Group Actions
Engagement
  • Author Author: Jan Cumps
  • Date Created: Date Created
  • Views 6341 views
  • Likes 12 likes
  • Comments 19 comments
Related
Recommended

Measure Unknown Inductor Value with Function Generator and Oscilloscope

Jan Cumps

There are affordable switch mode DC converters available.

One that you see often is this Benq 12 V to 5V converter.

image

I can find info on all components, except the inductor. And I don't have an inductance meter.

There are several other ways to measure inductance. In this exercise I'm using the method with triangle waveform as described on meettechniek.info.

 

Ingredients:

  • the inductor, desoldered
  • a function generator with triangle waveform option
  • an oscilloscope

 

The setup is straightforward.

Set the generator to 50 Ohm and tune it in to a reasonable frequency. I set it at 10 kHz (my inductor is made for the kHz range).

Put the output over the inductor, and the oscilloscope probes.

Then push the voltage of the generator as high as possible, while the waveform doesn't start to show saturation. See the yellow waveform below for an impression.

Then change the frequency so that the voltage drops as low as your oscilloscope can deal with. In my case, I turned it down to 5 kHz.

 

image

 

Then start measuring.

VR: delta of the ramp. I used the delta of horizontal cursors for that: 32 mV

VL: between 0 and halfway the ramp: 95.5 mV.

t: time of the ramp. I used the delta of the vertical cursor for that (and it's half the time of one cycle too): 100 µs.

Ri: impedance of the generator. We had set it to 50 Ohm.

Vg: peak-to-peak output of the generator. Unplug the inductor and measure it with the oscilloscope: 8.16 V.

(I also captured it with the white REF signal but note that it's not at the right horizontal location. The peaks should be at the same spot as the peaks in the yellow trace)

 

... and Calculate the Inductance

(the formula images are from the meettechniek.info site)

image

= 8.16 V / 50 Ohm = 0.16 A

 

image

= 0.095 V * (100 µs / 0.16 A)

= 0.0000594 H = 59 µH

In the common inductor tolerance series, that would most likely be a 68 µH one.

 

 

Validate with the Switch Mode IC specs

Let's check this off with the specs of the LM2576 5V IC that's used in this circuit.

The PCB has an input spec of 12 V. It claims to switch up to 3 A (I found one website that indicated this - most don't claim current).

On the Inductor Select graph, this brings us here:

image

image source: TI LM2576 datasheet

 

That is a match: 68 µH. The expected inductor is used on the PCB.

And the triangle waveform method turned out to be a valid way to get at the inductance of a coil.

 

The whole circuit is exactly the typical application schematic from TI, adjusted for the 12 V input and 3 A output.

image

image source: TI LM2576 datasheet

 

Input and output caps are 220 µF.

The Schottky diode is a On Semiconductor SS34.

 

Link to other magnetics related posts.

Parents Comment
  • in reply to Jan Cumps

    Hi Jan,

     

    What's the dimensions of the core? From that, and the number of turns, which looks about 30 in your photo, it is maybe possible to guesstimate the core type too : )

    From guessing from the photo approx dimensions, it may be a 'type 61' ferrite material, which gets into the mid tens of uH ballpark with 30-ish turns at about that size.

Children
  • in reply to shabaz

    You are right image.

    I have a known inductor from Velleman, the 50L1, a 50 µH at 1 A, 64 µH at 0 A inductor that has practically the same kernel dimensions, and 41 turns.

    image

     

    imageimage

     

    source: velleman

     

    The one one I reviewed in this post has 32 turns.

  • in reply to Jan Cumps

    I see.. perhaps FT-50-61 might be the core, which is 12.7mm dia, and 4.8mm high, which might be similar dimensions to the one on the board before the wire was wound on. It should be 68uH with 31-32 turns. There are possibly Farnell cores from a different manufacturer with similar core material properties to type 61, in that size.

  • in reply to shabaz

      wrote:

    ....

    From guessing from the photo approx dimensions, it may be a 'type 61' ferrite material, which gets into the mid tens of uH ballpark with 30-ish turns at about that size.

    image

     

    You were spot on! (also showing off my brand new LRC meter that arrived today).

    The triangle wave method was fairly close too. 8 µH off...

  • in reply to Jan Cumps

    Hi Jan,

     

    That looks like a decent LCR meter, it has separated source and sense connections which is really cool too.

    Does the Cal button request you to open and short the connections? If so, maybe it's possible to build calibration and test fixtures for this LCR meter (or maybe some are supplied), especially for SMD (but would also suit through-hole).

    As a suggestion, if it fits, they could be PCBs with tongues which insert into the device. You could have three (or more). One PCB would allow you to solder on a component (like an 0603 or 0805 location, but with no solder mask over a larger area so that it would be compatible with larger components even through-hole), and another PCB would be identical but kept as a open circuit with nothing ever soldered on it, and another with the same thing, but with a wire or zero-ohm SMD resistor permanently soldered across it. In other words, all the same copper layouts for all three boards, so that inductance and capacitance is about the same for all. Or it could be one board that you rotate around the sides of, to do the different stages. I don't know if that makes sense or if it's even practical.. I'm mentally picturing a PCB that looks like a jigsaw piece from all four sides.

     

    I did a similar thing for my VNA, however that's not got separate source and sense. In this case since it's RF related I had some constraints (and I copied a TI antenna development board and changed the gerber files to suit this different purpose). Here I've accidentally torn off part of the copper at the DUT portion, due to careless heavy use : (  but it's still usable, just not as accurate but 'good enough' for my purposes. I will have to make another, and also add another DUT portion for comparing two components more easily. So that would be another suggestion, to use all four sides, to have more DUT sections : )

    image

    EDIT: It could also be possible to have a breakout board, for converting to 4-wire clips : ) Maybe that's overkill though, it depends on what ranges are supported and if it would benefit from that.

  • in reply to shabaz

    Does the Cal button request you to open and short the connections? If so, maybe it's possible to build calibration and test fixtures for this LCR meter (or maybe some are supplied), especially for SMD (but would also suit through-hole).

     

    Yes yes and yes image

     

    image