Ugly distortion on '80 Fender Twin Reverb

[Sansteeth]

[EDITED THE Vac MEASUREMENTS)

Hey everyone,
I'm kind of going crazy over here on this '80 Fender Twin reverb (the one with the push-pull master). Beware it's a long read!
Came in for a full checkup, some of the mallory caps in the doghouse where either bulgy or open, I changed them all.
The 25uF bypass caps got the same treatment.
The 250pF at the tone stack of both channels were leaky and changed for 1kV ceramic caps. The owner wanted to go for a full retubing so that was done as well.



I'm having issues because the "Vibrato" channel starts crunching at 3 (out of 10 on the volume pot) and it's not a pretty distortion, push-pull thingy is disengaged.

The normal channel stays mostly clean up to 10 (out of 10). Tonestack on both channels at 5.5 out of 10.



BEFORE THE 3.3M/10pF mixer:

When I scope after the .022uF coupling cab on V2B plate of the VIB channel, I’m getting more or less the same signal as the NORM channel when I scope the .047uF coupling cap at the plate of V1B. Up to that point both circuits are the same (except for that last coupling cap). It is clean all the way to 10 on the volume, tonestack at 5.5/10. Getting around 130Vac Peak to peak

AFTER THE 3.3M/10pF mixer:

However, if I scope directly after the 3.3M/10pF mixer on the VIB channel (grid of V4B), I’m seeing a much much smaller signal, around 2Vac and it’s clipped pretty hard on the positive side. To go back to an unclipped signal at the grid of V4B, I need to turn the VOL pot back down to 4/10.


I’m thinking of grid current here, but the signal coming in the grid of V4B is so small (6Vac peak to peak) that I can’t see how grid current would develop there.


Here are the voltages (no signal):

V4B(Vk): 1.95Vdc

V4B(Vp): 252Vdc

V4B(Vg1): 0.13Vdc

What I did, from targeted and thought-out troubleshooting to desperate shotguning when madness kicks in in the workshop:

- tested the push-pull switch on the master volume pot, could be grimy and always on: NOPE

- replaced the tube: NOPE

- clean up the wax on that part of the board, I was getting stray voltages that could definitely mess up with the bias of V4.

- bypass the Reverb circuit entirely (go straight from the coupling cap of V2B plate to the 3.3M/10pF to the grid of V4B, with flying leads, not soldered to the board)

I got a big old signal coming into the grid of V4B clipping the heck out of the signal coming out of the V4B’s plate, made the problem worse.

- replaced the cathode resistor and bypass cap: NOPE

- replaced the tube socket: NOPE

- replaced the plate resistor: NOPE



It’s kicking my butt. Even with the funky push-pull thingy that makes it seem like they tried to push more crunch out a Twin reverb, I have a hard time believing the thing distorts at 4/10 on the neck pickup of a Stratocaster when the switch is disengaged.
 The circuit is also identical to a regular master volume SF Twin Reverb. but the push-pull does do something to the circuit even disengaged.
Any help on what's going on on the grid of V4B or people having experience with those particular Twin Reverb?
I know them to be extremely clean, heck that's the reason why people love/hate them (besides the weight), never heard of a model that gets fuzzy at 6/10 on the dial...
Thanks a lot folks!

[tubeswell]

What are the idle voltages on each HT supply node?

[Sansteeth]

Output TX CT: 533Vdc
Reverb TX: 430Vdc
PI node: 413Vdc
Preamp tube node: 379Vdc

On the plates of the preamp tubes I'm getting:
Norm channel
V1A: 241Vdc
V1B: 257Vdc

VIB channel
V2A: 259Vdc
V2B:236Vdc

Reverb driver
V3: 410Vdc

Reverb recovery
V4A: 250Vdc

VIB Chan 3rd gain stage
V4B: 254Vdc

Phase Inverter
V6A: 255Vdc
V6B: 248Vdc

[pdf64]

... if I scope directly after the 3.3M/10pF mixer on the VIB channel (grid of V4B), I’m seeing a much much smaller signal, around 2Vac and it’s clipped pretty hard on the positive side. To go back to an unclipped signal at the grid of V4B, I need to turn the VOL pot back down to 4/10.


I’m thinking of grid current here, but the signal coming in the grid of V4B is so small (2Vac) that I can’t see how grid current would develop there.


Here are the voltages (no signal):

V4B(Vk): 1.95Vdc

...

By a signal of 2V AC, do you mean peak (ie 4V peak to peak), or 2V peak to peak (ie 1V peak)?
Or 2V rms (5.6V pk-pk)?

If the cathode is at 2V DC, the grid signal will clip when the positive peaks approach 2V (4V pk-pk).

Has the V3 cathode been bypassed?

What exact model is it, 135W, TRII, evil twin?
If it's 80s it's not going to be the '100W with switch' schematic you've attached.
The 999K-HD pots used for volumes that I've investigated seem to be a linear taper, so with those, the preamp will clip at a lower number on the dial than with a 30-35% J taper pot.
https://el34world.com/charts/Schematics/files/Fender/Fender_twin_reverb_sf_135_schem.pdf

[Sansteeth]

By a signal of 2V AC, do you mean peak (ie 4V peak to peak), or 2V peak to peak (ie 1V peak)?
Or 2V rms (5.6V pk-pk)?

My bad, I don't know where those measurements come from  , with VOL pot on 10, I'm getting 133Vac peak to peak before the 3.3M/10pF mixer and 6.5Vac peak-to-peak, but clipped on the positive side) after it, on the grid of V4B. If I want to have an unclipped signal on the grid of V4B, I need to turn the VOL pot back down to 4.5/10. As I turned it down, only the negative side goes down (the clipped positive side stays clamped) until I get to around 4 or 5 on the VOL pot.

I will edit my measurements in the original post.

Has the V3 cathode been bypassed?

No, it's a plain 680R cathode resistor, but I bypassed the entire reverb circuit and I'm getting the same results

What exact model is it, 135W, TRII, evil twin?
If it's 80s it's not going to be the '100W with switch' schematic you've attached.
The 999K-HD pots used for volumes that I've investigated seem to be a linear taper, so with those, the preamp will clip at a lower number on the dial than with a 30-35% J taper pot.
https://el34world.com/charts/Schematics/files/Fender/Fender_twin_reverb_sf_135_schem.pdf

Fender Twin Reverb Silverface 135W with Ultralinear power section, I did attach this schematics to the original post, I felt that for all intent and puropose the "100W + switch" was easier on the eye.

The 999K-HD pots used for volumes that I've investigated seem to be a linear taper, so with those, the preamp will clip at a lower number on the dial than with a 30-35% J taper pot.

Interesting, I guess that could account for some, but that's a SF silverface that stays clean up to 10 on the Normal channel and starts crunching at 3 on the "vibrato"  channel, I'd be tempted to tell my customer: "lucky you! You got the amp that everyone wanted! A Twin reverb that crunches!" but the distortion is pretty bad haha.