Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Twisting/winding closed-loop wire essemblies - Litz -

  1. Jul 11, 2007 #1
    Consider a single length of 30 gage magnetic wire that is soldered to form a closed loop. Pinch the wire at either side of the loop between your thumb and index finger using both hands pull the wire away from each other, bringing the top and bottom, of the loop, closer to each other. Temporarily secure the left hand side of the wire so that you are able to twist the wire around itself. You now should have an assembly that looks very much like a common rope but in this case, each end of the rope is attached to each other forming the closed loop.
    Imagine an independent flux field that is now swiped buy the wire assembly so that the swipe is perpendicular to the twist. This would induce a maximum EMF in the sections of the 'rope' that are perpendicular and induce a minimal EMF in the sections of the rope that run in parallel with the swiping action. Since the wire is a closed loop I would expect that the EMF would be equal in strength regardless of the perpendicular or parallel orientation to the source flux field. Or, does it induce an EMF in half of the wire(twisted closed loop) causing a potential difference in only one half of the loop? What is it about litz wire that aides in maximizing an EMF? Does anyone have a link to help me learn how to make my own litz wire? Or, how does twisting wire aide in electromagnetic induction.
     
  2. jcsd
  3. Jul 11, 2007 #2

    berkeman

    User Avatar

    Staff: Mentor

    One key reason for twisting a pair of wires is to be able to reject external B-field interference. Since the orientation of the loops with respect to the external B-field flips for every other twist, the induced EMF cancels out for the full length of the twisted pair.

    Litz wire minimizes the series inductance of the wire by splitting the wire up into strands, maximizing the surface area for the current to flow. I haven't used Litz wire, so that's about all I know about it. Google or wikipedia should give you some links to better sources of info. I would guess that it's pretty hard to make without the proper weaving machine, though.

    http://en.wikipedia.org/wiki/Litz_wire
     
  4. Jul 13, 2007 #3
    Interesting, to say the least. If I understand you correctly. A straight run of wire would then be much more susceptible to B-field interferences. And two magnetic fields that are perpendicular to each others' spin domain will cancel each other out as well as, no EMF is realized. Can you enlighten me in regards to? If we take two, individual twisted/looped wires and lay them next to each other, in very close proximity, insulated, instead of twisting them around each other. Would this cancel out as well? As I see it, and I may stand corrected here, The two wires would be running parallel, relative to each other and be configured in a helix configuration. Much like two perhaps three springs coiled within each other. The B-field you mention may still be a factor. In order for me to close the loop, to achieve an EMF. I would need to engineer a return path. hummm? Local bus is the first thing that comes to mind......

    Thanks berkeman, You and xez are always good to keep the brain challenged.

    http://en.wikipedia.org/wiki/Litz_wire[/QUOTE]

    Thanks for the link, that's the first place I usually go to start my research.
    I'll need to look up 'series inductance', but am I to understand that electricity travels on the surface of its' medium, hence the thinner wire. So I would infer that Litz loops' minimize perpendicular electron spin, which we now understand to cancel EMF. This would also say that the closer the electron spin is in a parallel orientation relative to each other. the electrons would tend to skip over to the adjacent wire. Not good I would think. hummm?

    Litz wire weaving machine, I have something ridged up, a small scale weaver for textiles, but it fell short.

    Thanks again.
    robin07
     
  5. Jul 13, 2007 #4

    berkeman

    User Avatar

    Staff: Mentor

    The point you make about parallel twisted pairs having mutual magnetic coupling is an important one. This effect is why multi-pair twisted pair cable (like Cat-5 for Ethernet, etc.) uses a slightly differen't "lay length" (the distance between twists) for each of the twisted pairs inside the same outer sheath. Since they each have a different lay length, the twists won't line up in synch for very much distance, and the net effect is very little signal cross-coupling between the wire pairs. Reducing cross-talk is extremely important for maintaining signal integrity. You can look up terms like NEXT (near-end cross-talk) for more info on this subject.

    And as for the Litz wire, yes, at higher frequencies, most of the current will flow on the outer skin of a conductor (look up terms like "skin effect" and "skin depth" for currents). So more surface area will usually offer a lower series inductance for signals.
     
  6. Jul 13, 2007 #5
    http://thayer.dartmouth.edu/inductor/papers/stranded.pdf

    Great article, check it out, if you get a chance, and thanks for a direction of research. Sometimes all you need is the correct terminology to find what your looking for.
     
  7. Jul 13, 2007 #6

    berkeman

    User Avatar

    Staff: Mentor

    Wow, great paper, Robin. Thanks for the find.
     
  8. Jul 13, 2007 #7
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Twisting/winding closed-loop wire essemblies - Litz -
  1. Shielded twisted wire (Replies: 2)

Loading...