help for difference of pre-post simulation

hi, i'm confused by a problem like this:
i design a fully 2-stage opamp with continue cmfb. and the ac analysis show dc gain is 90db ,also have go0d phase response. but after post-simulation ,the gain is only 50db! and To make things worse, the phase have Obvious Curved-up before the gain to 0db ! my opamp layout is almostly Symmetrical. and have miller Compensation for both p and n Current Load drain in first stage.

fencl said:

hi, i'm confused by a problem like this:
i design a fully 2-stage opamp with continue cmfb. and the ac analysis show dc gain is 90db ,also have go0d phase response. but after post-simulation ,the gain is only 50db! and To make things worse, the phase have Obvious Curved-up before the gain to 0db ! .

Click to expand...

Just for clarification: post-simulation means "measurement"?

LvW said:

fencl said:

hi, i'm confused by a problem like this:
i design a fully 2-stage opamp with continue cmfb. and the ac analysis show dc gain is 90db ,also have go0d phase response. but after post-simulation ,the gain is only 50db! and To make things worse, the phase have Obvious Curved-up before the gain to 0db ! .

Click to expand...

Just for clarification: post-simulation means "measurement"?

Click to expand...

IT means simulation after layout, just add un ideal Parasitic Parameters such as R,C,C-C and so on..

fencl said:

IT means simulation after layout, just add un ideal Parasitic Parameters such as R,C,C-C and so on..

Click to expand...

How detailed (correct) is the rising red part of the phase response?
This could be an indication of instability (oscillation). This also could explain the low gain at lower frequencies.

How detailed (correct) is the rising red part of the phase response?
This could be an indication of instability (oscillation). This also could explain the low gain at lower frequencies.[/quote]

I'm sorry for that can't show my own simulation result, because the computer is isolate from interent. this show just the simular result by somebody else. and the red part of the phase response just means here different from pre-simulation with black part. and the db result show that here generate a zero , and different from pre-simulation also.i'm confused about it.....

U can only try to sim with C-C condtion which r effect is skipped.
This action can help to debug the problem from which part( r or c ).

BrianSu said:

U can only try to sim with C-C condtion which r effect is skipped.
This action can help to debug the problem from which part( r or c ).

Click to expand...

thanks for all help , but i still can't find the reason, without much time and too much associated signal ,so at last i change the whole layout , the result meet demand.