Compensation capacitor in two-stage OPAMP

[slchen]

Dear all:

I have a question about the compensation capacitor in 2-stage OPAMP.
As shown in the attached file, I used a PMOS capaciotr (M13) to realize the compenstation capacitor.
The gate of M13 is connected to node 5 and the source, drain and bulk are connected to the output node (Vo). This is the first choice.
Another choice is that the gate is connected to the output node (Vo) and the source, drain and bulk are connected to node 5.
I want to know which kind of connections is better.
Thanks for your help.

slchen

 

[OpAmp]

Hi,

I would say the second choice is much better. As you know, in two-stage miller compensation, dominant pole is at the output node of the first stage (node 5) and the none-dominant pole is at the output node of the second stage (Vo). So we like to minimize the parasitic capacitance at the output node of the second stage (to push the non-dominant pole to higher frequencies) and increase the capacitance seen at the output node of the first stage (that’s why we use miller capacitor).

Bulk of the PMOS cap has a lot parasitic cap to substrate, so the bulk node is better to be connected to output of the first stage (node 5). All the capacitance we see in the gate of the PMOS cap is just miller cap (almost no parasitic). So it is better to connect the gate to output node.

OpAmp