I've unsoldered the wiper before and still heard the signal, so it's not likely to be the pot. I'm definitely going to try a jumper from the pot output straight to V2, bypassing some of the tone stack to see if it helps.Phil_S wrote: ↑Thu Dec 08, 2022 1:56 am The pots are old. I wonder if 0 and 1 are the same. IOW, 0 used to really be off, but isn't any more. Maybe you can unsolder the pot and see what it meters when on the 0 position. Maybe the pot needs replacement? There are a variety of possibilities. If you want to solve the problem, I suggest you hunker down for the long haul. Eventually, you'll find it.
Help with signal bleed on a 1964 Gibson GA-30 RVT
Moderators: pompeiisneaks, Colossal
Re: Help with signal bleed on a 1964 Gibson GA-30 RVT
Re: Help with signal bleed on a 1964 Gibson GA-30 RVT
So a few years ago I built a custom jazz amp heavily based on a Gibson GA-75 for someone (single channel version with a number of extra features). There are some differences to your GA-30 RVT but also a lot of similarities. I had exactly the same problem that you are describing!
I was pulling my hair out trying to work out what it was. Like you, everything was pointing to preamp valve cross-talk and in my case, because I'd gone single channel it was between sections in the same glass bottle. I posted about it here and elsewhere, and as you say it seems to be not uncommon but no solutions. Just like in this thread, everyone told me it couldn't be cross-talk. Everyone pointed to how it happens in no other amp and how the cross-talk spec is so low it couldn't happen. I tested more and absolutely proved to myself that was the cause.
The GA-30 has a different tonestack than the GA-75, but it's still got a higher level of resistance between the volume/loudness pot and the second stage grid due to the mid T filter. Very often (not always) amp designs connect the volume pot wiper directly to the next stage grid, or might have a low value grid stopper resistance. So when the volume is turned right down, the next stage grid is directly or low resistance grounded, which prevents any stray pickup either from cross-talk inside the bottle or from nearby circuitry. In the case of these GA amps, volume turned down still has significant resistance to ground for the next stage grid and this leaves it susceptible to bleed through as you describe.
There are some other amps which have high resistance in series with leading into the next stage grid. I have another amp with this (6G6-B Bassman bass channel 4th stage grid has treble wiper, high resistance to AC ground unless you have the treble turned all the way up), so I got testing on it and found that there was actually very low volume bleed through that could be heard at the speaker that I'd never really noticed before. Measurement confirmed that it was the same phenomena but at a lower level. Turned out the difference in level was totally explained by difference in amp gain from the point of cross-talk injection to output to the speaker, and the cross-talk level I measured at only the preamp stage was the same. It is sensitive to preamp valve internal gain and internal construction. I had an old lower gain 12AX7 that was worse. New production with higher gain varied a bit across manufacturers.
I'm probably going to be one of the only people to tell you this, but your diagnosis is correct. After many, many hair pulling hours I came to exactly the same conclusion. It is a 'feature' of this family of amps.
As I was building a custom amp, my solution was to move the volume pot to downstream of the GA-75 James tonestack. I modeled both in LTSpice and convinced myself that the tone response was close enough that I could do this swap of order. Of course with a volume pot wiper now directly feeding the next stage grid, when the pot was turned to zero the amp became dead quiet as expected.
You could consider doing the same, to rewire the loudness pot to be after the bass and mid tone circuits. Or if you don't want to modify an old amp to that extent, accept the 'feature' and just never play the amp with the knob on zero.
In the case of these Gibson amps, there's nothing functionally wrong with it.
I was pulling my hair out trying to work out what it was. Like you, everything was pointing to preamp valve cross-talk and in my case, because I'd gone single channel it was between sections in the same glass bottle. I posted about it here and elsewhere, and as you say it seems to be not uncommon but no solutions. Just like in this thread, everyone told me it couldn't be cross-talk. Everyone pointed to how it happens in no other amp and how the cross-talk spec is so low it couldn't happen. I tested more and absolutely proved to myself that was the cause.
The GA-30 has a different tonestack than the GA-75, but it's still got a higher level of resistance between the volume/loudness pot and the second stage grid due to the mid T filter. Very often (not always) amp designs connect the volume pot wiper directly to the next stage grid, or might have a low value grid stopper resistance. So when the volume is turned right down, the next stage grid is directly or low resistance grounded, which prevents any stray pickup either from cross-talk inside the bottle or from nearby circuitry. In the case of these GA amps, volume turned down still has significant resistance to ground for the next stage grid and this leaves it susceptible to bleed through as you describe.
There are some other amps which have high resistance in series with leading into the next stage grid. I have another amp with this (6G6-B Bassman bass channel 4th stage grid has treble wiper, high resistance to AC ground unless you have the treble turned all the way up), so I got testing on it and found that there was actually very low volume bleed through that could be heard at the speaker that I'd never really noticed before. Measurement confirmed that it was the same phenomena but at a lower level. Turned out the difference in level was totally explained by difference in amp gain from the point of cross-talk injection to output to the speaker, and the cross-talk level I measured at only the preamp stage was the same. It is sensitive to preamp valve internal gain and internal construction. I had an old lower gain 12AX7 that was worse. New production with higher gain varied a bit across manufacturers.
I'm probably going to be one of the only people to tell you this, but your diagnosis is correct. After many, many hair pulling hours I came to exactly the same conclusion. It is a 'feature' of this family of amps.
As I was building a custom amp, my solution was to move the volume pot to downstream of the GA-75 James tonestack. I modeled both in LTSpice and convinced myself that the tone response was close enough that I could do this swap of order. Of course with a volume pot wiper now directly feeding the next stage grid, when the pot was turned to zero the amp became dead quiet as expected.
You could consider doing the same, to rewire the loudness pot to be after the bass and mid tone circuits. Or if you don't want to modify an old amp to that extent, accept the 'feature' and just never play the amp with the knob on zero.
In the case of these Gibson amps, there's nothing functionally wrong with it.Re: Help with signal bleed on a 1964 Gibson GA-30 RVT
I finally got around to trying out a couple of the suggestions, and here's what i've found:
There's definitely crosstalk in the first tube, with the input going into channel 2, i can see a large output on the plate of V1B and a small signal on V1A.
Unsoldering the connection from the loudness wiper to R44 and attaching it straight to the V2B grid didn't remove the overall signal, but i no longer had an output on the plate of V2B. This gave me the idea to do the same experiment to the other channel, which got rid of the signal completely when loudness was at 0.
With that modification the loudness pot behaved as expected: at 0 there was absolutely no sound coming through.
I guess ultimately it could be as katopan said: there's a design flaw and this is how the amp behaves. I'm not sure if some of them are wired slightly differently and it only happens on some of the amps, or if someone's rewired it already, but it seems clear now that the crosstalk happens in V1 and is then passed through by the tone stack, probably mostly through the mid pot.
There's definitely crosstalk in the first tube, with the input going into channel 2, i can see a large output on the plate of V1B and a small signal on V1A.
Unsoldering the connection from the loudness wiper to R44 and attaching it straight to the V2B grid didn't remove the overall signal, but i no longer had an output on the plate of V2B. This gave me the idea to do the same experiment to the other channel, which got rid of the signal completely when loudness was at 0.
With that modification the loudness pot behaved as expected: at 0 there was absolutely no sound coming through.
I guess ultimately it could be as katopan said: there's a design flaw and this is how the amp behaves. I'm not sure if some of them are wired slightly differently and it only happens on some of the amps, or if someone's rewired it already, but it seems clear now that the crosstalk happens in V1 and is then passed through by the tone stack, probably mostly through the mid pot.