frozenduck
Joined: 06 Oct 2008
Posts: 2
|
 06 Oct 2008 19:10 Some important questions about CMFB, Thanks! |
|
|
|
For a folded cascoded two stage opamp,
Q1:
Typically, we just add one CMFB from 2nd stage output to 1st stage current source. Why don't we add 1 CMFB for each stage? Is there any drawback?
Q2:
If we just need one CMFB, why do we sense the 2nd stage output instead of 1st stage output? The 2nd stage is just pseudodifferential, is that correct?
Q3:
For CT CMFB, we need to break the loop at the CMFB input and run AC analysis to check the PM and UGB for stability.
1. The UGB should be equal to or few less than that of the main amp. and PM≥60°. Is that right?
2. But how can we make sure it works during main opamp operation?
Q4:
Fro SC CMFB, it cannot be simulated in AC analysis.
1. Run Tran. and save the data, and then run AC.
2. Use PSS analysis.
Which one is correct? Is there any tutorial to do this? I can't find it in the Internet.
|
|
prince_123
Joined: 07 Oct 2008
Posts: 17
|
| 07 Oct 2008 9:50 Re: Some important questions about CMFB, Thanks! |
|
|
|
I guess if you are able to amplify the signal properly and predict the slewing of the load cap on the second stage accurately, they you can go for local CMFB.
But my experience says that if you have local CMFB for the second stage, then you cannot predict the current consumed by second stage accurately. Not good for class A operation where slew has to be taken care by setting the currents right.
I also have one doubt regarding switched common mode feed back.
How is mismatches in the replica circuits to generate bias voltage is taken care?
[quote="frozenduck"]For a folded cascoded two stage opamp,
Q1:
Typically, we just add one CMFB from 2nd stage output to 1st stage current source. Why don't we add 1 CMFB for each stage? Is there any drawback?
|
|
