>
I had no idea i would use anything differant than standard marshall style 3.3 kStandard Marshall is fixed-bias. The idle current can be small; power consumption will rise with signal.
Cathode-bias, the idle current and power consumption *must* be nearly as high as the full-roar power consumption.
So either you run a lower supply voltage, or a higher impedance.
>
pick one up that would give me around 400 VoltsOK. EL34 Pdiss is 25W-30W, 25W is safer. 50W per pair.
Since efficiency will not exceed 50%, we already know power-out will be around 25 Watts. Is that acceptable?
At 400V, 50 Watts is 40W/400V= 0.125 Amps or 125mA.
400V at 0.125A is like a 3,200 ohm resistor.
The push-pull plate-to-plate load should be around twice that, 6,400 ohms.
I can testify that 395V, 200 ohm Rk, large Rs, around 120mA, 6,600 ohms load, gives in excess of 17W clean and happy. (I'd put "24W" on a guitar-amp badge.)
>
I have a 4.2K,4K, and a 3.3kTake 4K as example. So we want a 2K load on the supply, and suck 50 Watts idle power consumption. V^2/R=P, P*R=V^2, 50W*2000r= 100,000, square-root of 100,000 is 316V. Cathode bias drop is likely to be near Vg2/
Mu or 300/10 or 30V. Total supply can be 316V+30V or 346V.
But why guess? Back in 1956 people were paid to *sell* these tubes, by making it easy to design-in to equipment.
EL34 data page 5 shows cathode-bias "AB" conditions for EL34. I _suspect_ there is a typo on this sheet: "Va+Vrk" should be "Va
-Vrk"? So 350-400V, 3k4 to 4k load, cathode resistor near 130 ohms (start above 150 ohms and work down). This heavy load and loose distortion spec allows current consumption to rise 26% from idle to roar, efficiency touches 49%, so power output is more than half the idle Pdiss. That's what hours of experimental work can do, better than cut-and-try or napkin calulations.