PE vs Poly wire isolation experiment with sound samples included

[Mike Sulzer]

One thing that concerns me, in some of the pictures I have seen with the driven coil/sensed coil measurements, the one coil simply sits on top of the other. This would allow capacitive coupling between the two coils to add error.

When I do driven/sensed coil measurements, I put a sheet of grounded nickle silver between the 2 coils to eliminate capacitive coupling.

But do eddy currents in the metal sheet affect the measurements?

[marku52]

Of course, but so would the C coupling. For my pickups, at least, the nickle silver case is a part of the pickup so should be included anyway.

Electrical measurements with/without case show pretty small deltas, a couple of percent or so in L and Q, so it is not very dramatic.

[Mike Sulzer]

Of course, but so would the C coupling. For my pickups, at least, the nickle silver case is a part of the pickup so should be included anyway.

Electrical measurements with/without case show pretty small deltas, a couple of percent or so in L and Q, so it is not very dramatic.

Thanks, that is good to know.

[Joe Gwinn]

Stainless steel foil should largely eliminate eddy current effects, and yet totally eliminate inter-coil capacitance. Not to mention titanium foil.

https://www1.mscdirect.com/CGI/NNSRI...MTERM=56484991

Amazon.com: Titanium Foil: Industrial & Scientific


Eddy Current Technology Incorporated Resistivity of Metals


Or, very thin brass foil strips arranged as overlapping leaves insulated from one another, and electrically connected in exactly one place will also do the job.

[David Schwab]

Stainless steel foil should largely eliminate eddy current effects, and yet totally eliminate inter-coil capacitance. Not to mention titanium foil.

https://www1.mscdirect.com/CGI/NNSRI...MTERM=56484991

Amazon.com: Titanium Foil: Industrial & Scientific


Eddy Current Technology Incorporated Resistivity of Metals


Or, very thin brass foil strips arranged as overlapping leaves insulated from one another, and electrically connected in exactly one place will also do the job.

Good suggestions! I hadn't thought about stainless steel foil. Both of those options are rather expensive though.

Another idea is to use brass mesh. I was using fine gauge brass screening for a while. I got the idea from dissecting an EMG pickup.

And here's a photo of some early Alembic pickups. Note that Rick Turner used foil strips with gaps in them. I'm guessing it might be a spiral grounded at one end.

Currently Alembic uses copper braided mesh as you would find on coax cable.

[Joe Gwinn]

Another idea is to use brass mesh. I was using fine gauge brass screening for a while. I got the idea from dissecting an EMG pickup.

This works, but there is some electric-field leakage through the mesh. How much and how important depends on details.

And here's a photo of some early Alembic pickups. Note that Rick Turner used foil strips with gaps in them. I'm guessing it might be a spiral grounded at one end.

A spiral wrap would work quite well. Best is where the brass strips overlap at the edges but are insulated from one another.

Currently Alembic uses copper braided mesh as you would find on coax cable.

That also works, but as with the mesh there is some leakage.

[FunkyKikuchiyo]

Not as much voting going on here as I thought.... just sayin' is all.

[MrCandy]

There should be a close relationship between the two kinds of measurements because the driver coil induces a signal in the coil and the response to this signal is a function of the coil impedance. However, we have to be careful in order to get a good correspondence. In particular, the comparison between the impedance measurement with 1 Meg and the driver with 1 Meg should be nearly the same.

That it is Mike.
Here are some measurements to compare(coil from experiments above was used):
1) Passive probe 1M

2) Passive probe with driver coil


If we understand you correct about scaling(translation diff),then you want to see the measurements for different coil load and current signal at the circuit,
for accuracy we up bins number from 512 to 4096. Also the measurements were carried out applying a chain with generator 250mV and 2V.
Active probe was used in parallel 1M and without. Measurement scheme you can see here pic №2.
1)250mV signal on circuit, active probe

2)2V signal on circuit, active probe

3)250mV signal on circuit, active probe 1M parallel

4)2V signal on circuit, active probe 1M parallel


Anyway effect of coil load is known for decades, for example EMG uses it for modeling AFC of their pickups,
because true RP of their coils are extremely high and they often use ceramic magnets as cores, permittivity of wich is like air.
Thats why for modeling AFC they use current amp that with its input loads the coil. They use different input R for in in each case from 47K-100K
in this way they gain "Frequency compression" in its kind, but many guitar players say that the sound is "lifeless' this way.
By the way the picture attached is AFC of EMG81 made with spectrum analyzer when there was no modern computer programs more then 10 years ago.

(the measurements were held using driver coil and special current amp)

[David Schwab]

This works, but there is some electric-field leakage through the mesh. How much and how important depends on details.

I found it worked well for my regular humbuckers, but the sidewinders were noisier around light dimmers and stuff, so I went to solid copper foil. Now they are dead quiet.

[MrCandy]

One thing that concerns me, in some of the pictures I have seen with the driven coil/sensed coil measurements, the one coil simply sits on top of the other. This would allow capacitive coupling between the two coils to add error.

When I do driven/sensed coil measurements, I put a sheet of grounded nickle silver between the 2 coils to eliminate capacitive coupling.

If you would ve closely look at measurements 1 and 2 here, then you would ve note, that data is almost the similar although measurements were made using different methodics. With driver coil we use specially designed amp with extremely high stabilization of current in driver coil in all significant frequency range.
We dont really know what driver coil you use, although when you claim something you would better to show measurement scheme and components char(measuring cable, probe and etc.) or we can have a casus like we had here.

About our driver coil - here is its Tec. Char. :

Test measurement coil TC:

Test measurement coil with Nickel Silver plate:


As you see inductance changed that influenced AFC

Active probe


Tthe same but with driver coil

Without nickel silver check


Here is the measurement of test coil with "driver coil on top" where driver coil is in short circuit...
Look post 2...

[MrCandy]

Here is the measurement of test coil with "driver coil on top" where driver coil is in short circuit


As we see effect as from the storm in the sun compare to what nickel silver do.

And for exact confidence you can see the measurements of Strat Pickup 57/62 measured with 2 methods
with and without driver coil:

(using driver coil)

(active probe)

If in your case you driver coil parasitically affect the measurements, then your coil is wrong, make a new one and special current stability amp with reverse connection in currentcy for it
Our driver coil here for example have 13 number of turns.

[Mike Sulzer]

If we understand you correct about scaling(translation diff),then you want to see the measurements for different coil load and current signal at the circuit,
for accuracy we up bins number from 512 to 4096.

I think I have not communicated this very well. It is a bit complicated.

The two ways of measuring a pickup are very similar in that they both show the resonance, but there are some differences, not necessarily all that large.

1. First consider an impedance measurement. You are doing that by connecting a signal to a 1 Meg resistor. You could do it with a current source also, or I prefer the IV circuit that Joe Gwinn suggested. At the low frequencies, the impedance of the inductance goes to zero, and you have the coil resistance. If you just look at an air core coil, you have a resonance, a simple one with the usual shape. A pickup with steel cores is more complicated. Eddy currents lower the impedance below the resonance and change the of response at the resonance, especially lowering the Q.

2. Second is measuring the response with a driver coil. If you hold the driver coil current constant as a function of frequency, then the voltage induced in the pickup coil rises with frequency (law of magnetic induction). This rising response is then modified by the pickup circuit. This circuit is a low pass filter: the coil is in series with the induced voltage, and the capacitor shunts the signal through the coil to ground. It is a resonant low pass filter: you see the same resonance as when you measure the impedance. The steel cores make this a bit more complicated.

The reason the responses you measure with the two different techniques look similar at the low frequencies, even though the circuit is a resonant circuit in one case and a low pass filter in the other, is that the voltage induced in the pickup coil falls at the low frequencies because of the law of magnetic induction. If you set up the driver coil so the the voltage induced in the pickup coil is constant with frequency, then the response you see does not fall off with decreasing frequency.

Or you can correct the response by "scaling it up" at the lower frequencies. This requires multiplying by a function of frequency. This function is 1/f, except at zero frequency, of course. I have done this with the plots I posted. That is why you can see the details of the response at the lower frequencies.

[MrCandy]

I think I have not communicated this very well. It is a bit complicated.

The two ways of measuring a pickup are very similar in that they both show the resonance, but there are some differences, not necessarily all that large.

1. First consider an impedance measurement. You are doing that by connecting a signal to a 1 Meg resistor. You could do it with a current source also, or I prefer the IV circuit that Joe Gwinn suggested...

Ok Mike now it makes sence.
The method you mention is good, but we need to go back to the start of discussion.To the measurement cable construction exactly.

Below I want to show measurement scheme suggested by you.


Here we can have а pitfall. Namely if you look at point B at the scheme it is important no less then point A when you measure.
If in point B it will be a parasitic capacitance then it will work like differential chain with its own settling time.
That will make phase to move and it will cause measurement error if you orient by ellipse.
Thats why Ideally you must measure in point B with min Rin. To achive this we used 600 Om load, that minimized parasitic capacitance influence and phase shift.

Further I show XY measurements where you can by the reversal of the ellipse estimate the resonance.
The reversal of ellipse is minimal on the same indexes of resonanse that we showed earlier by another methods(same coil).



Continue post 2...

[MrCandy]

Our data matched with high % of accuracy. Thats why mesurement with the help of active probe with minimal IN capacity we can and recognize as the most accurate.
We can even call it an INDUSTRIAL STANDARD for this type of mesurement.
Anyway XY measurements are the most useful when you measure 2 coils of humbucker in case when AFC of pickup have complex form.

The sience approach is good for the reason that everyone can test and have the same results for the same test, as a result we will stop to read catalogue crap data and start to experiment and develop industry

[MrCandy]

Thanks to every person who took an active part in this thread.
One can recognize that this test was not 100% correct,
because one wire gauge(PE) was thinner that was clearly heard by ear.
In another experiment we tried minimize wire gauge influence on the experiment
by winding the same pickup(dark horse one) with PE wire.

That was my guess exactly, actually. Poly has more weird uneven harmonics in the upper ranges - kinda reminds me of how ceramic magnets behave, in pickups or in speakers. I hear that a lot in the first one. It tends to sound like extra pick scrape, even though the difference is the wire rather than the technique. The second one had a more open, hollow sound which is typical of having multiple resonances, though still has those funny harmonics. The third sounds like PE to me; it is cleaner sounding and favors even order harmonics (to my ears anyway, I have no data to back that up), and as an added bonus it sounds more midrangey, which would be entirely explained by the thinner wire.

Here is the correct answers:

1) Poly
2) Dark horse(Poly wih the same number of turns)
3) PE

Сomments:
№1 and №3 sound pretty similar, typical sound with clothespin on the nose №3 especially(bacause wire gauge is thinner)
Nose sound happens due to phase distortion. Paradoxically, but phase shift of this two coils is negligible, its some kind of untuned "unison" in result.
90% of humbuckers have an effect like that.
In sample №2 we lowered this effect due to the large phase shift between coils, thats why it sound more clear. The same goes to PE "dark horse" we mentioned earlier.
And this PE sample sound completely different from sample №3.