Measuring inductance with standard multimeter

[David Schwab]

because the human ear does not sense phase either

Generally yes, this is true. But some people can hear the difference (I used to, but I haven't tried it in many years, so my hearing is probably not up to it anymore, and I couldn't tell you what I heard different. But I did.)

It seems that the high frequency stuff we assume is outside our range of hearing is a lot of phase related information, which might help us detect where things are around us by sound reflections.

[Joe Gwinn]

Generally yes, this is true. But some people can hear the difference (I used to, but I haven't tried it in many years, so my hearing is probably not up to it anymore, and I couldn't tell you what I heard different. But I did.)

It is true that we can sense phase differences between ears (at least at low frequencies), and this is used to determine the direction of the sound source. As is differences in intensity between ears (at higher frequencies).

However, we have drifted a bit. My point was that sound cards are typically not equipped to measure phase between output (to speaker) and input (from mic), because it isn't generally needed, and that this lack of phase information will prevent a sound card from making some kinds of inductance measurements.

But I bet there are industrial sound cards that are sensitive to phase. The giveaway would be a card that yielded I and Q (in-phase and quadrature) data on its input signal. A two-channel card could measure phase in and phase out, plus test signal frequency, from which one can calculate RLC as for a handheld meter.

I bet the Extech is cheaper than that industrial sound card. The advantage of the industrial card would be that one could work at any frequency.

It seems that the high frequency stuff we assume is outside our range of hearing is a lot of phase related information, which might help us detect where things are around us by sound reflections.

This happens too. There is a lot of research going on into how we can tell the elevation of a sound source, and the basic answer seems to be that we sense the delayed replicas of the source signal, the amount and pattern of the delays being the clue. The schemes I've seen are very complex; perhaps someone will find the simplicity within, but it hasn't yet happened.

Our ability to hear transients (like attacks) seems to well exceed the 20-20000 Hz bandwidth one gets by testing the young with sinewave tones, so I would design for response out to about 50 KHz.

[Steve Conner]

How about looping the output back to one of the inputs of the sound card? Then all it needs to do is measure the difference between its left and right inputs to get phase difference, or indeed a complete transfer function analysis.

There's a program called Spectrum Lab that can do I and Q measurements with any sound card. It just runs in stereo and assumes one channel is I and the other is Q. I used it to demodulate shortwave radio with a home-made Tayloe detector. I doubt that it measures inductance, but it proves that you can do things with phase...

[Yves]

And flood one with valuable free offers.

Most probably, but you don't need to use your primary email account for this purpose.

What is the basic measurement being made? This will help tell if the instrument can handle pickups, which have a large amount of DC resistance in series with the inductance. Most LCRs give wildly wrong answers on pickups.

My guess is that this sound-card based instrument will not work well with pickups for a very fundamental reason: sound cards ignore phase and measure only amplitude. For a sound card, this is reasonable, because the human ear does not sense phase either, and certainly does not care about relative phase between input and output.

.

How about looping the output back to one of the inputs of the sound card? Then all it needs to do is measure the difference between its left and right inputs to get phase difference, or indeed a complete transfer function analysis.

There's a program called Spectrum Lab that can do I and Q measurements with any sound card. It just runs in stereo and assumes one channel is I and the other is Q. I used it to demodulate shortwave radio with a home-made Tayloe detector. I doubt that it measures inductance, but it proves that you can do things with phase...

That's pretty much the principle here: one channel input is used as a reference, the other used for the measure, the output feeding the signal test (1000, 1250 or 2200Hz, which makes me think it should be usable for pickups).
AND it does test both amplitude and phase.

It's just silly as the fact that I don't have any spare minijack with me to run test at the moment (because there is also a way of using is with no IC buffer at all), and I don't way to dismantle some headphones for this...

I'll let you know (for those who are interested) how it does the job. I am personnally not looking for a precision lab tool at the moment but something to compare different pups or for trouble shooting.
No doubt an Extech is far more handy and reliable, but if this works as is, it would just be fine for me.

[Mike Sulzer]

I have a standard multimeter with an (supposedly) inductance meter position.
However I cannot get the inductance of any pickup with this. I believe this is "normal" since this is true for any pup I've tried to get the value.
Could someone give me the clue why this ? I suppose I'd need a proper LRC meter for this task, right? (But why then having an inductance meter on the multimeter if useless...is this just a crappy tool?)

Thanks,
Yves

Rather than attempt to measure the inductance directly, why not measure the resonant frequency with a typical amount of capacitance across the pickup? You need hardware and software on your computer so that you can output signals (I use random noise.), and you need an input and software to measure a power spectrum and accumulate over many passes to give good accuracy. I use a one meg resistor in series with the sound output so that you have close to a current source and measure the voltage across the pickup, using a high impedance preamp before the computer input. You can add capacitance across the pickup so that the total (cable + preamp input + added C) is a few hundred to 1000 pf to simulate actual conditions. (Or just use an actual guitar cable. You can measure its capacitance sort of accurately with you multimeter.)

[Joe Gwinn]

How about looping the output back to one of the inputs of the sound card? Then all it needs to do is measure the difference between its left and right inputs to get phase difference, or indeed a complete transfer function analysis.

There's a program called Spectrum Lab that can do I and Q measurements with any sound card. It just runs in stereo and assumes one channel is I and the other is Q. I used it to demodulate shortwave radio with a home-made Tayloe detector. I doubt that it measures inductance, but it proves that you can do things with phase...

The problem with using the left and right channels of a sound card is that they may not have time-aligned the sampling instants in the two channels. True I&Q requires that the samples be taken at the same instant in both channels.

That said, if one samples way faster than the signal, time misalignment isn't that much a problem.

Do these sound cards have four input channels? One needs two channels for I&Q of input to the pickup under test, and another two channels for I&Q of the output.

[Steve Conner]

Any half decent sound card should have time-aligned channels, otherwise it would have a lousy stereo image at high frequencies.

I have a M-Audio Delta 1010 in my home recording setup, so I'm never short of inputs and outputs

[Robert M. Martinelli]

Lutron LCR meter

What's the exact model number? I'd like to look it up.

Hi Joe, the model number is LUTRON 9073

Regards

Bob

[Robert M. Martinelli]

A simpler way to determine pickups' specs

I have developed a simpler way, a variable frequency sinewave generator and an oscilloscope will suffice.... here it is :

Salvage a relay with a small DC resistance coil ( and little inductance ) , and tie it to the signal generator, then hook the pickup ouput to the oscilloscope.

Start generating a signal and put the relay coil near the pickup under test - the sinewave will appear on the oscilloscope, so adjust the scope and the signal source for optimal measurement/visibility ....measure the amplitude, ensuring yourself the distance between the relay coil and the pickup poles remains constant - then simply increase the frequency measuring the amplitude as you go up, you' ll be able to determine the resonant peak, the Q factor and the bandwidth, which are determined by LC ( resonant peak ) and RLC ( Bandwidth - Q ). You can also evaluate the pickup' s behavior with a load connected ( real-world ) simulating an amp' s input circuit with a RC network, and also take into account cable capacitance ( added in parallel to the pickup' s output, say 100-120 pf/mt.

The low resistance, low inductance and equally low parasitic capacitance of the relay coil will ensure you accurate readings, it is true that they will generate a second resonant peak but since this will be located at a much higher frequency, it will not spoil your readings.

Hope this helps

Best regards

Bob

[Joe Gwinn]

Hi Joe, the model number is LUTRON 9073

Found it. Thanks. There is no mention of allowed Q ranges, or any way to select the series or parallel arrangement of inductance and resistance, so I'd worried that this instrument assumes pure inductances, and will not do well with impure inductances such as pickups.

[Robert M. Martinelli]

Found it. Thanks. There is no mention of allowed Q ranges, or any way to select the series or parallel arrangement of inductance and resistance, so I'd worried that this instrument assumes pure inductances, and will not do well with impure inductances such as pickups.

Hi Joe, that' s true, but I have verified the readings using the method I described in my last post, and they matched almost exactly.... since a pickup can be modeled as an inductance with a resistance in series and a capacitance in parallel, I measure R and L separately, then I apply a guess ( which has proved to be fairly good in several cases ) of assuming a parasitic winding capacitance of about 110 Pf for a 6 KOhm winding, measure the DC resistance of the pickup under test and estimate the parasitic capacitance accordingly; when verifying with the generator and oscilloscope, the actual resonant peak, Q and bandwidth were where I expected them to be....very close or matching with the calculated ones.

[Joe Gwinn]

Hi Joe, that' s true, but I have verified the readings using the method I described in my last post, and they matched almost exactly.... since a pickup can be modeled as an inductance with a resistance in series and a capacitance in parallel, I measure R and L separately, then I apply a guess ( which has proved to be fairly good in several cases ) of assuming a parasitic winding capacitance of about 110 Pf for a 6 KOhm winding, measure the DC resistance of the pickup under test and estimate the parasitic capacitance accordingly; when verifying with the generator and oscilloscope, the actual resonant peak, Q and bandwidth were where I expected them to be....very close or matching with the calculated ones.

This is pretty indirect, and is affected if the AC resistance differs significantly from the DC resistance, which is often the case.

DC resistance is what one measures if the test signal is DC, which is what ohmmeters do.

AC resistance is measured using an AC test signal, and is the AC voltage divided by the in-phase (with the imposed AC voltage) AC current.

AC resistance is not reactance. Reactance is the AC voltage divided by the quadrature (with respect to the imposed AC voltage) AC current.

The inductance causes the reactance, while the DC resistance is the DC resistance plus eddy-current loading.

The Extech measures and reports both reactance and AC resistance.

It isn't clear what the Lutron 9073 does, but a typical approach (for pure inductances and capacitances) is to measure AC voltage and resulting AC current (the current being the vector sum of the AC resistance and reactance), and assume that the measured AC current is entirely due to reactance. This works well enough for ordinary electronic components, which are pretty pure electrically.

What also works quite well is a Maxwell-Wein Impedance Bridge, which despite the big name is simple and cheap to construct from ordinary radio parts, and measures AC resisance directly. Bridges were the mainstay of national standards labs for a good century, partly displaced only by the rise of digital instruments.

Here is my report: http://home.comcast.net/~joegwinn/.

[Robert M. Martinelli]

Joe, all you stated is absolutely correct, I just wanted to point out that it is possible to get a pretty good idea of a pickup' s behavior using fairly cheap equipment and a little of experience....
Best regards
Bob

[Yves]

Joe, you were right: this setup lead me to nowhere (weird readings).
However, regarding the soundcard, I am convinced that nearly any card on the market has left and right channels of stereo inputs time aligned, or this would be a faulty design really. I can understand that multitrack digital systems may have time alignment issues depending of DSP load or plug-ins inserts, but stereo inputs must have both channels time aligned.
Besides the lousy stereo image that Steve mention, in the event of for instance a consumer that uses the card to capture a "fulltrack" mono (from say some old VHS for archiving), time misalignment would end up with nasty comb filtering. I guess most of the card are designed around integrated codecs that are time aligned.

Back to the RLC thing, I've bought this:
https://www.conrad.fr/pont_rlc_porta..._217179_217180

I believe this is the similar to the Extech (similar specs and look), and it is quite a handy tool indeed.
It was 35% off the price a few days ago: for anyone interested in ordering from Conrad, just Google "Conrad code promo" first and you most probably find some kind of special discount/offer (they always have some).

Thanks to all anyway for the advices on the measures.

[Joe Gwinn]

Joe, you were right: this setup lead me to nowhere (weird readings).

Bullet dodged.

However, regarding the soundcard, I am convinced that nearly any card on the market has left and right channels of stereo inputs time aligned, or this would be a faulty design really. I can understand that multitrack digital systems may have time alignment issues depending of DSP load or plug-ins inserts, but stereo inputs must have both channels time aligned.
Besides the lousy stereo image that Steve mention, in the event of for instance a consumer that uses the card to capture a "fulltrack" mono (from say some old VHS for archiving), time misalignment would end up with nasty comb filtering. I guess most of the card are designed around integrated codecs that are time aligned.

The R and L channels need time alignment only good enough to fool a human.

However, sound cards must sample fast enough to handle 20 KHz, so the misalignment probably isn't going to cause big problems for measurement of pickups at 1 KHz. I would just try it.

Back to the RLC thing, I've bought this:
https://www.conrad.fr/pont_rlc_porta..._217179_217180

I believe this is the similar to the Extech (similar specs and look), and it is quite a handy tool indeed.
It was 35% off the price a few days ago: for anyone interested in ordering from Conrad, just Google "Conrad code promo" first and you most probably find some kind of special discount/offer (they always have some).

Pont's website had useful information. To my eye, the Pont 4080 appears identical to the Extech. Probably comes from the same factory in Taiwan.

There is something else - the circuit diagram. It is as I had expected, being an "IV" design. The Plus terminal is driven by a voltage source, and the Minus terminal feeds a opamp transimpedance amplifier (which has ~zero input impedance and converts incoming current into voltage). It is easy to home-brew such things, and such a circuit could be wedded to the sound card, with the left channel measuring the imposed voltage and the right channel measuring the resulting current (as converted to a voltage by the transimpedance amplifier).