Slimelite BBS

d Phi / dt

Back to AMIGA_VIDEO

From Ron Freimuth · 1996-04-03 08:51 · #76652 in AMIGA_VIDEO

GH> RF>    I think I brought this up to Peter, but I'm not sure.
  > RF>
  > RF>    We know  Integ E dot dl =  - d/t  Integ H dot dS
  > RF>             closed path              surface
  > RF>
  > RF>
  > RF>    How does the voltage develop along a wire that only encloses,
  > RF> but doesn't otherwise see the flux?

GH>Good question!  I checked my texts.  My undergrad text doesn't go
  >into too much detail on transformers, and my grad text hardly even
  >mentions them!  Not surprising, I guess.  Even Landau's book doesn't
  >have it, which is pretty surprising since he tends to give some theory

    Herman's reply seems to be on the right track.  I can see that if an
 Electric Field integrates to V volts per loop just along the outside of
 the area containing a changing flux, then, a larger loop must develop
 the same voltage.

    One doesn't need to consider a transformer, a single winding
 inductor will do.  Remember, V = L di/dt.  So, the induced emf must
 show up on each turn.

    Consider a constant voltage placed across a winding that loops
 around a circular core (say a toroid).  Current increases linearly, a
 back emf is developed which just balances the applied voltage.

    E must exist in cylinders around the core.  Each loop adds dPhi/dt
 of potential.

    E would decrease as one moves from the core, but E dot dL would
 remain constant.  Similar to H dot dL around a filament of current.  I
 guess if I can accept H dot dL I should accept E dot dL.

    This still doesn't completely satisfy me.  I guess the idea of a
 conductor 'cutting flux' tends to create confusion in this case.


  >     There's another old text that I picked up really cheap that I
  >liked because it seemed to be full of physical examples.  (And
  >underlining was no better in the days before highlighters; give
  >someone a highlighter or pencil, and EVERY FRIGGIN' SENTENCE is
  >important enough to highlight or underline!)

    I hate marked up books.

  >     Anyway, it showed the transformer you're talking about with
  >primary and secondary "leakage" flux looping from inside the wire
  >and core to the outside, and the "mutual flux" that stays entirely
  >inside the core.  And I don't understand it any more than you do.

    Well, I understand that model of a transformer.  I measured leakage
 inductance for the switching inductors I designed for power supplies.
 It had to be as small as possible.

    Lm = k Sqrt (Lp Ls)

    0 < k < 1   ; k is the coefficient of coupling

    If every line of flux generated by one winding passes through the
 other winding, k = 1.

    One can model the effect of imperfect coupling as that book shows.

    Note that if one shorts out the primary, the impedance measured at
 the secondary is  Lleak.

    The inductance of two coils in series, with mutual inductance, Lm
 between them is:

    Ls = L1 + L2 +- 2 Lm

    If k is unity and L1 = L2,

    Lm = L1 = L2

    Ls = 2L1 +- 2L1  =  4, 0

    Double the number of turns, L increases by 2^2.

    I had to know formulas like this for getting my FCC RadioTelephone
 License 35 years ago.

---
 * OLX 2.1 TD *  The fault lies not in our stars but in ourselves
--- FidoPCB v1.5 beta-'j'
 * Origin: Lunatic Fringe * Richardson,TX * 214-235-5288 * (1:124/2113)