Help me simulate this OPA!

[wjxcom]

how to simulate this OPA

Hi, all: this is a full differential OPA which has CMFB. In this circuit, the CMFB use 4 capacitances and some switch controled by clock CKN2, CKN3, CKP2, CKP3 to supply common voltage. Now, I have a problem: how can I simulation this OPA, i.e. how can I do AC analysis for this circuit? because there have some clock singles, so I can not process AC analyse, but if I remove the clock singles, though I can do AC analyse, the CMFB may be wrong.

secondly, I find the drain of M52 and M53 connect with the gate of M51 and M50, at the same time, the source of M51 and M50 connect with the substrate of the M52 and M53, this kind of connection is so strange and I do not know why.

thirdly, what is the role of M56 and M55?

Help me, Please!!!

 

[wjxcom]

how to simulate this OPA

There has no body can explain why the drain of M52 and M53 connect with the gate of M51 and M50, at the same time, the source of M51 and M50 connect with the substrate of the M52 and M53, is it?

or who can introduce some data that explain the circuit?

 

[nxing]

how to simulate this OPA

(1) You can run PSS and then PAC to simulate this circuit.
(2): M51, M52 and M50, M53 are just two stage amplifier. There is nothing strange to connect the source of M51 and M50 connected to sub. or At least, I didn't see that's the strange.
(3): The Role of M55 and M56 are provide bias point for M52 and M53

 

[winsonpku]

Re: how to simulate this OPA

In fact,this is a two stage amp.the first stage is a folded amp.M13 and M14 are the current source load.M52 and M53 are the two inputs.M54 M55 M56 are the current source.the first stage is a fully differential amp,so it needs the CMFB.And the other trans excludes the CMFB block are composed of the second stage.the second stage is simply cascoded.M50 and M51 are the two inputs,but it's not fully differential.

It's clear?

But why not choose folded cascode architecture?High gain?why not gainboosted?

Hi, all: this is a full differential OPA which has CMFB. In this circuit, the CMFB use 4 capacitances and some switch controled by clock CKN2, CKN3, CKP2, CKP3 to supply common voltage. Now, I have a problem: how can I simulation this OPA, i.e. how can I do AC analysis for this circuit? because there have some clock singles, so I can not process AC analyse, but if I remove the clock singles, though I can do AC analyse, the CMFB may be wrong.

secondly, I find the drain of M52 and M53 connect with the gate of M51 and M50, at the same time, the source of M51 and M50 connect with the substrate of the M52 and M53, this kind of connection is so strange and I do not know why.

thirdly, what is the role of M56 and M55?

Help me, Please!!!

 

[wjxcom]

how to simulate this OPA

Hi, winsonpku: this is a two stage OPA? Do you mean that the first stage includes M13, M14, M52, M53, M54, M56 and M55; and the second stage includes the M15, M16, M17, M18, M48, M49, M51 and M50?

 

[winsonpku]

Re: how to simulate this OPA

yes,i think so

Hi, winsonpku: this is a two stage OPA? Do you mean that the first stage includes M13, M14, M52, M53, M54, M56 and M55; and the second stage includes the M15, M16, M17, M18, M48, M49, M51 and M50?

 

[wjxcom]

how to simulate this OPA

I am still not know how to simulate this circuit, i.e. how to process dc and ac analysis.
Somebody tell me that firstly I can process tran analysis, when the CMFB is stable, save the circuit's state. then I can process the ac analysis.

But I have already tried to do this, unfortunely I failed

 

[allan_guo]

Re: how to simulate this OPA

In fact,this is a two stage amp.the first stage is a folded amp.M13 and M14 are the current source load.M52 and M53 are the two inputs.M54 M55 M56 are the current source.the first stage is a fully differential amp,so it needs the CMFB.And the other trans excludes the CMFB block are composed of the second stage.the second stage is simply cascoded.M50 and M51 are the two inputs,but it's not fully differential.

It's clear?

But why not choose folded cascode architecture?High gain?why not gainboosted?


:Why the second stage is not fully differential?

 

[winsonpku]

Re: how to simulate this OPA

for the second stage,you analysis that:f you change the voltage of one input,then the other output is changed or not?if changed,then differential;not,sigle stage.
is it useful for you?

In fact,this is a two stage amp.the first stage is a folded amp.M13 and M14 are the current source load.M52 and M53 are the two inputs.M54 M55 M56 are the current source.the first stage is a fully differential amp,so it needs the CMFB.And the other trans excludes the CMFB block are composed of the second stage.the second stage is simply cascoded.M50 and M51 are the two inputs,but it's not fully differential.

It's clear?

But why not choose folded cascode architecture?High gain?why not gainboosted?


:Why the second stage is not fully differential?

 

[wjxcom]

Re: how to simulate this OPA

for ac analysis, I cut off the clock circuit, and add a DC power to the cmfb, then I think I can process the analysis. like the circuit in the attachment