Measue inductance of coils / marking?

Discussion in 'General Electronics Chat' started by spinnaker, Nov 28, 2012.

  1. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,887
    1,016
    I did some searching and found a few methods of measuring inductance of a coil using a function generator and a scope.

    I have a scope but I do not have a function generator. I do have pics on hand. Can this be done with a pic? If so how?

    I assume this would need a fairly stable clock? I do not have any high frequency crystals on hand. I only have 32.768khz xtals but I do have some 4mhz resonators. Would they do?

    Why aren't a lot of coils marked? I have a bag of 220uh coils, none of them seem to be marked. I have one on a board that I need to know the inductance and it is not marked. I have noticed this on a number of coils. What's the deal?
     
  2. DerStrom8

    Well-Known Member

    Feb 20, 2011
    2,428
    1,328
    There are so many different ways of marking inductors that with some you might not even realize they're marked. What type of inductor is it?

    In order to properly test an inductor, you'll need a clean sine wave. While I believe this can be done with a PIC which has a built-in DAC, it is a fairly difficult process. If you're up for a bunch of programming, i believe a lot of the dspics have built-in DACs.

    I did just find this project (scroll down to "Creating a Sine Wave with a PIC16F84") which uses the PIC and a resistor ladder network, but I'm not sure it'll be able to create a clean enough wave to accurately measure inductance. Might be worth looking into though.

    Hope this helps!
    Regards,
    Matt
     
  3. zero_coke

    Active Member

    Apr 22, 2009
    294
    1
    If you wanna measure inductance of a coil, do this:

    1) Connect your signal generator directly to your coil. Take note of your signal generator's internal impedance Rs (which should be noted on the signal generator somewhere)

    2) Note that maximum power transfer theorem states that the two impedances have equal magnitude when the voltage across the inductor is V_input / root(2). Try to achieve this voltage by changing the frequency dial on the generator since the inductor's impedance is a function of frequency.

    3) When you reach a frequency that makes the voltage across the inductor V_input / root(2), write the frequency down, and just apply this equality:

    2*pi*f*L = Rs

    Say your signal generator's internal impedance is 50 Ohms, then

    2*pi*f*L = 50

    Therefore, L = 50 / (2*pi*f) where f is the frequency seen on the signal generator when the voltage across the inductor observed is V_input / root(2).
     
  4. DerStrom8

    Well-Known Member

    Feb 20, 2011
    2,428
    1,328
    That's the whole problem--he doesn't have a signal (function) generator :p
     
  5. zero_coke

    Active Member

    Apr 22, 2009
    294
    1
    Oh!, I skipped directly to the solution then haha!

    Okay, I guess get an LC meter or do what DerStrom8 said I guess :)

    Or you can estimate their values with Wheeler's Formula if they are large copper coils:
     
  6. DerStrom8

    Well-Known Member

    Feb 20, 2011
    2,428
    1,328
    Haha, I almost did the same thing :D
     
    zero_coke likes this.
  7. Brownout

    Well-Known Member

    Jan 10, 2012
    2,375
    998
    All you really need is a square wave generator. The frequency isn't really critical, but the edge should be fast. Connect your inductor in series with a resistor of known value. You should experiment with different values of resistors, but they should be much larger than the output impeadance of your generator. Connect the circuit to your generator and measure voltage across the resistor, and plug into:

    L= tR/(ln(1/(1-Vr/Vg)))

    where R is the series connected resistor, Vr is the voltage measured across the series resistor, Vg is the voltage of the generator.

    t is the elapsed time from the generator pulse edge, to when Vr is measured.

    If you pick Vr/Vg to be 1/2, then the equation becomes:

    L=1.44*t*R.

    You can make a square wave gen from your PIC, but you might need a fast buffer to get the edge rate up. A fast schmitt trigger would be good here.
     
    Last edited: Nov 29, 2012
  8. takao21203

    Distinguished Member

    Apr 28, 2012
    3,577
    463
    You can actually use the coil itself instead of a crystal.
    No capacitors are required. This works downto about 12uH.
    HS mode must be used, while larger coils need XT mode (330uH, 1000uH etc.)

    Then you can use the 32KHz crystal as timebase.

    If you have a reference coil, you can easily get some idea about the actual value.
     
  9. nigelwright7557

    Senior Member

    May 10, 2008
    487
    71
    You can pick up a cheap LC meter for £15-£20.
     
  10. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,887
    1,016

    Can you give more details on how this is done?
     
  11. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,887
    1,016
    This is the coil


    [​IMG]


    The only marking on it is a W with a circle around it but I assume that is the manufacturer's mark.

    It is in circuit with a pt4115, supplying <500ma. According to the datasheet this should be 47-100uH coil.
     
  12. vk6zgo

    Active Member

    Jul 21, 2012
    677
    85
    Another,but "very rough" way ,if you have a DSO or an analog 'scope with delayed timebase,is to tack a known capacitor across the inductor,& connect it between the calibrate output of the 'scope & its vertical amp input.

    The ringing which appears on the top of the displayed signal will be at the resonant frequency of the LC combination.
    Knowing f & C you can determine L.

    To double check,try another known cap,calculate the resonant frequency using the value you just found for the L value.
    If the new ripple frequency agrees with your calculation,you have your value for L.

    Yet another way is to make a one transistor oscillator.
    Knowing the value of C,& measuring the frequency,you can determine L.
     
  13. takao21203

    Distinguished Member

    Apr 28, 2012
    3,577
    463
    You simply connect the coil to OSC1 and OSC1, where you normally would connect a crystal.

    I had 1000uH coils for this which needed no capacitor. Some larger coils may require one capacitor for startup (a few nF). Smaller coils need none.

    Then you also use the 32 KHz crystal for the secondary oscillator. By this you can use a timebase to count TIMER0 interrupts.

    You can use a small speaker to produce a tone relating to the frequency. You can blink LEDs, or you can drive a 7-seg display, LCD, etc.

    There might be the case when you don't need so much precision.

    Remember smaller coils must use HS mode, and larger coils must use XT mode.

    You do not neccessarily need any calculations, or absolute inductance calibration! If you like, as I pointed out, calibrate with a known inductor. It is however not guaranteed that the frequency chart will be linear. The scale for HS and XT mode also might be different.

    For sure this solution is useful to distinguish a 22uH or a 100uH coil, simply by experience, and the blink rate. If you want, display the counted values on a small LCD!
     
  14. timescope

    Member

    Dec 14, 2011
    298
    44
    I had a similar problem some months ago when I was designing a replacement circuit for a 24v to 5v @ 3A switching regulator module. I built this Inductance meter circuit, which converts the value of inductance to voltage, to sort out all my inductors (image attached).

    An improved version is Simple inductance meter.

    Inductance meter with 1nH resolution is more complex and will offer better performance.

    The Elektor rlc meter uses the sound card of a computer and a few external components to measure resistance, inductance and capacitance. You will have to register and subscribe to their mailing list to get the credits to download the documentation and software.

    Timescope
     
  15. jt6245

    Member

    Feb 18, 2011
    21
    3
  16. #12

    Expert

    Nov 30, 2010
    16,338
    6,823
    My blog #4 shows how to do this with a function generator but you can use the same circuit to do a one shot measurement. My circuit repeatedly loads current into the coil, shuts it off, catches the resultant energy in a capacitor, then dumps the capacitor. You could use this arrangement to do a one shot load and catch, then measure the result with a peak reading DC meter.
     
  17. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    The inductor on your board could be anywhere from 22uH to 220uH depending on the core material and wire gauge used. I have some 100uH that are about that size which are rated for 1.8A in a SMPS.

    I used the scope + generator method for finding inductance, but once I got an LCRZ meter, I only use the scope method in a pinch. The biggest benefit of an LCRZ meter is that you also get the Q factor, as well as variable frequency measurements.

    This is very useful for capacitors as well, The "Q" for Capacitors is Dissipation Factor, which is also a method of finding the ESR of a cap at various frequencies. For resistors, any stray inductance or capacitance is easily found as well for when working with very high frequency circuits.

    A single or dual frequency LC meter can be found for under $100, and will provide results at least as accurate as the scope method, in much less time.
     
  18. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,887
    1,016
    What is the aprox. size of the inductors on those 34063 based USB car chargers. They seem to be tiny.
     
  19. takao21203

    Distinguished Member

    Apr 28, 2012
    3,577
    463
    by the way most coils work somehow. Even 1n4148 diodes are good.
     
  20. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,887
    1,016
    What does this mean?
     
Loading...