"Walter Maxwell" wrote in
:

Sorry about the 'long dissertation on Class C amps', Owen, but I
thought it appropriate to include it in view of Richard's similar
discussion on the automotive engine analogy to the RF tank circuit.
I'll try to keep my comments shorter from now on.

Walt, it wasn't so much that it was long, but it was long and for all
that was said, it didn't address the linearity issue.

I understand your position to be that the behaviour of the tank circuit
is independent of the transfer linearity of the active device... but
asserting that 'things' are linear because there are no harmonics is
wrong and being so, is no support for your argument.

I am wary of analogies, the switch analogy that was raised is not a good
approximation and I haven't even thought about the car engine.

I am genuinely insterested in your argument. I don't accept it (yet?) as
you know, and I have spent some time over the last 18 months or so
exploring the concept you describe.

Fundamentally, I am trying to reconcile what you say with the techniques
commonly accepted for designing such a PA. Those design techniques give
us a method of predicting power output at different load impedances, and
the E/I characteristic for different loads is not always a straight line
(as it would be if a Thevenin equivalent circuit exists), though it might
appear fairly straight over a narrow domain. Since working from
characteristic curves is so prone to error, my modelling has been based
on an idealised triode transfer characteristic, but with similar
behaviour to an 811A. The analysis is waiting for me to build the
analytical equations for the negative feedback due to cathode
degeneration in a grounded grid configuration. I need to apply more time
to it, and the revived discussion might focus me for a bit!

Owen