PWM Control for switching power supply with arduino

Other Hardware Motors, Mechanics, Power and CNC
1

Hi fellows,
I'm here because I couldn't find a project of a switching power supply controlled with arduino.
I want to learn how to structure the software....I mean, I want to learn how to use asm code and c code togheter to make a PWM swtiching control..

I want to control a mosfet of the converter with arduino and make the control loop (voltage loop or current loop...etc) with arduino.

If somebody knows some project, please share with me.

thanks
Alex

2

The main issue with switching power supplies that will catch the unwary is saturation
of cored inductors. If you don't catch that fast you'll just be frying components left
right and centre. When a cored inductor saturates the nice steady dI/dt rise will
jump as the inductor suddenly behaves like a piece of wire (ie a short circuit).

Software detection of overcurrent is unlikely to be fast enough, some comparator
based method of capturing hard over-current will be required I think. (Air cored
inductors never saturate but are huge and unwieldy by comparison to ferrite-cored
inductors.

I presume you're comtemplating a few switches, capacitors and an inductor as the
basic skeleton and software control of the switches?

You know about using gapped-cores for power inductors and non-gapped for
transformers?

3

MarkT:
The main issue with switching power supplies that will catch the unwary is saturation
of cored inductors. If you don't catch that fast you'll just be frying components left
right and centre. When a cored inductor saturates the nice steady dI/dt rise will
jump as the inductor suddenly behaves like a piece of wire (ie a short circuit).

Software detection of overcurrent is unlikely to be fast enough, some comparator
based method of capturing hard over-current will be required I think. (Air cored
inductors never saturate but are huge and unwieldy by comparison to ferrite-cored
inductors.

I presume you're comtemplating a few switches, capacitors and an inductor as the
basic skeleton and software control of the switches?

You know about using gapped-cores for power inductors and non-gapped for
transformers?

In my head, I've thought pwm control.
I mean a cicle by cicle regulator... I'm thinking about...
An pwm interrupt to read (without analogread()) input voltage, output voltage and inductor current.... make a little PI control and apply the correction on next pwm cicle then in the mainloop do the other low priority things. with this method there isn't any inductor saturation...

4

So long as you monitor inductor current fast enough, then yes - which ADCs are you
using (fast enough?). But then what switching frequency are you considering?

5

MarkT:
So long as you monitor inductor current fast enough, then yes - which ADCs are you
using (fast enough?). But then what switching frequency are you considering?

I think slow engough to manage all the acquisitions and the PI control inside on 1 PWM interrupt.