3-phase Inverter (MOSFET Si & SiC & IGBT) losses calculation

Hello everyone,

I'm looking for some ressources on how to calculate all the losses (are there others than switching & conducting?) in a 3-phase PMSM motor inverter.

Went through the web looking for it but couldn't find anything...

Should I make a simulation model ? If so, which free/open-source software would you recommend?

Thanks for the help, I'm not an EE by trade but trying my best here... :-)

 

A. B

Hi,

I'm not sure if this would be enough, but might help, http://www.egr.msu.edu/~bingsen/files_publications/C-12_IPEMC_2.pdf

For the control system, look for application notes on MOSFET and IGBT losses in switching applications, as you say, they will be switching and conduction. You will get values from the data sheets for a particular part. You have to be careful how you interpret the data sheets, different vendors may present the data with different control parameters.
I have just done the design for a SMPS, the loss components and how they are applied are available in application notes.
As an example, for calculating the total losses in a MOSFET, you need to know the frequency, gate drive voltage and current, total gate charge and gate to drain charge, RDson, Vds, Ids etc. I loaded the equation into a spreadsheet, together with the parameters, by tweaking the parameters I could observe the impact they had on losses. From my point of view, a lot easier than spice, since I would have needed complete models for quite a few parts,
cheers,
Richard

Hello Richard and A [not the asker's real name I guess?],

That's amazing! Do you use BSIM3v3?
I'm guessing level 3 or 4 only?
I heard they just released level 11 and the equations were really big.

Anyway, I'm really interested in designing circuits.
Do you have any advice or rules of thumb on switch mode power supplies design?
I'd be very grateful if you could share some.

Thanks.

Regards,
Justin

Hi,
as I said, the manufacturers of MOSFETs and IGBTs have application notes covering this issue, just be prepared to do a lot of reading.
Spice can be useful at times, and in this case it may help, but it relies on having adequate models for your power switch, and the circuit which drives the gate, since it has finite switching time and current drive capability. Inductive loads may muddy the waters as well. Again, the manufacturers will have some models available, and others that will be close enough. I would still do a sanity check, spice versus the algebraic method if you decide to use spice. Spice won't help with layout, which is critical in the design of any high power switching circuits, where you are dealing with fast switching times, and large circulating currents.
Again, applications notes will point you in the right direction on this issue.
In my case, most of the calculations were trivial, but the issue of power dissipation is not, and I wanted to use the smallest power switch that was possible, so a spread sheet was an ideal solution. It still involves some guestimates, so go for the worst case parameters,
cheers,
Richard