I'm trying to simulate the electro-thermal behavior of an IGBT in a Boost DC-DC Converter (100 Vin - 400 Vout, 8kW Output Power) using Proteus VSM/ISIS. The electro part was easy as pie. For the thermal part, however, I couldn't figure out how to extract the active power dissipated by the IGBT. I'll be using the real-time value of dissipated power by plugging it in a current source, which will be used in an equivalent thermal circuit like so:

I tried:

- The Laplace Multiply Operator to multiply voltage and current. I really don't know how it works, but someone who's supposed to know told me that it multiplies voltages. So, I figured if I could put a Current-Controlled Voltage Source on where I want to get current from and then link its positive terminal, along with the voltage across the IGBT, to the Multiply Operator, I'll get a voltage equivalent to power in value. TL;DR Here's a screenshot of what I did (but this is just a test to get the output power):

Then all I had to was to hook its output to a Voltage-Controlled Current Source...

- A Multiplying Voltage Source ("Ideal 4 Quadrant Multiplier"). Same logic..

When I turned to the IGBT to get the real-time value of power, using the first suggestion, I got this:

As you can see, I am trying to calculate P=Ic*Vce.

I hope I am not wrong with my wiring, but I suppose that the AVCVS (Arbitrary Voltage-Controlled Voltage Source) is going to give me Vce and the ACCVS (Arbitrary Current-Controlled Voltage Source) will give me Ic.

I also got the same result without the AVCVS:

But, if you take a look at the results, it looks like the power consumed by the IGBT reaches 4-6 times the value of power at load??

There must be something wrong here...

Someone from Labcenter told me:

But, that someone couldn't help me further... Because I am using the Demo version...That's wrong! What you are calculating here is the instantaneous power Pswitch = dV/dt * dI/dt while the output power at load is effectively P=VOUT*IL or VOUT^2/R1 that's the same.

Put in another way, you are comparing apples with oranges - which are good to make a fruit salad.

In order to make the two things comparable you must integrate the instantaneous input power.

From my calculation which I verified in Proteus the input power should be less than 200 Watt at 8KW output.

http://support.labcenter.co.uk/forums/viewtopic.php?f=9&t=6466&p=21466#p21466

I know it is possible (and much easier) in LTSpice (or other circuit simulators) using an Arbitrary Behavioral Current Source.. Too bad I'm stuck with Proteus ISIS...

Any help will be much appreciated!

Thanks!

NOTE: This thread is a modified copy of my thread in the Labcenter Support Forums.