Switching transition time and loss discrepancies - datasheet and simulation

General
1

I am trying to get a better understanding results of a simulation I am running in PLECS Blockset, particularly switching loss as I am need to quantify the conditions for soft switching operation on a circuit.

From the switch loss calculator block in simulation, the turn-on loss for a single switch is about 9.2W, and rough calculation - from the expected currents and voltages and switch datasheet - gives me about 8.9W. That part seems fine.

However, the switch loss calculator block shows switching losses that my switch waveforms do not quite correspond to; the switching waveforms in my results show switching (both turn-on and turn-off) transitions occurring in the order of 10^-17 seconds. I exported the datapoints and created a turn-on transition plot showing what I mean. A close inspection of the data tips for the ‘Current’ shows the transition occurring in about 6*10^-17 seconds. If I am to use this switching time from the waveform, I will grossly underestimate the loss.

image
image2423×1391 357 KB

I am not sure how to get over the discrepancy with my switch transition times in simulation and those what I would expect from the datasheet (also attached: https://assets.wolfspeed.com/uploads/2025/11/Wolfspeed_E4MS025120K_data_sheet.pdf).

I am using thermal models from the manufacturer last updated about a month ago. It doesn’t seem to be a solver error to me. What is the best way to resolve the problem?

2

PLECS uses ideal switches and does not include physical models of components as you would find in SPICE (note this is changing in the near future!). The vendor’s thermal models do not describe the electrical dynamics of the switch and you therefore will not see the typical switching waveforms during transitions. The thermal model uses a look-up table based approach to accurately capture switching and conduction losses without having to model the electrical dynamics of the switching event. See the Thermal Modeling concepts of the PLECS documentation, including the Switching Losses subsection.

Wolfspeed also has a nice write-up on the differences on the PLECS models and SPICE models they provide: