The following picture has two circuits that contain the different way of compensation for polos & zeros. As I know that the 1-st(left)one has a zero on the LHP that approximate equals gm/Cc, the 2-nd(right)one has a zero on the RHP that approximate equals gm*rds*gm/Cc. I may ask that which compensition is better?
thands~ [img][/img]

In general a LHP zero (if properly placed), has a positive influence on stability, as it 'pulls' your root locus to the left. And vice versa for the RHP zero of course.

raja.cedt wrote on Nov 9^th, 2011, 1:00am:

I don't think the small signal equivalent is the same for both.I draw it as following:
I have calculate the transferfunction Av=Vout/Vin with the help of Maple.It shows that it really has LHP zero as I mentioned before.

raja.cedt wrote on Nov 9^th, 2011, 1:00am:

[img][/img]
This is the small signal of the 1-st ckt.

hello,
i am sorry for my previous post. So if you have razaavi could you please verify page no..371. He has derived TF for the left ckt, it is gm*rds*gm/Cc. You get gm/cc only if there only cap feedback aorund the transister. By refering razavi page no 176, it has simple method to estimate zero location, i would suggest you to estimaate with this method without deriving full equations.

And comming  to smal signal models, whats the diference between both of these ckts, in one case you have PMOS commongate amplifier and in another case you have NMOS commongate amplifier, so small signal model don't care about NMOS and PMOS. One more thing both circuits have zero in RHP, because forward path an feedback path are having oppsite polarity gains.

Once again sorry for the previous post.

Thanks,
Raj.

raja.cedt wrote on Nov 9^th, 2011, 5:56am:

I have read the pages of Razaavi you mentioned, but Razaavi says"this topology contains a zero in the left half plane" just up his equation 10.42.
I tried the method on the page 176 for the left ckt.In this way I got a RHP zero. It is ovious that the two results are in contradiction.

By the way, I think the forward path of the left ckt is somehow weird,since the signal will go from the drain to the souce of the common gate PMOS?

Thanks~

In fact, this two sch are the same in small signal analysis, because both of them are providing a feedback signal path with  gm if PMOS is almost the same with NMOS.

and the difference between them is:
at high frequency, the second sch may cause a straight path from input to output, which lower the gain at high frequency.

Hi,
if a step-up signal is assigned to input, the voltage of output will increase almost as input, this could pose a stress at the source and gate of MOSFETraja.cedt wrote on Nov 14^th, 2011, 12:17pm: