what's Wu comes from? It's caused by which pole?
Wu is unity gain bandwidth which determines your phase margin
| H(Wu) |= 1 and in order to see phase margin we look to the point where gain is 1 or 0 in dB domain.
what's g1 and gL
g1 is 1/R1 and gL is 1/Rl
why introduce a zero to the LHP(left land Plane) which is bad.
compensating adds zero to circuit and it is inevitable.
Is that adding a CM, a internal capactior?
its an additional capacitor which u add to ur circuit to get better phase margin.
- - - Updated - - -
How compensation works:
with out compensation you have 2 poles close to each other so when your magnitude reaches 1 your phase will be more than -160 or so ...
because of 2 poles = 2 * -90 deg phase
so what we do in compensation :
we try to pull p1 back and p2 forward. when the second poll is far away until the unity gain frequency, the magnitude drops with -20 dp per dec and the phase drops -90 because there is only the effect of first pole near unity gain frequency (Wu) however, the second pole is not so far and its is close so we have a bit phase drop like -40. so the total phase drop is like -130.
Then our phase margin is 180 - 130 = 50
How we do pole splitting :
when we add CM, using miller effect the Cap in left side is Cm* gm3 *Rl (Cm is multiplied by gain of second stage) so we have total cap C1 + gm3.Rl.CM then our cap is increase so it pulls the first pole to lower frequencies.
Second pole: when we go to higher frequencies CM will be short and it shorts the gate and drain of M3 so the output pole will be gm3/(C1+Cl) and this will push the pole to higher freqs.