aaron_do wrote on Oct 23rd, 2008, 1:17am:Hi all,
I'm trying to analyze inductive degeneration (LNA) using basic feedback analysis. Assume i break the loop at the source terminal of the transistor. So the open-loop transconductance is simply gm, while the feedback factor should be sLs making the loop gain gmsLs. Therefore, the overall transconductance is,
Gclosed = gm/(1+gmsLs)
which seems to be correct. .
I did not recalculate all the formulas, however, I think the overall transconductance is not correct as you did not take Cgs into consideration. If you like to consider Cgs during Zin calculation you should do this also for the gain resp. overall transconductance as a part of the current through the source terminal originates from the current through Cgs. May be that is the cause of the discrepancy.
Added some time later:
In the mean time I have thougt a little about the problem - and I think there is another reason for the discrepancy (but it is connected with the above mentioned arguments):
The input impedance is increased by the same factor which decreases the gain (resp. transconductance) due to feedback only in the case that there is a CLEAN voltage feedback without any disturbance.
But in your case, the element Cgs disturbs this feedback - and therefore the situation is a bit more complicated and we cannot expect that there is a common factor (1+gmwLs) for the main parameters.