Circuit for measuring switching losses of CAB450M12XM3

Przemek

Hello,

I am trying to make a simulation model of SiC MOSFET from CAB450M12XM3. I want to validate it among others by comparing the switching losses from the simulation and datasheet. However, I need help finding the information on the measuring circuit in which Eon and Eoff's values were measured in the datasheet. Was it the double pulse test setup? If yes, what was the freewheeling diode? Was it a body diode of the second transistor in the module or an external diode? Thank you for your answer.

Best regards,

Przemek

Forum_Moderator

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

BlakeNelson

Hello Przemek,

The switching losses of all our power modules are characterized by double pulse testing using our CIL evaluation kits (e.g.https://www.wolfspeed.com/products/power/evaluation-kits/kit-crd-cil12n-xm3/). For modules, we always use the inactive switch position as the freewheeling path for the inductor. So when testing the low side, we use the high side MOSFET's body diode as our freewheeling path.

If your goal is to match the datasheet's switching loss values, you will also need an appropriate model of the PCB and gate drive (the gate drive used for testing XM3 switching losses is https://www.wolfspeed.com/products/power/gate-driver-boards/cgd12hbxmp/). You can find the PCB's parasitic inductance in the RBSOA plot (Figure 19) of our datasheet (6 nH for the CAB450M12XM3). You could instead get the layout details for the KIT-CRD-CIL12N-XM3 (under technical documents on the kit's web page) and use an analysis tool (e.g. ANSYS Q3D or COMSOL) to create a more sophisticated model of the system parasitics, though I would claim that probably isn't necessary to accurately model the switching losses. Finally, for the gate drive model, I would recommend using an ideal voltage source with 0.5 Ω ESR as a starting point. You could also add 10 nH of gate loop inductance.

Finally, we do try to match the datasheet switching losses (typical target of +/- 20%), but given the range of possible inputs there will be areas that are less accurate. Our emphasis is on nominal voltage (600 - 800V), current (~450 A), and temperature (125­ °C).

Let me know if you have any additional questions!

Blake

Przemek

Hello Blake,

Thank you very much for your answer. It will be very beneficial for me. Right now, I have no more questions. Thanks!

Przemek