3-Phase PFC Rectifier

Hello,

I have been simulating the example "3-phase PWM Rectifier with Power Factor Correction" and it's able to track the DC setpoint and the currents look sinusoidal with a resistive load. I am trying to convert the blocks to the digital domain to be able to generate code, but am not getting similar results. I have attached some images showing the blocks that I've converted. Any help would be greatly appreciated.

inner-current-loop-id.JPG

inner-current-loop-iq.JPG

outer-DC-voltage-control-loop.JPG

PWM.JPG

Find more posts tagged with

English

PSIM

Accepted answers

All comments

Albert_Dunford

Hi Esha,

you are likely encountering a few problems

s-domain to z-domain translation of coeffiecients

This example in the digital control example folder is a great resource

along with this conversion tool

That being said you are going to have problems as you will likely need to redo the control design entirely to account for digital delay and its impact on the phase margin! please see this tutorial video.

Also this tutorial video does a full 3 phase bidirectional converter you should find it very useful:

Also this webinar covers the entire code process for a 3 phase PFC and does verification with typhoon HIL.

Please follow up with additional questions

Community-Pasted-Attachment-1723825498604.png

eshea

Thanks for the links Albert. I've been watching the webinar and you do an AC sweep when inverting power to tune the current controllers. Can I use the same approach, but tune the current controllers when rectifying power? I'm curious why you tuned them while inverting power.

Albert_Dunford

Thanks for the links Albert. I've been watching the webinar and you do an AC sweep when inverting power to tune the current controllers. Can I use the same approach, but tune the current controllers when rectifying power? I'm curious why you tuned them while inverting power.

It has been a while since I did that. I recall it was easier to do it that way to keep track of the sign convention for dq vs power flow vs input voltage phase vs theta generation/detection. The controllers will work in both direction the tricky part is the sign convention.

If you are manually tuning since you don't have smartctrl it might not matter. You just need to be careful to about your conversion.

additionally an different method to convert the controller, you need to find a crossover frequency that has at least 45 degree phase margin

- do a loop gain sweep of the analog system - this tells you phase margin at crossover

- do an open loop sweep of the digital and of the analog system - this gives you the difference in phase response between analog and digital

- compare the phase difference in open loop between analog and digital at the analog close loop system, this should give you the expected phase margin of the digital system. From what you describe this will likely be negative or very close to zero.

- if you make the "P" of the PI smaller directly you will just shift the cross frequency lower without impacting the phase response, so if there is a stable phase margin of the digital open loop at a frequency you might be able to just make the P smaller of the analog PI.