The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> CMFB stability fot a differential diffence OTA
https://designers-guide.org/forum/YaBB.pl?num=1231308845

Message started by seahs on Jan 6^th, 2009, 10:14pm

Title: CMFB stability fot a differential diffence OTA
Post by seahs on Jan 6^th, 2009, 10:14pm

The following is the schematic of my design.

The left one is the differential difference OTA with gain enhanced, the right one is the CMFB circuit for pmos.

I breaked the cmfb, and replaced the Vcmfb in OTA with a dc bias and an AC signal(Vc).
When I measured the AC responce of the CMFB(Vcmfb/Vc), the phase is positive, which is quite strange. The simulation wave is as follows;

When I checked the ac response of (Vo+)/Vc , it is also postive in magnitude and phase.
while as Vc increases, the Vo+ should decrease actually. :-[

Thanks for any advice.

Title: Re: CMFB stability fot a differential diffence OTA
Post by ACWWong on Jan 8^th, 2009, 1:07am

some of the detail of the loop is missing off the side in the schematic so its difficult to diagnose the circuit... when connected does the cmfb circuit give the correct dc levels in dc analysis ?
Anyway breaking the loop is not a very good way to do this analysis, even if you tweak the dc level of the broken node to match that of the disconnected end... the common way to analyse loops is using stability analysis.

assuming your using Spectre, try
spectre -h stb
or your simulator documentation to learn more.

There are plenty of further infomation on this site if your use the search facility. hope this helps

cheers
aw

Title: Re: CMFB stability fot a differential diffence OTA
Post by sheldon on Jan 8^th, 2009, 8:25am

Seah,

One additional comment, there is a special probe for use with
fully differential designs in analogLib called cmdmprobe. You can
select either common mode, 1, or differential mode, -1, operation.

Best Regards,

Sheldon