So with my makeshift L8 choke in there ran more tests with very low excitation in the tuned input mode. Things appeared stable at about 100W out of the amp, except the right side tube was glowing a pretty good RED color, while the other was not. I put in another that I felt was a working tube, and that glow went away and I gradually worked my way up to about 500W output, and made several long ssb transissions on 80m last night. The tubes showed barely any red at all after all that work, so I believe things are stabilized.

L8 may have been getting too hot because testing the chassis without the proper air flow for the compartment fan may have been asking too much. So I limited my tests to strictly the TUNE mode and short transmissions no more than 300w. Then I put the covers back on, made sure the fans were working and upped power levels.

Also painted the cover flat black:
https://www.mediafire.com/view/r8gw3...t_blk.jpg/file

I went in and put the plate switch back into the front panel and filled the lamp lense with a NE2 neon lamp and dropping resistor. The switch now takes 115Vac, and switches it
back through an unused fuse holder through the main input AC receptacles non-used pin, which leads back to the plate relay in the power supply. So now its easier to shut the plate supply on or off, which comes in handy when not being used, and if something goes wrong, that switch is very close by.

Couple of other things still need to be done, like placing the 115Vac on the 115 primary winding for T103, instead of the 110 winding, and perhaps some of the decoupling caps
on the 400A sockets ought to be replaced. They seem to be drying out because of the heat nearby and that usually means they will eventually short.

I may also try another pi-type L8 and ditch my home brew sub, but I am tempted to leave it as-is because things are working well.