10KW Buck Converter using Arduino & IGBT

Projects General Guidance
1

Hi,

I'm interested to build 10KW buck converter with following parameters:

1-Input voltage range; DC 150V ~ DC 400V
2-Output voltage fixed: DC 56.0V
3-Use of IGBT STGW80H65DFB - In parallel configuration.

I've following considerations in mind already.

1-N Channel IGBT will require additional gate driver as arduino can't drive it alone.
2-Feedback will be made available making sure output is stable.
3-Freqency of PWM will be changed to high side
4-LCD display showing delivered power will be add-on
5-Efficiency show on LCD will be a good addition too.

Please share any similar project for me to as starting point. or share your inputs making it error-free design.

2

Hi, @maqsolarenergy
Welcome to the forum.

Can you please tell us your electronics, programming, arduino, hardware experience?

This is not a small enterprise.

10KW at 56V.
Output Power = Output Volts x Output Current.
Current = Power / Volts.
Output Current = 10000 / 56 = 178 Amps.

What is the converter for and are you aware of the high output current.
Not only circuit but physical properties have to be considered.

How many of these do you aim to use?

image
image1002×728 57.9 KB

Do you have a basic circuit?

Thanks.. Tom.... :smiley: :+1: :coffee: :australia:

3

Thanks for the response.

I don't have much experience with designing power circuits as of now. but we've been using them in out solar deployments as (MPPT or PWM Controllers).

Now with my custom requirement; i want to build my own.

I intend to use 8 of these IGBTs in parallel with 300W dissipation as heatsink design in mind.

This will be used as Buck dc-to-dc converter from solar panels and will be charging batteries at fixed volts. (CV only).

4

Please buy something off the shelve...
This is not an easy task.
It may well cause a fire...

5

:rofl: - trying smaller in lab. like 6 to 8kw delivery. then will proceed for 10 to 14kw DC to DC delivery. - might take some time; as I get little time from main routine tasks.

6

How you prevent panel voltage collapsing when your pv power is below "nominal" if you only have CV output.

7

It never ceases to amaze me that amateurs are confident that they can "design" and build something as advanced as an extremely high power buck converter.

@maqsolarenergy Best to wear a face shield and body armor when testing, to avoid severe injuries from molten metal splatters.

8

will use feedback mechanism to alter the width of pulse. making sure o/p remains constant.

9

Noted :rofl: -

10

"Switching Regulators for Poets"
This app note was written by a well known app engineer at Linear Technology (now owned by Analog Devices). The App note is decades old and is completely out of date but you must read at least the first page. Believe me, what he says is absolutely true.
So you have a long road ahead of you.

an25fa.pdf (330.9 KB)

I've done several buck designs but all were below 1kW.

11

And I'm interested in Alessia Crippa being my girlfriend or wife, but not everything is possible in this life.
There are those who ask for daring designs, but you have gone too far.

12

Thank you sir. I'll check and revert.

13

As per document share & others; Most of the issues are related to side-effects of being operating at higher frequencies; I wonder if physical size is not a constraint. have you ever heard of 1khz buck converter? - Just for my knowledge. I mean ... I'm at a stage where I can afford a bit (or byte) larger size but to understand the concept well. then at later stages I will increase the frequency to optimize the physical & electrical efficiency of the circuit.

14

image
image605×628 29.3 KB

can we also try h-bridge + step-down transformer design. your thoughts...

15

What do you have in mind?

16

Stage 1. Capacitors bank to stabilise voltages.
Stage 2. H bridge at 1khz (or higher)
Stage 3. Power step down transformer and filters. Dc and stability
Stage 4. PWM only to meet load requirements

17

I'm still mainly interested in this one, you have some examples?

18

No.
Even if they were available, the size of the inductors and capacitors required would make such a design impractical.

19

A transformer for such a design does not exist. You would have to design and build your own transformer.

20

This is just speculation but given that power conversion at 50Hz or 60Hz uses reasonably sized transformers it doesn't seem to me to be unreasonable to build a buck converter using 1kHz. The only reason I can think to attempt this would be to satisfy interest and curiosity, I can't imagine a practical use that couldn't be done better with higher, more usual frequencies.