spice transformer model using physical parameters

[PlasticIsGood]

I haven't seen this done before. Might be of some interest before I delete the pics.

I made these models originally for valve output transformer experiments a few years ago. They are in Proteus, a 3f4 variant. I assume similar can be done in any version because the idea is so simple, although I must admit I don't know how spice manages to resolve the circular reference in the functions.

The original idea was to take into account hysteresis and varying mu, but I never got that far.

The models, one for a core and one for a coil, use voltage and current-controlled voltage and current sources, together with capacitor differentiation, to model the magnetics as an electrical circuit.

The magnetic circuits of coils and core can be linked up as shown to make a transformer. Any number of different coils can be wired in parallel to the same core. Flux density can be monitored as shown.

Parameters appearing in <brackets> should be set up so they can be entered when the models are used. Hopefully the meaning of each is obvious...Rw is winding resistance, etc...

One such parameter is mu, which unfortunately remains constant. I couldn't think of a function to model the BH curve. In consequence, hysteresis is not modeled either.

First pic is of a transformer in operation, using the two models, at top. Other simple models show output is same for all. Dunno what the one on the right was for.

Second and third pics are the models underlying the coil and core. The winding resistance, leakage inductance and winding capacitance don't have to be split up like I've done...I was just messing about.

In functions including the likes of I(A,B), the A and B refer to the controlling inputs of the voltage or current source.

[jcx]

from Ltspice inductor help description:

There other non-linear inductor available in LTspice is a hysteretic core model based on a model first proposed in by John Chan et la. in IEEE Transactions On Computer-Aided Design, Vol. 10. No. 4, April 1991 but extended with the methods in United States Patent 7,502,723. This model defines the hysteresis loop with only three parameters...

probably more on Ltspice Yahoo group (have to sign up)

I also like cruising Intusoft site, newsletters for spice hints

[PlasticIsGood]

Interesting, thanks. At the time I was working on the idea, models including hysteresis were defined only for a particular constant frequency. From memory, they used several capacitors with leakage resistances and zeners, such that the caps would absorb energy beyond certain flux densities, and not return all of it.

I'll take another look.