Non-ideal inductors in SPICE

[tb1990]

I need to simulate a non-ideal inductor in SPICE (I'm using LtSpice) so that I can test the saturation behaviour of the inductor. I've used this link: The Arbitrary Inductor model - LTwiki and have been able to make the inductor saturate. But I was wondering if there are any better ways to simulate the saturation as the method in the link is rather simple.

Thanks.

 

[FvM]

Did you notice, that the LTSpice help file is referring to two different types of non-linear inductors? The tanh() characteristics discussed in the LTWiki seems to correspond to the simple first varaint. There's however a second built-in model with hysteresis. Other methods from literature are basically available with LTSpice behavioral simulation features.

 

[tb1990]

I did read the about the built in model, but I'm not sure how I implement it and can't seem to find any documentation for using it.

Edit: I've seen the documentation about the second non-linear model. But I don't know how to apply it in LTSpice, as the example uses netlist, but I can't seem to edit the netlist in LTspice.

 

[FvM]

Use the model in schematic entry.

 

[tb1990]

If I was to simulate this core... http://www.elfa.spb.ru/uploads/tdpdf/05875885.pdf

Would these be the right parameters?
Hc=52. Bs=.36 Br=.31 A=0.0000309 Lm=0.0237 Lg=0.0131 N=1

 

[FvM]

Lm is specified as 55.8e-3, Lg should be assumed 0. I would also use the typical Hc value from the hysteresis loop, about 35.

I'm curious, if the model is representing the behaviour of the hard magnetic material correctly.

 

[tb1990]

I've been able to get inductor to do what I wanted it to. Thanks for your time and help!

 

[tb1990]

I take it this isn't right?


Also, if I was to change the current at which the the inductor saturates how would I achieve this?

 

[FvM]

The curve is a correct result of the rectangular hysteresis curve, I assume. To get a B/H curve, you need to plot integral V against I.

 

[atripathi]

You can easily accomplish this in PSpice. In PSpice you can model core, to support BH curve non linearity, saturation, effect of air gap etc.