Electronics > Beginners

Litz wire vs common link wire.

(1/2) > >>

Many individualy isolated strands vs many non insulated strands. Due to surface effect in medium/high frequency coils Litz wire is preferred. But this insulation is really needed? Just use outer layer insulation for many non isolated wires. Did current flow can be pushing out from inside to outside strands?

There has been discussion about Litz wire here before.
Recently, see
or insert "Litz" into the search window on this sige.


--- Quote from: analityk on December 03, 2022, 10:02:49 pm ---But this insulation is really needed?

--- End quote ---

yes, it is needed and it is very important. Each thread have to be insulated and they all have to be twisted in a certain way. Regular stranded wire (where each thread have no isolation) will perform the same as a solid wire.

But note, such kind of wire is useful at high frequency AC current (about 100 kHz and above that). It is useless for DC.

I can't understand mechanisms why multiple strands without insulation work as solid wire. (have only surface area from diameter of wire). If all thread have this same diameter it will contact each other in six place (not entire surface) and current should flow whole surface of each thread. How it can escape to surface of threads on external layer? And why it can not be induced in internal threads?
Upper linked thread did not gave me answer.

Yes, exactly: there is some conductivity between strands, and therefore we can model the stranded conductor as an anisotropic conductor, which therefore has skin effect that depends upon direction.  The effective skin depth, in the, well, depthwise direction, will be deeper -- as it turns out, it gives about 20% better overall resistance in typical circumstances.

Mind, compare this to the almost unlimited advantage of insulated litz: with insulation between strands, the sideways conductivity is essentially zero and therefore the skin depth is essentially infinite.  The ratio of solid (or stranded uninsulated) to litz resistance, for identical (metal) cross section, can be tens, hundreds, thousands even!  The ~1.2x advantage of stranded is much less impressive in comparison.

The reason skin effect occurs in the first place, is that the outermost layer of a solid conductor carries some current, and therefore magnetic field; this field opposes the current flow, so the next layer down carries slightly less current, and so on and so forth.  This effect occurs regardless of how the current is carried, or through what resistances -- indeed the resistance doesn't need to be purely resistive at all, it just needs to have a loss component.  We can express the skin depth of a complex material like ferrite (which has ohmic as well as hysteresis losses, and has magnetic permeability), or even an optical material like glass or water (indeed, light only penetrates water to some 100s m).

Granted, it may not be obvious how to justify averaging over a very-obviously-discontinuous material like stranded wire.  Well, suffice it to say, the fields work out the same.  (A less satisfying answer is that I'm not sure I could explain in any satisfying (basic enough?) manner.  Ultimately these are complex problems -- EM fields are defined by boundary conditions, and, clearly, we have a LOT of boundaries inside the cross-section of a stranded cable.  It is far from trivial to solve this problem exactly.  But, well, one can make compile measurements, and build a model in this way [the anisotropy], and get good agreement -- predictive value.  Presumably, one could also prove such a model from the fields directly, but solving over that many boundary conditions -- even just to cross them all out, down to an average -- is not likely to be any easier.)



[0] Message Index

[#] Next page

There was an error while thanking
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod