Glen you want a load bank that is wound as non-inductive. Wire wounds resistors are usually quite inductive (especially these big ones) which is why I do not use them for the emitter resistors (although PL did). Wire wounds can be had in non inductive but they are specially wound to cancel the inductance normally created when you coil up (spiral up) a lot of wire around a tube.
You probably do not have an LCR bridge to test the load resistors. As a quick test, see if a 0.1uF mylar or ceramic type cap tacked right across the resistor itself makes the situation better or worse to give us a clue.
How long are your test wires that go out to the load resistor (these are also inductive if not paired closely together)?
Joe the 300 watt had inductance of 2 uh. The 100 watt is 14.5 uh
I put the 0.01 uf cap across the resistor and it made it worse
I then undid the small call and the single was good. Some time had past from powering on
I then left it with smooth sine at 15 volt current. I had 2 amp fuses installed and then after about 3 minutes the b- fuse when out.
I guess it was close to load but should not have gone I thought. I wil put in some three amp and run again at 15 volts to see.
So it looks like for now this signal distortion settles done after some ideal time, powered on
Still looking into things. I will retest the cap over the resistor again , as well inverting the leads,
Still missing row six transistors.
Joe how about the .01 caps on the bridge, could one of those influence the output?