Power losses in SiC MOSFET switching cells.

Discussion in 'Homework Help' started by KvotheKote, Apr 1, 2016.

  1. KvotheKote

    Thread Starter New Member

    Mar 11, 2016

    I'm a student of Electrical Engineering, and I chose to specialize in Power Electronics. I have a base in the theories about the subject, but now I've a practical assignment to do and it is causing me a lot of brainstorm. I have to calculate and measure the power losses in a SiC MOSFET to be used in a synchronous BUCK converter. The conduction losses are easy to measure, but the switching losses are a lot trickier. I did a bibliographic research, and found 3 common methods to measure those losses. They are: the Calorimetric Method, the Opposition Method and the Double Pulse Method.

    About the Calorimetric Method, I have some issues because I'm not able to separate the Eon from the Eoff, and I need each one separately. Other problem is that I need to keep the temperature quasi-constant and that's not possible with this method.

    Now, the other two are confusing me. First off, the Opposition Method is highly dependent in Rdson, and its value is highly dependant in the temperature and if the wrong value is used, a huge error is propagated to the result. This method is good because we don't insert any parasitic impedance in the switching loop.

    In the Double Pulse Method we have to use probes to measure the current in the transistor, and as you know, the SiC MOSFET have very little impedances, so any parasitic impedance will change the switching process. Other problem is the need for high bandwidth probes. I have available a Rogowski probe with 30 MHz bandwidth, but I don't know if that's enough. Another problem of this method is the synchronization of the voltage and current curves. But for this, I can use some MATLAB scripts and synchronize both curves.

    I have a ready to use model to do the Double Pulse Method, so I think I'll just make the tests and analyze the results. But are they reliable? Did I think of any possible factor that can cause me problems and errors? Should I make a model for the Opposition Method? Is there any other solution for this problem?

    I would be really grateful for any input you guys can give me.

    Sorry for the English errors and thank you all.