Interesting compensation scheme

Hello All,

Recently I came across a nice little 2 stage op-amp which intrigued me.
Basically the the 2nd stage of it is a source follower where primary compensation is achieved
using a big compensation cap between the 1st and 2nd stages.
Further compensation however seems to be achieved by making the bias of the source
follower programmable.

I guess the reason is - increasing the source followers bias effectively reduces its output
impedance thereby pushing the output pole further out.

I guess this would be effective if you have large cap varying loads on the output and
may not be achieving adequate pole splitting with the one comp cap at the output of
the 1st stage.

First of all:

1. Is my understanding of this additional compensation correct?
2. Is it not a little inefficient - increasing current to push a pole out?

Thanks,

Diarmuid

Post here the circuit diagram in order to illustrate the concept that you're describing


+++

diarmuid said:

1. Is my understanding of this additional compensation correct?

Click to expand...

Yes, in principle. But if the current increase works proportional to the cap load increase, this (probably dominant) pole won't really get pushed out (RC ≈ const.). Actually it shouldn't get pushed out, in order not to shift it too close to the non-dominant pole, which would jeopardize stability resp. PM.

Such load-dependent dominant pole compensation is often used for LDO regulators, too.

diarmuid said:

2. Is it not a little inefficient - increasing current to push a pole out?

Click to expand...

Might seem so, at the first glance. But to keep the necessary slew rate, you need more current for larger cap, anyway.