Calculating Output Transformer Primary Impedance

[Bob M.]

Hi jazbo8,

I'm not really sure what my bias is set at but knowing my proclivities, I would say with 2x 6L6GC at 440Vdc plate voltage, I might bias somewhere around 30mA per tube, maybe a little less, depending. However, I'm not sure or don't necessarily see the connection between bias and which output transformer tap I'm using. With the output transformer I had on hand, 6,600 at 8 ohms is a little high for the textbooks (using 6L6 type tubes) so I experimented to see what difference the next lowest tap (3,300) would make. After one year's playing time, I was unable to discern any real-world difference between these two impedances choices. I'm not really asking a question, per se, just adding my experience in with the other opinions about how finely to match tubes, impedance, transformer taps and speaker ohmage using any particular output (audio) transformer. I was hoping for some great opinions and insights like we've had here in the previous posts on this excellent thread. In short, I don't have a question. I'm just soliciting informed comments. Thanks for contributing.

Bob M.

[jazbo8]

The bias has an effect on the plate resistance thus the amplifier's output power, plate dissipation and distortion level. I drew two load lines and as you can see with the 3.3k, the plate dissipation is quite high whereas the 6.6k looks just about right. And I am still somewhat surprised that there's not much difference between the two settings - are you typically playing at low or high level?

[Old Tele man]

This might help somewhat: http://thermionic.info/mccaul/McCaul...nRule_2008.pdf

[Gingertube]

That helps a lot.
I had always assumed that the "70% Rule" was to allow the use of a X2 higher Rg1 so as to not load the phase Splitter too much - it seems that that is ONLY half (or less than half the story).

Cheers,
Ian

[guitician]

I think old fender power supplies would load down and drop voltage as the power tubes drew more current. So, the tubes never really burned up with a fixed resistor bias voltage common to those amps, which were loaded with whatever tubes the radio store sold.

[J M Fahey]

FWIW I design and build my own, and afterwards *measure* them to verify that real world matches math.

I begin with published load lines which already give me a working result, and then tweak to improve it, with the real tubes I bought over the counter and not expensive NOS.

FWIW with modern generic tubes , 400 something +V (say, 420 to 450) and 1K "safe to 6L6" screen resistors, 5500 ohms give me best power (above 40W RMS *clean*) from almost any easily available 6L6 pair.

Which would be around 50W if you cheat a little and let tops reach visible rounding and a little clipping or some wave kinking because you reach power tube grid rectification.
Scale to taste to 4 or 6 6L6.

That said: 4 x 6L6 into 4 ohms? ... look no further than "anything that fits a Twin Reverb"