feedback instead of negative. Swap your OT primaries (typically blue and brown, not the center tap), reconnect NFB, and reevaluate.

Be ready for an extremely loud howl the instant you flip the standby if you do this!

[Tallablokki]

Amp no longer oscillates when gain and volume on full and activating boost switch. But it still oscillates when V1 (the boost tube) is pulled out.

This may indicate you have your wires reversed, and you're getting positive feedback instead of negative. Swap your OT primaries (typically blue and brown, not the center tap), reconnect NFB, and reevaluate.

I reversed the wires and it started to oscillate immediately. Clean and distortion channel started oscillation.

I tried to measure the phase with oscilloscope before I reversed the wires and weirdly the output was in the same phase that phase inverter v4b side. I measured from C30 with oscilloscopes second channel. I was very suprised after reversing wires as the oscillation started.

I hae 400Hz sinewave feeded into the input.

Before that I measured Voi from output jack with nfb and without nfb. With nfb connected I measured 2,64Vpp and without nfb it measured 1,52Vpp. So I thought that it was positive feedback and disconnecting nfb wire dropped the output Vpp.

But It started to oscillate. After I connected the OT wires back I checked the nfb Vpp again. With nfb it was the same 2,64Vpp. Then I disconnected the nfb, but I didn't want to solder/heat the nfb terminal again. So I disconnected the 47k resistor R45. And after that the output was Over 5Vpp.

There is a big chance that I don't know what I'm doing with my oscilloscope.

Well after all was like before I wanted to check if the 12Vdc V+ line was ok. V+ feeds the relays and also preamp tube heaters. I feeded 12 Vdc from external powersupply. There was no change in hiss, so I think the 12V regulator works fine.

But what I don't understand is that now the output and v4b are in the same phase and v4a is 180 degrees from that. And if I change the polarity, the amp oscillates wildly. But again it could be me.

[Tallablokki]

Something that has been in my mind but haven't tried yet are the power tube heidän resistors. In the schematics they are 5,6k resistors but in the layout they are 1,5k. The designer had changed the resistors to 1,5k so it was ment to be that value. But I am thinking that could this amp (my build) be so unstable that 5,6k would be better.

Is there so me kind of signs of power tube HF oscillation that I could see with scope?

[Tallablokki]

How did the oscillation manifest?

The oscillation was audible through the speaker when both the Gain and Master Volume were set close to maximum. Engaging the Boost could trigger it, and lowering either control stopped it. I could also see oscillation on the oscilloscope, but I did not measure its exact frequency, so I cannot say whether the actual oscillation was within the audio range or at a higher frequency with an audible side effect.

[Helmholtz]

The oscillation was audible through the speaker when both the Gain and Master Volume were set close to maximum. Engaging the Boost could trigger it, and lowering either control stopped it. I could also see oscillation on the oscilloscope, but I did not measure its exact frequency, so I cannot say whether the actual oscillation was within the audio range or at a higher frequency with an audible side effect.

It is easy to read the oscillation frequency with a scope when the time base setting is known.

[Tallablokki]

The whole amplifier was very unstable and would start oscillating quite easily. I did various measurements and somehow ended up in a situation where, with the Boost engaged, the amp would start oscillating when both the Master Volume and Gain were turned almost all the way up. Removing the V1 tube did not prevent the oscillation from occurring in this case either.

This made me start wondering whether there might be some kind of local positive feedback somewhere in the circuit. The frequency of the oscillation was affected by the Gain setting, and it also seems that the oscillation frequency varies from one occasion to another.

I decided to separate the hiss problem from the effects-loop problem, so I removed the effects-loop PCB completely. This did not eliminate the loud hiss, but the amplifier became noticeably more stable and no longer seems to oscillate, at least not at any frequencies that I can hear.
I tried using the FFT function on my oscilloscope, but I am not really sure how to use it properly or how to interpret what I am seeing on the screen. If the hiss is actually caused by HF oscillation, or perhaps by oscillation at several different high frequencies, should I be able to identify those frequencies using an FFT measurement?

I still believe that the source of the hiss and interference is located before the Gain potentiometer, because turning the Gain all the way down eliminates all of the hiss. The oscillation itself may then be generated later in the signal path as a result of some local positive feedback.
I also checked the hiss with an audio probe. The plate resistor definitely produces noise, but I do not know whether I am hearing the resistor's own noise or noise/interference coming from the B+ supply.

Could the 12 VDC supply used for the relays and the preamp tube heaters somehow couple noise into the audio signal path? Could the relay driver transistors or their switching also be a possible source of the hiss?

[Tallablokki]

The oscillation was audible through the speaker when both the Gain and Master Volume were set close to maximum. Engaging the Boost could trigger it, and lowering either control stopped it. I could also see oscillation on the oscilloscope, but I did not measure its exact frequency, so I cannot say whether the actual oscillation was within the audio range or at a higher frequency with an audible side effect.

It is easy to read the oscillation frequency with a scope when the time base setting is known.

The frequency of the audible oscillation seems to be changing. And I am also a beginner with oscilloscope. There is so much happening in the screen and I don't know how to isolate things. The noise is so messy hairy signal in the screen.

I'm hoping I could measure things and fix the problem (or problems) according the results and not just randomly change parts. Still it might be my only way to do this

[Fischerman]

Just a guess but maybe try separating that blue wire that is above the board and those other wires zip-tied to it. Move them around and see if it affects the oscillation.