Output Transformer/Power Tube impedance pairing

[jlboxlc]

If a perfectly impedance matched OT is not an option for 6L6GC build, is it not correct to go with a higher OT primary impedance ? I know "correct" may be a matter of opinion, and the effects of higher or lower are different.

[hasserl]

Fender used a higher impedance OT in the tweed 6L6 (6k6) amps than they did in the black face amps (~4k). I think most people would say that it worked out OK for them.

[Old Tele man]

...usually, you go slightly above optimum Zo for more power, and slightly below optimum Zo for lower distortion.

[Jerry Glass]

And to further add to the confusion...Fender BandMasters had a primary impedance of 20K plate to plate.

[Keyser Soze]

...usually, you go slightly above optimum Zo for more power, and slightly below optimum Zo for lower distortion.
:agree

The key word being slightly, you probably have a little more room when going lower than optimum. This often appears in the lower budget amps from the 60's and 70's, probably contributing to some of that lo-fi sound. But, go too far in either direction and you are heading for trouble.

[jlboxlc]

Looking at tube data sheets, my assumption is that effective plate to plate load resistance appears to vary with voltage level/class (class A1 push-pull vs class AB1 push-pull). So if you were going to run push-pull power tubes at a voltage somwhere in between class A and class AB, your effective load resistance would be somewhere in between, is this correct ?

[donnyjaguar]

...usually, you go slightly above optimum Zo for more power, and slightly below optimum Zo for lower distortion.

There are tricks you can do to the output section to move the impedance around. Trioding brings it right down, as does higher idle current (bias). You can of course convert from fixed to cathode bias or vice versa. You can also split the cathode resistors if cathode biased (usually bad idea). Other tricks with where you place the voltage on the screen (and suppressor) grids.

[jlboxlc]

Is it correct then that cathode biasing will "use some of the available voltage" therefore reducing the effective voltage, and therefore reduce the effective load resistance ?

[pdf64]

For a given B+ and OT, cathode bias will have a lower effective voltage, therefore lower output power. My understanding is that it would not change the effective load resistance (though I'm not totally sure what you mean by that?)