loop stability in CMFB loop

Status
Not open for further replies.
M

manissri

Full Member level 5
Joined
Apr 16, 2005
Messages
281
Helped
9
Reputation
18
Reaction score
3
Trophy points
1,298
Activity points
3,279
dear all
i am desing ing a fully differential with gain boost opamp in which i used the continuous time CMFB circuit.
my question is what should be the gain of CMFB circuit ( i want to know this gain is differential r common mode gain) .
and the other query is how i check the loop stability of the loop containing the opamp outpur bracnh and CMFB circuit.
waiting for ur reply
manish
 

Basically the CMFB gain should be high to get a good response.
For the stability, there is a simple method.
- Apply a pulse to the reference voltage of the CMFB (of about 200mV) and see the transient response of the output common mode. If there is ringing, it means there is a risk of instability. It also allows you to see of the loop response is fast enough.

I think u can also do an AC analisys by applying an ac source at the CMFB input, but it's more tricky.
 
Hi Skal81,

"I think u can also do an AC analisys by applying an ac source at the CMFB input, but it's more tricky."

What u mean by this statement? When you say "tricky", what are the effect you are expecting to the circuit or measurement?

Thanks,
suria3[/quote]
 

hi skal ..
actuallly as u said that the cmfb gain should be high but i want to know the whether the gain is common mode r differential gain ..
actually the cmfb circuit play with the common mode signal...
so plz clearify that the gain is common r differential..
manish
 

To manissri:
I was refering to the differential gain. CMFB sense the difference between common-mode and reference and feedback the error. So it's the differential gain.

To suria3:
When you need to do the open loop analisys, you need to set the DC point. That's the tricky part. In fact if you use spectre, there is a device available to do that. It's a switch which value depends on the analisys type. So you put ON for DC so that the loop get close and you can work around the bias point and OFF for AC analisys. If I'm correct it's in the analogLib.
The problem is that for Hspice or other simulator I'm not sure there is such a device. So you might need to write a script or something. That's what I meant.

Good luck to you
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…