In another question someone asked about using a transformer (in current transformer mode) to detect current through a power line. The purpose of circuit was not to measure the current, but was only intended to detect current. The threshold current in the power was not defined, but let's assume the power line current is either 0 or some known value. If the current is present, the circuit should light an LED. Again, the LED is not specified, but lets assume 20-40mA.

Now, if a current transformer were to measure, say a 20A power line, and give an output of say 20mA, then the turns ratio would need to be somewhere around 1:1000. That is a rather tiresome ratio to wind by hand. It could be split into 2 transformers one 1:31 and the other 1:32, and that would be approximately 1:1000. However, it occurred to me that in an application where we only need to detect current, and not necessarily measure it at all accurately, one might exploit the saturation of a core to limit secondary current.

Out of curiosity, I began poking around to see how cores are spec'ed and to see if I could find an appropriate core to detect, say 20A, and generate 20-40mA at a max of 2-4V, with a much smaller turns ratio.

Of course, one of the things that needs to be known is the maximum ampere-turns for the core to saturate. Ok, choose a material, and the dimensions of the core, and go from there. However, I ran across this article which claims that saturation properties have considerable variation from manufacturer to manufacturer, and furthermore, may vary from one batch to another because manufacturers farm out their production to other manufacturers.

Now, the article in question suggests checking saturation current with an LRC meter together with a DC bias current. OK, good advice. But my question is, whether saturation ampere-turns really varies a great deal for identically dimensioned, and supposedly similar material cores, or whether the article was just giving friendly advice, and possibly also promoting the sales of LRC meters. :-). (A specific brand is mentioned). If anyone has experience using magnetic cores in saturation mode, I would appreciate knowing your experience in this area. Is it reasonable to expect that saturation ampere-turns to be within, say +/-40% of calculated values, based upon nominal saturation characteristics of magnetic material?

  • 1
    \$\begingroup\$ I would expect core saturation B field to vary, but not very much, less than your 40%, as it's a material property. The H field, the ampere turns, might vary more than that if the core is split, so has a gap that could reassembled with a little air in there, which can increase the reluctance by a large factor. A toroid is likely to be more stable. \$\endgroup\$ – Neil_UK Jan 4 at 18:54
  • \$\begingroup\$ Backing up for a moment, an off the shelf current sense xfmr such as this is quite straightforward for power line frequencies, and useful even for debugging things like stepper motor applications etc. (add resistive load, hook it right up to a scope, no fuss). I keep one around the lab. \$\endgroup\$ – Pete W Jan 4 at 19:00
  • \$\begingroup\$ I haven't ever actually done this, but to detect saturation, I believe one would add a high frequency test current. If the core is saturated, inductance drops, voltage at test frequency goes down (?). \$\endgroup\$ – Pete W Jan 4 at 19:05
  • \$\begingroup\$ Look into fluxgate current sensing. This is essentially how they operate \$\endgroup\$ – JonRB Jan 4 at 20:19

Your Answer

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

Browse other questions tagged or ask your own question.