A driver coil approximating the string magnetization

[ms]

The purpose of the exciter coil described in this discussion is to make a magnetic field very similar in shape to that produced by the strong.  This would, for example, excite a similar pattern of currents in the metal parts of a guitar pickup.

When a guitar string is magnetized, it is like two long thin bar magnets along the string with like poles facing each other over the pickup pole piece.  The vibrating string changes the flux from these magnets through the pickup coil.  This driver coil uses a section of a string as a core magnetized by coils to simulate this field. Below is a photo of a coil setup for driving a pair of humbucker poles.  The winding direction changes on opposite sides of a pole gap.  The string has a dimeter of .042".  The wrap is nickel, which has a permeability similar to carbon steel




The coil is glued into a plastic (3D printed PLA) holder.




This is mounted over a humbucker located in a holder.




There are a couple of possible problems. First, there are eddy currents induced in the string core.  Are they a problem? Second,  Does the dc magnetization induced in the string by the pickup reduce the effectiveness of the excitation?

For the second problem, the string can be demagnetized by a magnet(s) over the string.  This makes some difference for a humbucker, but it is not really necessary since the string does not appear to be fully saturated by the pole pieces.  This could be more important for a Fender type single coil pickup.  However, even then there is a significant output since the saturation does not occur right over the pole piece since the magnetization goes through zero right over the pickup pole.

The current through the exciter coil is  from the headphone amp.  About 0.1 amp gives a pickup output (SD SH1N) of about 100 mV at 1 KHz.

The response of an SD SH1N is shown below.  

The red line is the response as slightly reduced at high 
frequencies by eddy currents.  The green line is the corrected response.  The correction factor is generated from a measurement made with a small low inductance coil, which has very flat response.



Below is this measurement.  The exciter coil is down a bit more than 1 db at 10 KHz from eddy currents (assuming that there's not other effects).


Below is a measurement of a 2.5 H strat type pickup.  The higher Q is apparent.

[antigua]

This is a cool idea. Could it will model eddy currents in metal covers more effectively by having the magnetic lines intersect with the metal cover more like a real guitar string? Some portion of eddy currents are due to the cover, and some the pole pieces, maybe this configuration will place blame more accurately that an exciter that is focused more directly at the pole piece. 

Even if the string causes its own eddy currents, isn't that a realistic fact of what would happen with moving guitar strings? Maybe a toothpick could be used to avoid eddy currents if string eddy currents don't belong in the measurement, but guitar strings aren't made of wood. 

A big upside also seems to be that the pickup will magnetically couple with the steal string the way a real pickup would, and that might give a more true output measurement comparison between steel pole pieces versus AlNiCo, and other pole piece configurations, as compared to an air core exciter.

[ms]

Jan 2, 2023 21:14:34 GMT -5 antigua said:

This is a cool idea. Could it will model eddy currents in metal covers more effectively by having the magnetic lines intersect with the metal cover more like a real guitar string? Some portion of eddy currents are due to the cover, and some the pole pieces, maybe this configuration will place blame more accurately that an exciter that is focused more directly at the pole piece. 

Even if the string causes its own eddy currents, isn't that a realistic fact of what would happen with moving guitar strings? Maybe a toothpick could be used to avoid eddy currents if string eddy currents don't belong in the measurement, but guitar strings aren't made of wood. 

A big upside also seems to be that the pickup will magnetically couple with the steal string the way a real pickup would, and that might give a more true output measurement comparison between steel pole pieces versus AlNiCo, and other pole piece configurations, as compared to an air core exciter.

Yes, that is what am thinking about a conductive cover like brass. Currents induced in the cover by the time varying field from the string reduce the pickup output.  If the field is not right, the loss could be wrong.  I do not know how big the effect is, but I think it cannot hurt to get it right, or at least close.

The eddy currents in the string that I am concerned about result from the current in the coil around it.  This is different from anything that might get induced by the string vibration.

Your final point is subtle; I do not know if that makes a difference, but it is something to think about.

[aquin43]

An interesting development. Zollner mentions, but declines to illustrate in the pdf version of his book, a tripole coil which I would guess to be the single pole version of what you have produced.

More accurately modelling the string field should excite the pickup in a way that better reflects the cover shielding losses and also might better model the overall response of a humbucker which comprises the sum of two different coil responses. Have you compared the result with the more common method of putting a larger exciter between the two sets of poles?

How necessary is the string to achieving a good level of output?

[ms]

The toothpick coil has about 8 db less output than the string coil, still good.  The tooth pick diameter is a bit more than twice that of the string.  Not sure how that affects the field outside the solenoid.

[ms]

Jan 3, 2023 11:47:21 GMT -5 aquin43 said:

 Have you compared the result with the more common method of putting a larger exciter between the two sets of poles?

You have shown in another current discussion that different screw heights can make a difference in frequency response when using a small exciter coil.  It appears that a very conductive cover (brass), although it can nearly remove the resonant peak, does not cause a significant difference in frequency response between a KW type of coil, and the coil described in this discussion.

In this comparison I use a "toothpick coil", that is, air core, but different from the one described above, using #41 wire for more turns per millimeter.

First plot the difference between the amplitude in db of the response of the tc minus that of the kWc for an SD SH1 N with no cover.  There is some difference age very high frequencies, where it does not matter.  The difference at 5 KHz is about 0.06 db and less at lower frequencies.  This is not significant.


Then plot the same difference with a brass cover in place.  The difference at 5 KHz increases to just under 0.5 db, still not significant.



I will attempt aquin43's measurement with the SH1 instead of a mini humbucker.