Switchable additional winds

[Helmholtz]

You said further up that you tested foil tape around a single coil, and did not get as high of a capacitance as a was apparent with the tapped coil, but isn't a secondary winding of a tapped pickup extremely close to the primary coil? Even foil tape will contain a little more distance between the windings than will the continued winding of the secondary coil. Being able to model this in LTSpice and see the "apparent" high capacitance without actually assigning a high capacitance would be very informative.

My measurements with the foil (I have measured several PUs) show that it has no noticeable effect on the resonance of the active winding, as long as the foil is not grounded (or shorted). While the interwinding capacitance between the two winding parts will be somewhat higher than the capacitance between coil and foil (my guess: around 50pF, can be easily measured in a 4 wire PU; in a 3 wire PU it is shorted anyway), there will be no influence on the active coil's resonance as well, as it acts like a capacitor with one lead connected and the other lead up in the air. The active coil's resonance can only be influenced by capacitances that take effect to ground.
Please note that in the pdf article on coupled windings the interwinding capacitance is not even mentioned, even though there is some also in toroidal transformers with separated windings. The reason is that it has absolutely no effect as long as not more than one of the windings are grounded.
IMHO the increased effective capacitance across the active coil is not caused by the interwinding capacitance but by reflected capacitance.

I may assist with the LTSpice model but fear that this would mean hijacking this thread. We seem to have lost the OP a while ago.

[bbsailor]

I'm the last one to impart magical properties to shape. However, coil shape is a "marker" for other things that can be predicted to affect the electrical properties, as in Helmholtz' example of resonance. Certainly the properties of the magnet or polepieces (or slogs) play a significant role, however their role is not in spite of the coil properties, but in concert with them.

I am assuming that if I used normal Strat polepieces, and a conventional number of turns with very conventional wire and tension, but inserted a spacer between the polepieces and coil, such that the coil really only "started" 1/8" or 3/16" out from the polepieces (which, of course, would not fit inside the usual plastic pickup cover), it would likely end up sounding different, if only because of the distance of so many turns from the polepieces themselves. If a pickup has 800 additional turns that are switch-selectable, I would not expect that pickup to sound much different when those turns are added from the location closest to the polepieces vs the outside of the overall coil, simply because so few turns creates a negligible difference in distance from the polepieces.

But when one has a flatter pickup, like a Jazzmaster or even P90, it doesn't take as many turns to create distance from the polepieces (or center bar and slugs). And I would think that where perhaps it might not make that much difference if additional turns were situated on the inside or the outside of a Strat-type pickup, they just might for a flatter wider profile. Again, taking into consideration that there are more, and less, consequential numbers of additional turns.

Mark,
Do not forget to take into account the strength of the magnetic field on the inner turns versus the outer turns voltage generated. Using a metal plate to help spread the magnetic field under the coil is one way to make the outer winds more efficient voltage generating turns.

The ultimate test is to wind Jazzmaster or P90 type pickup with an inner coil with independent start an ending leads or connections and then add an outer coil with the same number of winds with independent start and ending leads. With no metal plate under the coils attach the inner and outer coils to a dual trace oscilloscope and visualize the output levels of each independent coil. Now add a metal plate to help spread the magnetic field and do the same thing to see the output levels of each coil change due the this added metal plate variable being observed.

Wider coils detect a wider string length window and will have a different harmonic component than a narrow coil with a shorter string length window due to harmonic cancellations of certain frequencies over the pickup coil area.

If you do this experiment you will see results and add to a scientific answer to this question.

Joseph J. Rogowski

[Helmholtz]

I'm the last one to impart magical properties to shape. However, coil shape is a "marker" for other things that can be predicted to affect the electrical properties, as in Helmholtz' example of resonance. Certainly the properties of the magnet or polepieces (or slogs) play a significant role, however their role is not in spite of the coil properties, but in concert with them.

I am assuming that if I used normal Strat polepieces, and a conventional number of turns with very conventional wire and tension, but inserted a spacer between the polepieces and coil, such that the coil really only "started" 1/8" or 3/16" out from the polepieces (which, of course, would not fit inside the usual plastic pickup cover), it would likely end up sounding different, if only because of the distance of so many turns from the polepieces themselves. If a pickup has 800 additional turns that are switch-selectable, I would not expect that pickup to sound much different when those turns are added from the location closest to the polepieces vs the outside of the overall coil, simply because so few turns creates a negligible difference in distance from the polepieces.

But when one has a flatter pickup, like a Jazzmaster or even P90, it doesn't take as many turns to create distance from the polepieces (or center bar and slugs). And I would think that where perhaps it might not make that much difference if additional turns were situated on the inside or the outside of a Strat-type pickup, they just might for a flatter wider profile. Again, taking into consideration that there are more, and less, consequential numbers of additional turns.

I don't want to question your findings regarding the coil to polepiece distance. But from the point of physics the only things that change with a wider "blind" core are an increased inductance because of greater mean circumference of the coil for a given number of turns, resulting in a lower resonant frequency and a slightly lower resonance peak because of the increased resistance. Altogether this will make the sound somewhat mellower. But there is another good reason to not let the wire touch the polepieces, i.e. prevention of chemical corrosion.
Again, different PU designs should be compared via samples having identical inductance/resonance (and not identical resistance!) to avoid that one only hears the sound difference caused by different resonance. Resonance/inductance of most designs can be easily adapted by the number of turns.

[Mark Hammer]

That's pretty much what I was suggesting, though perhaps in a manner that wasn't as clear as it needed to be. In essence, you have confirmed my initially query: that if one placed the additional turns closest to the polepieces - without any other modifications to compensate (as Joseph suggested) - there could be a change in the inductance and resonance of the default/basic coil, compared to if the basic coil was on the inside, by virtue any difference in the circumference of the turns in that basic coil.

Again, all of this is qualified by whether the number of turns making up the "additional" part yields any sort of appreciable difference to the circumference of the majority of turns on the basic coil. I say "majority of turns" because, even if the additional turns constituted 2000 out of 8000, and were situated on the inside rather than outside, many of the 6000 turns would be the same distance from the polepieces, and have the same circumference they would have had if the coil were only 6000 turns. Some, however, would be farther away from the polepieces than they might have been without those inner 2000.

Joseph's point about the influence of polepiece proximity suggests that placing additional turns on the inside mght result in a more noticeable difference in tone if one switch-selects those additional turns, compared to if the switch-selected addition was on the outside. Or have I misunderstood?

[Antigua]

I am well aware of all this and if the loudness step is what you are after, my proposal won't do.

My point is, that the unavoidable inductive coupling between the winding parts prevents a tappable overwound SC from convincingly reproducing the brighter sounds a a lower or "standard" wound PU, no matter how you split the coil or separate the windings, as the unused part of the coil always increases the effectice capacitance across the active coil by a considerable amount. As a result the two sounds available won't differ much - except when using very long guitar cables. But find out yourself.

I dunno, you know you stuff, I won't deny that, but I'm getting pretty close to winding a four conductor tapped single coil to see for myself. I only wind pickups for testing, so it's an ordeal to get everything set up.

In my experience the sound and dynamics of a good P-90 cannot be emulated by a Strat type design with alnico cores. The passive filter characteristics of a PU cannot explain the total response of a PU, even if the aperture is equal. But this would be a completely different topic.

I'm willing to entertain that topic. The AlNiCo poles have strong magnet pull, but a lower permeability. The lower permeability means the magnetic reluctance is higher, despite the stronger pull of the magnetic pole pieces. I don't believe magnetic reluctance really effects the tone, I don't see any plausible cause for that to be the case, but it is a fact that stronger magnetic pull upon the string interferes with how the string vibrates. I think the key to making the Strat pickup sound like a P-90 is, aside from matching the RLC characteristics, is to set it very low, so that the magnetic pull between the Strat pickup and strings are similarly weak. The overall output might be lower, but that should only effect the S/N ratio, and not the transfer function. A complication with testing this notion is that few Strats have P-90s, and few Gibsons have Strat pickups in them, so it's hard to compare apples to apples.

[Antigua]

Wider coils detect a wider string length window and will have a different harmonic component than a narrow coil with a shorter string length window due to harmonic cancellations of certain frequencies over the pickup coil area.

The coil can't be the window, because the coil doesn't determine what flux change occurs. The coil size and shape only matters for how efficiently it sees flux change that is cast off by the interaction between the pole piece and the guitar string. If a wide section of the string is magnetized, it will cancel higher harmonics as you said, but the coil doesn't determine that, it's merely a receives what has already been determined, with varying degrees of efficiency.

For example, with a coil as wide as the Jazzmaster pickup, it's not only seeing flux change from from the primary magnetic path between the guitar string and pole piece, but the wider loops of the coil are also receiving a flux change from the return path of that same magnetic field. Though the flux density of the return path is weaker than the primary path, the return path is opposite polarity, so it represents a cancellation of signal, and so the wide coil is just an inefficiency, and little more. The job of the coil is just to take the flux change and turn it into a voltage. The question is just how efficiently it does the job.

[Antigua]

That's pretty much what I was suggesting, though perhaps in a manner that wasn't as clear as it needed to be. In essence, you have confirmed my initially query: that if one placed the additional turns closest to the polepieces - without any other modifications to compensate (as Joseph suggested) - there could be a change in the inductance and resonance of the default/basic coil, compared to if the basic coil was on the inside, by virtue any difference in the circumference of the turns in that basic coil.

Again, all of this is qualified by whether the number of turns making up the "additional" part yields any sort of appreciable difference to the circumference of the majority of turns on the basic coil. I say "majority of turns" because, even if the additional turns constituted 2000 out of 8000, and were situated on the inside rather than outside, many of the 6000 turns would be the same distance from the polepieces, and have the same circumference they would have had if the coil were only 6000 turns. Some, however, would be farther away from the polepieces than they might have been without those inner 2000.

I just tested a tappable SSL-4 with the DE-5000, and got these values from the two halves of the pickups (Ls at 100Hz, Cp at 100kHz). Note that this is a three conductor tapped pickup, so the both coils are always connected at at least one end.

white to blk (whole coil) DCR: 13.62k Ls: 6.833H Cp: 179pF
white to red (inner coil) DCR: _6.57k Ls: 1.675H Cp: 384pF
black to red (outer coil) DCR: _7.05k Ls: 1.992H Cp: 310pF

As for how the inner verses outer coil geometry impacts the inductance, the outer coil has both a higher DC resistance and a higher inductance, but the inductance is not that much different, considering.

Regarding the capacitance, the DE-5000 is showing different values from what I calculated from the f and L using an oscilloscope, but it is still showing a higher capacitance for either individual half, compared to the whole.

[Mark Hammer]

Thanks for your diligence. The extra mile is much appreciated.

True, the differences are not dramatic. But then I suspect the differences existing between pickups X, Y, and Z, in those same parameters, that lead a player to prefer one over the others may be similarly "undramatic". Sometimes, what looks small on the measurement devices is a nuance that attracts a player. It may be small, but if they hear it (reliably), and it makes a difference to them, then I suppose we can say that it matters. Maybe more than it should, but it matters.

[Helmholtz]

The coil can't be the window, because the coil doesn't determine what flux change occurs. The coil size and shape only matters for how efficiently it sees flux change that is cast off by the interaction between the pole piece and the guitar string. If a wide section of the string is magnetized, it will cancel higher harmonics as you said, but the coil doesn't determine that, it's merely a receives what has already been determined, with varying degrees of efficiency.

This is correct. German electroacoustics professor Manfred Zollner could show by lab measurements that the string lenght window (aka aperture) only depends on the distribution and strength (to a lesser extend) of the PU's magnetic field. No influence of the coil's shape. The effective aperture length of a single coil is around 0.5''max. Humbuckers have two discrete windows, one over each coil. Humbuckers are "blind" right in the middle over the coils, because in this region the string saturates and its a.c. permeability drops to 1.
He also showed that the inner and especially the upper windings (closest to the strings) contribute most to the signal voltage, as the a.c. magnetic field is strongest in and around the upper part of the poles. The alternating part of the magnetic flux takes paths different from the permanent flux. The reason for this is that the permeabilities and consequently the magnetic conductivities of ferromagnetics (strings, pole pieces) differ vastly between a.c. and d.c. fields, a.c. permeabilities beeing generally much lower.

[Helmholtz]

Regarding the capacitance, the DE-5000 is showing different values from what I calculated from the f and L using an oscilloscope, but it is still showing a higher capacitance for either individual half, compared to the whole.

As mentioned earlier, calculating the capacitance from the resonance peak of a PU's impedance is not very precise.

1)There are additional parasitic capacitances from the setup: Wiring, input capacitances of scope/probe/meter etc.
2)The peak frequency of the impedance of a parallel resonant circuit is always somewhat lower than the theoretical resonant frequency given by 2*pi*fres = (L*C)^(- 0.5), depending on damping. If you use the peak of the transfer response instead of the impedance response, the error will be even greater.
3)The inductance drops with increasing frequency and is always lower at resonance than at 100Hz. This effect depends on core materials and eddy current paths. It is typically small for strat type PUs with alnico cores not having metal covers.

But - apart from these more academical considerations - I don't think an error of 10% or +/- 20pF really matters. A difference of more than 100pF might be audible, though.

(Did you try the "Automatic Voltage Scale" option? It will allow you to see the straight high frequency -12dB/octave slope in the transfer response (resp. -6dB/octave in the impedance response) which is masked by the noise floor in your plots.)

[Helmholtz]

The ultimate test is to wind Jazzmaster or P90 type pickup with an inner coil with independent start an ending leads or connections and then add an outer coil with the same number of winds with independent start and ending leads. With no metal plate under the coils attach the inner and outer coils to a dual trace oscilloscope and visualize the output levels of each independent coil. Now add a metal plate to help spread the magnetic field and do the same thing to see the output levels of each coil change due the this added metal plate variable being observed.

Can you post scope pictures?
I don't have tappable SCs, but I am curious because lab measurements (Zollner) have shown that the a.c. magnetic field at the bottom of a PU is only a very small percentage of the a.c. field at the top, which means that there is not much to be distributed. The baseplate of a tele bridge PU increases inductivity by 6% but the output is increased only by 0.6dB, which is close to nothing.

[Helmholtz]

I'm willing to entertain that topic. The AlNiCo poles have strong magnet pull, but a lower permeability. The lower permeability means the magnetic reluctance is higher, despite the stronger pull of the magnetic pole pieces. I don't believe magnetic reluctance really effects the tone, I don't see any plausible cause for that to be the case, but it is a fact that stronger magnetic pull upon the string interferes with how the string vibrates. I think the key to making the Strat pickup sound like a P-90 is, aside from matching the RLC characteristics, is to set it very low, so that the magnetic pull between the Strat pickup and strings are similarly weak. The overall output might be lower, but that should only effect the S/N ratio, and not the transfer function. A complication with testing this notion is that few Strats have P-90s, and few Gibsons have Strat pickups in them, so it's hard to compare apples to apples.

I agree, a true comparison is hard to do.
As an experienced player I notice major differences in the attack response between different core materials, even with equal B values at the strings and equal resonances. I think the clue is that the effective a.c. permeability of steel is not a constant but depends strongly on a.c. and d.c. field strenghts as well as on frequency. I have been doing measurements for some time that seem to support my theory. Don't want to go into more detail at the time.

[Antigua]

I agree, a true comparison is hard to do.
As an experienced player I notice major differences in the attack response between different core materials, even with equal B values at the strings and equal resonances. I think the clue is that the effective a.c. permeability of steel is not a constant but depends strongly on a.c. and d.c. field strenghts as well as on frequency. I have been doing measurements for some time that seem to support my theory. Don't want to go into more detail at the time.

Attack response would generally correspond to transient signals. You know a lot about transformer issues, is there any analogous consideration is transformer design, especially when you choose to use steel core versus ferrite? Modeling the guitar pickup and string as a loosely coupled transformer seems to draw an endless number of parallels otherwise.

[Antigua]

Can you post scope pictures?
I don't have tappable SCs, but I am curious because lab measurements (Zollner) have shown that the a.c. magnetic field at the bottom of a PU is only a very small percentage of the a.c. field at the top, which means that there is not much to be distributed. The baseplate of a tele bridge PU increases inductivity by 6% but the output is increased only by 0.6dB, which is close to nothing.

I've done similar tests with Tele base plates using the Vellemen and an inducer coil over head, and found the different with and without the base plate to be in the area of 1dB, give or take. One thing I did not try, whoch I suspect would make a difference, is if the pole pieces were to be replaced with steel instead of AlNiCo, there would be a lower magnetic reluctance between the strings and base plate, and so it might be realistic to see the output climb higher when the pole pieces are steel.

Regarding the magnetic path though, the metal base of the P-90 is either brass or nickel silver, which are not magnetic / permeable, so it does not have much effect to speak of on the magnetic circuit anyhow, and a Jazzmaster pickup of course uses fiberboard typical of Fenders.

[Antigua]

As mentioned earlier, calculating the capacitance from the resonance peak of a PU's impedance is not very precise.

1)There are additional parasitic capacitances from the setup: Wiring, input capacitances of scope/probe/meter etc.
2)The peak frequency of the impedance of a parallel resonant circuit is always somewhat lower than the theoretical resonant frequency given by 2*pi*fres = (L*C)^(- 0.5), depending on damping. If you use the peak of the transfer response instead of the impedance response, the error will be even greater.
3)The inductance drops with increasing frequency and is always lower at resonance than at 100Hz. This effect depends on core materials and eddy current paths. It is typically small for strat type PUs with alnico cores not having metal covers.

But - apart from these more academical considerations - I don't think an error of 10% or +/- 20pF really matters. A difference of more than 100pF might be audible, though.

(Did you try the "Automatic Voltage Scale" option? It will allow you to see the straight high frequency -12dB/octave slope in the transfer response (resp. -6dB/octave in the impedance response) which is masked by the noise floor in your plots.)

Regarding point #1, I determined the rig capacitance to be 10pF, so I subtract that from the final value.

I haven't set the Velleman up again yet, but I will try automatic scale when I do. I'd like to create a four lead tapped pickup this weekend, and do it all at once.