A problem in 2-stage op design

[Alan_Nesta]

when i design 2-stage op, i find in the mag-f curve there is a peak under 0db
doe it matter?
how does it appear?
in the circuit, the PM is about 65

 

[jiangwp]

Do you use the R&C Miller way to compensation your opamp?

If you use only C- Miller way, i think a Resistor has to be add to cancell or shift the location of zero which lies on the right half plane.

If you use them like above, i think you have to regulate the value of resistor and capacitor to get a match to shift the loaction of zero in RHP.

 

[Alan_Nesta]

I used both miller compensation and cascode miller compensation,
and I have tried add resistor in it and sweep it, still have the peak.
Is there anything generated by 2-order system's damping factor?

 

[sunking]

would you mind upload the schematic?

 

[Alan_Nesta]

ok

 

[jiangwp]

Maybe, a compensation capacitor has to be added for the second stage for the NMOS amplifier side of the first amplifier transistor. Because the NMOS side has the two stage amplifier for the second stage amplifiers of overall op.

i have update the schematic about it .

have a try, just a ideal.

 

[scottieman]

ok

How large is your output capacitance to be driven?
The 2nd-order response is mainly due to the cascode miller compensation.
To sliminate the 2nd order response, pls just use miller compensation.

My question is, what makes you using both miller and cascode miller at the same time?

Scottie

 

[wjx197733]

what's the ratio of w/l?

 

[cliffj]

It should be a zero around your GB. It helps in increasing the PM.

 

[Alles Gute]

I am also interested in why you use both miller and cascode miller in this way cause I've never seen it before.

 

[xwcwc1234]

Is this a class-AB output stage or class-A ? What are the advantages of this circuit ?

 

[einfrage]

also , what models are you using for simluation , and what
is the simulator