Cathode Follower PA Driver question

[stephenl]

I've been following a couple of related threads and I'd like to give this a try on one of my existing amps.

The donor amp has a hole for another tube and I have plenty of filament amps so I want to go with a triode instead of a mosfet in this case.

The amp is cathode biased and I haven't found any examples of a driver circuit used in one.

My question is, how do I correctly bias the CF to directly couple to the PA grids? What extra circuitry do I need to add?

[martin manning]

The problem is pretty much the same as in the MOSFET thread, except that you want the cathode of the driver to be at zero volts since the power tubes are cathode biased. That means the driver's cathode resistor needs a negative voltage source with a magnitude around 3x the power tube's cathode voltage.

R.G.'s MOSFET Follies shows a FET driver with a cathode biased power stage. Take a look at an SVT schematic for a tube driver example (t's a fixed bias power amp, of course).

[stephenl]

Thanks Martin.

So I'll need approx -100v supply with pots to set the cathodes at zero at PA quiescence, right?

[martin manning]

And a branch off the screen node to provide about +200V for the anode of the driver. You need to draw a load line. I took a quick look for a 12AU7, and I think -150V and +200V with a 33k cathode resistor looks pretty good. The grid wants to be at about -8V.

[jazbo8]

Take a look at the version of the Ampeg SVT with the 12BH7 driver or the Dumble SSS, both use CF drivers to drive the PA's.

[martin manning]

Combining this type of driver with a cathode bias power stage seems kind of odd, no? The cathode voltage (and bias) will shift up when signal is applied the same as it would if the grid were just grounded at zero volts as usual. With the addition of the driver circuit, you have the option to ground the cathode and switch to fixed bias and the possibility of class AB2 operation.

[stephenl]

Martin, as I was mulling over a few different example circuits over coffee this am, I was thinking the same thing...

[martin manning]

If you go fixed bias, you could do this with a 12AU7 and the same bipolar supply (+/-160V) I proposed for the MOSFET version. Add a couple of bias circuits to it, and you would have a self-contained driver board. See below...

[stephenl]

Thanks for the design Martin.

The donor amp has 12vdc heaters w/ -75vdc elevation. Do I need to do anything different with the heater for the driver tube?

[martin manning]

The driver's cathode is going to be at the power tube bias voltage, so ~50V below ground. That will discourage elevating filament voltage too much. I assume you have it elevated it because of a CF in the preamp?

[stephenl]

Yes, the third stage has a CF. i also have 6vac , that isn't elevated, for the power tubes. Since the drivers are push-pull halves, heater hum should be nulled out if I use the 6vac for their heaters, right?

[martin manning]

Yes I think using the 6VAC would be the way to go.

[stephenl]

Would this circuit also work in a cathode biased PA?

[martin manning]

It would, if you adjust the bias point such that the output is at 0V. To do that you would bring the 10k bias pots up to 0V at the top (delete the 33k's above them), and increase the resistors below to 91k or 100k. The bias voltage on the 12AU7 grids wants to be about -8V.

[Bombacaototal]

Hi all, for a SSS type built should I use 12AX7 or 12BH7?