Hi all!

Well, I got my interest in the nonlinear dynamics and bifurcation diagrams and I am trying to simulate a buck converter with output voltage feedback control.

I am using opamps in order to make my feedback equation which is [gain*(Vreference-Vout)=Vcon] and Vcon is compared with a ramp voltage to make the PWM pulses.
I am reading from literature that you just have to adjust the parameters correctly but I see it is quite a difficult task to do...

Well another problem is that LTSpice isn't responding in the beginning of simulation telling me something about "stepping source blah blah press escape to quit". After pressing escape though the simulation is running OK.

The steady state seems to run OK but I can't achieve the period doubling or the chaotic operation. (Or I achieve it and I can't recognize it!)
I am uploading as well the simulation file so you can play with it as well! I am pretty sure my design is tremendously poor therefore I am asking for your help! Thanks for the favor everyone!

 

Hi chao,
You have two Vout labels, change the lower right side one to say V0, also in the Transient analysis, set to 'Skip inital operating point.
E

 

Hi,
I have moved the circuit around, no actual changes.
E

 

Hi,
I have moved the circuit around, no actual changes.
E

Hi Eric, thank you for your reply. The tip about the initial point worked pretty well!
I am new at LTSpice and I didnt really understand the labeling in your first reply, I think it is no problem, right?

Anyway, did you get to see the period doubling process somehow?
Does anybody has the experience to explain it to me? Thanks a lot again!

 

Hi all!

Well, I got my interest in the nonlinear dynamics and bifurcation diagrams and I am trying to simulate a buck converter with output voltage feedback control.

I am using opamps in order to make my feedback equation which is [gain*(Vreference-Vout)=Vcon] and Vcon is compared with a ramp voltage to make the PWM pulses.
I am reading from literature that you just have to adjust the parameters correctly but I see it is quite a difficult task to do...

Well another problem is that LTSpice isn't responding in the beginning of simulation telling me something about "stepping source blah blah press escape to quit". After pressing escape though the simulation is running OK.

The steady state seems to run OK but I can't achieve the period doubling or the chaotic operation. (Or I achieve it and I can't recognize it!)
I am uploading as well the simulation file so you can play with it as well! I am pretty sure my design is tremendously poor therefore I am asking for your help! Thanks for the favor everyone!

What evidence do you have that these behaviors occur in a buck converter system? Stable oscillatory behavior is possible, even in non-linear systems. They certainly wouldn't be much use if it were otherwise. I don't think they are chaotic in the sense of the Lorentz Attractor or the Rössler Attractor.

https://en.wikipedia.org/wiki/Lorenz_system
https://en.wikipedia.org/wiki/Rössler_attractor

 

What evidence do you have that these behaviors occur in a buck converter system? Stable oscillatory behavior is possible, even in non-linear systems. They certainly wouldn't be much use if it were otherwise. I don't think they are chaotic in the sense of the Lorentz Attractor or the Rössler Attractor.

https://en.wikipedia.org/wiki/Lorenz_system
https://en.wikipedia.org/wiki/Rössler_attractor

Hi Papabravo, thank you for your reply!
Yes!! Power converter like the buck converter can behave chaotically, isn't it amazing? If it is of use under this state is a whole other topic! But they do have strong nonlinear phenomena such as period doubling and chaotic motion.

Please see attached paper for your reference.
Here is the link also if you wish to download it:
http://www.ncnsd.org/proceedings/proceeding05/paper/132.pdf

 

Hello Chao,

Have you got any further work on Period Doubling and Chaotic motions in LTSpice?

Thank You.

 

Hello Chao,

Have you got any further work on Period Doubling and Chaotic motions in LTSpice?

Thank You.

Nothing on those two topics, but the new thing that I have recently discovered is subharmonic oscillation. This is where the system has oscillations at frequencies that are sub multiples of the switching frequencies. They cause a dramatic increase in the output ripple, and they don't seem to die out on their own. This happens when the duty cycle in a buck converter exceeds 50%. There is however a solution which is slope compensation. None of this is new news for those involved in power supply design. I only became aware of it because of a deep dive I have recently undertaken.