distortion on leading edge of note

[Firestorm]

Seems like the normally rejected 120 Hz ripple is modulating your output. There aren't that many ways for it to get in. I might suspect that it is getting in via the bias.

How big is your bias supply filter cap?

[martin manning]

Seems like the normally rejected 120 Hz ripple is modulating your output. There aren't that many ways for it to get in. I might suspect that it is getting in via the bias.

How big is your bias supply filter cap?

You can measure the ripple on the bias supply to see what it is, but it is half-wave rectified and therefore any ripple would be 60Hz. How much ripple voltage is on the plate and screen nodes again?

[Firestorm]

Seems like the normally rejected 120 Hz ripple is modulating your output. There aren't that many ways for it to get in. I might suspect that it is getting in via the bias.

How big is your bias supply filter cap?

You can measure the ripple on the bias supply to see what it is, but it is half-wave rectified and therefore any ripple would be 60Hz. How much ripple voltage is on the plate and screen nodes again?

True, but I wasn't thinking of the ripple. When the grids go positive, won't they try to source current from this node?

[martin manning]

Yes, but that wouldn't be a source of 120 Hz. That is only going to come from a full-wave rectified source.

[pula58]

Do you still have 100K resistors on both PI plates?

100K and 82K, with 20K trim-pot (also had 10K trim pot, but it was not enough to get the right balance).

[pula58]

Seems like the normally rejected 120 Hz ripple is modulating your output. There aren't that many ways for it to get in. I might suspect that it is getting in via the bias.

How big is your bias supply filter cap?

The first power-supply filter cap is a 500V 47uF F+T, then the choke (4H, 90mA, 108 ohms), then the screens-supply cap is another 500V 47uF F+T, the rest of the caps down the line are 500V 22uF F+T's.

[pula58]

Seems like the normally rejected 120 Hz ripple is modulating your output. There aren't that many ways for it to get in. I might suspect that it is getting in via the bias.

How big is your bias supply filter cap?

100uF, 100V F+T.
Then, the bias output is further filtered by a 10uF 100V cap.

BW, The schematics show all of this..

[pula58]

Seems like the normally rejected 120 Hz ripple is modulating your output. There aren't that many ways for it to get in. I might suspect that it is getting in via the bias. That would be the most sensitive place. I would look at the coupling caps as a possible way that could happen. Seems like the only component that you haven't tested. I would also look at the bias supply/power tube grids with the scope during the distortion.

You could put a pi filter before the power tube plate tap to significantly reduce the ripple like on a single ended amp. I don't see that as a solution but more as a test. You might also temporarily convert to cathode bias to see if that eliminates the issue.

There is a ground missing on the schematic for the pt cathodes?

Oops, yes, the schematics had a missing gnd for the output tube cathodes. But,the actual circuit does have the cathodes grounded (star gnd with PT center tap and gnd of 1st power supply filter cap)

[cxx]

Looking at some of the fender schematics that have been modified to eliminate the normal channel, some maintain the mixing resistors in front of the PI input, grounding the one that was for the normal channel, attenuating the input to the PI from the vibrato channel by 50%. You have a master at that point. Might be that the 50% wouldn't let you drive the grids to conduction.

[pula58]

Looking at some of the fender schematics that have been modified to eliminate the normal channel, some maintain the mixing resistors in front of the PI input, grounding the one that was for the normal channel, attenuating the input to the PI from the vibrato channel by 50%. You have a master at that point. Might be that the 50% wouldn't let you drive the grids to conduction.

Hi CXX.
True, the AB763's have the two 220K mixing resistors, that cut each channels output in half. But that wouldn't that just mean that the amp user would have to turn-up the vol knob more to get the amp to the edge of breakup? In my case, the amp sounds fine with max output, but as soon as the output tubes begin to distort (as the vol is turned-up from there) I get the sidebands. I am sure that if I turn the master to 1/2 and then turn up the volume control (before V1B) I'll get the same nasty effect ...but...I shall go try that experiemnt. Thanks!

meanwhile, here are some up-to-date gut shots. Maybe one of you folks will spot something. Firestorm already spotted (Thanks !) some crossed output tube heater wiring!

[Tillydog]

.. as soon as the output tubes begin to distort (as the vol is turned-up from there) I get the sidebands.

You sure it isn't the PI clipping, rather than the output grids being driven positive?

What voltage do you have at the PI anodes (plates) under no signal? What is the B+ here?

[pdf64]

Consider re-arranging the B+ grounds, such that the screen grid B node cap ground goes to the star ground, rather than being linked to the pre-amp grounds.
Also there was a suggestion on P2 about taking the B+ winding CT to the reservoir cap (then going to the star from there).
The biggest 120Hz ripple currents will be between the CT and the reservoir cap neg terminal, so it may be beneficial to avoid sharing that with anything else.
Pete

[cxx]

Looking at some of the fender schematics that have been modified to eliminate the normal channel, some maintain the mixing resistors in front of the PI input, grounding the one that was for the normal channel, attenuating the input to the PI from the vibrato channel by 50%. You have a master at that point. Might be that the 50% wouldn't let you drive the grids to conduction.

Hi CXX.
True, the AB763's have the two 220K mixing resistors, that cut each channels output in half. But that wouldn't that just mean that the amp user would have to turn-up the vol knob more to get the amp to the edge of breakup? In my case, the amp sounds fine with max output, but as soon as the output tubes begin to distort (as the vol is turned-up from there) I get the sidebands. I am sure that if I turn the master to 1/2 and then turn up the volume control (before V1B) I'll get the same nasty effect ...but...I shall go try that experiemnt. Thanks!


meanwhile, here are some up-to-date gut shots. Maybe one of you folks will spot something. Firestorm already spotted (Thanks !) some crossed output tube heater wiring!

I built the a normal channel blackface without the divider and was surprised at how much gain was provided by the two gain stages with a TMB tonestack between and the PI. More than I could use actually. Then again, this was a low voltage implementation of about 350 on the plates.

The express only has the 3 unattenuated gain stages with a fendery TMB tonestack.

[pula58]

Some more lab data:

1) the ripple on B+ ("A" in schematics) is 13.2V p-p (seems about right)
2) the ripple at output tube plates (at the sockets) = 13.3V p-p
3) the ripple at bias voltage (where the two 220K output tube grid R's meet) is 12mV p-p @ 60Hz
3) the ripple at the screen supply is 120mV p-p (using 4H choke, 47uF at 120Hz this is consistent with ripple at B+)
4) chaning the grounding of the screen supply filter cap to the star ground connection made no difference in the sidebands
5) adding 47uF to bias voltage on the 220K output tube grid R's made no difference in the sidebands.
6) the phase inverter is not clipping, it is definietly the grids of the output tubes going into conduction that clips the phase inverter output. The PI outputs can swing all the way to the PI supply voltage (currently at ~ 350V), but it is getting clipped long before it gets there.

[JazzGuitarGimp]

Have you tried replacing or re-tensioning the output tube sockets? I am wondering if perhaps one of the tubes is not operating because of an open connection in the socket. If this were the case, you'd lose the benefit of the self-canceling push-pull topology.