Cathodyne follower? :)

[mashaffer]

Another weird idea that I got while musing.

I was thinking about the difference in output Z of the two sections of cathodyne when a strange thought occurred to me. Could we use only the follower side as output (for low Z) and use the plate signal for negative feedback for response shaping.

Visualize a standard cathodyne but instead of taking the plate signal to a power tube connect it via a capacitor to the grid. and take the signal from the cathode resistor to feed the following circuitry. So we could define one pole using the feedback path and then use the low Z of the CF to feed additional passive eq.

Could it be done? Would it work as outlined? Is there any advantage of such an arrangement over other more conventional approaches?

mike

[smoking-amp]

Sounds like the usual partial feedback scheme: plate thru resistor/cap back to grid (or driver grid?) but with output from the cathode here. If you have something else in mind, please post a schematic since its not clear what you mean.

Don

[SY]

Except that the tube already has 100% feedback and the source impedance of whatever is driving that grid shunts whatever it's getting from the plate. I'd have to work out the equations to be sure, but I think all you'll have is a cathode follower with a lousy input impedance.

[smoking-amp]

Probably not what was originally intended, but how about using a cathode follower to drive the output tube grid and the plate feedback from the output tube comes back to the cathode follower's plate.
That way the CF tube provides 1/Mu feedback attenuation approximately.

Don

[revintage]

Originally posted by mashaffer
I was thinking about the difference in output Z of the two sections of cathodyne mike [/B]

If you assume that each side is driven one at the time you are right, but..........................when driving a PP-load you are wrong as both sides have about equal(and very low) Zout. Well below 3kohm/side for a 12AX7 with 1,5mA Ia. On the downside is that a high B+ is needed to get full swing.

[mashaffer]

I appreciate your looking at this. This is what I have in mind. It would be used at line level not as a driver or PI.

I suppose one might also conceivably want to replace the cap with and LRC tuned circuit for BP or BR filtering.

mike

[smoking-amp]

Putting a frequency selective network in the feedback will cause the input impedance to vary with frequency, besides its effects on the CF stage gain. You may get more poles in the frequency response than you expect. I would simulate any design like this (including the previous driving stage) to check the results.

Don