Solar buck converter

Author Message
Senior Member

Joined: 17/10/2017
Location: Belgium
Posts: 258
Posted: 12:10pm 12 Nov 2019      

Hi there,

Yes that's a 3D printed former, a colleague of mine has one. I bought those big ferrite cores, but couldn't find the formers. I'm going to build a new series inductor for my ozinverter.

I combine the wire myself, I used 0.56mm wire, so @ 4A/mm² that is about 1A per wire. I then put them in a vise and twist the wires togheter with a drill.
A few years ago bought a small litz wire bobbin, think it was 200gr for 30€!
I'm not doing that anymore! There is a small company (Bobimat) that's selling all sorts of wire and spools for making transformers nearby. There I payed about 140€ for 12kg enamelled wire rated for 180°C.
For my inverter I did this too but I had to buy a spool of 42kg... that was 10mm² wire.

I don't know if you could build this cheaper then a commercial controller.

I made a schematic, normally I just do it on paper, it's faster that way, but I need to do it more in the future. For more complex pcb's there's no other way...
I didn't used parallel diodes on my gate resistors, and no snubbers on diodes and mosfets. I don't think they are needed, mosfets and diodes are overrated for my use (48V 30A).
With some inductors I experimented with, there were some really nasty spikes on my input and output capacitors. But ones the inductor was well designed (overdimensioned maybe) they were as close to zero.

This one for example worked verry good, but gave a lot of high spikes:

I had it for a half year in a test rig, maybe it would be better with higher frequencies.

Not sure about all the values, the filter on the ouput of the current measuring is not tested.

All measured values are going in a differential amplifier before they are going in the uC. There was something about the outputcurrent of a µC, about sinking and sourcing capability.  It could be important when driving an optocoupler @ high speed.

Edited 2019-11-12 22:34 by nickskethisniks