Diagnosing noisy Marshall amp that got worse and worse. Fixed, or is it????

[greengriff]

This is related to another thread, but I thought the issue deserved its own thread. If you're curious you can read the other thread here:

https://music-electronics-forum.com/...ing#post910033

Basic background info necessary to know how we got here:

I have a Marshall 2103 that had a broken presence control. It turned out to be a broken lead on the cap (hidden by a blob of solder). After repairing the circuit I wasn't happy with the tone of the amp (the presence circuit in question turns out to be always 'half on' even with the knob on zero). Chuck H suggested switching to the original Marshall presence circuit, which I did, and was happy with the result. However switching the amp off and turning the presence knob made very loud unpleasant noises, but the amp was perfectly usable. That's where the old thread ends.

Now it gets weird: The noises progressed within a few uses of the amp from being only when the standby switch was turned off to all the time. God knows why but I now could not switch the amp on without hideous loud noises, even with all controls on zero. With the phase inverter removed the amp was silent. I tried another known good valve in the PI slot. Still hideous noises.

So I decided to reverse the presence circuit to it's original (and broken) state, one step at a time, in order to try to find out where the issue was. I disconnected the presence cap. No difference, still hideous noises. I put the 4.7k resistor back. No difference, still hideous noises. I lifted the tag of the presence pot that I'd soldered to ground on the casing of the pot.... and bingo! Silence - and normal amp operation (with no presence control) restored.

Out of curiosity I removed the 4.7k resistor again, and the amp still continued to work perfectly (at least at the very, very quiet low gain setting I am using at home). So now something puzzles me - at this point the purple feedback wire and the ground wire are the only things connected to the presence pot. With no other resistor in place and no tag of the pot connected to ground, how is the phase inverter working? I thought the pot acted as the shunt resistor?

So why dig grounding the tag of the pot cause the progressive noise problem? I'd really like to get to the bottom of this issue, so any suggestions would be very much appreciated.

[Helmholtz]

With the phase inverter removed the amp was silent.

This seems to confirm my oscillation theory.

What do you mean with "tag of the pot"? Please post schematic drawings of what you did.

[greengriff]

Schematic attached. As you see when I modified the circuit one of the tags (is that the right word) of the pot was soldered to ground. This caused the issue. When that tag was grounded through the .1uf cap in the original circuit there was no issue, so what is it about grounding it directly that could cause the noise issue? Also why might the noise issue have been progressive (i.e. started with just switching off the amp, then spread to every time the amp was turned on)?

[Leo_Gnardo]

Maybe that explains why someone in the past disconnected the presence capacitor? I know you want to maintain "all original" but perhaps replacing that cap is in order. Worth a try anyway. FWIW I have on rare occasions seen those "mustard" caps fail - sometimes one end develops an obvious crack or breaks off completely. An intermittent internal break in that cap would explain your symptoms, plus why they go away when you disconnect the presence system.

[greengriff]

Maybe that explains why someone in the past disconnected the presence capacitor? I know you want to maintain "all original" but perhaps replacing that cap is in order. Worth a try anyway. FWIW I have on rare occasions seen those "mustard" caps fail - sometimes one end develops an obvious crack or breaks off completely. An intermittent internal break in that cap would explain your symptoms, plus why they go away when you disconnect the presence system.

That can't be it though, because when I repaired the broken presence cap the amp worked fine. I could have left it at that but I didn't like the change in the tone that the 'always at least half way on' presence circuit made. It was only when I changed the circuit to the old style where the tag of the pot is soldered directly to ground that the problem occurred. And only when I lifted that tag from ground that the noise problem disappeared.

[g1]

This thread would be a lot more useful if there was a schematic instead of a layout.
As far as I know, you are now using an older presence version like in the early1959T. In that layout pic above, the 4K7 resistor is disconnected.
So the only way for the PI cathode to get ground reference is by that ground connection you are removing. I don't see how the PI can work without it. You say the amp is working when you disconnect it, or does the noise just stop because there is no sound anymore?

https://drtube.com/schematics/marshall/1959t-66.gif

[Helmholtz]

So the only way for the PI cathode to get ground reference is by that ground connection you are removing. I don't see how the PI can work without it.

If neither the 4.7k or the 5k pot provide a DC path to ground for the PI cathodes, the PI could still work somehow, using the DC path via the 100k NFB series resistor and the OT secondary.

[drewl]

On older Marshall's the grounding bar soldered to the pots can cause hum over the years and can be fixed by reworking the solder connections.

[g1]

Any chance the primary OT wires have been reversed?

(edit: I guess that would not apply anyway if presence is working correctly)

[greengriff]

This thread would be a lot more useful if there was a schematic instead of a layout.
As far as I know, you are now using an older presence version like in the early1959T. In that layout pic above, the 4K7 resistor is disconnected.
So the only way for the PI cathode to get ground reference is by that ground connection you are removing. I don't see how the PI can work without it. You say the amp is working when you disconnect it, or does the noise just stop because there is no sound anymore?

https://drtube.com/schematics/marshall/1959t-66.gif

This schematic is 'nearly right' in that it shows the presence circuit as it was factory wired on my amp, but the schematic is for 6550s rather than EL34s, so shows the NFB wire coming from the 8ohm tap rather than the 4ohm. https://drtube.com/schematics/marshall/2203u.gif

You are correct in your assertion that I changed it to the older spec as per the 1959T.

The amp is still working with no path to ground for the NFB wire!!!

[greengriff]

On older Marshall's the grounding bar soldered to the pots can cause hum over the years and can be fixed by reworking the solder connections.

I connected the 4.7k resistor back to ground after I'd lifted the grounded tag of the pot, and that did not bring the noise back. The noise problem isn't a hum, it's out of control crackles and pops..

[pdf64]

Perhaps that ‘soldered to the back of the pot case’ connection just isn’t a reliable chassis connection. Better to trust higher torque fasteners with star washers.

[greengriff]

Perhaps that ‘soldered to the back of the pot case’ connection just isn’t a reliable chassis connection. Better to trust higher torque fasteners with star washers.

That's very interesting. I will try remaking the ground connection with a different method.

[greengriff]

OK, so the problem seems to be solved: The issue was a bad connection from the back of the presence pot to the bus bar. So when I'd grounded the tag of the pot to the casing it wasn't actually properly grounded. I don't know why that would have caused such horrendous noise, because when it wasn't connected at all there was silence, and when it was properly connected there was silence! So why would a bad/intermittent connection cause such a problem?

So now I have a fully functioning (old style) presence circuit, and a quiet amp!

One thing that has now appeared that was not there before: the background hum has changed character. When I turn the master volume up above half way there is a noise a bit like a 2 stroke engine on a fast idle. Previously there was just a steady hum. Could some of the shocks that when through the amp when the grounding wasn't work have damaged something?

[Helmholtz]

So why would a bad/intermittent connection cause such a problem?

The noise produced by the intermittent connection shows that the points to be connected have different potentials. With a good contact, the potential difference must be zero but a current will flow through the contact. When the contact resistance varies, the current generates a noise voltage across the contact.