Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Is it good to have complex-conjugate pole pair as first on-dominant pole?

Status
Not open for further replies.

samiran_dam

Full Member level 2
Joined
Apr 17, 2010
Messages
122
Helped
3
Reputation
6
Reaction score
3
Trophy points
1,298
Activity points
2,419
Dear all,

In opamp design, if there is a complex-conjugate pair as 1st non-dominant pole in the frequency response plot, is the design considered to be reasonable? Because of the complex-pair, phase margin is good ( > 60 deg.). But in closed loop gain-frequency response plot, I am observing a little inductive peaking around the unity-gain-bandwidth frequency. So I am having a doubt whether the design can be considered good or not?

Can somebody guide me understanding the pros and cons of the complex-conjugate pair as 1st non-dominant pole?

regards
Sam
 

I think it is rather uncommon to have a complex pole pair as the "first" pole - and I only see disadvantages as far as bandwidth and stability properties are concerned. All designers try to place the first real pole at a frequency as large as possible - with respect to stability margins for closed-loop operation.
 

Re: Is it good to have complex-conjugate pole pair as first non-dominant pole?

I may not be clear in my last post. I mentioned that complex-pair was the first "non-dominant" pole. The dominant (or "first") pole is a real one. In the open-loop frequency response, the complex-pair is appearing after the gain-crossover frequency. In the closed loop response, I am observing a slight inductive peaking. Just now I have observed that the Gain margin is only -8 dB. Is the low GM forcing the peaking?
 

Are you really sure that there is a complex pole pair in the open-loop frequency response?
How did you proof this?
Or is the occurence of peaking (what is "inductive peaking"?) in closed-loop operation the only indication?
 

For my opinion, you have a beautiful open-loop frequency response with a phase margin of app. 60 deg and a gain margin of app. 15 dB.
What else do you want?
These data are in full accordance with the closed-loop peaking of less than 1 dB. This value is nearly identical to the theoretical peaking that is calculated for a 2nd order response.
Question: Where (or why) do you detect a complex pole pair?

---------- Post added at 13:20 ---------- Previous post was at 13:19 ----------

 

The complex pole pair is: (-19.2 ± j20.4) MHz (as I have already mentioned in the previous post). I got this value from a PZ-analysis of the opamp circuit. The occurring of complex-pole-pair is also reflected by the fact that around the second pole frequency the phase drops sharply.

Actually the root of my doubt is the transient response of the circuit. I feed a pulse [pulse_width=5us, period=10us] function to the opamp. The output is shown below:


View attachment transient.bmp

I am not able to explain the occurring of the glitch near the falling edge of the output.
 

The complex pole pair is: (-19.2 ± j20.4) MHz (as I have already mentioned in the previous post). I got this value from a PZ-analysis of the opamp circuit. The occurring of complex-pole-pair is also reflected by the fact that around the second pole frequency the phase drops sharply.

But, for my opinion, it cannot be concluded that this phase drop ic caused by a complex pole pair. It could be also caused by 2 real poles. A compex pole needs a LC resonance or internal feedback effects. But as this gain and phase drop occurs beyond the transit frequency I would not care to much about it. I repeat: The closed-loop frequency response looks very good.

Actually the root of my doubt is the transient response of the circuit. I feed a pulse [pulse_width=5us, period=10us] function to the opamp. The output is shown below:
I am not able to explain the occurring of the glitch near the falling edge of the output.


I don't trust this simulation result. Did you have enough points (time resolution)?
In any case, I think it cannot be caused by a stability margin that is to small.
 
Yeah you may be right. I guess, I have to dig some math....

Anyway, your view has given me some confidence. Thanks....if you find any further explanation, please let me know.

regards
Sam
 

Hi sam,

I again had a look on your tran results.
Am I right that the voltage applied at the input is in the kilo-volt range???
 
:)

I was so careless....while specifying the positive voltage of the pulse instead of writing 900m, by mistake I wrote 900.....thats's why all these happened....

Thanks a lot....

anyways the discussion was good for me. It cleared some of my doubts regarding the frequency response [which is indeed correct :)]...

regards
Sam
 

Slew rate leads to delay too.
However, transient result shows low level is not followed.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top