> Isn't there a problem with running tube power amps without load (or with a higher impedance than intended)?
Unlikely.
Yes, with old shellac insulation, if you left the transformer unloaded, and drove the signal up past clipping, you "could" punch-through insulation. And yet, I never have, in decades of abusing tube and transformer amps. I think modern poly insulation is tough enough.
Also you are probably only going to have 300VDC on that transformer. And using triodes and the soft-pentodes like EL84, the plate resisance will damp any high-voltage ringing.
> trans will be for pentode operation, triode changes impedance to lower value. How low? I need help on that one.
With pentodes, the plate resistance does NOT matter to the optimum load impedance. With pentodes you pick a Plate and Screen voltage, draw a load-line up to the knee of the zero-bias curve, that's your max-power load.
With triodes, your max-power load-line and best efficiency and lowest THD is always the highest plate voltage you can stand and a high load resistance. For low-Rp triodes, you don't have to go to extremes. 6EM7 or 2A3 will be good 2K and up. For high-Rp triodes, which includes most audio pentodes triode-strapped, higher voltage and loads are needed for good power and low THD. (That's the big difference between the tubes optimized for audio-triode use, and ones optimized for pentode mode. OTOH, some of the TV H-sweep pentodes make nice low-Rp triodes.)
EL84 triode curves... uh.... hmmm. There ISN'T any good loadline. Plate voltage rating is low for the plate resistance. 3K5, 5K, 35mA, 50mA, all about the same power and similar THD: 1.5W and 6%. In Pentode, it can make twice the power at the same THD number, four times the power at 10%THD. EL84 is a great little pentode for the same reasons it is a lame triode.
