6
$\begingroup$

Is there a simple empiric way of determining the permeability (mhu) of a ferrite object (be it a ring, rod or any form)?

I've been thinking this for a rod: if I wind coils with different number of turns (same wire, just wind or unwind it) and measure the inductance, then given the number of turns, diameter of the rod and "height" of the coil I would be able to determine the permeability. I tried a couple of times but I'm getting inconsistent results. My LCR meter can measure only upwards of 0,01mH (or 10uH). Is the meter the culprit or is my method flawed?

$\endgroup$
5
$\begingroup$

you're on the right track to measure/calculate effective permeability.

However, you need to use an inductance meter that can measure up to 1 nH accurately or so. If you're trying to use one of those $25 LCR meters , it won't have the accuracy or precision you require, plus you won't be able to zero it. Another issue is that you'll have a wide variance in results due to the wire, winding, lengths, etc - see Silvio's comments at http://www.sklaic.info/forum/index.php?topic=175.0 .

The other method is to measure frequency resonance, by using the loop as the inductor in an LC circuit. Put a known capacitor in series with the loop, and then use a frequency generator and frequency counter to measure the resonance (peak) frequency of the LC circuit.

$\endgroup$
  • $\begingroup$ the resonance method pretty certainly is what those LCR meters do ;) $\endgroup$ – Marcus Müller Aug 1 '17 at 21:16
  • $\begingroup$ Weeeell you got me :). I'm using a 20€ (Europe yay!) meter that doubles up as transistor and capacitor tester. Nice all around kit but the accuracy is what it is. Therefore I didn't expect to get super good measurements but at least a ballpark. I will give a try to the resonance method; I read about it but never tried. $\endgroup$ – Luca Aug 2 '17 at 10:35
6
$\begingroup$

Following up on @dfannin's excellent answer:

The most intuitive way of dealing with this would be:

  1. wrap your ferrite rod in halfway stable paper or so, something that certainly doesn't have high $\mu_r$ (gut feeling: baking paper is nice as it is very "slippery" on flat surfaces)
  2. make as many turns as you want around that; you're building a coil now, with the ferrite core.
  3. Measure ferrite core coil's inductance by one way or another
  4. carefully slip the core out of the paper. Now you have exactly the same coil, but with an air core
  5. repeat measurement
  6. since air has $\mu_r\approx1$, the ratio between the ferrite core inductivity value and the air core inductivity value is the $\mu_r$ of the core.

Downsides:

  • Since ferrite's $\mu_r$ is going to be impressively high, you'll need a measurement method that can span easily 3 orders of magnitude
  • high-inductivity coil (which would allow measurements with your >1 mH LCR meter) might mean that even at relatively benign currents, you might be saturating the core and get it to nonlinearity (that's not all that likely, but don't rule it out before knowing how much power your meter uses)
  • might waste a lot of copper :)

Things that can generally go wrong:

  • core saturates
  • core too small for coil, not (nearly enough) all field gets concentrated into core, so that measurement isn't proportional to core's $\mu$
  • ohmic losses make measurement hard
$\endgroup$
  • 1
    $\begingroup$ Good idea. I'd also add that the permeability and loss tangent can be frequency dependent, so without knowing what frequency the LCR meter is using, it may not be accurate to translate the results to some other frequency. $\endgroup$ – Phil Frost - W8II Aug 1 '17 at 21:36
  • $\begingroup$ Nice idea with the ferrite/air swap but for that I will need a new meter :) $\endgroup$ – Luca Aug 2 '17 at 10:33
  • $\begingroup$ don't know – what is the maximum value - to - minimum - value relationship on that meter that you have now? $\endgroup$ – Marcus Müller Aug 2 '17 at 14:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.