The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> loop stability transient
https://designers-guide.org/forum/YaBB.pl?num=1379601945

Message started by circuit_cook on Sep 19^th, 2013, 7:45am

Title: loop stability transient
Post by circuit_cook on Sep 19^th, 2013, 7:45am

Hi, I am simulating a control loop stabillity by running transient simulation. Because this loop is a little complicated, and I don't see I can change the architecture. The ota plus input is loaded with current source, so it is a very high impedance node, and some ac positive feedback also from this node. So it is very easy to oscillate.
My transient shows there are some initial ringing, but it settles OK.
see the attached graph, but the rings seems to have some high frequency ringing attached. about 60MHz
My question: Is this loop stable?
What could cause that high frequency component. high frequency gain peaking ? pole, zero pair?

Thanks.
Judy

Title: Re: loop stability transient
Post by harpoon on Sep 20^th, 2013, 1:44am

hi circuit_cook,

Just a few comments ...
1. It looks stable ... but for only 1 simulation point and 1 corner. I am sure if you put it through input freqiency, Process, Voltage and Temperature, you will find that it may start to oscillate or ring.

2. Do a PXF or AC analysis to find out the frequency response and see if there are any peaking

3. Do a transient analysis with signals going through and see if there is any unwanted artifacts

4. You will have to be very careful in layout. Any feedback parasitic coupling to this high impedance node of yours will result in ringing !

5. As a saving grace, you may well find that the final layout adds damping to your circuit ...

Is this a filter stage ? Switched cap application ? Would love to hear more about your design, the challenges and how you overcame them !

good luck ...

Title: Re: loop stability transient
Post by circuit_cook on Sep 20^th, 2013, 7:39am

Harpoon:

Thanks for your reply. In fact, I ran through statitistics simulation and corners, it all seems stable.
I have also tried pulsing the supply a little, and changing the input code during transients. Both turns out to be fine.
Just don't know if there is any artifacts from simulator or what is causing this. stb simulation shows PM about 50, and GM about 20. But stb doesn't seem to explain the transient.
I will definitely check it after layout.
This is part of circuit in a filter tuning circuit.

Thanks.
Circuit Cook

Title: Re: loop stability transient
Post by boe on Sep 23^rd, 2013, 3:21am

Circuit cook,

please note that most common solvers for ODEs tend to dampen oscillations (esp. Gear's method).

- B O E

Title: Re: loop stability transient
Post by circuit_cook on Sep 24^th, 2013, 7:21am

Hi,

Thanks for the reminder. I checked the log file, it was using gear2only, and I read that gear2 and BE could be overly stable. I changed the method to traponly, and reran these transient. The results seems similar.

circuit_cook