Question about MOSFET switching loss with external Coss and custom commutation current in PLECS 4.7.4

PLECS
1

Hello,

I am using PLECS 4.7.4 to simulate an SRDAB converter. The switches use the MOSFET with Diode component together with Infineon thermal descriptions.

I added an external Coss across each MOSFET because I want to model the dead-time commutation and ZVS behavior electrically.

After adding the external Coss, I noticed that the PLECS Device current can show extremely large current spikes during switching transitions, while the actual resonant/commutation current of the converter remains normal.

These spikes can also trigger thermal-model warnings such as:

Loss variable limit exceeded

For example, the thermal description may have an upper current limit of 14 A, while the instantaneous Device current detected by PLECS can reach hundreds or even thousands of amperes during the switching transition.

To prevent these current spikes from affecting the Eon/Eoff lookup, I modified the thermal description:

  • I added a custom variable called Icomm.
  • Icomm is connected to the actual resonant/commutation current of the converter.
  • I copied the original Eon and Eoff lookup-table data into custom lookup tables.
  • The custom Eon/Eoff formulas use Icomm instead of the built-in device current for the current-axis lookup.

For example, the modified turn-off calculation is essentially:

Eoff_base(Icomm, v, T) × Rg_correction(Icomm, v) - Eoss(v,T)

instead of using the built-in device current for the Eoff lookup.

However, I found another issue when checking the switching losses separately:

  • If I set the Turn-on loss formula to 0, the total Device switching loss becomes 0.
  • If I set the Turn-off loss formula to 0, the Device switching loss still has a non-zero value.

Therefore, in my current simulation, almost all of the calculated switching loss seems to come from turn-on loss, while the turn-off loss is zero.

I also checked the MOSFET signals around turn-off. Near the falling edge of the gate signal, the Device conductivity sometimes briefly behaves like:

1 → 0 → 1 → 0

during the commutation in

SRDAB_PD324热模型修改版.plecs (239.7 KB)

pd324低压侧人改.xml (31.4 KB)

pd324高压侧人改.xml (25.6 KB)

terval.

The resonant current is also changing through this interval, so I suspect that the body diode and Coss commutation may affect how PLECS determines the actual turn-off event.

Could you please clarify the following?

  1. How does PLECS determine the turn-on and turn-off loss events for a MOSFET with Diode?

  2. If the MOSFET gate turns off but the body diode subsequently conducts during the commutation, can this cause the effective turn-off event to occur at or near zero current, resulting in Eoff = 0?

  3. Is it valid to use a custom commutation current (Icomm) instead of the built-in Device current for custom Eon/Eoff lookup tables when an external Coss causes very large instantaneous Device-current spikes?

  4. What is the recommended way in PLECS to combine:

    • an external Coss for ZVS/dead-time commutation,
    • a manufacturer thermal description containing Eon/Eoff/Eoss data,
    • and switching-loss calculations,

    without double-counting Coss-related energy or allowing unrealistic current spikes to corrupt the thermal-loss calculation?

  5. Should the very large current spikes generated when an ideal MOSFET switches while an external Coss still has residual voltage be regarded as numerical/ideal-switch artifacts, or is there another recommended modeling approach in PLECS?

Thank you very much for your help.

2

The short answer is that you should not use an explicitly modeled Coss in conjunction with a PLECS thermal model as it invalidates the assumptions of the thermal model - namely an instantaneous transition from the off state to the on state.

You end up chasing various artifacts 1) the instantaneous current discharge of the Coss through the switch on-state resistance and 2) the parallel capacitance means the voltage across the switch voltage cannot change instantaneously at the switching moment. These distort the inputs into the look-up tables. You can use the workarounds you mention in your post like using the tank current, perhaps eliminating the artificial power losses from discharging Coss, but this isn’t a consistent engineering workflow one can make broad recommendations on.

The typical recommendations are twofold:

  • Clearly separate the thermal model used for loss analysis (requires some assupmtions, for example ZVS is always achieved or ZVS losses) from the model used to analyze ZVS/dead-time commutation. This is explained further in this post.
  • Use PLECS Spice which, with the appropriate model of the semiconductor device, can explicitly model the Coss, ZVS behavior, and power loss.