Is this a good meter?

[madialex]

Extech 380193 LCR Meter - instrumart.com

It's an Extech.. I think this is the one Possum speaks highly of.

[Joe Gwinn]

Extech 380193 LCR Meter - instrumart.com

It's an Extech.. I think this is the one Possum speaks highly of.

Yes, it is a good meter, and it works well with pickups. (Many LCR meters do not work properly with low-Q inductors like pickup coils.)

The Extech 380193 is the standard LCR meter used by guitar pickup makers.

[madialex]

Thanks Joe, looks like this is in my near future

[Spence]

I have one. I sits in a drawer. You'll never need it to make a good pickup. I wouldn't waste your money Ben. I feel like I wasted my money on one. I only bought it because people may ask what the inductance is etc. They never have.

[salvarsan]

Yah, it's the one.

Just so you're clear on this, an LCR meter is for quality control.
It tells you how consistent your pickups are for a particular design and winding,
and helps identify shorted coils whose DC resistance is only slightly off spec.

-drh

[Possum]

They are useful, to a point. If you're designing a new pickup and can't quite tell the difference between 2 prototypes the LCR meter can spot small differences, which can be helpful. Also, if you have a questionable pickup that doesn't sound right yet the "ohms" reads right the LCR meter can spot a coil short. I don't use mine as much as I used to, but when prototyping it gets used alot. Been using it recently on some new pickups. I also look at it as a learning tool for learning the physics of magnetic coils, it taught me alot along the way. What it totally fails at though, is what happens in the high treble range up past the 1khz test signal. You can have 2 pickups that read nearly identically, yet the treble range can be noticeably different. I haven't quite thought it important enough to buy one with a 10khz test signal though, the AC resistance on one like that could be useful maybe. Bottom line is what your ears tell you.

[Joe Gwinn]

They are useful, to a point. If you're designing a new pickup and can't quite tell the difference between 2 prototypes the LCR meter can spot small differences, which can be helpful. Also, if you have a questionable pickup that doesn't sound right yet the "ohms" reads right the LCR meter can spot a coil short. I don't use mine as much as I used to, but when prototyping it gets used alot. Been using it recently on some new pickups. I also look at it as a learning tool for learning the physics of magnetic coils, it taught me alot along the way.

This seems to be the consensus.

What it totally fails at though, is what happens in the high treble range up past the 1khz test signal. You can have 2 pickups that read nearly identically, yet the treble range can be noticeably different. I haven't quite thought it important enough to buy one with a 10khz test signal though, the AC resistance on one like that could be useful maybe. Bottom line is what your ears tell you.

While I don't doubt that LCR meters using a 1 KHz test signal tell one not quite everything about treble response, I have to mention that because many pickups resonate below 10 KHz, a 10 KHz test signal won't help. There are units that can test at multiple frequencies between 1 and 10 KHz, but they are quite expensive. (The slow but cheap approach is to build a Maxwell-Wein Impedance Bridge. See Joe Gwinn's website homepage for more information.) This issue has been discussed at length in other threads.

[Possum]

Your website says the file doesn't exist. A 3 or 5 khz test signal would be real useful, I've used your peak resonant method but it gets a bit cumbersome and slow, maybe because my frequency generator is an old dial analog type, hard to dial in exact frequencies. And then I don't think you can see harmonics. Really the hard rule is put the pickup in a real guitar, keep the wiring harness standard and play. It sometimes take 3 or 4 days playing a pickup to get a sense of what it'll do, and then you have to get past the "settling in" effect of putting a new humbucker together. From what I'm seeing there's some kind of settling from the coil and also it seems the magnetic domains need time to all come together, most of it happens over 2 days, but it really takes about a month for a new humbucker to stabilize....

[Mike Sulzer]

What does the Extech do?

It is important to understand what the Extech does. It is a two step process:
1. It measures the complex impedance at either 100 Hz or 1 KHz. That is, it measures the magnitude of the ratio of the voltage across the unknown impedance to the current through it, and the phase angle between the two.
2. It interprets that measurement in a way determined by the function setting.

For example, if you want to measure inductance, you also get to choose whether you want to measure a resistor in series or the equivalent of a resistance in parallel. The measurement is only going to be accurate if one situation or the other is a physically accurate description of the unknown impedance and you choose the right one.

Pickups in general have both kinds of loss. The series resistance is really important, but eddy currents in various metal parts of the pickup can also matter at some frequencies. The loss due to eddy currents is in parallel with the coil, but it can be more complicated than just a resistor, but that is the best first approximation.

To see how having both types of loss screws up step 2, the interpretation, first consider an inductor with a series resistor alone. The Extech measures some magnitude and angle. Draw this as an arrow on an x-y plot. The tail of the arrow goes in the middle at (0, 0). The length of the arrow is the magnitude. The angle between the x axis and the arrow is the angle of the impedance. For example, if the arrow points straight up along the y axis, then the impedance is a pure inductance. If it points along the x axis it is a pure resistance. If it points somewhere in between, one gets the size of the inductance by the projection on the y axis, and the size of the resistance by the projection onto the x axis.

Then we are interpreting the y axis as the "imaginary" part and we can write the impedance as R + j*2*pi*f*L, where j is the square root of minus one and L is the inductance. R and 2*pi*f*L are the lengths of the projections on the x and y axes.

What happens if we now add a resistor in parallel with the inductor that had only a series resistance? Since it is in parallel this reduces the magnitude of the impedance. That is, it makes the arrow shorter. The projection on the y axis decreases, and so the apparent inductance decreases even though the inductance has not really changed at all.

Also the angle of the impedance gets smaller. This is because the resistor in parallel make the impedance more like a resistor. This also causes the apparent inductance to decrease since it also shortens the projection on the y axis.

So if you measure different inductances at 100 Hz and 1 KHz, are the inductances really different, or are the errors in the measurement different?

None of this matters if you are just using the meter to look for possible problems in a batch of pickups. But if you need accurate inductances?

[Joe Gwinn]

Your website says the file doesn't exist.

I just looked, and it's still there. What URL did you use?

A 3 or 5 khz test signal would be real useful, I've used your peak resonant method but it gets a bit cumbersome and slow, maybe because my frequency generator is an old dial analog type, hard to dial in exact frequencies. And then I don't think you can see harmonics.

The Maxwell-Wein Bridge is not especially sensitive to frequency, far less so than for resonant frequency measurement, so it would suffice to set the analog generator to a nominal frequency and measure the actual frequency with a digital voltmeter (many will measure frequency in the audio band).

The M-W Bridge does work best with fairly low distortion test signals, for an odd reason. While the M-W Bridge is frequency-independent, iron-cored inductors like pickups are not, so one will get slightly different nulls at fundamental and at harmonics. The inductance varies with frequency. One can see this using the Extech, switching between 60 Hz and 1000 Hz test frequencies.

Really the hard rule is put the pickup in a real guitar, keep the wiring harness standard and play. It sometimes take 3 or 4 days playing a pickup to get a sense of what it'll do, and then you have to get past the "settling in" effect of putting a new humbucker together. From what I'm seeing there's some kind of settling from the coil and also it seems the magnetic domains need time to all come together, most of it happens over 2 days, but it really takes about a month for a new humbucker to stabilize....

Three to four days? I would hope that the M-W Bridge would be far faster than this.

[Mike Sulzer]

The inductance varies with frequency. One can see this using the Extech, switching between 60 Hz and 1000 Hz test frequencies.

You have not shown that the inductance really varies with frequency. When you determine inductance from complex impedance assuming a series resistance, there is an error from any parallel resistance. This has different effects at different frequencies.

[Joe Gwinn]

You have not shown that the inductance really varies with frequency. When you determine inductance from complex impedance assuming a series resistance, there is an error from any parallel resistance. This has different effects at different frequencies.

Mike, not again. We thrashed this out ad nausem, in what must have been a few hundred posts, some quite strident.

If you like, read Terman (the precise reference being in one of my posts), where he gives the method for determining self-resonant frequency of a coil, and comments that the line curves if the inductance varies with frequency. The implication being that Terman certainly believed that the inductance varied with frequency. As did all of his colleagues. Your own measurement data showed this curvature. And, no Extech was involved in generating that curve, so the Extech cannot be cause or cure.

Perhaps Terman and his colleagues meant some kind of "effective inductance", versus some kind of "true inductance", but they never made such a distinction, probably because circuits containing such an inductor could not make the distinction either. So Terman et al believed their instruments and got on with their lives.

[Mike Sulzer]

Perhaps Terman and his colleagues meant some kind of "effective inductance", versus some kind of "true inductance", but they never made such a distinction, probably because circuits containing such an inductor could not make the distinction either. So Terman et al believed their instruments and got on with their lives.

I think that is a very good point about effective inductance versus true inductance; it is likely that that is exactly what Terman et al. meant. And if they could explain the circuit operation in that way, that is all very good.

But that approach is not very useful for understanding the physics of a humbucker guitar pickup. I think we all recognize that eddy currents load the pickup resonance where the impedance is high and so is easily affected, but the effect below resonance seen in Lemme's measurements of pickup response and my measurements of impedance are more subtle but nonetheless important for understanding the pickup.

A physical model of such effects does not agree with your contention that the pickup inductance drops at the mid-frequency point measured by the Extech. This effect can instead be explained by the fact that an instrument measuring complex impedance and interpreting the result as an inductor and series resistor gives a low value for the inductance when parallel loss is also present.

It is necessary to reject explanations that do not agree with measurements as interpreted by theory and accept those that agree.

[Joe Gwinn]

I think that is a very good point about effective inductance versus true inductance; it is likely that that is exactly what Terman et al. meant. And if they could explain the circuit operation in that way, that is all very good.

But that approach is not very useful for understanding the physics of a humbucker guitar pickup. I think we all recognize that eddy currents load the pickup resonance where the impedance is high and so is easily affected, but the effect below resonance seen in Lemme's measurements of pickup response and my measurements of impedance are more subtle but nonetheless important for understanding the pickup.

A physical model of such effects does not agree with your contention that the pickup inductance drops at the mid-frequency point measured by the Extech. This effect can instead be explained by the fact that an instrument measuring complex impedance and interpreting the result as an inductor and series resistor gives a low value for the inductance when parallel loss is also present.

Eddy currents can exist in the absence of eddy-current loading. The classic example is the superconducting slug pole. A copper slug is a pretty good approximation. This was all explored at length in my posts from the prior thread. Many references were cited as well.

I don't have the energy to re-fight that war, and no new issue has been raised.

Your theory must encompass all the (gedanken) examples and experimental results reported over at least a century.

It is necessary to reject explanations that do not agree with measurements as interpreted by theory and accept those that agree.

Umm. Not quite that simple. The theory and/or the interpretation can be wrong, and the explanation can still be correct. Or the theory can be misapplied. Or all can be wrong.

[Mike Sulzer]

I have shown why measurements made with the Extech are in error when both series and parallel loss are present. You have not demonstrated that there is anything wrong with what I have shown, nor have you presented any evidence that the Extech is correct. Neither now nor before.

If you want to recommend it, you should show that the errors are small enough not to matter. Long ago I suggested this. I am certainly not going to to do it for you. You are the one who says its use is OK.