Designer’s Guide Community

hiSkill_11

Community Member

Offline

Posts: 35

A small ringing riding over outputs when simulating step response
Apr 22^nd, 2014, 11:01pm

Hi,, Guys,

I can't explaine why there are small ringing(several mv) riding at the outputs of the PGA(class AB opamp with gain resistors) when I simulated step response. Please kindly see the attached. Von is basically bais at half VDD(2.5v), while Vop is the true output of the amplifier, when Vop transition, Von will ring but dampled within milli voltages. What can casue that? nonstable circuit or simulator setting? I used default setting for tran analysis in spectre.

Regards,
hiSkill

Back to top

stepResponse.png

IP Logged

boe

Community Fellow

Offline

Posts: 615

Re: A small ringing riding over outputs when simulating step response
Reply #1 - Apr 25^th, 2014, 11:48am

HiSkill,hiSkill_11 wrote on Apr 22^nd, 2014, 11:01pm:

... What can casue that? nonstable circuit or simulator setting? I used default setting for tran analysis in spectre. ...

Both. The default tolerance in Spectre is 0.1%, making 2.5 mV an acceptable numerical error for a 2.5V signal.
- B O E

Back to top

IP Logged

Ken Kundert Global Moderator Offline Posts: 2386 Silicon Valley	Re: A small ringing riding over outputs when simulating step response Reply #2 - Apr 25^th, 2014, 11:01pm That is not simulator error. The circuit is under damped. -Ken
Back to top	IP Logged

hiSkill_11

Community Member

Offline

Posts: 35

Re: A small ringing riding over outputs when simulating step response
Reply #3 - Apr 27^th, 2014, 7:55pm

Hi Ken,

I also doublet that is due to undamped. But my question is that, why so small ring(2.5mv) over von can lead to instablility. I have simulated stb for von loop, it is stable with pm >60. Do I have some tricky methods to validate your guess such as running some other simualtions or modify some simulator option.
BTW, does this small ringring can lead to problem in real silicon?

thanks
Ken Kundert wrote on Apr 25^th, 2014, 11:01pm:

That is not simulator error. The circuit is under damped.

-Ken

Back to top

IP Logged

hiSkill_11

Community Member

Offline

Posts: 35

Re: A small ringing riding over outputs when simulating step response
Reply #4 - Apr 27^th, 2014, 7:58pm

Thanks boe,, it sounds reasonable. But how can I fix it?
Do you have any suggestion? thanks.

boe wrote on Apr 25^th, 2014, 11:48am:

HiSkill,hiSkill_11 wrote on Apr 22^nd, 2014, 11:01pm:

... What can casue that? nonstable circuit or simulator setting? I used default setting for tran analysis in spectre. ...

Both. The default tolerance in Spectre is 0.1%, making 2.5 mV an acceptable numerical error for a 2.5V signal.
- B O E

Back to top

IP Logged

Ken Kundert

Global Moderator

Offline

Posts: 2386
Silicon Valley

Re: A small ringing riding over outputs when simulating step response
Reply #5 - Apr 28^th, 2014, 1:15am

It is really hard to say what is causing the problem without seeing the circuit. The reason why you might see a small but highly underdamped ringing is either that you have a unstable or underdamped mode that is not associated with the main feedback loop and the signal you applied is only lightly coupled this this unstable mode. This mode may be due to a second loop or it might be due to a tuned circuit (an inductor and capacitor or perhaps a source/emitter follower driving a capacitor).

-Ken

Back to top

IP Logged