current feedback

[mashaffer]

I thought I had posted here asking about current feedback (for use in current source power amp) but I can't seem to find it. I remember someone suggesting adding extra resistance in the output tube's cathode for additional cathode degeneration. I was also thinking about using an approach similar to Schade but taken from the output pentode's cathode and summed at the grid of the driver. Like Schade the stage before the driver would need to have stable and rather large Zout of course but initial simulations seem to be promising.

Something like this is what I have in mind.



I set the voltage source to 68k Z out in order to simulate a previous tube stage. The simulation seems to indicate that it works to reduce HD. Unfortunately it seems to be primarily HD2 that is reduced with much less effect on HD 3 and HD5. However, I don't fully trust distortion results from a sym. And in any case it should also be increasing Zout too which is what we want in this case.

Any thoughts on this approach

[DF96]

Bear in mind that for a tetrode/pentode output you are not sensing the output current. Negative feedback controls what you sample, in this case the cathode current. For small signals the cathode and anode currents will be roughly proportional, but when negative peaks send the anode down below the screen you will get a kick in g2 current so the proportionality will be lost. If you want low distortion output current then use a triode, or use triode connection.

[mashaffer]

Interesting point. I hadn't thought of that. I wonder if UL would work as well as triode mode since the two taps are tracking each other and both contributing to the total current through the primary...

[smoking-amp]

If you return the screen supply negative side directly to the cathode, then the screen current will stay out of your feedback. (a common trick used for curve tracers)

[DF96]

UL will be better than pentode mode, but it does not guarantee that the output current is equal to, or proportional to, the cathode current. Only a triode does that.

The screen-supply trick is another way to do it, as it effectively forces the g2 current to just circulate within the valve. OK if your signals don't go down to DC.

[mashaffer]

Cool. Amazing what one can lear here.