3rd quadrant operation power loss of power switch

Question

Hello,

I have some basic entry-level questions regard the power loss of MOSFET with bodydiode and hope someone could provide some clarification - thank you in advance for patience!

1, When both current and voltage are reversed, it is called "3rd quadrant" operation, correct?

2. My understanding is that when 3rd quadrant operation, only bodydiode will conduct the current(reversed) and thus has CONDUCTION loss. But why the manufacturer given thermal model/datasheet also include the CONDUCTION loss of the MOSFET itself? So, MOSFET itself is also able to conduct the reversed current? How?

3. Even if MOSFET itself can also conduct the reversed current, why there's no SWITCHING loss (but only CONDUCTION loss) involved in the given thermal model?

4. The given thermal model of the bodydiode has no SWITCHING loss. Why is it so?

Thank you for patience.

Answer 1

Hello.

Here are answers to your questions:

1. Actually 3rd quadrant operation is when the current through the device is negative (either through the channel or body diode). So, the voltage could be positive or negative.

2. See above- Yes, MOSFETs are bi-directional and conduct current in both directions when sufficient gate voltage is applied (unlike the IGBT which can only conduct current in the forward direction).

3. What thermal model are you speaking about? The conduction loss will correspond to third quadrant operation, not switching loss. Turn-on and turn-off losses for the MOSFET channel could be for (before or after) forward or reverse conduction.

4. Diodes typically have negligible turn-on losses, but have more substantial turn-off losses, e.g., due to the reverse recovery effect. If your diode thermal model has no switching loss it may have been left out or not considered substantial enough.

If you reference a specific part I can give more feedback.

Regards,

Kris

Comments

Thanks a lot Kris for the response. I actually took my time to make a video about "How I manage to add thermal models" for two real SiC Mosfets. Here's the video link and I'd like to have some inputs/corrections on what I did and how I understand them.

Thank you very much again!