Chip Directory Host site

Advertisement provided by Adprov

Inductors

Here are several emails about measuring the inductance of inductors.


Date:    19990528
From:    Declan Moriarty <genius@esatclear.ie>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: Inductance testing

I very occasionally want to find the value of an inductor. This does not justify the purchase of dedicated space consuming equipment. I'd like to make a simple circuit to allow for an educated guess to the nearest value, maybe even using a 'scope or frequency meter. Any suggestions or urls?


Date:    19990528
From:    Clint Hamilton <orpheusaudiovideo@home.com>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: Re: Inductance testing

Inductance meters are very cheap. I have one I got from Parts Express that is a capacitance meter, and checks ohms, hfe on xistors, diodes, and inductance. Costs about 85 bucks US. There are many out there a little cheaper and for a whole lot more too. www.partsexpress.com


Date:    19990528
From:    John E Freeman <edfree@concentric.net>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: Re: Inductance testing

Declan Moriarty wrote:

I very occasionally want to find the value of an inductor. This does not justify the purchase of dedicated space consuming equipment. I'd like to make a simple circuit to allow for an educated guess to the nearest value, maybe even using a 'scope or frequency meter. Any suggestions or urls?

This may work, don't know for shure though. Lets say that the inductor is fairly large so we will use 60(50) Hz line frequency as a standard. Using a filament type transformer with a safe secondary voltage of 6-24 RMS volts, place in series to the secondary a resistor and the inductor in question (the resistor should be great enough to limit the current to a safe value) Measure the voltage drop across the resistor, then calculate, using Ohms law the RMS current, then measure the voltage across the inductor and calculate, using Ohms law and the RMS current, the effective impedence, now with the frequency and impedence known you should be able to determine the inductance. for smaller inductors a higher frequency (must be a known frequency) will be needed.
Let me know if this works.


Date:    19990528
From:    Bob Gordon <kb2zgn@mindspring.com>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: Re: Inductance testing

If you have a function generator, you have all you need to determine the value of any inductor when used in conjunction with the o'scope and frequency counter/meter. The added benefit is that you can also get the Q of the inductor as well with that set-up.


Date:    19990528
From:    Bob Gordon <kb2zgn@mindspring.com>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: Re: Inductance testing

This may work, don't know for shure though. Lets say that the inductor is fairly large so we will use 60(50) Hz line frequency as a standard. Using a filament type transformer with a safe secondary voltage of 6-24 RMS volts, place in series to the secondary a resistor and the inductor in question (the resistor should be great enough to limit the current to a safe value) Measure the voltage drop across the resistor, then calculate, using Ohms law the RMS current, then measure the voltage across the inductor and calculate, using Ohms law and the RMS current, the effective impedence, now with the frequency and impedence known you should be able to determine the inductance. for smaller inductors a higher frequency (must be a known frequency) will be needed.
Let me know if this works.

I'm afraid that this method will not work. The impedance of an inductor is not directly proportional to the frequency of the voltage applied, otherwise this would be a valid test. An inductors impedence is at it's lowest at a frequency (resonant) determined by several factors including wire size, coil diameter, and core material and is not linear over a wide frequency range. The "Q" or bandwidth is determined by finding the points on either side of the peak frequency where the loss reaches the 3db point and varies widely depending on the construction of the coil. This can be easily plotted using a function generator or signal generator, o'scope, and to make frequency measurement simpler, a frequency counter. Frequency can be determined from the scope display, but it is not as precise and assumes that the scope is calibrated. With the advent of cheap inductance meters available, this procedure is time consuming and the only benefit is in finding the natural "Q" of the inductor. Save yourself a load of work and get an inexpensive inductance meter. They have the generator, counter and can calculate the inductance instantly. The time alone that it saves will pay for itself.


Date:    19990528
From:    Andrew Ingraham <Andrew.Ingraham@compaq.com>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: RE: Inductance testing

Of the three basic passive electrical components (resistors, capacitors, inductors), inductors are the least ideal.

They tend to have sizable parasitic components (series resistance, shunt capacitance). Their electrical characteristics may depend more on those parasitics than on the inductance, over wide frequency ranges. That makes measuring their inductance more difficult; you need to find the right frequency range (for each type of inductor) where you are actually able to see and measure the inductance. Too low a frequency, and DC resistance swamps it. Too high, and you're above resonance where capacitance dominates; or you may be into a region of second-order effects, like lead inductance which could be a lot smaller than the low frequency inductance. The "right" frequency range depends greatly on the inductor model ... or on what frequency range is of interest to you.

Many inductors also are very nonlinear in the sense that inductance varies greatly (orders of magnitude) versus DC current. Also they are affected by placement, nearby iron, nearby planes, wires, etc.

So take any inductance measurement with a few grains of salt.


Date:    19990528
From:    Kerry Berland <kerry@siliconengines-ltd.com>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: Re: Inductance testing

I'm sure you'll get some excellent suggestions on a roll-your-own approach. But we use a Wavetek DM27XT multimeter. In addition to the usual-- volts, ohms, amps--it also measures capacitance and inductance. Takes very little space, and not much more money than a standard multimeter.


Date:    19990528
From:    Logger <wireline@mindspring.com>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: Re: Inductance testing

We use the same meter and like it alot.



Date:    19990531
From:    Herne Renaud <Renaud.Herne@UCB-Group.com>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: RE: Inductance testing

What about building an LC oscillator? This circuit has an unique resonance frequency. So it will oscillate with a frequency related to L and C. C is known or can be measured easily (by measuring this way a known inductance ie: calibrate). The problem is resumed to a frequency measurement. The correct frequency is automatically selected provided. the feedback amplifier of the oscillator has sufficient bandwidth. The advantage of the frequency measurment is that it has a very wide dynamic response (1Hz-10MHz is common) I think that this method of measurement should work
... but I never tryed.
Any comments?


Date:    19990531
From:    Declan Moriarty <genius@esatclear.ie>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: RE: Inductance testing

On Fri, 28 May 1999, you informed me:

Of the three basic passive electrical components (resistors, capacitors, inductors), inductors are the least ideal.

They tend to have sizable parasitic components (series resistance, shunt capacitance). Their electrical characteristics may depend more on those parasitics than on the inductance, over wide frequency ranges. That makes measuring their inductance more difficult; you need to find the right frequency range (for each type of inductor) where you are actually able to see and measure the inductance. Too low a frequency, and DC resistance swamps it. Too high, and you're above resonance where capacitance dominates; or you may be into a region of second-order effects, like lead inductance which could be a lot smaller than the low frequency inductance. The "right" frequency range depends greatly on the inductor model ... or on what frequency range is of interest to you.

Many inductors also are very nonlinear in the sense that inductance varies greatly (orders of magnitude) versus DC current. Also they are affected by placement, nearby iron, nearby planes, wires, etc.

So take any inductance measurement with a few grains of salt.

Can I refine my query? I am amazed by the huge response, and thanks to you all. I do not have a function generator - I do have an RF generator (valve operated Heathkit ;-).

The main inductors I come across are small ones from switched mode power supplies, 20-400 uH. These may have a lead broken. Apart from actual value in Henries or fractions thereof, nothing matters. I will not be able to specify all design details; But I can replace with one from the correct range and hope that most things turn out equal. It's called making money, not ethical, or perfect. I had thoughts about a few things (part of a D.C. current transformer) to do this, a square edge into a capacitor with the inductor inline. Give me a few days, and I'll draw up and submit the idea on a .gif.


Date:    19990601
From:    Andrew Ingraham <Andrew.Ingraham@compaq.com>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: RE: Inductance testing

What about building an LC oscillator?
This circuit has an unique resonance frequency.

A possible problem is that the real circuit DOESN'T have an unique resonance frequency, due to all the parasitic effects I've mentioned. If your amplifier has enough gain at high frequencies, it is possible to excite one of those modes rather than the primary parallel resonance. I don't know how often this happens, but it is a potential pitfall.

Also you should account for the amplifier's capacitance, which might be significant compared to the inductor's own.

As a sanity check, connect a known capacitor in parallel, repeat the measurement, and see if the oscillation frequency changes the right way by the right amount. If not, maybe you caught one of those parasitic modes.

Anyone remember a Grid Dip Meter? In the old days, no radio engineer would be without one. One use for those things was to "measure" inductance, by finding the lowest resonant frequency, with a known capacitor in parallel. The idea is similar to making an oscillator, but a whole lot less automatic. I imagine it was tedious, because you had to sweep the meter over several ranges, and you needed to make sure you found the lowest frequency where there was a dip ... not one of those higher modes.

Some coils (ferrite beads, but also some leaded components) are made to appear resistive rather than inductive (much less than 90 degrees phase shift between voltage and current) over a wide frequency range. You might not be able to get them to oscillate readily. But with such a device, maybe you don't care so much about the inductance anyway.


Date:    19990606
From:    Declan Moriarty <genius@esatclear.ie>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: Inductance Testing Notion

I have submitted a file called induct1.gif. Those interested can collect it. It contains my notion on Inductance Testing which I threatened you all with last week. This is part circuit, part block diagram, part omitted. Here is a decscription: (Fairly boring for everyone else :-)

Transformer 1 is a ferrite pot core; It has one feedback winding, and one centretapped main winding: the centretap goes to +Vcc*; One end of the main winding goes to T1, and the other end to T2, via a n/c switch across which the device to be tested is connected. The n/c switch is simply to facilitate zeroing, if needed, and to keep tempratures equal.

T1 and T2 have 10 ohm resistors R1 & R2 on the emitters, and are driven such that when either T1 or T2 are on, the other is off; they switch over when the voltage across the emitter resistor reaches a threshold of 1V, so the current rises from 0 to 100 mA. With me so far? Good.

With the test/zero button closed, the inductance of the pot core governs frequency, and, all things being equal, the dutycycle of the waveform across R1, and R2 should be 50% 0V, followed by a sawtooth with a peak of 1V, as ahown.

If another inductor is added to the equation by opening this test/zero switch , duty cycle should change, and T2 will have a longer sawtooth rise than T1 when the test button is open circuit. That facilitates measurement. The math would involve L1(the pot core) * L2 / L1 + L2 type sums, something even I can manage. Another and perhaps better approach would be to measure duty cycle on T1 & T2 - the longer the duty cycle, the bigger the device.

I like the idea because

  1. Because it's mine :) - funny how that happens always, isn't it?
  2. Because it measures inductance only. It would have to have fairly large parasitic capacitance to affect things at the currents used.

Date:    19990607
From:    Radek Soltys <rspublic@friko6.onet.pl>
To:      Multiple recipients of list CHIPDIR-L <CHIPDIR-L@fatcity.com>
Subject: Re: Inductance testing

I very occasionally want to find the value of an inductor. This does not justify the purchase of dedicated space consuming equipment. I'd like to make a simple circuit to allow for an educated guess to the nearest value, maybe even using a 'scope or frequency meter. Any suggestions or urls?

Here's a simple circuit that should allow you inductance approximate recognition. I haven't tested it yet, but it is very simple, thus worth trying. If you have any questions about its operation or assembly - contact me personally and I'll refer to the Polish article where I found the schematic.


Ad for PCI video capturing boards by Dektec.
Goto: Main Mirror About Author
Register: Yourself Company
Feedback: Correction Addition Question Forum
Order: Chips (Deutsch) Chips (English) Chips (Nederlands)

Advertisement by Adprov

Viewable with any browser