AIMS 6550 2 channel, reverb vibrato amp

[Bruce / Mission Amps]

I have one in for a complete refurb and a distortion complaint. It was dead stock.
I suspect it is just another version of a Twin Reverb but with two 6550s and big ass power and output trannys.

I've changed out ALL the original E-caps, installed a new matched pair of Svet 6550s, built an adjustable bias supply for it and fixed a few minor issues... yes, vibrato ticks like a Fender.
The amp still has an ugly wave form at medium or full level output.
This odd figure on the scope looks what you'd see with a bad power tube on one side of the OT center but after swapping three sets of tubes in with little change, I started thinking I just have a bogus OT.
So, I swapped out the stock output tranny with another big OT of close primary Zed... yup, the ugly wave form is still very similar to the amp with the original OT.
I'm thinking some odd NFB issue with oscillations and all that because it does this off lopsided clipped thing even more with the NFB loop disconnected.
It has a typical Fendery NFB configuration with 100hm/820ohm pair off the OT's 4 ohm secondary.
What's weird is that it shows clean sine waves coming out of the phase inverter coupling caps and on into the two 6550 grids though.

At 120vac input, allowing for up +10% distortion, it is only making about 65-75 watts with 535vdc on the plates and the two power tubes idling at about 18-20 watts each.
Then it gets ugly on my scope, lopsided and has the look of peaky parasitic sq wave oscillations.
This is only the second one I've ever seen but the owner is convinced it should be making at least 120 watts clean. I doubt it as the original OT is only about 20% bigger then a 45 watt Super Reverb or 50 watt Bassman OT.
Anybody have a schematic for this?
The one at Schematic Heaven is not this amp.

[Enzo]

120 watts clean with two 6550s?

Well, just a thought, how about the screens? What is happening there, and how are the screen resistors? Someone didn;t stick in a 100k in place of a 1k or something.

If the signal is clean and more or less equal at the control grids, that part would seem OK.

The NFB would try to correct distortion unless it was the cause of it, so it seems OK that disconnecting it increases the problem. But disconnecting it also prevents the NFB from obscuring anything, so leave it off for tests.

Assuming the plate voltages are about the same, how does the wave look at the plates?