Post by exp on Apr 12th, 2016, 6:07am
Hi,
I've a pretty simple question: How do I map an AM/AM curve, such as the Rapp model to an RF signal?
Yo my understanding, these AM/AM (and AM/PM) curve models are supposed to be applied to baseband signals (i.e. to their envelope). Hence they also describe the input/output relationship in terms of the absolute value (i.e., there is no negative).
However, when I want to apply such a nonlinearity to a modulated RF signal x(t), the signals swings between Vpk and -Vpk. Suppose the RF signal swings between -0.2 and 0.2 V. Then I could "shift" the curve on the x axis (or shift the signal) similar as for the operating point on a class A amplifier. Example:
But this does not sound "physical" to me. Alternatively I could extend the model to the negative axis and let the signal swing around zero, for example:
This sounds a bit more physical to me but since the function is odd symmetric it does only generate odd-order harmonics. However, I need to generate even order harmonics too.
There's gotta be a simple way to model such an amplifier at RF - at least to first order??