Few questions concerning the Miller Op amp

miller op amp

1) Is Miller op amp still used by today's analog designers? or are the folded cascode and telescopic op amps the standard now? I mean what makes it so special?

2) Why are the input transistors of the miller op amp preferred to be pmos?

3) Why does the common source amplifier that follows the diff amp has to be the opposite type. i.e if the input transistors of the diff pair are pmos then the CS has to be nmos and vice versa?

noise in miller opamp

1) Yes, it still used. All depend on for what you will use your opamp.
2) It's not preferred to be pmos. If input signal changes from 0 to x*VDD, where 0<x<1, it must be pmos. If input signal changes from x*VDD to VDD input transistors must be nmos. Otherwise it will not work correctly.
3) Because V4 is small and yo can't make common source amplifier with pmos transistor.

miller opamp common feedback

Warlike said:

1) Yes, it still used. All depend on for what you will use your opamp.
2) It's not preferred to be pmos. If input signal changes from 0 to x*VDD, where 0<x<1, it must be pmos. If input signal changes from x*VDD to VDD input transistors must be nmos. Otherwise it will not work correctly.
3) Because V4 is small and yo can't make common source amplifier with pmos transistor.

Click to expand...

Warlike :

Thanks for your answer but concerning the first point can you elaborate on it because i don't understand why V4 would be small.Also, I am following Berkeley's EE140 and I've passed by several CS's with PMOS transistors. So could you re-explain.

millers op amp

If the differential amp is completely balanced and no offset, the currents through T3 and T4 are the same and it is supposed that the drain voltage of T3 is equal to that of T4, which is also it's gate voltage. The gate of T5 is connected to the drain of T3 which is virtually the gate of T4 - something like a current mirror if the W/L of T5 is proportional to those of T3 and T4. Even if it is not, the Vgs of T5 is still in the order of a Vgs of another NMOS and it is ok for biasing. If T5 instead were PMOS, then it is not quite appropriate to bias it that way, because you'll have Vgs between Vdd and the output of the diff. amp and it will probably tend to put one of the PMOS transistors (input and tail current) in triode. So, T5 has to be NMOS. The opamps work with feedback and it will adjust the output voltage of the diff.pair so that it will bias the second stage and not put the PMOS of the first stage in triode.

amplifier with pmos input

1) Is Miller op amp still used by today's analog designers? or are the folded cascode and telescopic op amps the standard now? I mean what makes it so special?

There are several advantages of folded cascode and telescopic opamps that makes it win over miller opamp. Some of them are low power , low noise, high gain.
The main advantage of miller opamp is the high signal swing which is difficult to get in other two.
Power dissipation in miller is high because of extra compensation capacitor.


2) Why are the input transistors of the miller op amp preferred to be pmos?
Not only in miller they are also preferred in other opamps

1. It has low 1/f noise, which is requirement for audio applications.
2. PMOS fabricated in NWELL, hence will contribute to low substrate noise.

3) Why does the common source amplifier that follows the diff amp has to be the opposite type. i.e if the input transistors of the diff pair are pmos then the CS has to be nmos and vice versa?

Just elaborating "Warlike" reply
Generally it depends on common mode output of the stage, and common mode input of the next stage.
Another good reason to have NMOS in the next stage to have high gm, since gmn> gmp. This will cause the non dominant of pole to be pushed outside.
fnd = gm/(2*pi*Cl)

pmos common source amplifier dc analysis

you can find good explanation regarding the questions in the book
Analog Integrated Ckt design by John and ken martin......