Using PSPICE to model magnetic hysteresis

[X. WANG]

Hi, all friends
I'm trying to use pspice to build a model for component with magnetic
hysteresis (actually that's a transformer core). PSPICE Manual book says
that hysteresis model in PSPICE is based on Jiles-Atherton model.
However, there is one thing confusing me.
In pspice, for a magnetic core, there is only four parameters to describe
hysteresis. They are: Ms, A, C, K
I read some academic papers about Jiles-Atherton model. Actually in original
Jiles-Atherton model, there should be five parameters: Ms, alpha, a, c, k
So, does anybody know the relationship between these two set of parameters?
BTW: The pspice I'm using is the one in Orcad 9.2

Thanks a lot.
X. WANG

 

[Jim Thompson]

Hi, all friends
I'm trying to use pspice to build a model for component with magnetic
hysteresis (actually that's a transformer core). PSPICE Manual book says
that hysteresis model in PSPICE is based on Jiles-Atherton model.
However, there is one thing confusing me.
In pspice, for a magnetic core, there is only four parameters to describe
hysteresis. They are: Ms, A, C, K
I read some academic papers about Jiles-Atherton model. Actually in original
Jiles-Atherton model, there should be five parameters: Ms, alpha, a, c, k
So, does anybody know the relationship between these two set of parameters?
BTW: The pspice I'm using is the one in Orcad 9.2

Thanks a lot.
X. WANG

Are you reading the reference guide? I don't see "alpha" there.

...Jim Thompson

 

[X. WANG]

Jim,

It's ture that the reference guide of pspice uses four parameters: Ms,A,C,K
as pspice does. And it says that this is based on Jiles-Atherton hysteresis
model.
But if you read the original paper from Jiles-Atherton, for example,
D.C.Jiles, D.L.ATHERTON, Theory of ferromagnetic hysteresis, Journal of
Magnetism and Magnetic Materials 61(1986) 48-60
you will find it diffrent from reference guide of pspice.

In original Jiles-Atherton model, there are five parameters with parameter
ALPHA.

I think the difference between original Jile-Atherton model and the one in
pspice is the way to describe anhysteric magnetization.

In pspice:
Man=Ms*H/(H+A)
where Ms and A are two of parameters

In original Jile-Atherton model, modified Langevin function is used
Man=Ms*(coth( (H+Alpha*M) / a )- a / (H+Alpha*M))

then there is Alpha parameters.

Actually I can get the parameters of some material in the latter form (with
alpha), but the difficulty is that I don't know how to reasonably convert
them into the form in pspice without alpha

thank you, Jim

Richard

 

[Jim Thompson]

Jim,

It's ture that the reference guide of pspice uses four parameters: Ms,A,C,K
as pspice does. And it says that this is based on Jiles-Atherton hysteresis
model.
But if you read the original paper from Jiles-Atherton, for example,
D.C.Jiles, D.L.ATHERTON, Theory of ferromagnetic hysteresis, Journal of
Magnetism and Magnetic Materials 61(1986) 48-60
you will find it diffrent from reference guide of pspice.

In original Jiles-Atherton model, there are five parameters with parameter
ALPHA.

I think the difference between original Jile-Atherton model and the one in
pspice is the way to describe anhysteric magnetization.

In pspice:
Man=Ms*H/(H+A)
where Ms and A are two of parameters

In original Jile-Atherton model, modified Langevin function is used
Man=Ms*(coth( (H+Alpha*M) / a )- a / (H+Alpha*M))

then there is Alpha parameters.

Actually I can get the parameters of some material in the latter form (with
alpha), but the difficulty is that I don't know how to reasonably convert
them into the form in pspice without alpha

thank you, Jim

Richard

[snip]

I'm not a magnetics expert, but can't you "fit" the PSpice equations
to the original Jiles-Atherton.

Or build your own subcircuit using behavioral modeling?

Last night, in a fit of exasperation with libraries in HSpice "crap"
format, I developed a resistor subcircuit that will take HSpice
parameters directly and then clean it up to PSpice RES model
equivalency, without me having to do the conversions manually every
damn time :-(

...Jim Thompson

 

[Charlie Edmondson]

Jim,

It's ture that the reference guide of pspice uses four parameters: Ms,A,C,K
as pspice does. And it says that this is based on Jiles-Atherton hysteresis
model.
But if you read the original paper from Jiles-Atherton, for example,
D.C.Jiles, D.L.ATHERTON, Theory of ferromagnetic hysteresis, Journal of
Magnetism and Magnetic Materials 61(1986) 48-60
you will find it diffrent from reference guide of pspice.

In original Jiles-Atherton model, there are five parameters with parameter
ALPHA.

I think the difference between original Jile-Atherton model and the one in
pspice is the way to describe anhysteric magnetization.

In pspice:
Man=Ms*H/(H+A)
where Ms and A are two of parameters

In original Jile-Atherton model, modified Langevin function is used
Man=Ms*(coth( (H+Alpha*M) / a )- a / (H+Alpha*M))

then there is Alpha parameters.

Actually I can get the parameters of some material in the latter form (with
alpha), but the difficulty is that I don't know how to reasonably convert
them into the form in pspice without alpha

thank you, Jim

Richard



04:48 >>>

Definitely not an expert in magnetics, that was Brian. I do know that
there were some definite deficiencies in the Jiles-Atherton models in
PSpice, which is why in the latest versions they now have a magnetics
model editor and some new models. For 9.2, you just have to work with
what you have, or go looking for some of the ABM models for magnetics
out there...

Charlie