I am working on the 2A3 circuit. With the power Iron, a suitable choke must be specified. I don't know if an off-the-rack Hammond would be a better deal than a custom from Heyboer. Going to have to examine that one a bit...GZ34 rectifier, and one motor run ought to take care of the rest. A bias supply capable of -75 volts is easy enough. A votlage doubler with 30vac input and decent caps will be adequate I think.

And then there is the OPT design. I think a 6k6 with 10-20% E-Linear taps would be an ideal sol'n. Something like an Acrosound TO-300 with a modified tap location. I did some work with a PP 6A5 amp and found that a bit of proper NFB was better than none. The usual doctrine says no NFB with Triodes, but I have reached other conclusions.

We need a driver which can do 120V p-p per phase. Out of a 250-260 VDC B+, this puts some restriction on the plate resistance( lower plate z valves have a more nearly vertical V-grid=0 plate line). The closer to vertical it is the greater percentage of B+ you can swing. Look at 6SL7 and 6SN7 curves, try a 60k plate load and see which can swing more AC.

A 5687 is a likely suspect. It would be a bit light on the gain, but with Guinevere at the input, I think it can be done quite well. It is also a bit simpler with one stage. ECC99 is slightly higher gain, and is reputed to be a good sounding valve too.

The grid choke is available from a few suppliers and is a *BIG* benefit to sonics. Every single amp I have wired one into has been improved drastically over resistive grid circuits. Even fairly low value ones ( down to 47k v. the usual 100k, or 270k for a Dynaco St.70 ).

that's the bones of it at this time now. You could also do a SE 300B at a fairly conservative OP point with the Guinevere power Iron. Finding OPT Iron for SE is not something I have any practice doing.
regard,
Douglas