This is what I suspected. It does not account for the behaviour of RC coupled amps using the GM70. 4-600 V of B+ and reasonable disipation leave a lot of the available bias taken up by the 20vac filament. Yet they seem to survive and sound good, not to mention put out power like they had unipotential cathodes and could use all the bias V...So, now I have been left teh problem of computing the g1 Z and see what the curve looks like.
regards,
Dogulas