will my circuit oscillate ?

[aaronwlee]

I designed an OPAMP and test it in unity gain structure with a -1~1 step voltage input , I get the transient response ( in the attached) 5% overshot

So , will this circuit oscillate?

 

[LvW]

I don´t understand your question.
The diagram clearly shows that it does not oscillate - instead it exhibits rather good stability properties as the overshoot is small. The design looks good - be happy.

 

[aaronwlee]

thanks

well, in fact, if taking the process variation or other uncertainty into account, I just want to know whether is it risky to fabricate this circuit into a chip, in real tape out, maybe the circuit would have a larger ripple and longer settling or even oscillation

Aaron

 

[Blackuni]

Hi,

if u r sending your design to fab. then i recommend to all this.

1. do ac analysis both closed & open loop, get gain margin, phase margin. make sure its well below the requirement for oscillation.
2. do the same for different process corners(FF, TT, SS ...)
3.do layout extract parasitics and repeat 1 & 2
4. You can also try varability analysis to model process varations effectivley
by doing all this u can make sure that u design works as expected (>80%)


Hope this helps.

thanks,

 

[aaronwlee]

thanks

1 do both close and open loop ac analysis for PM and GM for open loop case, I draw the gain vs frequency and phase vs frequency curves and find the phase of the frequency at which the gain is unity, then PM=180 - that phase, but what about the close loop method ? can you describe it

Thanks

Aaron

 

[rfsystem]

To add something ...

An opamp could have different load current conditions, output voltages and input common mode voltages.

For a general purpose opamp these items add as further 8 usability corners to your checks. In practice these tend to vary more than process, temp or supply and is often a fault situation because the concrete embedding was not foreseen.

 

[aaronwlee]

Sorry to bother you all

I did the close loop AC simulation of the gain and phase vs frequency (unity gain buffer connection of my OPAMP) and i put them in the attached

can you help me confirm is it OK?

In fact, I do not quite understand how to judge the stability directly from the close looped configuration such as unity gain buffer


Thanks

 

[linxiang0431]

I think you'd better show the open loop gain of the OP instead of close loop gain.

 

[LvW]

thanks

1 do both close and open loop ac analysis for PM and GM for open loop case, I draw the gain vs frequency and phase vs frequency curves and find the phase of the frequency at which the gain is unity, then PM=180 - that phase, but what about the close loop method ? can you describe it
Aaron

Are you sure that you have performed a correct simulation of the open loop behaviour ? It can´t be true that the PM is 180 deg. because that is the ideal value and your closed loop response clearly indicates that the PM is in the region of app. 50...70 deg.

Added after 2 minutes:

A short comment to the closed loop response: It looks somewhat strange that the gain does not fall below 300 mvolts for frequencies above 100 MHz. Perhaps the result of parasitics ?