Hello everyone,

I would be happy to receive your feedback about my comments on equation (2.90) of this precious book by Razavi. This equation gives the Rout of a feedback amplifier with input match.

I cannot find the same result for the output impedance. Small signal modeling gives me a totally different result, i.e. a very small Rout vs a big one predicted in (2.90).
But before doing the math, intuitive analysis also predicts a small Rout:

1. Shunt-Shunt Feedback:
I see shunt at the output node (i.e. voltage sampling) and shunt at the input (current mixing). Thus the feedback will tend to reduce the input impedance (to make it a more ideal input for current) and it will ALSO tend to REDUCE the output impedance (the ideal voltage output stage has zero Rout).
So, equation 2.89 indeed predicts a small Rin. BUT equation 2.90 predicts a quite BIG Rout, which seems inconsistent to me.

2. M1 is effectively diode connected in AC
M2 is a source follower so practically vg2=vs2. Due to the gain of M2 source follower = 1, the gate and drain of M1 are effectively shorted in AC.
This means that M1 is DIODE CONNECTED in AC. Thus, the Rout should be 1/gm1 (in parallel with the bigger Rd which can be ignored). That's exactly the result I take after doing the math on the small signal model.

Any comment on my thinking is welcome.

Regards