Can't shake volume bleed on Deluxe Reverb pull boost

[Diverted]

I have an annoying issue with a Deluxe Reverb pull boost owned by a friend of mine.
This is a late 70s pull boost Deluxe Reverb. He said it was having intermittent volume drops, phantom reverb bursts etc.
I've eliminated all that stuff and the amp is playing nicely but I can't seem to root out a volume bleed in either channel when volume is set to 0. In the normal channel the bleeds sounds clean; in the vibrato channel it's overdriven (extra gain stage?). In both pots, there's a sweet spot just up from 0 where the volume drops off a bit but still doesn't completely go away.

I've tightened all the grounds in the amp, reflowed the ground solder joints on the volume pots, swapped myriad tubes, checked pots' resistance to ground (0 ohms) played around with the lead dress hoping I could reduce it ... nothing doing.

Here is the list of things I've done on the amp:

1. Stock power rail resistors are 2.2K and 10K. Someone had recapped power supply and used 1K and 4.7K in those spots. I removed them and put stock values back in.

2. Replaced 250pf cap in vibrato channel tone stack. First one was bad which was causing the tone stack to behave oddly (bass behaved like a volume control, treble control almost non-existent). It works as it should now.

3. Replaced all cathode bypass caps with correct values, and replaced bias cap as well.

4. Previous to me someone had clipped the 6.8K midrange resistor to ground and put in 5.1K. I took it out and put a 6.8K back in.

Right now the voltages are all about 30 or so volts high but I've been running it at wall voltage which is about 124vac. The amp is biased at around 62-65 percent.

I have a hard time believing that the pots on both channels are bad?

[shooter]

there's like 700 schematics to sift n I'm camping 

the one I looked at, I would disconnect the "top" of both volume pots, disconnect the TS and use a cap to "bypass" the TS, so V1a out, test cap, top of vol pot.  I'd probably just do both at once n make sure it "changed" the problem, then put one channel back to normal, test.

[Diverted]

Oh, sorry about that. Here it is:

https://el34world.com/charts/Schematics/files/Fender/Fender_deluxe_reverb_boost_schem.pdf

Didn’t quite follow your suggestion. Anyway with no changes made:
V1 out, bleed still there on vibrato channel.
V2 out, bleed still there on normal channel.

Thanks, have fun camping!

[Diverted]

I can mask the issue by putting a 5751 in v1 or V2. But that’s not solving the issue, just hiding it.

With 12AX7s in, voltages are way high at:

V1a
Plate 224
Cathode 1.7

V1b
Plate 242
Cathode 1.8

V2a
Plate 230
Cathode 1.6

V2b
Plate 233
Cathode 1.5

This is with a b+ of 420 (very close)
B node 408
C node 392
D node 331

Should I try upping the power rail resistors to bring D node voltages down closer to spec?

[sluckey]

I think you should bridge a 20µF cap across each filter cap (one at a time) before doing anything else.

[Diverted]

What’s the reasoning? More filtering, less ripple to all the nodes? And how would that possibly affect this issue? Don’t doubt at all, would just like to understand what’s going on here better. Thanks! will give it a go.

[pdf64]

If each power supply node isn’t adequately decoupled (eg due to ecap with high esr), signal will end up in unexpected places, as per the symptoms here.
Another thing to try is a completely different brand / manufacturer of 12AX7. Once in a while, for reasons unknown, some batches seem to have excessive coupling between sections.

[Diverted]

Interesting, that’s great to know. I’ve swapped out 12AX7s ad nauseum. I’ll beef up those electrolytics, thank you.

[sluckey]

I’ll beef up those electrolytics

My suggestion is not about "beefing up" the caps. If you bridge a cap and that cures the problem, then replace the bridged cap.

A 40µF @ 450V cap with a couple gator clip leads soldered to it was a very common troubleshooting tool back in the '50 and '60s. If you are gonna work on this stuff you need one too.

[Diverted]

I did try bridging them one by one with an FT 20@500v. Unfortunately it did not cure the issue.

Thanks for the tip though. In all I’ve read I have never heard of that tip.

[sluckey]

Well now you know it's probably not the filter caps.

[Diverted]

I’m looking over the schematic and see a couple of things that don’t add up/jibe with the amp:

1. The 1200 pf caps from 6v6 control grids to ground have either been clipped out or were never put in.
2. The 47 pf cap in the vibrato channel tone stack has been clipped out or was never installed.
3. The 330k resistors in the phase inverter are 220k.

I’m pretty sure I’m looking at the pull boost PR as it has the pull boost circuit built into the reverb circuit and looks definitely original. I’m wondering if there are other versions of this amp around that have these differences in them.

Edit: pic of chassis before cathode bypass and bias caps replaced

[PRR]

> a volume bleed

Since the dummy clean and tighten first-fixes aren't helping, maybe use ears. What is this "bleed"? Is it full fidelity? Tinny? Just the boom?

Did we cover the stupid common cathode cap on V2 point [A] and V4 point [E]? That's one bad idea which is easily fixed with two separate 1.5k+22uFd R-C networks, maybe taken to different grounds. In fact point [A] may be the root of the problem. Or the solder-blobs where they go to chassis.

[Diverted]

Sounds normal in the normal channel and overdriven a bit in the vibrato.

I was laying in bed last night trying to think of where to look next and I considered separating those cathodes. It’s on the list, taking day off👍
Thanks!

[pdf64]

It's probably not that the cathodes need separating per se, rather that the design choice of sharing cathodes results in poor resilience to aging ecaps (ie failure mode of excessive ESR).

[Diverted]

I separated the cathodes and that got rid of the issue in the vibrato channel, but not normal. So spent the next hour rolling tubes and playing with lead dress, and was finally able to get normal channel quiet.
I went over all the numbers and everything is about 20-25 percent high with 123VAC going in. It sounds good.

Had a little startle when I measured voltages on V5A plate and got 430-odd. This was with tremolo off (not grounded). Turned tremolo on and plate voltage fell to around 200-300 fluctuating. I'm assuming this is because with tremolo off circuit is not loaded down? I have not measured voltages in this type of tremolo before with circuit on and off so I don't know if this is normal. I'm assuming it is. Tremolo sounds good.

Thanks for the tips so far!

[sluckey]

Had a little startle when I measured voltages on V5A plate and got 430-odd. This was with tremolo off (not grounded). Turned tremolo on and plate voltage fell to around 200-300 fluctuating. I'm assuming this is because with tremolo off circuit is not loaded down? I have not measured voltages in this type of tremolo before with circuit on and off so I don't know if this is normal. I'm assuming it is. Tremolo sounds good.

That's quite normal.

[Diverted]

Thanks, and thanks for the schooling on ESR etc.

[PRR]

> V5A plate and got 430-odd. This was with tremolo off

This one has the negative 60V on the oscillator grid for "off". -60V turns the 12AX7 OFF. Plate passes zero current through plate resistor. Plate should rise to supply voltage (B), noted as 415V, but possibly 430V on yours. Normal.

Trem "on", LFO grid is biased to zero volts via 1Meg. With these circuit values we would expect plate to fall to 250V-300V. But then it starts oscillating (quick because of the 60V jump). If it does not clip, oscillation increases. When it clips, grid diode drives grid negative across the 1Meg. We expect to eventually end up with a small negative grid voltage and the *average* plate voltage about as noted. (Many DMMs won't read the average DC properly with the huge trem swing on top.) Of course usually we don't investigate if the trem is working.