a general question about CMFB circuits

[jimito13]

Hello everyone,

I have a very general question,as the title of this thread indicates.Suppose that i have a fully differential OpAmp or OTA and i want to setup its CMFB.What are the criteria to choose the point (the node) of the circuit that i will apply the Common Mode Control (CMC) Voltage? (CMC Voltage is the voltage that comes back to the OpAmp from the CMFB circuit).I cannot locate an appropriate answer in several bibliography i have gone through and i am seeking for an explanation,so i would appreciate any helpful answer.Thanks in advance.

Best Regards,
Jimito13

 

[erikl]

Usually the CMC voltage is applied to the gate of the tail current source. But there exists also another CMFB topology, where the CMC voltage is applied to both gates of the cascode transistors (which normally are supplied by a fixed bias voltage). This topology can avoid the danger of instability which might arise from the tail current source CMC. A 2003 IEEE paper by Arturo Tauro et al. "COMMON MODE STABILITY IN FULLY DIFFERENTIAL VOLTAGE FEEDBACK CMOS AMPLIFIERS" studies and interpretes both CMFB topologies.
It's not permitted to publish IEEE papers on this forum, but if you don't have access to it and want it, I could send it to you personally.
Regards, erikl

 

[jimito13]

Thank you very much for your answer erikl.I just downloaded the paper from IEEE with my subscription and i am gonna study it.I have one more question...if it is not answered in the paper i will come back to you

Best Regards,

Jimito13

 

[jimito13]

Hi erikl,

I read the paper you suggested me a couple of days ago and i found some criteria from the analysis the authors are performing and i understood them,but my initial question was very generic about any op amp topology and not just for the classical 2-Stage Miller.So,are there any other criteria you are aware of or those that the paper suggests are unique at the moment?

I want a favor from you if it is possible...My subscription to IEEE doesn't cover the references of the paper you proposed to me,could you please send them to me via PM (if you have them and if your subscription allows you to download them ofcourse)?

Many thanks in advance.

Best Regards,
Jimito13