Ran across this vid the other day :
https://www.youtube.com/watch?v=kzuOezZ_quI

Without much concern or schematics he just swapped the output devices and seems to have made some gains.

Thoughts?

Those SIC devices are more efficient than the existing IGBT's especially at higher voltages, as in the case in that video, so the efficiency will be better; however SIC devices generally require a higher gate switching drive voltage to get that better efficiency, so just swapping out the IGBT's for SIC may not make a huge difference, SIC's also cost a lot more than IGBT's.

It is interesting that the alcheapo Chinese unit in the video uses "old" but less expensive tech; for instance the IGBT drivers are all FOD3120 opto isolated drivers, these are quite old now, but cheap. What is great to see there was every IGBT had its own opto driver chip, gives confidence in their product design.

Cheers
Mike

The inverter in the video is quite unlike the inverters built on this forum.  In the video, he is replacing high voltage IGBT's with Silicon Carbide FETs. Silicon Carbide Fets have voltage ratings ranging from 400v to 2000v, while most of us are using FETs rated at 80v.  So an interesting new technology, that would probably be applicable to say a high voltage MPPT, but not the LF inverters we build here.


On another note, I noticed that the guy up in port Hope(near you) is selling more A123 battery systems.  For $300 you get a box with 16x 756w rated packs (plus HV DC relays, fans, heaters, wire etc).  They are very used but you can still net ~ 10KW with 16 packs.  Last time I did this I got around $120 after recycling the unused steel, copper and cast aluminum.  So that works out to $200cad for 10kw (plus a bunch of labour).

BAE Battery units

On the downside, these are 36v packs, which is why I decided to make my system 72v rather than the more normal 48v.

As the video shows, direct replacement of IGBT with SiC MOSFET in a circuit designed for IGBT only provides small gains at a much higher cost.  Not really worthwhile.  However, in a properly designed circuit topology (e.g. dual active bridge) for SiC MOSFET that is switched at a much higher frequency (e.g. 10x+) compared to IGBT inverters there could be significant power density gains due to the size, weight and cost reduction of other components (e.g. inductors, transformers, etc.).  Unfortunately, none of it is DIY friendly.
Edited 2025-07-23 02:33 by analog8484

Interesting, 123A cells, wonder how old they are, this would have been build before the bankruptcy the first time?
That becomes quite a project.

Focusing on finishing my small build of LEV60's (x16) which I brought back while on  southern vacation. Not having to ship across borders and catching the last price drop
made them attractive $22/ea.

Given their reputation for high current hoping to end up with something that can deal with a well pump start.