Wolfspeed Power Applications Forum

Wolfspeed Technical Community:

SpeedFit™ Design Simulator

Simulate and evaluate the performance of SiC-based power circuits and determine the right SiC device in seconds.

Start a Simulation

Categories

PLECS thermal model transient thermal impedance, is it correct?

Options
Mikesim
Mikesim Contributor Level 1
edited November 2023 in General

Hi, I have a question regarding the thermal chain in the PLECS thermal description e.g.

CAB425M12XM3.xml

CAB425M12XM3_bodydiode.xml

It appears to me that the thermal chain has been added to the body diode as well as the FET. I don't think this is representative as in reality, the bodydiode is closely thermaly linked to the FET and share a common thermal chain. In most applications this will not matter much as the diode losses should be small. However, in pulse power applications or when modelling fault conditions the "paralleling" of the thermal chains will significantly underestimate the diode junction temperature. I can think of a workaround for this in PLECS but I'd have thought it would be better to use the FETD model so the losses ascoceated with both the FET and the body diode go through the same thermal chain.

I'd be interested to hear other opinions on this.

Cheers

Mike

Tagged:

Comments

  • Forum_Moderator
    Forum_Moderator Wolfspeed Employee - Contributor Level 5
    Options

    Thank you for your post, it has been approved and we will respond as soon as possible.

  • BlakeNelson
    BlakeNelson Wolfspeed Employee - Contributor Level 3
    Options

    Hi Mike,

    I definitely agree, the body diode should be at the same temperature as the MOSFET, and the models as-is do not accurately reflect this. Unfortunately, the FETD also has a limitation: by default you can only give the losses for reverse conduction when the part is not gated on (it also lacks Err, but that is less significant). Most of our customers take advantage of synchronous switching to reduce 3rd quadrant conduction losses, so not simulating the 3rd quad losses correctly would be unacceptable. Of course this leads to the body diode error you have mentioned.

    We recognize that it would be best to have a model that has a single thermal node for both MOSFET and body diode while not losing any model features, so we have been developing just such a model (it also has a few other improvements). I can't promise a release date at the moment, but hopefully this tradeoff will be moot in a few months. We aren't ready to share a beta, so I wish I was able to give you a more immediate solution. If you have a need for a "FETD" model of the CAB425M12XM3, it would be possible for us to make one.

    Thanks,

    Blake

  • Mikesim
    Mikesim Contributor Level 1
    Options

    Hi Blake, thanks for your reply. This is not an immediate problem for me as the diode losses are very low in the applications I'm looking at.

    It's good to know it's being worked on though.

    Cheers

    Mike

  • BlakeNelson
    BlakeNelson Wolfspeed Employee - Contributor Level 3
    Options

    Ok, great. Glad I could help.

    I will close this discussion for now but if you have a follow up question, please "Start a New Discussion" and I would be glad to support you further.

    Blake

This discussion has been closed.