Inductance went down on homemade toroid - magnetized?

Discussion in 'General Electronics Chat' started by Roderick Young, Feb 24, 2015.

  1. Roderick Young

    Thread Starter Member

    Feb 22, 2015
    408
    168
    Hi, any inductor or switching power supply experts here?

    I salvaged a toroidal core from a transformer. The transformer had markings of 115/230 volts AC in, and various outputs totaling about 25 watts. The core is about 3" outer diameter, heavy like metal, painted solid green, and a magnet easily sticks to it.

    With 12 turns of #12 wire wound around the core, my inexpensive LCR meter read 234 uH. It was more than I would have liked, but was what it was. The turns must be a multiple of 12, for reasons that would take another thread to explain. I then put the inductor, L1, into the test circuit below:[​IMG]
    SOLAR_POSITIVE is a +24 volt supply relative to ground, with a large stiffening capacitor. You can ignore the unconnected parts of L1, just consider it to be a single inductor. Also disregard Q2 and C1, they not important to this question. The PUMP signal pulses high for 20 uS, once a second, causing Q1 to apply voltage across L1. With an inexpensive oscilloscope, I monitored the voltage on both sides of R1 with 10 meg probes.

    While the circuit was running, I saw 12 volts develop across R1 during each pulse, suggesting that 12 amps were flowing through L1, and that the inductance was really more like 30 to 50 uH. There was also a soft clicking sound at a 1 Hz frequency coming from the inductor. I turned off the lights in the room, but couldn't see any arcing. At one point, the clicking stopped, and when I looked on the oscilloscope, the current was only developing to about 2 amps during the pulse, implying the original 240 uH or so inductance. I let that run some more, and eventually, the clicking started again, and the 12 amps were back.

    At that point, I unsoldered the inductor and reversed the leads, so that current would flow in the other direction through it. When I looked at the scope again, I saw 0.3 volt developed across R1, implying that the inductance was now more than 1000 uH.

    Okay, here are the questions. Opinions and guesses are welcome. Thanks in advance.

    Q1: What material is the core likely to be made of? Could it just me a solid piece of iron or iron alloy?

    Q2: What happened to drive the inductance down? Did the toroid get magnetized, like core memory? If so, is that normal?

    Q3: I would love to use the lower inductance in my design. Is it ok to intentionally magnetize the inductor in this way (assuming that's what caused the change), and will there be more losses of any kind if I use an inductor operating in this way?

    Q4: What's my next move? Should I just buy a new toroidal core of known material? I dislike spending money as much as the next guy, but do have the resources to buy a brand new core if I have to.
     
  2. alfacliff

    Well-Known Member

    Dec 13, 2013
    2,449
    428
    the core was designed as an ac power transformer, your putting dc through it. look up magnetic saturation, a saturated core does reduce its inductance. and since the current is not reversing, there is nothing to unsaturate it.
     
  3. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,004
    1,525
    Roderick Young, cmartinez and Bernard like this.
  4. alfacliff

    Well-Known Member

    Dec 13, 2013
    2,449
    428
    a simple test to do, hook the transformer to your inductance meter and move a magnet near it. watch the inductance change as the core flux changes.
     
  5. DC_Kid

    Distinguished Member

    Feb 25, 2008
    638
    9
    what does C1 do exactly? C1 has it's neg side connected to +24v ??
     
  6. Roderick Young

    Thread Starter Member

    Feb 22, 2015
    408
    168
    Thanks, Alfacliff. After your tip, I looked up saturation in Wikipedia, and it does sound like the core got into a semi-permanent saturated state. This led me to other pages to learn about permeability, inductance, and core memory. Possibly this gives an answer to Q1. Given the low-current inductance, it appears that the permeability of the material is on the order of 2.13 x 10^-3 H/m, suggesting that the core might be made of of some kind of annealed steel, but can't be any kind of ferrite. Given that permeability, it appears that I was developing a field of 1.9 Teslas in the core at one point - possibly enough to leave a significant residual magnetism.

    I'm ok with operating with a saturated core, unless that means it's not going to be an efficient inductor, anymore. Hopefully someone can advise me on that, otherwise, I think I'll chicken out and purchase a core of known material.
     
  7. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,558
    2,527
    shortbus likes this.
  8. Roderick Young

    Thread Starter Member

    Feb 22, 2015
    408
    168
    Good eyes! That's actually intentional. This topology is called an inverting switching supply (not to be confused with an inverter). When Q1 turns on, current builds up in L1, like water rushing through a pipe. When Q1 turns off, that current has no where to go but through the diode at Q2, charging up capacitor C1. So the positive side of C1 is actually at a higher voltage than 24 volts, referenced to the ground in the picture. In fact, if the ground in the picture was also the ground reference for the output, this would be called a boost switching supply.

    This is, however, not intended as a switching supply. Some 98% of the energy is lost in the 1 ohm resistor. This is a test circuit for the inductor. As such, you can think of C1 as simply a means to prevent the voltage across L1 from spiking too high when Q1 turns off. It's like putting a diode in parallel with an inductor to prevent a kickback, the only difference being that there is a big capacitor in series with that diode in the circuit shown.
     
  9. Roderick Young

    Thread Starter Member

    Feb 22, 2015
    408
    168
    shortbus likes this.
  10. Roderick Young

    Thread Starter Member

    Feb 22, 2015
    408
    168
    Wow. I thought a toroid might have some immunity to an external magnetic field, but boy was I wrong. In putting the magnet in various places, I was able to influence the inductance by a factor of 6 or so! After seeing that, I decided not to mess with this unknown material toroid any longer. I ordered a T-300A-26 core from Amidon, along with some other cores to make the minimum order. If I have trouble with that one, I'll be back!
     
  11. alfacliff

    Well-Known Member

    Dec 13, 2013
    2,449
    428
    you just demonstrated what was known as a magnetic amplifyer. normally used to control ac current before more modern things came along.
     
    shortbus likes this.
Loading...