Pole Shading

[aquin43]

Shading a pickup pole.

Sometimes the top string of a strat can be too shrill or lack the meat desired for certain types of music. Turning down the tone control corrects this but at the expense of the desired edge on the tone of the rest of the strings. A low pass filter just for the one pole can be constructed by shading the pole with a copper shim placed between it and the strings. The impedance of the pickup and its basic response are barely affected by this.

The diagram below shows the effect of a 0.8mm copper shim made by flattening a piece cut from plumbing pipe. It represents the difference between the unshaded and shaded responses. The effect is as a low pass filter with -3dB just below 2kHz.

Note how the phase shift starts to return to zero at high frequencies. This is because, with this particular setup,  the attenuation bottoms out at -12dB.

The shim is thinner than the maximum pole projection on a staggered pole pickup and so doesn't interfere with picking.

Arthur

[newey]

Wondering what you use to hold the shim in place, and whether whatever the substance is might also affect the freq response?

[aquin43]

Mar 26, 2020 12:45:41 GMT -5 newey said:

Wondering what you use to hold the shim in place, and whether whatever the substance is might also affect the freq response?

Thin double sided tape is all that is necessary. Magnetically neutral.

Arthur

[ms]

Is the 12 db limit a result the characteristics of the shim?  That is, does a larger or thicker shim make a difference?  How about using a small ring rather than a shim?

[aquin43]

Mar 27, 2020 7:20:03 GMT -5 ms said:

Is the 12 db limit a result the characteristics of the shim?  That is, does a larger or thicker shim make a difference?  How about using a small ring rather than a shim?

Good question. Using a much larger piece of material, the attenuation continues to increase with frequency to beyond 40kHz, with some suggestion of a very slightly increased slope beyond about 8kHz. Obviously what I was originally measuring included overspill from the exciter coil into the next pole. The coil is about as wide as the pole and very slightly shorter than the width of the pickup, while the shim is shaped to fit the end of the pickup cover and half way to the next pole.

The string will be interacting over a smaller area than the coil so the attenuation depth for the string signal will probably be deeper than 12 dB. The 12 dB measurement shelf occurs outside the usual guitar amp bandwidth, so the original measurement should be reasonably representative.

A ring would be a possibility, but it would be much harder to make it look pretty. The pickup I fitted the shim to has the first pole magnet reaching to just below the cover, so a neat installation was easy.

Arthur

[antigua]

I did a test with eddy currents and copper shields a few years ago, they're some of the earlier threads on this forum, and found that most of the attenuation caused by brass covers (which are mostly copper) owed to the geometric relation between the cover and the coils, not the string and the cover. For example, if you take a Telecaster neck pickup with a cheap brass cover, it will register a heavy attenuation, and it doesn't matter if you induce a magnetic field from the top of the pickup or the bottom, it's about the same either way. The technical explanation for this must be that the coupling coefficient between the coils and the brass cover are of a much higher value than is the coupling between the cover and the guitar string, or an excitation coil, as it were. The coupling between the coils and the wimpy guitar string is still sufficient for a usable voltage, because the coils are so physically large compared to everything else, the exciter coil, real guitar strings or the cover. But the magnetic relationship between the guitar string and the cover is going to be next to nothing. IIRC, the LtSpice created by Tele Tucson which modeled the eddy currents as transformers bore this out.

Based on this understanding, I don't think shading of this sort would work, but maybe your plot demonstrates something contrary to my understanding. How did you derive the plot?

The only way I know of to really impact the strings on an individual basis is not eddy currents, because that interaction is too weak, but instead, altering the magnetic coupling value, which can be done by raising and lowering pole pieces, or mixing different grades of AlNiCo, or even mixing steel and AlNiCo slugs, which I don't think I've ever seen anyone do. 

[aquin43]

Mar 27, 2020 22:18:23 GMT -5 antigua said:

I did a test with eddy currents and copper shields a few years ago, they're some of the earlier threads on this forum, and found that most of the attenuation caused by brass covers (which are mostly copper) owed to the geometric relation between the cover and the coils, not the string and the cover. For example, if you take a Telecaster neck pickup with a cheap brass cover, it will register a heavy attenuation, and it doesn't matter if you induce a magnetic field from the top of the pickup or the bottom, it's about the same either way. The technical explanation for this must be that the coupling coefficient between the coils and the brass cover are of a much higher value than is the coupling between the cover and the guitar string, or an excitation coil, as it were. The coupling between the coils and the wimpy guitar string is still sufficient for a usable voltage, because the coils are so physically large compared to everything else, the exciter coil, real guitar strings or the cover. But the magnetic relationship between the guitar string and the cover is going to be next to nothing. IIRC, the LtSpice created by Tele Tucson which modeled the eddy currents as transformers bore this out.

Based on this understanding, I don't think shading of this sort would work, but maybe your plot demonstrates something contrary to my understanding. How did you derive the plot?

The only way I know of to really impact the strings on an individual basis is not eddy currents, because that interaction is too weak, but instead, altering the magnetic coupling value, which can be done by raising and lowering pole pieces, or mixing different grades of AlNiCo, or even mixing steel and AlNiCo slugs, which I don't think I've ever seen anyone do. 

When you measure a pickup in a cover, the cover surrounding the pickup is strongly coupled to the coil.
This case is different and, as far as I can see, can only be modelled by the shielding parameters in my new model. The copper shading the pole is barely coupled to the coil but, being between the pole and the string, it shields the pole from the string in a frequency dependent fashion.

The TeleTucson model provided a valuable insight but it is incomplete in that alongside the direct coupling it only considers the parts of the path from string to coil that are coupled via the same eddy current losses that control the coil impedance.

The measurements were made using my digital frequency response measuring system which allows me to zero out a measured response so that subsequent responses are shown relative to it. Using an exciter coil of length about the same as the strat pickup width and about as wide as the pole, the response is measured and zeroed out without the shim and then the shim is introduced.

Arthur

[antigua]

Thanks for the pictures, it makes it a little more clear. So if you excite another pole piece other that the shielded one with the shield/shim in place, do you observe the same attenuation as it expected of an unshaded pole piece?

[aquin43]

Yes, only the shaded pole is filtered, the others have the full response of the pickup.

Arthur

[antigua]

Mar 28, 2020 13:20:40 GMT -5 aquin43 said:

Yes, only the shaded pole is filtered, the others have the full response of the pickup.

Arthur

I'll have to give this a try, because my earlier experiments suggested it wouldn't quite work out this way. 

[antigua]

I was wrong, I was able to get the same outcome with a little copper plate, about 1mm thick, like a square penny. The attenuation at resonance looks to be about 5.7bB







Also I tested a brass covered Tele neck pickup, the attenuation was less from the underside, but still attenuated more than if there were no cover at all: