N-Channel MOSFET Body Diode Tabulated I-V Curve Convergence Error

In Simscape, I am switching in an ideal 500 VDC source to an RC load using the following N-channel MOSFET: Wolfspeed CAB450M12XM3. The circuit is relatively simple.

I added a body diode based on data points extracted from the body diode I-V plot in the datasheet. Voltages and currents in the Simulink component block parameters (below) came from the QIII body diode I-V plot in the datasheet. The plot from the datasheet is given below the Simulink Body Diode parameters. How did this addition cause an unconverged equations error at the MOSFET, and what can I try to fix it? Give more datapoints to the block parameters? Some kind of numerical fix?

The entire error message is reproduced below the body diode plot from the MOSFET datasheet. Thank you for any help that people can provide.

An error occurred while running the simulation and the simulation was terminated

Caused by: Here is the set of components with unconverged equations:

'LSS_v3_LS_Only/N-Channel MOSFET1'
Equation location is:
'C:\EngApps\MATLAB\R2021b\toolbox\physmod\elec\library\m\+ee\+semiconductors\diode_base.sscp'(no line number info)

[Rick Hyde]

Hi Brian

If you could send us a copy of your model, we can take a look. One thing you might try is adding some more points to your diode I-V curve for lower forward voltages e.g. for voltages of 1, 2 and 2.5 volts. With the values you have, interpolation will be occuring between points at 0 and 3 volts which will not give you the typical diode I-V curve. However, I would still expect the circuit to simulate.

Rick

[Brian Malone]

Thank you, Rick. Please see the model attached.

[Brian Malone]

Has anyone had a chance to look at this body diode issue? Thank you.

Answer 1

[Sabin]

The model although simple poses some numerical challenges for the solver. Due to recent improvements the model can simulate without errors since R2024a release.

I can reproduce the error in R2022b release. The model is a bit on the edge considering the parameterization and the mix of discrete and continuous states. Because there is a discrete-time logic controlling the voltage source one possibility is to enable local solver and run it in discrete time with Backward Euler.