The inner working of `Magnetic Resistance`

PLECS
, ,
1

Hello all,

I am interested in the inner workings of Magnetic Resistance component, trying to decide if I should use it in series or in parallel with Saturable Core. The help on Magnetic Resistance states:

Magnetic resistances (analogous to electrical resistors) are used to model frequency-depending losses in the magnetic circuit. They can be connected in series or in parallel to a permeance, depending on the nature of the specific loss.

If we connect it in parallel to a saturable branch they will “see” the same MMF and split the flux based on the time-dependent value of permeance. Is the permeance of magnetic resistance varied through time or does it remain constant? If constant, what is the expression for instantaneous power: F(t)^2/Rm? If we now drive the saturable core deep into saturation, the MMF will increase significantly which will, in turn, significantly increase losses on magnetic resistance. Which “nature of the specific loss" is this a good representation for?

The other option is to connect magnetic resistance in series to a saturable core. In this case they “see” the same rate of change of flux, and most of the “MMF drop” occurs across the saturable core with the balance across the magnetic resistance. I confirmed in simulations that this arrangement holds core losses proportional to the square of driving voltage, which matches the conventional representation of Xmag || Rmag of a transformer’s magnetizing branch, but I do not have a simple way to explain why this is the case.

Based on this work the first approximation for a transformer magnetizing branch Xmag || Rmag should be a series combination of magnetic resistance with a saturable core.

I recap my key questions for convenience:

  1. Is the permeance of Magnetic Resistance held constant or varies with F or Φ’?
  2. Is the instantaneous power calculated as F(t)^2/Rm or differently?
  3. When would it be appropriate to use a parallel combination of magnetic resistance and saturable core?

I acknowledge that sophisticated loss modeling as a function of frequency requires setting up a network of Pm+Rm branches. Another option is to follow the approach from the paper referenced in this post. The approach is physics-based and able to match experimental data nearly-perfectly, but it comes at the expense of significant simulation complexity.

Thanks in advance.

2

The Magnetic Resistance value is held constant and doesn’t vary with F or Φ’. I’d also note that Permeance is a property of the magnetic core elements, not the magnetic resistance which has different units (Amp-turn seconds / Wb for magnetic resistance vs. Wb / Amp-turn for permeance).

  • Is the instantaneous power calculated as F(t)^2/Rm or differently?

Yes that is the instantaneous power equation.

  • When would it be appropriate to use a parallel combination of magnetic resistance and saturable core?

I’m honestly not sure what physical loss mechanism it would accurately represent. The statement about “depending on the nature of the specific loss” may be overly generous; perhaps empirical curve fitting in certain cases. One should certainly be cautious using it in conjunction with a non-linear core element due to the behavior you identified.

Modeling the resistive loss in series with the core is the correct approach, as detailed in the reference you have already linked.