Notes on Magnetic Wire Insulation

[Tom Foy]

I've recently considered these three wire insulation materials:

Plain Enamel
Solderable Polyurethane (Electrisola P155, NEMA MW 79-C)
Solderable Polyurethane w/ Nylon Overcoat (Electrisola PN155, NEMA MW 80-C)

I don't want to use Plain Enamel, because it's too expensive, and I'm concerned about the durability of the insulation as it ages. I've chosen the Polyurethane over the Polyurethane w/ Nylon Overcoat, and here's why.

I've concluded, based on extensive mind-numbing internet crawling, that the difference in sound when using different wire insulation materials (using the same wire gauge; what happens when you change wire gauge is much less mysteriously discussed online) can be understood as differences in capacitance. I found this page on the factors affecting capacitance helpful: Factors affecting capacitance : Capacitors

None of these differences may be audible at all; I'm assuming here that they are.

So, scatter winding puts more space in the coil than machine winding, causing less capacitance, allowing more high frequency output. That's pretty much accepted in forums. So is the idea that a thicker wire insulation (Heavy Formvar, for example) allows more high frequency output than a thin insulation. These explanations agree with the what the allaboutcircuits page says about plate spacing in capacitors.

Each insulation material has its own dielectric constant (relative permittivity); a higher value means higher capacitance, and reduced high frequency output. Here are some relevant links:
Relative permittivity - Wikipedia, the free encyclopedia
http://202.194.4.88:8080/wulihx/data...nt%20Table.htm

I haven't found the dielectric constant for Plain Enamel (oleoresinous enamel), but you can sort of guess at it by looking at some of the oils and stuff in the Dielectric Constant table. You can look up the definition of oleoresin and resin - that should give some clues about Plain Enamel's composition. I've seen some online talk suggest it's similar to spar varnish, which includes boiled linseed oil. Who knows? Anyway, it seems that the dielectric constant would be pretty low, like between 2 and 4.

The dielectric properties of polyurethane and polyurethane with nylon overcoat seem close, until you consider that the nylon overcoat is hygroscopic, so it's relative permittivity will increase as it absorbs water.
Electronics Reference: Wire Chart
Litz Wire - Litz Film Insulations

I don't intend to pot my pickups, so it seems moisture in the air will cause my pickups to slowly lose treble response if I use the poly/nylon, so I'll use polyurethane. I have seen polyurethane referred to as a hygroscopic material, online, but not by any wire companies. I don't know why that is; it could be that the polyurethane formula used for wire is not hygroscopic.

[Mike Sulzer]

I've recently considered these three wire insulation materials:



I've concluded, based on extensive mind-numbing internet crawling, that the difference in sound when using different wire insulation materials (using the same wire gauge; what happens when you change wire gauge is much less mysteriously discussed online) can be understood as differences in capacitance. I found this page on the factors affecting capacitance helpful: Factors affecting capacitance : Capacitors

Sure but how big an effect can it be? Cable capacitance is quite a bit bigger than pickup capacitance, and so you really cannot change the total capacitance very much.

[Tom Foy]

Considering we're talking about something like 8000 windings, I can see how capacitance could make a difference, but I really have no idea whether it makes any difference in practice. I've been gathering information before winding my first pickups, so I don't have anything empirical to point to.

[David King]

The reason cable capacitance is so much greater even with the relatively short lengths of wire involved is that a cable looks a lot more like a capacitor than a pickup does.

[ScottA]

I can't back this up with hard data (yet), but I believe there is a fundamental difference in losses that exist in the pickup itself and are then, in effect, incorporated into the tone as the signal is being generated, and losses that are introduced downstream.

At least, I think we could agree that the pickup as a sensor should have as pure a response as possible, and capacitive losses should be minimized.

[Mike Sulzer]

The thing that is different about losses in the pickup, those related to its capacitance, is that the pickup really does not have a capacitance, but rather an effective capacitance. You can think of it as kind of like a huge number of really small capacitors connected together in a complicated circuit. At audio frequencies the net effect is very very much like a single capacitance, but not exactly the same. This difference is a tiny effect in a circuit with other larger effects, and I doubt that it matters at all in a practical sense.

I can't back this up with hard data (yet), but I believe there is a fundamental difference in losses that exist in the pickup itself and are then, in effect, incorporated into the tone as the signal is being generated, and losses that are introduced downstream.

At least, I think we could agree that the pickup as a sensor should have as pure a response as possible, and capacitive losses should be minimized.

[bbsailor]

The thing that is different about losses in the pickup, those related to its capacitance, is that the pickup really does not have a capacitance, but rather an effective capacitance. You can think of it as kind of like a huge number of really small capacitors connected together in a complicated circuit. At audio frequencies the net effect is very very much like a single capacitance, but not exactly the same. This difference is a tiny effect in a circuit with other larger effects, and I doubt that it matters at all in a practical sense.

Mike,

The net effect of having a pickup with a Q of from 2 to 5 at resonance is to have a very high impedance that is affected by the resistive load of the pot value in parallel with the typical 1M ohm amp input impedance for an effective load of about 400K for guitars using 500K volume and tone pots. Placing a buffer in the guitar end of a guitar plug or building it into the guitar powered by a 9V battery effectivly elimates the effect of the capacitance added by the guitar cable and the 1M input impedance loading, assuming you make the input impedance of the FET buffer in the 3 to 5 meg range.
JFET Buffer and buffer cable

When you use the FET buffer in the guitar end of the plug you can hear a definate lift to the high frequencies compared to using a passive guitar cable. If you build the buffer right after the pickup but before the volume pot and tone pot, the buffering effect is even more as you are eliminating or minimizing volume and tone pot loading also. It is like having a tone pot that can extend the normal tone range up to 12 or 13. Next comes using you ear to decide what sounds better. You can always turn down the tone pot a bit but the ear is the final decider.

Every pickup experimenter should have one of these buffer cables in their tool box to listen to the raw output of a pick unaffected by cable capacitance. Pickup winding capacitance alone can range from 30pf up to 120pf and eliminating the 300 +pf of the guitar coax cable will change the pickup voicing by raising the resonance frequency.

Joseph Rogowski