I'm about out of things to check. Both have a problem where when the input is brought up to near 1vrms they start to hiccup. I don't think it has anything to do with the layout because it's much better than the original. It's better but not that it had to be. There is an analog common that is 10 ohm from the chassis. The resistance from the supply to the common is low single digit mOhms and nearly a zero inductance.
I changed the feedback wire back to a wire (from a shielded cable) and ran it directly. That made no difference.
I, and others, have sanity checked the wiring and we're sure it's wired correctly.
1) In your specs you mentioned 1Vrms in = 125W out. That implies a gain of about 31. My gain is 19 on both. The gain is the same when just two power tubes are installed. Does that give us a hint of what may be going on? Could a low gain be screwing up the stability?
2) I removed the power tubes and measured the output of the driver board. I adjusted the AC in for a 475Vdc B+. The drivers were happy with even 2Vrms input where they both output about 105Vrms. Is this correct?
3) What is the nominal target B+ voltage?
4) It's starting to smell like the output transformers. I made measurements on the windings but because their inductance is so high I don't think the readings are reliable. Have you ever had trouble with the output transformers?
This is a pic of the output while it's hiccuping. It cycles at about 10Hz. Anybody have an idea of what could cause this?
Also, a pic of how they are wired now.