0

I'm an amateur/beginner, (maybe asking somewhat advanced questions) following a rather unorthodox learning curve, so please provide context if you can. I'm curious about the retardation of the pick-up coil as a source of self (opposing) inductance, formulas would assume instant inductance, just dependent on the differential of flux over time, but haven't found anything about propagation time.

Joining 2 100mH inductors I feed the primary with a square signal and probe the signaled leg of the primary (CH1) and the secondary. With a 110MHz scope I can see a 100nS (pic shows 150, but with better zoom it's 100) delay from the beginning of the square signal to the beginning of the ringing on the secondary (that's about 30m at c, if you'll jump to "magnetic flux propagates at the speed of light").

I-Shaped 100mH Inductors used and disposition: enter image description here

Probing setup: enter image description here enter image description here

Perfectly aligned scope probes on same pin, connected to fn generator: enter image description here

Inductance retardation (actually 100nS) enter image description here

Primary pulse rise time (750nS) enter image description here

So, questions:

  • Are there formulas that account for causality & propagation more realistically?
  • And can we assume that after ~100nS the primary would start to oppose its own current flow? (maybe a bifilar coil would serve better, but in that case would the (dragging) dielectric field from the primary induce current on the secondary before magnetic coupling?)
  • What would the voltage rise look like without self-inductance? (would it last less than 750nS?)

Possibly related papers I can't read:

https://link.springer.com/article/10.3103/S1068371210090117 https://ieeexplore.ieee.org/document/8329715/figures#figures

juanmf
  • 113
  • 5
  • 2
    Show a schematic please and check both probes on the same signal to make sure one doesn't produce an unacceptable delay. – Andy aka Oct 12 '22 at 17:12
  • @Andyaka will do. Thanks for the suggestion. – juanmf Oct 12 '22 at 18:09
  • If the equivalent circuit is a transformer made of two inductors coupled together, then the transient response is given by the RLC equivalent circuit of both parts. Or if identical, the one part squared, if you will. The equivalent in turn is driven by the wire length on the winding, which will be considerable at a value of 100mH. – Tim Williams Oct 12 '22 at 18:50

1 Answers1

1

Really interesting questions and here are some observations that might be contributing to what you are seeing.

First of all, remember that the output voltage on your secondary is related to the rate of change of the magnetic field, not its amplitude. So, looking at your waveforms you can see that not only is the voltage low, the slope of the yellow line is smaller between your oscilloscope cursors, meaning that the primary inductor current is small during this period.

Secondly, each of your cores has a "magnetizing force," which is the magnetic field strength required to get the core magnetized. The primary coil current must be high enough to achieve magnetization of the primary core before the secondary core can be magnetized in its turn.

Depending on the primary wire size and your current, you might be experiencing the effect of conductor skin depth during the rise time. When the current in a conductor changes rapidly, the opposing field forces the current to the outside edge of the conductor, which increases the apparent instantaneous resistance. You could also be inducing eddy currents in the core itself.

I think you are on the right track. A bifilar wind does introduce some capacitance between primary and secondary, but each of the inductors already has a capacitance effect between its own adjacent winding layers. The bifilar wind will reduce the self inductance which should help the most from your present example. So look for a core with a high initial permeability or low hysteresis, and wind bifilar with a heavier gauge wire or litz wire to increase your conductor surface area. For high frequencies, the core resistivity must also be high to inhibit eddy currents.

Good luck!

John Birckhead
  • 10,524
  • 1
  • 11
  • 27
  • Much appreciated, just opened many new dimensions to pay attention to. Waiting for a 10 gauge speaker wire delivery. The inductors I used are I-shape from assortment kit, taped head to head. (Will add image). – juanmf Oct 12 '22 at 19:13
  • So… just got a new signal generator (rise time 10ns) up to 25V. And saw a delay of about 5-10 ns if not less on induced voltage. I guess the current pick up/magnetization threshold was at play as you said. Still I’d expect a delay. But seems out of reach to my devices. – juanmf Oct 14 '22 at 20:19