• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Radiotron A530 6J7G-6V6G-2A3 PP Amp

I found this unusual Willamson schematic using a 6V6G as the cathodyne splitter. I have a few questions about the circuit.

There’s a voltage source from the center tap of the power transformer for bias and the .01 (10k ohms, I assume) resistor of the 6V6G. I’ve never seen this done on a Williamson before.

Is the negative supply really needed for the 6V6G? If so, what voltage range could I use to supply the 6V6G cathode resistor?

Thank you


  • Radiotron A530 6J7G-6V6G-2A3 pp.jpg
    Radiotron A530 6J7G-6V6G-2A3 pp.jpg
    182.3 KB · Views: 180
It looks to me like that negative supply is also the grid bias for the PP 2A3 tubes. My best guess is that it would be about -50V DC.

It's not really a Williamson circuit. Maybe you could call it 3/4 of a Williamson. That pentode voltage amp to cathodyne splitter to PP outputs is really close to the Dyna MkIII or Stereo 70 circuit, but without the feedback loops (and with triode outputs instead of UL).

Looking at the schematic, the 6V6 is set up as a concertina (cathodyne) phase splitter, with 10k in the plate and 10k in the cathode ('split load phase inverter').

The 3k resistor is a cathode bias resistor, bypassed by the 25uF capacitor. I guess the 6V6G plate current is set to only about 7 mA, so if the B+ is about 350V, then its bias would be about -20V (grid to cathode), so the Ebb (supply) voltage would be about 330V. Each 10k resistor would drop 70V, so 330V - 140V = 190V plate to cathode on the 6V6G. That works out using the GE 6V6GTA triode curves.


I'd suggest re-drawing the circuit in a more familiar layout, so you can see what's happening. It's drawn in an unusual way.

Thank you.
I agree that the circuit looks a little odd.
I was think of running the first stage as a triode and DC coupling it to the 6V6. My thinking being the fewer parts, the better. I am no expert, so am not sure if the bias circuit on the 6V6 affects symmetry. That is why I am inclined to using only a 10k on the cathode and eliminating the 3k bias resistor and cap. Is this frivolous?
The bottom 10k resistor from the 6V6 (the cathode load) doesn't go to ground, it goes to the negative grid bias supply for the 2A3's. This adds about 50V to the total voltage available for the 6V6.

On the 6V6, since the 3k cathode bias resistor is bypassed, it does not present an AC load to the cathode (at least not within the audio band). There is about 22V dropped across that 3k resistor. You could DC couple the first stage to the second stage, but that would decrease the available voltage from plate to cathode for the 6V6, which would worsen performance in this case.

I drew up the circuit in LTspice, and at low output power (1 watt) it looks like most of the distortion from this circuit would come from the input stage 6J7. The simulation predicts that it should make 10 watts RMS out with some headroom, but distortion will be kind of high by modern standards.

Re: 6V6 -- A cathodyne phase splitter operates with 100% negative feedback, which is why it has no gain. (Actually, it will have somewhere around 0.9X gain.) You won't get the 'sound' of a 6V6 there. If you like the sound of 6V6 tubes, why not use those tubes as the outputs? Or use a 6V6 strapped triode, as a driver tube for a bigger output tube?
This design has the look of a struggle to get sufficient gain and 2A3 drive. The 6V6 bias arrangement doesn't affect symmetry. It might be possible to direct-couple the 6J7 to the 6V6 by tweaking 6J7 G2 voltage, but you definitely won't have enough gain if you convert it to a triode. I would build the first two stages on a breadboard and play with 'em before committing to a full build.