Hello guys,

What brings me here is the simulation of compensator. I am using a VCVS with some passive components to get the desired frquency response.

So, liek I said, I am trying to obtain the frequency response of a compensator and for that I am using a configuration and the components values from a book that I am reading and what I have is this:



and the result:



I configured the AC analysis like this:



I know that I can't expect an accurate result because I am using a VCVS that is an ideal OPAMP.

The shape of the frequency response matches the frequency response presented on the book. However the values are slighty different.

What I'd like to know is if I am doing well this simulation. Do I have everything well configured?

One more question: Is it possible to do a PSS + PAC simulation using this circuit?

How can I do that?

Kind regards.

I get the same freq response with your circuit. What do you mean with
"slightly different". Can you share the graph from your book?

Why do you want a PSS+PAC simulation?

Hi guys, thanks for the reply.

I will explain what I am doing.

I am designing a buck converter in voltage mode control.

The power stage, the drivers and the non overlapping circuits are already designed. Now I am trying to close the loop.

Because I want to keep things simple (because I am switching at very high frequency > 300MHz) I will be using PWM control, in voltage mode (as I said) using a type compensator designed using the K factor approach.

I managed to simulate the buck converter using PSS + PAC to get the open loop converter frequency response. I did that because it is need to get the gain at the desired crossover frequency.

Now that I have the type III compensator designed I wanted to simulate the circuit alone, but before that I tried to simulate a circuit from an application note to see if I could get the same frequency response (at least the shape) because I am using VCVS (ideal OPAMP). Well, I did that and the result is that one that you guys can see in my previous post.

I have simulated that circuit using the AC Analysis and the PSS + PAC analysis and I got the same results (between analysis).

When I said that the results (values) were a bit different what I meant was for example, the DC gain in my simulation you can see something around 24dB. In the application note they have a different value and so on. I could post here the application note but I can't find it right now. If you guys wish I can post another result from another application note that I have with me where the same thing happens, the shape is pretty much the same but the values are slight different. I understand that this might be happening because I am using a VCVS (that is a ideal OPAMP) and the author in the application note might have used another OPAMP or the results might have been simulated using MATLAB. Don't know. What I know is that the values are slightly different and that maybe is because I am using a VCVS.

After that I used the components values that I got using that K factor approach and I have simulated in the same way I did with that compensator from the application note.

The compensator:



The simulation results:



Well, what I can say is that at least the frequency response of the gain has the shape that it should. We can see the effects of the pole at the origin and then the zero, pole, etc. The phase I can say that it looks like what it should. We can see that phase boost.

Then I plugged this compensator in my buck converter and I tried to simulate using PSS + PAC.

I want to use the PSS because this is a time varying non linear circuit which the PSS analysis fits very well. That's the point. I want to explore this analysis on the buck converter.

I saw two articles where the authors have used this analysis to simulate the converter and to extract the frequency response from it:

http://edn.com/design/analog/4327245/Periodic-steady-state-and-small-signal-anal...

and

https://www.cadence.com/cdnlive/library/Documents/2009/EMEA/CD12_WerthT_iasrwth....

The circuit that I used is here:



From what you can tell, I have written the purpose of each source.

First I ran a transient and the circuit has stabilized pretty well.

If you look at page 9 from the PPT presentation, you can see that the author used a iprobe. I used too and I inserted that iprobe in the feedback look (after the buck output - low impedance) and before the error amp (compensator, which has high input impedance).
This corresponds to the current probe at page 10 of the presentation.

I can't see the testbench/setup from the author, but what I did was to configure the PSS in the VPULSE source (shaped to a triangle) giving there a name to that source so that in the PSS analysis could appear the frequency that I am working on. I did the normal configuration on the PSS.

Then I tried to perform the PSS + PSTB and the result was:



It looks fine, but I don't know if I am supposed to get this or not. I don't know if I am doing right and if the testbench/setup are right.

I could do a MATLAB simulation through the transfer function of each block but I would spend lots of time on this. (get the SSA of the circuit, compensator, etc)

After getting the result from the PSS + PSTB, I tried the PSS + PAC. For that I used a VDC source (and tried too with a VSIN) in series with the feedback like the author of the presentation and the article, and what I got was this:



and the frequency response:



Now the problem is that if we If we compare the results from the PSTB and PAC analysis with the PSS + PAC we can see that they don't look similar and from what we can see from the presentation that I posted the author says that results from both analysis were equal in his experiment.

I don't know if I am doing something wrong.

Hope made myself clear.

I'd appreciate any help.

Kind Reg