# Interpretation of conduction losses

360 views
Hi there,

I'm pretty new with PLECS and wanna try to understand how conduction losses are calculated in a MOSFET.

I’ve checked that it’s calculated as (in watts) as Vth * Ids, where Vth is the MOSFET voltage drop and Ids is the MOSFET current (from D to S). Apart from that, Vth is calculated as a function of the Ids and Tj.

However, I’ve a couple of questions:

- Is there any way to set the Rdson value? I’ve checked that there is a parameter called Ron. Is this the same? Can I have set the curves that relate the Rdson value (in p.u.) depending on the Tj and Ids?

- Why does not conduction losses consider also the term Rdson * Ids^2?

Best regards,

You can set the electrical on-resistance, Rdson, however, this is not directly linked to the lookup table data for conduction loss, and is anyway fixed value (non-temperature varying/dependent).

We use lookup tables to calculate the conduction loss (or using an equation defined in the thermal description), so while a physical/electrical on-resistance could have a small but non-zero influence on the currents that a device "sees" such that the resultant conduction loss lookup can change slightly with different electrical resistance values, the electrical and thermal domains are effectively parallel. The electrical on-resistance rather could be useful for looking at current sharing conditions or providing dampening, and if one doesn't provide a thermal description (or specifically data for the conduction loss), the on-resistance is the only way that a device on a heat sink component can have its conduction loss calculated.

But I don't fully understand your question about the Rdson curves, and wonder what your ultimate goal is? Are you just saying you have Rdson vs. Tj and Ids data, but not IV curves directly?  We may be able to provide some suggestion if you clarify what you really want to achieve.

And let me know if anything I said was unclear.
answered Jun 12, 2023 by (1,575 points)
Hi Kris,

Many thanks for such a detailed feedback.

My ultimate goal is to model de conduction losses as the result of Vth * Ids + Rdson*Ids^2. That’s why I asked before about the Rdson parameter.

However, I’m interpreting from your feedback that the look-up-table defined in the conduction losses considers directly the Vd as Vth*Ids + Rdson*Ids so that the conduction losses are directly derived as Vd*Id, isn’t it?

Regarding the Rdson, yes, my goal was to know if it’s possible to define a look-up-table to define the dependence of the Rdon of the Tj and Ids. Nevertheless, this wouldn’t be needed if the assumption of the previous paragraph is true.