6AS7GA as push pull output?

[kodabmx]

Yes, I do. I'll PM you.

[kodabmx]

Originally Posted by PRR

Which means HUGE drive voltage. For audio, it tends to need a driver supply B+ higher than the power stage B+, an awkward combination.

This is my solution to that. This is the PSU for the amp schematic I posted above.

[jhstewart9]

Originally Posted by kodabmx

The VA/PI/Driver in my 6P36S "Black Edition" amp will easily drive 6AS7G to full output with room to spare for a gNFB loop.

I have just made the VA/PI and Driver sections of this into separate circuit boards. Let me know if you're interested in a set

This copied from RCA Receiving Tube Manual RC16 published in 1950. Not much is new in audio under the Sun or anywhere else.
Bootstrapping avoids the HV PS.

[rongon]

Originally Posted by kodabmx

The VA/PI/Driver in my 6P36S "Black Edition" amp will easily drive 6AS7G to full output with room to spare for a gNFB loop.

I have just made the VA/PI and Driver sections of this into separate circuit boards. Let me know if you're interested in a set

Wow, the Super Williamson Driver (sorta). Interesting what you did with the 6SN7 (6N8S) differential driver stage. Don't the unbypassed cathode resistors introduce some degenerative feedback, raising the output impedance from that stage? Although I'm sure it also reduces distortion there, which is a good thing.

For an amp that big, it makes sense to have a separate plate supply transformer for the output tubes. The driver stages don't need that much of a transformer for their plate supplies. I did it that way once, to get a 440V B+ to a driver stage for a pair of 300Bs with 360V B+. I think the separate power transformers helped the sound quality (but I can't be sure of that).
--

[kodabmx]

They raise the Zout, but the grid leak for the following stage is 330k so it doesn't matter, and there's plenty of gain already Depending on the amp, I either use DC boost for the front end, or a quadrupler off of the doubler as in post 252. One older design used a choke input for the output stage, and cap input for the front end giving 360V and 560V respectively from an 800VCT transformer.

[rongon]

They raise the Zout, but the grid leak for the following stage is 330k so it doesn't matter, and there's plenty of gain already

You know, that's one of the problems with the Williamson topology.

If using triode output tubes like 2A3 or triode-wired EL34 in Class A, I find the sensitivity is a bit too high, like only 200mV to full power, when I'd prefer something like 1V to full power. Reducing the gain at the diff driver stage could work nicely for that.

Also, I've noticed that the diff driver stage is where the highest signal levels are, and consequently where distortion levels are highest. Reducing the distortion there should make a pretty big difference in the distortion performance of the whole amp.

Nice trick, I'm gonna steal it from you!