0
\$\begingroup\$

I have a question regarding the selection of transistor modules. These modules often have a parallel diode. When calculating the losses and thus the expected heat development, I am not sure how these diodes have to be considered. The maximum allowed power dissipation is (according to this Infineon application note, page 12 equation 2):

enter image description here

I know that both the transistor and the diode has a thermal resistance. Still, I don't see how it affects the temperature rise since, e.g. in a buck-boost converter, the diode and transistor in the module will never conduct at the same time.

\$\endgroup\$
2
1
\$\begingroup\$

Still, I don't see how it affects the temperature rise since, e.g. in a buck-boost converter, the diode and transistor in the module will never conduct at the same time.

Just because two circuit elements are electrically connected in series or parallel doesn't mean their thermal paths are connected the same way. In fact, normally we'd treat the thermal paths as independent, unless the parts are actually physically stacked on top of each other or something.

The concept of thermal resistance is independent of electrical resistance. We call it a resistance because the equations for thermal conduction (\$W=\frac{\Delta T}{\theta}\$) has the same form as the equation for electrical conduction (\$I=\frac{V}{R}\$). But that doesn't mean that the thermal path and the electrical path are the same.

Generally the thermal path for a semiconductor device is from the junction area through the semiconductor, through the package leadframe, to the surrounding air. That is why your thermal resistance formula has two resistance terms, the first one, \$R_{thJC}\$, from junction to case and the second one, \$R_{thCA}\$, from case to "ambient".

So the fact that the transistor and diode are not electrically conducting at the same time has no relation to the fact that they are both conducting heat from their active regions to the air around them pretty much all the time.

These modules often have a parallel diode.

In the device you provided a link to, it's not clear that there is any separate diode. The diode shown in the equivalent circuit diagram most likely is just the body diode formed between the gate and drain implant regions and the bulk or body of the transistor.

More details about the behavior of the body diode of SiC MOSFETs can be found in a Rohm application note.

\$\endgroup\$
3
  • \$\begingroup\$ Thank you for your answer. In the datasheet, it is written: C2M MOSFET and Z-Rec Diode. Then how should I consider the Diode when I choose a module? Would it be: RthJCM + RthJCD? \$\endgroup\$
    – Noah
    Oct 26 '20 at 16:48
  • 1
    \$\begingroup\$ If they're actually separate, you calculate the temperature of the diode using the heat generated in the diode, the \$\theta_{JA}\$ of the diode and the ambient temperature and you calculate the temperature of the FET using the heat generated in the FET, the \$\theta_{JA}\$ of the FET, and the ambient temperature. And you make sure neither of those exceeds the operating limits. If they're actually thermally coupled, then you need to get a thermal model from the manufacturer that tells you how they're coupled. \$\endgroup\$
    – The Photon
    Oct 26 '20 at 16:51
  • \$\begingroup\$ Thank you very much! :) \$\endgroup\$
    – Noah
    Oct 26 '20 at 16:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.