cancel
Showing results for 
Search instead for 
Did you mean: 

Theoretical issue transistor switching loss academic project

hugo41
Associate

Hello,

I'm a french student in a 2 years intensive class to prepare engineering school entrance exam and I have to present a scientist project to this exam. I choose to work on transistor conduction and switching loss.

 

My problematic is  " To what extent are switching and conduction losses depending on switching frequency ?"

 

I'm willing to establish a "simple" theoretical model of a transistor able to link power losses to switching frequency.

Moreover I'll do some measure and simulation to compare the results.

 

Currently i decided to assimilate the transistor during the conduction phase to a high value resistor when its Off and to a low value resistor when it's On. This two value will be measure by an experiment.

 

However i struggle to find an accurate model to estimate the power loss during the switching phase.

 

 

Do you have any idea on how to do it ?

Or any document dealing with the subject ?

 

Thanks for your help

Hugo 

 

 

 

 

 

 

 

1 REPLY 1
AScha.3
Chief III

Hi,

> However i struggle to find an accurate model to estimate the power loss during the switching phase

Thats to be expected - because ther is no "accurate model" for switching ...what? mosfet, IGBT, SiC-mos..?

All models are more or less good approximations to the real world device.

Just because real mosfet or igbt exibit a very complex behavior in switching, changing/altering when  current, voltage, temperature or gate drive is changed. And even different for switching on and off !

 

But for your question: > To what extent are switching and conduction losses depending on switching frequency ?

I would say, basically: no changes at all. Just in a certain time intervall , 1 sec or so, you have the ON-time loss ( x time ) and the switching loss (x events/time ).

On sate loss and switching are not changing by itself, just both will change with rising teperatur of the device.

So to get it exactlx for a certain device, you have to measure it in real world.

And adjust your "model quality" to the deviation you want to accept.

 

About real switching...just ask google and read...like this:

https://www.infineon.com/dgdl/Infineon%20-%20AN2003-03%20-%20Switching%20behaviour%20and%20optimal%20driving%20of%20IGBT3%20modules.pdf?fileId=db3a304412b407950112b40ee84712c4

 

 

 

If you feel a post has answered your question, please click "Accept as Solution".