Designing Magnetics

I was working recently on a design where I needed custom high-temperature magnetics for a new application. Nothing like what I needed was available off the shelf. My vendor had some different high-temperature magnetics, but they weren't in the correct size or shape. The task was to come up with a design that met my requirements using the factory's tooling and cores. Then came the hard part: coming up with a design that would meet the new operating inductance and frequency requirements on the old core.

Coming up with the design (wire size and turns) that would do this without sending the production department through a batch of trial-and-error tests was more difficult than it first seemed. I took a couple of approaches. First, I used the magnetics design software described below to take a pass at the problem. I then compared this answer to the results from a set of formulas from a paper from the University of Colorado's Department of Electrical, Computer, and Energy Engineering.

The software I used is Magnetics Designer from Intusoft. The company says the software is intended for designing single- and multi-layer inductors and transformers. It will also generate a Spice model. Intusoft goes into lots more detail on its site — core material, bobbin size, wire types, PC board trace dimensions, etc. This quote pulled from the site should provide a bit more background.

Typical design applications include high frequency switching regulator transformers and output chokes for off-line, full-wave and forward converters, 60 Hz single-phase line transformers, AC inductors, planar magnetics and 400 Hz aircraft transformers. Virtually any single phase, layer wound inductors or transformers that range from 10Hz to over 5MHz can be synthesized with Magnetics Designer.

It looks like the program gives you plenty of flexibility to change every design parameter you could conceive and a few more that you couldn't.

And it provides reasonable information for the simulation.

After some back and forth, the two answers were reconciled, and I completed a successful first pass prototype.

What is your experience with inductor design?

6 comments on “Designing Magnetics

  1. eafpres
    July 5, 2013

    Hi William–I'm guessing in your application the core was a metallic core material?  There are also a lot of wound inductors made on ferrite toroid cores.  A problem would be getting good properties and whether the simulation tool accurately accounts for the behavior of the core.  Can you expand a bit on what core you used and if it was easy to get the properties for the simulation?

    In a comapny I worked for previously (Laird) they make a broad line of soft ferrite products, including bare cores (toroids).  I believe they provide Spice models for all of them.  I'm not sure how broadly standard that practice is in the magentics industry.  What has your experience been there?

  2. D Feucht
    July 5, 2013


    The Caltech guys in CO put much emphasis on equating magnetic (core) and electrical (winding) losses because, according to the max power-xfr theorem, this transfers the most power. However, this is a broad optimum and for linear magnetics is at equal power loss in each. (Erickson & Macsimovic's book also work out the optimal values for nonlinear magnetics but the optimum is broad and if the magnetics becomes too nonlinear, there are other bigger problems.)

    However, nobody seems to have realized that to maximize magnetic (core) utilization and hence minimize core size is to operate the core, within acceptable design margins, near both saturation and max power loss. These are the conditions under which you can achieve the maximum power transfer from a core, and it depends on waveform parameters. I have worked all this out in detail at

    and it might be of interest in your efforts at magnetics design optimization. Also, see

    for a different look at the same optimization.

  3. Davidled
    July 6, 2013

    Simulation design provides the good guideline at the initial stage. I am wondering how the simulation model is fitted into high nonlinear hystersis as the polarity of particles is changing. I guess that there is some comparsion between actual data and simulated data in the magnetic system.

  4. BillWM
    July 6, 2013

    All — I used Laminated Iron Core, because Ferrite would disintigrate and lose it's properties at 300 deg F. (Design operating temperature for the PWM filter inductor)

  5. Brad Albing
    July 6, 2013

    Ah – so this is for a down hole drilling application then. From my previous experience, those apps are some of the most stringent.

  6. SunitaT
    July 30, 2013

    The requirement of optimal magnetics design is loss minimization. The magnetics components are most costly as they are not properly designed. If removal heat from the device is possible, lower losses will always advantage the design.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.