JCM 800 Ghost In The Machine

[Randall]

WOW. Am I confused! This JCM 800 Lead came in for a long time noisy hum. Attempts to remedy it have been made in the past.

While checking it out I discovered plugging a dummy plug in the FX Return caused it too go ballastic. The hum got much louder. I measure it as 60Hz. The 3 wire bundle from V4A to the board were very touch sensitive and noisy when probed or moved. I removed C26 on V4A grid, no change. Grounding V4A grid stops the noise. Tried a different tube, no change. The Pl,Gd,Ck wire bundle wraps around the reverb transformer, so I disconnected the HV to it, no change. With V4 removed I see 1.5K ohms from cathode pin to ground, 1M ohms from grid pin to ground, and 100K ohms from plate pin to R58 (10K 1 W).

I swapped out V5, no change. I removed C25 at the grid of V5 (second page left) to isolate stages, and the noise stopped. With the V4 stage isolated, I see noise on it's plate and with my scope I can see that the wire bundle is still sensitive. I see the noise on the pin side of the plate resistor, but not the feed side, and Y goes to nothing, so where could this noise be coming from? I'm at a loss.

I also notice that V4 socket has been replaced, so it looks like someone else had a shot at this at some point. I looked it over real close, and it LOOKS okay. Oh, and the filter caps have been replaced as an attempt to cure this as well.

Marshall-JCM800-Splitch-50W-2205-Schematic.pdf

[The Dude]

It seems you have it isolated pretty well to something in the V4 stage, so can you list the voltages on V4? Also, aside from the hum, can you hear that the reverb is working? And, check your effects send/return jacks- both connections to the circuit and that the return switching jack is indeed switching. If you jumper send to return does the hum go away?

[Randall]

V4

1. 206.8
2. 0v
3. 1.5v

6. 175v
7. 0.02v
8. 2.06v

The cap at the grid C26 is still out so I verifed reverb with the FX send jack into another amp. It's quiet.

The cap at the grid of V5 is still out, so I can't hear anything right now, but my meter tells me the jack is switching, Remember, the loudest noise and sensitivity was with an open dummy plug in FX return. Wires were really touch sensitive, and it got louder if I touched them with a chopstick or put my hand near them.

[Sowhat]

I also notice that V4 socket has been replaced, so it looks like someone else had a shot at this at some point. I looked it over real close, and it LOOKS okay. Oh, and the filter caps have been replaced as an attempt to cure this as well.

If the V4 socket has been previously replace that's where I would look first, and by look, I don't mean look, really I mean remove and replace all the solder joints on that socket with fresh solder. You can't always visually detect cold solder joints but that sounds like what you got going on there.

Oh yeah, before you suck out and replace the solder joints make sure the V4 tube has been removed, not doing so will certainly trash the tube.

[Randall]

I've hit all those joints already. I'm good enough at soldering that I do not worry that it will "certainly trash a tube".

[Sowhat]

I've hit all those joints already. I'm good enough at soldering that I do not worry that it will "certainly trash a tube".

Hey, O.K. Randall, I didn't mean to ruffle your feathers. That type of precaution is standard operating procedure in my shop.

[nickb]

V4

1. 206.8
2. 0v
3. 1.5v

6. 175v
7. 0.02v
8. 2.06v

The cap at the grid C26 is still out so I verifed reverb with the FX send jack into another amp. It's quiet.

The cap at the grid of V5 is still out, so I can't hear anything right now, but my meter tells me the jack is switching, Remember, the loudest noise and sensitivity was with an open dummy plug in FX return. Wires were really touch sensitive, and it got louder if I touched them with a chopstick or put my hand near them.

The dummy plug test just leaves that grid signal open to capacitively couple to any noise so I'm not surprised it's louder. I don't think it tells you much, if anything.

Measure the resistance for tip to sleeve of the send jack to check the 12K resistor is good. Does the MV and/or reverb level affect it? Do the jumper test suggested by the Dude as a high resistance switch contact will make that wire to V4b grid more sensitive. Also check that the filter cap C44 ground is really making contact .

[Randall]

"Measure the resistance for tip to sleeve of the send jack to check the 12K resistor is good. Does the MV and/or reverb level affect it? Do the jumper test suggested by the Dude as a high resistance switch contact will make that wire to V4b grid more sensitive. Also check that the filter cap C44 ground is really making contact ."

Nick, Please notice that the coupling cap on the grid of V4A is removed, but I do remember that when everything was intact that the reverb vol had no effect. Also, I have already proven the 3 resistors on V4A are working. I measure 1.5K from cathode pin to ground, 1M from grid to ground and 100K from plate to R58. All filter cap grounds are solid. Jumpering send and return jacks makes no difference.

[nickb]

"Measure the resistance for tip to sleeve of the send jack to check the 12K resistor is good. Does the MV and/or reverb level affect it? Do the jumper test suggested by the Dude as a high resistance switch contact will make that wire to V4b grid more sensitive. Also check that the filter cap C44 ground is really making contact ."

Nick, Please notice that the coupling cap on the grid of V4A is removed, but I do remember that when everything was intact that the reverb vol had no effect. Also, I have already proven the 3 resistors on V4A are working. I measure 1.5K from cathode pin to ground, 1M from grid to ground and 100K from plate to R58. All filter cap grounds are solid. Jumpering send and return jacks makes no difference.

The point, that you seemed to miss, is that the 12 k resistor provides significant insurance against interference. We will still don't know if it's OK as you dismissed the test out of hand. I think we need as answer to that to make nay further meaningful deductions. By removing the coupling capacitor you have created a high impedance on a sensitive node.

Not having a JCM800 in front of me I have no idea where the coupling cap is physically located, but if it is remote from V4a grid then you still have a long piece of wire which can couple noise into the grid. You already established that is where the noise is getting in. A related question is how is that noise coupling into that wire. It wasn't clear from the opening post what the wires in the mentioned bundle are actually connected to. 'Pl,Gd,Ck" doesn't covey much useful or clear information so we are left guessing. So am guessing that one of the wires is a HV supply and that leads me to wonder if the the filtering is OK and hence the question about the filter cap ( it may be a different cap depending on which HV node it is, again I'm guessing) ground since the cap has already been replaced.

A piccie would help.

It might also help if we knew of if this is buzzy ( i.e. HF) at 60/120Hz or hummy (LF). This gives a clue to the source of the noise and the nature of the coupling into the V4A stage.

[Randall]

"Measure the resistance for tip to sleeve of the send jack to check the 12K resistor is good. "

It measures 9.28K. If by "that 12k resistor" you are referring to R29, it is actually 10K.

"It wasn't clear from the opening post what the wires in the mentioned bundle are actually connected to. 'Pl,Gd,Ck" doesn't covey much useful or clear information so we are left guessing. "

The 3 wire bundle is about 6" and is connected to V4A plate, grid and cathode. And yes it seems pretty clear that this is where the noise is being induced. I measure it at 60Hz as mentioned in OP.

UPDATE: The customer brought me his other JCM800, same year, and it does the exact same thing with a dummy open plug in the FX Return jack! So it seems I have been chasing my tail for many hours on this. I have never seen an FX return jack act this way, and I check each amp that I see with this same dummy plug.

[nickb]

"Measure the resistance for tip to sleeve of the send jack to check the 12K resistor is good. "

It measures 9.28K. If by "that 12k resistor" you are referring to R29, it is actually 10K.

"It wasn't clear from the opening post what the wires in the mentioned bundle are actually connected to. 'Pl,Gd,Ck" doesn't covey much useful or clear information so we are left guessing. "

The 3 wire bundle is about 6" and is connected to V4A plate, grid and cathode. And yes it seems pretty clear that this is where the noise is being induced. I measure it at 60Hz as mentioned in OP.

UPDATE: The customer brought me his other JCM800, same year, and it does the exact same thing with a dummy open plug in the FX Return jack! So it seems I have been chasing my tail for many hours on this. I have never seen an FX return jack act this way, and I check each amp that I see with this same dummy plug.

The schematic is blotchy - it looks like 12K. So there you have it. If you plug an open jack into the FX return you break the connection to the 10K resistor and turns that wire into an antenna. When you plug a pedal into the return, it will have a fairly low impedance and will prevent the antenna effect.

The hum description is important. 60Hz? Do you mean pure 60Hz with no harmonics - it sounds like a low hum and not buzzy at all- is an indication of AC from the PT getting in, usually from the heaters. OTOH, if it 60Hz with lots of high harmonics then that is likely from the pulsing rectified current and sounds buzzy. Also, capacitive coupling into the antenna wire favors high frequencies so the presence of such HF is suggestive of said coupling method.

[Enzo]

When I want someone to plug into the FX return , I usually say plug a guitar into it and turn the volume to zero. That results in a shorted plug. a shunt. An open (unwired) plug into any input is an invitation for noise. A shorting plug will insure no signal coming in from the outside.