[SOLVED] SiC Diode for anti parallel diode across IGBT

[rmorris]

Hi. I am looking at using a Silicon Carbide diode as the anti parallel diode across an IGBT.

This may be a bit of a simplistic question but:

At the moment I am using an IGBT Co-Pack where the IGBT has an 'Ultrafast' Silicon Diode as the anti parallel diode.
Can I put the SiC diode in parallel with this to evaluate or do I need to replace the co-pack with an equivalent non-co-pack IGBT.

Basic reason for asking is that it would be a lot of bother to change the IGBT and change it back again due to the hardware configuration.
But I am wondering about the internal co-pack diode turning on first and hogging current etc.

Thanks for reading.

 

[FvM]

As a first step your would want to review the diode data. There is good chance that the SiC diode doesn't share significant current due to it's higher forward voltage. At best, part of the current will flow through the SiC diode, but never the major part.

 

[mtwieg]

I assume that you're using SiC because of its fast operation. If the IGBT has an integrated diode then the external one might not see much use during switching transitions due to higher parasitic ESL, so it will appear that the SiC diode doesn't help at all.

 

[dick_freebird]

SiC is good for very high voltages, high temp, etc. but
is inferior to a silicon Schottky pretty much wherever
you can actually find a suitably rated silicon Schottky
and apply it within its limits.

I think the IGBT run upside-down will see only -1.4V or
so and you'd want Vf to be well less. But most wide
bandgap Schottky devices have high Vfs.

 

[rmorris]

Thanks to all for input. Yes - I think your basic points are correct. Pretty much along the lines I was thinking but was hoping for a quick way to evaluate.
No big problem - will just take a bit longer to implement on the hardware and compare.
for info - it is too high a voltage for standard Schottky ( around 400V - using 600V and 1200V rated semiconductors ). Looking to see if the SiC would increase overall efficiency through less switching loss in both the diode and IGBT at the expense of extra conduction loss in the diode. IGBT gate is driven periodically by a modulated pulse train so not so simple to calculate as a fixed PWM application.
Thanks again