The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> stability of class-ab opamp
https://designers-guide.org/forum/YaBB.pl?num=1245256806

Message started by subgold on Jun 17^th, 2009, 9:40am

Title: stability of class-ab opamp
Post by subgold on Jun 17^th, 2009, 9:40am

hi circuit virtuosos,

i am wondering the criteria about the stability of an opamp used in switched-cap circuits.

for a class-ab opamp, when it is slewing, its biasing condition is dynamically changing, and completely different compared with that at dc operating points. so the pole-zero locations are also moving around. for example, i have a class-ab opamp with only 20degree phase margin from ac simu, but it has a perfect settling behavior from transient simu. in this case, since the opamp is not used to buffer any signal, but only to settle a charge for the switched-cap circuit, can i say it is stable by seeing the transient results, or do i really need to take care of the ac result? what would be an efficient standard to judge its stability?

thanks.

Title: Re: stability of class-ab opamp
Post by vivkr on Jun 17^th, 2009, 11:58pm

Hi,

While it is true that the pole-zero locations will change with the signal level and hence in a switched-cap circuit over the clock phase, I think that the common stability criteria still need to be applied. However, you are generally forced to use transient analysis (step response) under all possible conditions (proces, voltage, temperature, large & small signals) to verify that settling behavior is OK.

When you do a small-signal stability analysis, then you are going to see results for the stability of the opamp towards the end of the clock phase (when all large transients have died out). If you have insufficient phase margin, then it would mean that any disturbance anywhere, it will potentially start another round of ringing and so your settling performance will be ruined. So, stability margins are not something you can compromise on. That bein g said, it is generally possible to use single-stage class-AB amps in switched-capacitor applications (depends on the gain you are trying to achieve) which are inherently more stable than multi-stage class-AB.

Regards,
Vivek

Title: Re: stability of class-ab opamp
Post by thechopper on Jun 18^th, 2009, 7:22pm

Hi subgold,

What Vivek pointed out is totally correct. In particular, as you suggested when you consider a different from the dc operating point biasing conditions for the class AB stage you will see the zeros and poles moving around and - as Vivek said - a susceptible condition to generate ringing at the output if any disturbance shows up.
You do not see the effects of the low PM in your step response since such disturbances do not exist in your simulation and your starting point I guess is your dc operating point
You can check the transient response is not as good as what you are getting if you run another transient analysis starting at a biasing condition close to the final value of your original step response.

Regards
Tosei