[Amps] Designing the Cleanest Linear with RF Negative Feedback

[Gary Schafer]

Bill Turner wrote:
> At 09:09 PM 12/14/2005, Tom Cathey wrote:
> 
> 
>>It was pointed out to me that the cathode resistor for negative feedback is
>>to be used only for a grid driven [grounded cathode] amplifier to generate
>>degenerative [negative] feedback.
> 
> 
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 
> This is a question I've often wondered about.
> 
> Does a cathode resistor in a grid-driven amp actually produce 
> negative feedback, or does it merely reduce the gain?
> 
> My understanding is that for true negative feedback, some of the 
> output must be fed back to the input out of phase. In the case of the 
> cathode resistor this does not seem to be happening. In fact, the 
> output could have a wide variation in gain without affecting the 
> input circuit at all. As an extreme example, suppose the plate load 
> circuit dropped to a very low impedance for some reason. With true 
> negative feedback, the circuit would automatically compensate and the 
> output would remain the same (within the capabilities of the feedback 
> loop of course). With only a cathode resistor, the output would fall 
> and no compensation would occur. To me, this does not sound like true 
> negative feedback.
> 
> Is this correct?
> 
> 73, Bill W6WRT

An unbypassed cathode resistor in a grid driven amp does in fact provide 
negative feedback. It is often done in audio amplifiers to improve 
distortion.
As the grid is driven positive the plate voltage goes negative because 
of increased plate current. With increased plate current the cathode 
voltage goes positive because of the resistor. The higher you allow the 
cathode voltage to rise the more it counters the grid drive. In other 
words the positive cathode voltage reduces plate current.
This is a direct feedback method that is out of phase with the input as 
required for negative feedback.

As far as automatically compensating for plate loading, it will to some 
extent depending on the amount of feedback selected.
In your example of plate load dropping to a lower value, that would 
normally allow more current to be developed in the plate circuit which 
would produce more output.
But with the cathode resistor in place the cathode voltage rise will be 
greater in this situation which will develop more feedback voltage at 
the cathode because of the greater cathode current which reduces plate 
current (because the grid to cathode voltage is being biased negative) 
which reduces output. All this happens in real time.

If you use enough feedback the output could remain nearly constant 
regardless of plate load applied but overall stage gain would be quite low.

73
Gary  K4FMX