## Question - Find inductance of unknown coil

36 posts / 0 new
Author
Message

Hi all, I feel really stupid for asking this, but something is whispering in my ear that I'm wrong, so I would appreciate a second (third, fourth, etc...) opinion.

I'll make a coil (air core inductor) from a coil winding formula and would like to measure it's inductance.

My plan is to place the unknown coil in series with a known, non-inductive (carbon comp) resistor, then apply a sinusoidal signal of known F from my function generator. Then measure the signal amplitude across the coil and resistor (oscilloscope) and derive the "resistance" (inductive reactance) of the coil, finally plugging the value into "L=Xl/2PIF" and get the inductance. The inductance will be in the tens of uH, so the scope probe and test leads will not influence the result (much).

It seems right to me, but something says I'm missing something. Comments please?

Thanks!

Gentlemen may prefer Blondes, but Real Men prefer Redheads!

Last Edited: Thu. May 13, 2021 - 11:04 AM

Neither inductor, or resistor are ideal.
Resistor may have inductance, yet the carbon one is right choice, 50 ohm, though- for that L expected.
Inductor may have both resistance and capacitance. To prevent capacitance influence, do use the frequency just big enough to measure.

Your formula is not applicable (and the scope should have 2 beams, guess).
Regards
Milan

grohote wrote:

Neither inductor, or resistor are ideal.
Resistor may have inductance, yet the carbon one is right choice, 50 ohm, though- for that L expected.
Inductor may have both resistance and capacitance. To prevent capacitance influence, do use the frequency just big enough to measure.

Your formula is not applicable (and the scope should have 2 beams, guess).
Regards
Milan

I know the R and L are not ideal, I just want to see if, for example, I wind what I think should be a 33 uH coil, it tests around 33 uH. Might end up being 30uH or 40uH, I just want to know if it's "in the ballpark". If it tests to be, say 3uH, then I know I screwed up. That's only as good as it needs to be.

BTW, the scope is dual trace.

However, the formula is just the standard inductive reactance formula re-arranged to solve for L.  I'm using L (in Henrys) equals inductive reactance (derived from the known R, the measured E across the R, giving me the current I through the series circuit and the reactance from Xl = Measured E of coil * the known current I. Then, divided by 2 * PI * Freq. Have I done this part wrong? If so, what should it be?

Gentlemen may prefer Blondes, but Real Men prefer Redheads!

My multimeter has an inductance test capability. Maybe check yours for same? Or see if a friend has one that has it as well?

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

"The critical shortage here is not stuff, but time." - Johan Ekdahl

"Step N is required before you can do step N+1!" - ka7ehk

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

And of course there is always the cheap and handy eBay offering.

https://www.ebay.com.au/itm/1337...

Yes Ozzie dollars.

Ross McKenzie, Melbourne Australia

Krupski , how yould know the known current I. You may also try phase measuring approach.

valusoft Common, Ross, did you ever try this one, it is dumb on uH range, practically invalid under 100uH.

grohote wrote:
valusoft Common, Ross, did you ever try this one, it is dumb on uH range, practically invalid under 100uH.

Specs say it can go from 0.01mH to 20H.  While I would say that there might be a large tolerance, theres no reason that you cannot get a decent idea of the coils value under test from this device.  I have purchased many devices from ebay whose price/performance could be called into question, but the device worked just fine and whose accuracy was good enough for what I/we needed.

JIm

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

"The critical shortage here is not stuff, but time." - Johan Ekdahl

"Step N is required before you can do step N+1!" - ka7ehk

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

If you have an accurate (known) C, you can make an RLC & find the resonant freq (suddenly get a sharp peak at the right freq).

Or if you apply a pulse of voltage to the inductor, the slope of the linear current ramp (meaning meas current with current probe, or near zero ohms) )gives the inductance  V=L(deltI/deltT),   L=VdeltT/DeltI

here is a textronix way

https://www.tek.com/document/app...

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Thu. May 13, 2021 - 04:35 PM

jgmdesign , you are right, and I apologize to the Ross, too. Just measured two chokes with result:

1.4ohm L=.03H  (27uH)
1.8ohm L=.05H  (47uH)

Sorry to blame innocent EB product, really.

I'd be tempted to do it as a bridge. Leg 1 has two equal value resistors and goes to scope channel A. Leg 2 has your inductor and a resistor equal in value to what you expect the inductor to be at some frequency, and goes to scope channel B. Use the scope to invert one channel and sum them. Adjust the frequency of the generator for a null on the scope, upping the channel sensitivity as needed. Note the frequency and derive L knowing what ZL is (the same as the resistor in that leg).

This method eliminates trying to make accurate level measurements on a scope, assuming your function gen has a digital readout.

#1 Hardware Problem? https://www.avrfreaks.net/forum/...

#3 All grounds are not created equal

#4 Have you proved your chip is running at xxMHz?

#5 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand."

Less expensive than a NanoVNA.

Start measurement | NanoVNA

[1/4 page]

• REACTANCE : The reactance component of the measured impedance

"Dare to be naïve." - Buckminster Fuller

Hi Krupski, Sorry for typos would, and H instead of mH, I should be less hurry.
Googled and found How To Measure Inductance By Tim Daycounter: do connect R and L in serial. Just adjust the signal-generator amplitude that you can use this formula:

When VL/VIn=1/2 then L=R/(2*pi*f*sqrt(3))

grohote wrote:
I should be less hurry.

+1000000000

We all should take a step back once in a while.

Jim

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

"The critical shortage here is not stuff, but time." - Johan Ekdahl

"Step N is required before you can do step N+1!" - ka7ehk

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

... which, with 50 ohm and 100kHz sine-wave will measure 46uH.

Think, that instead of amplitude only, you may vary the frequency, too.

When VL/VIn=1/2

I wonder why the H they don't simply say:  When Vl  = Vin/2  or maybe say: when the inductor voltage equals half of Vin   In any case, seems like a doable approach

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Thu. May 13, 2021 - 09:16 PM

Initial formula is unnecessary complicated. Way of saying, math, semantics... avrcandies  I agree, your objection is reasonable, formula clean and definitely better.

Let wait the result, connection and measure may not be difficult.

I think I just (re)posted this image a week or two ago...

This was from one of my son's classes a few years ago.

They had to make their own coil of whatever-the-value-was.

We fed the RC circuit with a (slow) square wave, (Baseline offset to 0V, so not a bi-polar signal).

Then we measured the RC time constant, Tau.

When the voltage drops to 36.8% of its original value, you measure the time for that to happen.

L = Tau * R

Easy.

The scope shows a few of our his coils, and a line across the scope at 36.8 % of the Vmax.

(I think that line was the second channel trace, IIRC)

That made it easy to read the time, and calculate the L.

Photos, always a challenge these days, will follow.

JC

Last Edited: Thu. May 13, 2021 - 10:47 PM

Hi DocJC, your inductors are massive, but imagine 30uH: it may be just 0.5mm wire, 30 turns, 3cm diameter. Output voltage will drop quickly on any R.

Personally, for that L-range,  I do like more what avrcandies in #8 suggest. Add known capacitor in parallel, square wave of SG, and use scope to see the resonance frequency. Or, use signal generator sine-waye and swap to find resonance peak /and, with some skills, also Q-factor/.

Brian Fairchild wrote:

I'd be tempted to do it as a bridge.

Interestingly I was looking for LC oscillator circuits and (on an 90s style Russian site) found one that can be viewed as an "auto-tuned" bridge:

I redrawn it to make it clearer that it is a bridge

The way I interpret this, the opamp tries to balance the bridge by making the circuit oscillate at such a frequency that the LC circuit has 10k impedance to maintain the impedance ratio on both sides (this will happen close to resonance, where impedance is theoretically infinite).

Anyway, if you know C and measure the frequency it should be easy to calculate L.

I haven't tested this circuit (yet) but I want to - more work for the AVR-DB opamps

But of course since the OP has a signal generator, a frequency sweep should allow the resonance frequency to be found as previously noted.

Double supply... can it be any simpler. There are schematics of "LC meter" similar to those, running on 5V with LM311 if I remember well.

Here is a follow on to El Tangas good idea...Note the 741 is dirt slow & should be upgraded if you expect higher freqs  (low inductance).

Of course getting a high-accuracy C value, might not be a "junk drawer" component.

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Fri. May 14, 2021 - 07:01 PM

Interesting, that limiting network. I think I'll experiment a bit this weekend

Like the OP, I also have some inductors to measure and no specialized equipment for that.

I even found some fast opamps in my collection, probably overkill  http://www.3peakic.com.cn/En/product/inproduct/catid/149/id/283.html

The diodes on the limiting network probably also need to be fast to keep up an not cause frequency distortions.

I'll just use the AVR-DB internal opamp (no speed demon but still 20x faster than the 741...) and try to capture the resonance frequency using an internal timer.

avrcandies, this circuit also requires double power supply, 2x6 or more V. Agree with El Tangas, it is interesting one. The diodes- do not mind, any two 1N4148 can do it. Cap 1 is usually 330-1000 pF, try the most accurate one. Pot of 100K, resistors of 100K, do put cap blocking in + and - pins.

In days of old, one trick to linearise the sine wave output (if you see what I mean) was to include an incandescent bulb in the feedback loop (don't ask me for circuits, it's been fifty years!). As the output increased in value, the non-linear resistance of the bulb would increase with its increasing temperature and reduce the gain.

Neil

Remember those articles, the lamp was tiny 3V 50mA type. Not remember Nyquist, that was for the HF oscillations.

Often used in Wein bridge oscillators.

Ross McKenzie, Melbourne Australia

And, the real reason why "to linearize" may be that such circuits, when used as signal generator does needs to produce constant amplitude at output. For, example, make cap variable, then, normally the amplitude will not be linear from min to max, unless linearized.

don't ask me for circuits, it's been fifty years!

Ask Hewlett or Packard...one of them wrote their thesis around the method and, I believe, they started the whole company with that innovation (not sure if they were the first to use this).

ah..here we go..

After applying his restless intellect to the project, Bill combined these principles and came up with something that was elegant, simple and new. He inserted a light bulb in a Wien bridge oscillator circuit. With the introduction of an incandescent lamp into the design, Bill solved the problem of regulating output of the circuit without causing distortion. He had simplified the circuit and improved its performance. It was a good start on the road to perfecting tools engineers could use to make precisely accurate measurements. Today, the original prototype resistance-capacity oscillator Bill built as a graduate student resides in the HP Archives.

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

not sure if they were the first to use this

Not the first, but their signal generator was the perfect one, and, as you said the HP started with this product.

The moral is: all we do should be perfect, as much as possible.

My multimeter does C, but not L. :(

Gentlemen may prefer Blondes, but Real Men prefer Redheads!

And your SG, does it have 100kHz, do you progressing- tell us what you did last days.

Krupski wrote:

My multimeter does C, but not L. :(

Does it measure L as -C? <grin, duck, run...>

Neil

Does it measure L as -C? <grin, duck, run...>

Looks like it is time to throw away the inductor forget about it!! Yeah trashcan....sit down, you will be building a gyrator inductor alias circuit today.

No turning back on ELI the ICE man!!

Well, since you are making a LOW inductance part, this is probably not for you (Gyrators are aimed at high inductance replacement), but fun to consider!

https://www.epanorama.net/docume...

https://www.hewlettpackardhistor...

Here is  Bill Hewlett's letter to Dave Packard talking about producing their first oscillator

from a very interesting site

https://www.hewlettpackardhistor...

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Sat. May 15, 2021 - 06:44 PM

You means  the op-amp based inductor approximation, heard about, yet the application is extremely rare. Simulating large C, however, is quite possible.

Observation:

Subject says "unknown inductor". Yet the OP says:

Krupski wrote:
from a coil winding formula

If that is so, then it is not unknown. Maybe not known to 0.001% but none of the methods suggested will do that, either. The coil winding formulae are quite good, actually.

Also, an op-amp oscillator is mentioned. Maybe a few MHz with a reasonably good op-amp, maybe a bit more for a good high bandwidth amp. The limiting method is critical for distortion and amplitude stability. You may find significant variation over frequency as the op amp gain varies. The Wien Bridge has been the subject of a huge amount if study over the years because it is so challenging. My grad office buddy (btw, an HP employee) spent over 2 years analyzing the Wien Bridge oscillator and still ran into distortion and stability and other limits he could not solve.

Jim

Until Black Lives Matter, we do not have "All Lives Matter"!