[Amps] Designing the Cleanest Linear with RF Negative Feedback

[Steve Thompson]

Gary Schafer wrote:

> I am guessing that the 3:1 ratio for IM3 happens in pure class A?
> Also from what I have read recently that the IM5 is a 5:1 ratio.

Yes - although IMD5 (or higher order) is rarely significant with class A 
amps running low IMD3.
> 
> The data that I cited earlier about the IM products being lower at 
> higher power levels was for AB1 operation. Apparently IM performance is 
> entirely different in AB1 and A.

Very! It's a whole new world where driver stages are much bigger than 
you think they need to be. Say your output stage will deliver the wanted 
power at -50dBc IMD3 - if the driver also runs at -50, the overall 
result will be -44dBc. The driver has to run at -56 or better to affect 
the overall IMD by 1dB or less. You churn the linearity numbers, and 
it's a bigger device than you might think at first. With older 
semiconductors where each stage has a gain of 8-10dB, a typical lineup 
would have 1 driving 2 driving 4 identical devices, all running the same 
class A dissipation - massive overkill by class AB standards (imagine 
4CX250 driving 3-500 driving 2x3-500).

Steve