Amplifier Stability & PZ analysis

[europus2002]

Hi all,
I was wondering what techniques do you guys use to check the stability of low frequency , very high gain AC coupled amplifier made using Opamps? I ran some simulations for Pole-Zero analysis and the simulator gave output in the form that i never expected....It says "Imag" over Y axis varying from -12.50 to +12.50 and "Real" (rad/sec) over X axis swinging between -3.000T to +3.000T.I expected it to show Gain and phase margins with pole zero location ???.. I don't know what to make of this...anyone?? All my poles and zeros lie exactly at 0 of Y Axis and between 0 to +1.3T (to the right) and this circuit i am so far not able to get it oscillating on board.....I could post a jpg but am unable to find the link on procedure...thanks in advance

 

[Skogsgurra]

Do you want it to oscillate? If there are no Im poles, it shouldn't. I think that you are describing a stable amplifier. Isn't that what you wanted?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[europus2002]

You are right skogsgurra, Amplifier seems to be pretty stable and that's what is needed.This is designed for biometric signals. I did supply switching , squarewave testing , HF saturating but nothing seems to drive amp in oscillations...Hence i wanted to cross check with pole zero analysis if that balances with the practical outcome...but the simulator output is quite weird...So it looks like you can conclude something ....As far as i know, all the poles and zeros should lie withing your circle of unity gain for absolute proof of your stability....but the units on the output plot as explained above, is confusing me. Do you have any idea what that T on X axis (real) axis is ??? when i reduce series path resistors in signal path, it changes to G......not sure what that is....Thanks for your help

 

[Skogsgurra]

It is kind of difficult to visualize your charts. Use faq238-1161 for information on how to post pictures and diagrams.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[jimkirk]

Giga? Tera?

 

[europus2002]

http://i10.tinypic.com/4hip94y.jpg

Here's the link to the screen shots....Yeah Jim, it seems like Giga, Tera but what does such mega units have to do with phase angle, i have no idea....Look up the image and send me your suggestions....Thanks in advance to all who help....

 

[jimkirk]

One quick thought is that in at least some versions of SPICE, you have to be careful with suffixes. 1MEG is 1,000,000 but 1M is 0.001. (case insensitive) I've occasionally missed that, and had results 9 orders of magnitude off.

Are you sure your model component values are what you think?

 

[Skogsgurra]

Jimkirk's suggestion is absolutely worth following up. As long as we don't have standardized prefixes (or better, no prefixes at all) in all engineering programs, we will have this kind of problems.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[europus2002]

Yes Jim/skogsgurra,
Fortunately the simulation package i use has facility of defining miliVolt as "mV" and so on. So there is hardly anything that can go wrong there. Moreover , i am getting perfect outputs for Output waveforms / Transient response and frequency response plots...So component models should be fine....But just still wondering about the G & T units for phase...Thanks for the help