High current high frequency transformers and inductors are wound with copper foil. Sometimes, foils also have to be paralleled to handle the rms current. Thus, the inner foil always carries the highest current since it has the smallest resistance and inductance. Transposing will balance the currents, but how to do it with foils.
Sujit,
thank you for this interesting point. If I understand you right, you have chosen a core with a certain winding breath that defines the width of the foil you use. Then, you defined the thickness of the foil, likely that there are only one or two standard-thicknesses. I guess that 100 µm is one of those standards. Now, you find that the cross-sectional area won't be able to handle the rms current.
You are right, transposing would be a way to equalize the electrical properies of each of multiple, insulated foils. The question is - do you really need insulated foils paralleled? In fact this inslulation won't bring you too much since the proximity effect will come into play. I suggest the alternative: paralleling without insulation. In that case, you can easily wind them, you safe the space of the otherwise necessary insulation layers. Hence, you get a high copper filling factor.
There may be different reasons why engineers use foil windings. But one of those reasons is the possibility of advanced cooling. The foils are not only a good electrical conductor, they may also be a material of choice to transport the heat out of the winding. Maybe it is possible in your design to allow the foil to be wider than the winding breath where it is possible. You could then form some fins that transfer the heat to ambient air.
What do you think?
Thanks Michael. Your assumptions are correct with regard to foil placement.
We use foils with the intention to control the ac resistance of the high current winding which would have otherwise risen due to skin effect at the high operating frequency (>10kHz). Since one foil cannot be too thick, multiple foils in parallel are sometimes needed. Mutual insulation is necessary to give the effect of 'Litz wire' so as to control the ac resistance. I appreciate your point of proximity effect but do not know which will give lesser total loss: with mutual insulation between paralleled foils or without.
If you wind with 2 concentric foils in parallel, the difference in total length between the 2 foils is minimal and circulating current will be negligible. I concur with Michael Meisser that insulation between the paralleled foils may not be necessary. But you can easily test it. First wind without insulation and check the performance. Then unwind and rewind with -very thin - insulation between the paralleled foils. Then test performance again and see if there is a noticeable difference. If there is, and you can explain and quantify it, it could make a nice paper for publication.
Dear Sujit, dear Jacob,
experimental evaluation seems to be a very comfortable solution since it provides you with real data. However, it may turn out that it is impossible to fit the experimental results with the assumptions and hard to identify the real physical soucres of either higher or lower losses. Hence, it may be advisable to run simulations of limited accuracy first in order to get an idea about the interplay of proximity and skin losses in your intended design. A 2D-Simulation of the core cross-section with the windings represented as circular stripes may be sufficient. I suggest COMSOL to build and simulate such an model.
Eddy-current losses can be calculated. Marian Kazimierczuk's book, High-Frequency Magnetics Components, Second Edition, (Wiley), ch. 5 has the design equations. Also, my book, Power Magnetics Design Optimization (www.innovatia.com) derives the optimum thickness for foil for minimum resistance.
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering