Buck,boost and buck-boost with PID controller problem

[tester]

Hi!

I try to simulate buck,boost and buck-boost converters in Matlab/Simulink with PID controller but I don't know how calculate PID parameters Sad Anybody knows how I calculate this parameters ? Or maybe someone haves buck,boost or buck-boost example circuit in Matlab/Simulink ?

Or maybe someone simulate this converters with another method of control like hysteresis controll, current programmed control or sliding mode control ?

Plz help me :(

PS. Sorry for my language mistakes :(

 

[berkeman]

What power supply textbook are you using? And what power supply design website resources have you found so far? How detailed do you want your simulation to be? Like, do you want just simplified cycle-by-cycle simulations, or do you want SPICE level accuracy, including parasitics?

 

[tester]

I looked to Simulink Help and I found model of boost converter : http://www.mathworks.com/access/helpdesk/help/toolbox/physmod/powersys/igbt.html

But I want to build buck, boost and buck-boost converters using an Ideal Switch (not IGBT or MOSFET) and I want build my converters with one of control methods in closed-loop circuit (not using block Pulse Generator). So I have decided to build this converters with PID regulator and I don't know how design this regulator.

I read this pages :
http://services.eng.uts.edu.au/~venkat/pe_html/pe07_nc8.htm
http://www.powerdesigners.com/InfoWeb/design_center/articles/DC-DC/converter.shtm


Maybe someone earlier try to simulate this converters in Simulink and have some examples circuit ?

 

[berkeman]

The last link you show is a good one. It doesn't talk much about the feedback loop, however. Your PID circuit will need to use a sample of the output voltage, and determine the cycle-by-cycle pulse width to use in turning on your ideal switch. Quiescently, the PW will just be related to the ratio of the input and output voltages (assuming an ideal switch and a zero volt drop ideal diode). But when the output current varies, your control element needs to use PWM to keep the output voltage within regulation.

What have you used PID loops for so far? How much control loop theory have you had?

 

[berkeman]

I googled "buck converter" +feedback +tutorial, and got lots of good hits. See if some of these are helpful:

http://www.maxim-ic.com/appnotes.cfm/appnote_number/2031

http://www.smpstech.com/map.htm

http://www.google.com/search?hl=en&q="buck+converter"++feedback++tutorial

 

[tester]

Heh :(

Maybe you take a look for my Simulink model and maybe you try fix this ? Here is link : http://gacolek.wz.cz/buck_PID.zip

I wait for answer :)

 

[berkeman]

I'm only able to read the Simulink file as a text file -- I don't use Simulink.

What compensation method are you going to use for the feedback loop? Are you using lag or lead compensation? What is your phase margin? Are you using a state space approximation in your calculations? What minimum current are you going to allow on the buck output? Are you going to include discontinuous mode operation (and the different loop characteristics), or are you going to require a minimum Iout that keeps your buck running in continuous mode?

 

[tester]

I try to build a buck converter with follow parameters :
Vin = 24V
Vo = 12V
Ro = 13Ohm
L = 69uH
C = 220uF
fs = 100kHz
it's work with CCM mode.

My feedback signal witch I use to control is voltage. In feedback loop I added PID regulator, where the lead compensator is derivative action (D) and lag compensator is integral action (I).

 

[berkeman]

Good so far. Now what is your feedback equation? What gain-phase plot does that give? What is your phase margin?

To test your control loop, you will need to vary the output current and see what the loop response is. You generally test a power supply circuit to determine its load response (how well the output voltage stays regulated when the output current changes) and its "line" response (how well the output voltage stays regulated when the input voltage varies). Your simulation should test both of these responses, and your calculation of the phase margin is an important measure of stability.