CCB021M12FM3 PLECS model: no body diode recovery loss Err chart
Hi, there is no body diode recovery loss of CCB021M12FM3 PLECS model, could you comment how to add this Err loss in the PLECS simulation?
Comments
-
BlakeNelson Wolfspeed Employee - Contributor Level 2
Hello KNJ,
The value of Err is often not included in the PLECS models as it can cause simulations issues. Specifically we see at higher Rg's the energy loss becomes quite small leading to occasional convergence issues. Therefore we have traditionally left it out in the name of making our models as robust as possible.
To add it yourself, in PLECS you can add an image from the datasheet to the Turn-off loss section of the model.
You can then curve fit the losses in the datasheet shown in figures 11 and 12 (you will obviously need to fit each figure individually, which is accomplished by adding a second image):
Plexim has tutorials on model development if you have not done this in the past (https://www.plexim.com/sites/default/files/tutorials_categorized/plecs/thermal_domain.pdf).
Of course, you may need the model to account for Err's dependence on gate resistance or temperature, both of which can be accomplished by using a table + formula model (rather than lookup table only). For gate resistance, I would look at figure 16 and multiply E by the ratio Err(Rg=your Rgon)/Err(Rg=0 Ω). For example, this could look like:
For an Rgon of 6 Ω. (Remember, Err is turn off of the diode, and the di/dt is affected by the turn-on gate resistance of the opposite switch position in the switching leg). Of course, adjust the numerator to the value at your Rg of interest (or skip this step if you are using 0 Ω).
Temperature could also be done with a fixed scaling constant from Figure 14, or you could curve fit Err(Tj). E.g.:
You of course need to normalize the fitting formula by dividing by the 25 °C value (use 25 °C because figures 11 and 12 are defined at 25 °C). Making the final formula:
Let me know if you have any additional questions!
Blake
-
KNJ Contributor Level 1
Hi Blake, thank you for the thorough reply!