Is it possible to accurately describe the tone of magnets and/or pickups?

[Antigua]

there is some truth to what he is saying - you can't fool my ears......

Yeah there is some truth, but then you have the remainder, which is not truth.

[Antigua]

Its not possible to describe the tonality of any alnico magnets. Why? Because no two alnico manufacturers use the same formulas. So you think alnico 2 is warmer, less trebly? Nope. I have different company A2 magnets that are super bright and really high gauss readings when hot charged. Its the same for all the alnicos, A4 from one company won't sound like A4 from another company.

I've tested both the magnetic and electrical properties of a few dozen AlNiCo 2 magnets that came with pickups sourced from all over the place, I've never seen much variation. There is room for creativity in the production of AlNiCo, but they also adhere to specs that ensure that they will not be exceedingly different, either. Otherwise it must be said that the magnets are poorly made. For example, AlNiCo 2 should not have "really high gauss" like that of AlNiCo 5, and if it does, there is something wrong with that magnetic alloy, and it should be sent back for a refund.

Alnico is mostly IRON.

It's not as simple as the composition, the structure of the alloy plays a big role. That's why electrical steel has a permeability that's many times higher than that of AlNiCo, even though both are primary composed of iron.

And iron sucks up treble, so the more massive a magnet is, the less brightish it will be. Some TTop magnets were super thin and short, those things are really shrill.

To say this is an oversimplification would be an understatement. All of these same effects manifest in inductors and transfomers, where they are categorized as wanted or unwanted inductance, core losses and capacitance, and so while there might be limited resources with respect to pickups specifically, reading up on inductor and transformer design reveals nearly all there is to know about the electrical side of pickups, too. Treble is lost to inductance, capacitance and conductivity, due to eddy currents. The fact that iron is involved is almost incidental, for example, you mention a pickup with a thin AlNiCo magnet was shrill, in your subjective estimation, but were you to add a few thousand more turns of wire, or put a high value capacitor across that pickup, these factors could also reduce the treble content to such an extent that it could not be described as "shrill". You mention a large magnet subtracting from the highs, but as an example, testing shows that a Filter'tron's treble attenuation is caused, almost exclusively, by the twelve fillister screws. But of course Filter'trons tend to be bright pickups none the less, and that owes to its very low inductance.

[Tone Cam]

Its not possible to describe the tonality of any alnico magnets. Why? Because no two alnico manufacturers use the same formulas. So you think alnico 2 is warmer, less trebly? Nope. I have different company A2 magnets that are super bright and really high gauss readings when hot charged. Its the same for all the alnicos, A4 from one company won't sound like A4 from another company.

I don't understand why it is not possible to accurately describe the tone of materials from different vendors. If you take a scale with ‘super bright’ on one end and 'really dark' on the other, add additional variables that talk about the smoothness or harshness of the high frequencies, the overall harmonic density (‘fat’ or ‘thin’), the relative magnitude of other frequency bands and the overall loss (sharp vs dull tone??) I think that you would have a description that could be used to accurately differentiate the Super Bright Al2 material from Al2 samples that were cast at other foundries.

I agree with the later statement (Post #117) that impedance measurements are poor predictors of pickup tone. A magnetic pickup and the ferromagnetic strings of a musical instrument in which it installed can be accurately modeled as a magnetic circuit with a flux that varies in response to string vibration. This formalism was used, for example, by Lemarquand in “Calculation Method of Permanent-Magnet Pickups for Electric Guitars,” (IEEE Transactions of Magnetics, Vol. 43, No 9, pp. 3573-3578, 2007) to describe the harmonics that are produced by the string as it moves in the asymmetric field generated by the poles of single coil pickup. The flux variations in the permanent magnets that are part of a magnetic circuit are described by minor hysteresis loops that are centered along recoil permeability lines. (See, for example, pages 483-484, Introduction to Magnetic Materials, 2nd Ed., B.D. Cullity and C.D. Graham, Wiley, Hoboken, 2009). The area of a minor hysteresis loop is much smaller than the area of the major loop that describes the magnetization (and demagnetization) of a material but the nonlinear relationship between B and H around a minor loop generates harmonics. These harmonics vary significantly with the small scale structural properties of a material and contribute significantly to the tonal differences that are observed in samples of the same alloy from different vendors.

Impedance and other parameters that have been put forth to explain alloy-dependent tonal variations cannot explain the fact that I can significantly alter the tone of a pickup by attaching components to it that do not affect its impedance or alter the spatial distribution/strength of the magnetic field at the strings. The recoil permeabilities of my insulator-bound alnico powder materials (see US Patent #8,853,517) are near unity and their eddy current losses are much lower than bulk samples of the materials from which they are made. They are, however, quite effective in changing the tone of a pickup in both magnetized and unmagnetized conditions.

Mike is correct in stating (Post #119) that the reluctance of the magnetic circuit that is formed by the pickup and strings is dominated by large air gaps but incorrect in stating that the small gaps in the circuit do not affect tone. I can hear the effect of the small gaps that are created when I glue two pieces of material together to form a composite pole piece (see US Patent # 8,415,551) and sometimes use the surface quality of the mating surfaces to adjust tone. The output tone of a pickup is a very sensitive meter that responds to small parameter variations that have little/no effect in other applications. That is the reason that trial and error is the only effective way to optimize the tone of a pickup. Possum’s technical explanations (Post #115, for example) are less than accurate but the multiyear trial and error optimization of his PAF pickups appears to have yielded some pretty good results. (See review of Stephen’s Design PAF at https://www.guitarplayer.com/gear/fi...reviewed-video).

[Mike Sulzer]

"...incorrect in stating that the small gaps in the circuit do not affect tone."

Obviously there are certain cases in which contact matters. For example, if you space the magnet from the pole piece, the latter is less magnetized, and that certainly has an effect. But that is nothing like what Possum was saying.

[Antigua]

The area of a minor hysteresis loop is much smaller than the area of the major loop that describes the magnetization (and demagnetization) of a material but the nonlinear relationship between B and H around a minor loop generates harmonics. These harmonics vary significantly with the small scale structural properties of a material and contribute significantly to the tonal differences that are observed in samples of the same alloy from different vendors.

These are just assertions. I'm not sure how those small hysteresis loops are supposed to produce harmonics, and even if they do, how loud are they? Could you even hear them? If everything were audible, you'd probably hear the sound of magnetic domains re-orienting too, but in the end, only the most prominent string movements manifest in something that we hear.

Impedance and other parameters that have been put forth to explain alloy-dependent tonal variations cannot explain the fact that I can significantly alter the tone of a pickup by attaching components to it that do not affect its impedance or alter the spatial distribution/strength of the magnetic field at the strings. The recoil permeabilities of my insulator-bound alnico powder materials (see US Patent #8,853,517) are near unity and their eddy current losses are much lower than bulk samples of the materials from which they are made. They are, however, quite effective in changing the tone of a pickup in both magnetized and unmagnetized conditions.

Mike is correct in stating (Post #119) that the reluctance of the magnetic circuit that is formed by the pickup and strings is dominated by large air gaps but incorrect in stating that the small gaps in the circuit do not affect tone. I can hear the effect of the small gaps that are created when I glue two pieces of material together to form a composite pole piece (see US Patent # 8,415,551) and sometimes use the surface quality of the mating surfaces to adjust tone.

Without knowing what you're talking about, I still do know that claiming you can alter the tone does not mean you actually alter the tone. You're mixing objective claims and subjective outcomes. Not only should an outcome be quantifiable, but the quantity itself should be known, again, to figure out whether or not we might actually hear it.

That is the reason that trial and error is the only effective way to optimize the tone of a pickup. Possum’s technical explanations (Post #115, for example) are less than accurate but the multiyear trial and error optimization of his PAF pickups appears to have yielded some pretty good results. (See review of Stephen’s Design PAF at https://www.guitarplayer.com/gear/fi...reviewed-video).

The reason why trial and error is necessary is because ultimately it's an aesthetic decision. No musician is ever going to say "I hate the sound, but he mathematical formula says this is the sound I must use." It's the same reason you use a swatch to choose a color of paint instead of a calculator.

[Mike Sulzer]

In reference to the bolded statement: Impedance measurements characterize the pickup-cable resonance, and this is without doubt the most important factor in determining the tone of the pickup. They can even describe the deviations from the two pole resonance of a simple linear circuit, that is, the more subtle effects introduced by the frequency dependent losses due to eddy currents.

I have not been able to get access to the full text of the paper you cite, but the abstract indicates that it presents an analysis using the law of magnetic induction and calculates in detail that the response is nonlinear. The well known causes of this nonlinearity are the nonlinear part of the variation of the flux through the coil resulting from the change in distance of the string, the frequency doubling resulting from string motion parallel to the face of the pole piece, and the variation of the strength of the permanent field with distance from the magnet. I do not see any indication that hysteresis enters into this. The "minor loops" that you describe are extremely close to straight lines, and so I do not see how you get an effect from them.

Certainly materials such as you describe with low conductivity and permeability make a pickup with a higher resonance frequency and Q, and I think this would be very similar to using neo magnets with no additional pole piece. Neos are of course available in a vast range of sizes and shapes, and so the field strength variations and spatial variations due to magnet shape are easily produced. However, you seem to be describing changes other than those I mentioned, and I do not see what they can be.

I agree with the later statement (Post #117) that impedance measurements are poor predictors of pickup tone. A magnetic pickup and the ferromagnetic strings of a musical instrument in which it installed can be accurately modeled as a magnetic circuit with a flux that varies in response to string vibration. This formalism was used, for example, by Lemarquand in “Calculation Method of Permanent-Magnet Pickups for Electric Guitars,” (IEEE Transactions of Magnetics, Vol. 43, No 9, pp. 3573-3578, 2007) to describe the harmonics that are produced by the string as it moves in the asymmetric field generated by the poles of single coil pickup. The flux variations in the permanent magnets that are part of a magnetic circuit are described by minor hysteresis loops that are centered along recoil permeability lines. (See, for example, pages 483-484, Introduction to Magnetic Materials, 2nd Ed., B.D. Cullity and C.D. Graham, Wiley, Hoboken, 2009). The area of a minor hysteresis loop is much smaller than the area of the major loop that describes the magnetization (and demagnetization) of a material but the nonlinear relationship between B and H around a minor loop generates harmonics. These harmonics vary significantly with the small scale structural properties of a material and contribute significantly to the tonal differences that are observed in samples of the same alloy from different vendors.

Impedance and other parameters that have been put forth to explain alloy-dependent tonal variations cannot explain the fact that I can significantly alter the tone of a pickup by attaching components to it that do not affect its impedance or alter the spatial distribution/strength of the magnetic field at the strings. The recoil permeabilities of my insulator-bound alnico powder materials (see US Patent #8,853,517) are near unity and their eddy current losses are much lower than bulk samples of the materials from which they are made. They are, however, quite effective in changing the tone of a pickup in both magnetized and unmagnetized conditions.

[Tone Cam]

In reference to the bolded statement: Impedance measurements characterize the pickup-cable resonance, and this is without doubt the most important factor in determining the tone of the pickup. They can even describe the deviations from the two pole resonance of a simple linear circuit, that is, the more subtle effects introduced by the frequency dependent losses due to eddy currents.

Impedance measurements are easy to perform but provide limited info on some of the physical mechanisms that are important contributors to pickup tone. I don’t see how an impedance measurement is going to provide much info on the nonlinear core and wire losses that are significant sources of harmonic generation in a pickup.

Impedance effects cannot explain the significant tone-modifying effects of a ¼ inch wide x 1/32 inch thick, unmagnetized strip of insulator bound alnico granules that is taped to the outside surface of a humbucker coil. In this position it adds hysteresis loss to the magnetic circuit that is formed by the pickup and strings but does little else.

Impedance measurements also provide very little info on the changes in pickup tone that are associated with variations in wire tension, insulation type and scatter in the coils.

I think it is an oversimplification to claim that the pickup-cable resonance is the 'most important factor' in determining pickup tone. Pickup cable resonance does not generate harmonics - it filters them. Nonlinear loss processes, including hysteresis, eddy currents, anomolous ferromagnetic losses and proximity losses in the coils, are harmonic generators. The time dependent spectrum of the signal that gets to an amp is determined by the harmonics that are generated by the pickup and the filter characteristics of the circuit that includes the pickup, tone circuit, and cable.

[bbsailor]

In reference to the bolded statement: Impedance measurements characterize the pickup-cable resonance, and this is without doubt the most important factor in determining the tone of the pickup. They can even describe the deviations from the two pole resonance of a simple linear circuit, that is, the more subtle effects introduced by the frequency dependent losses due to eddy currents.

I have not been able to get access to the full text of the paper you cite, but the abstract indicates that it presents an analysis using the law of magnetic induction and calculates in detail that the response is nonlinear. The well known causes of this nonlinearity are the nonlinear part of the variation of the flux through the coil resulting from the change in distance of the string, the frequency doubling resulting from string motion parallel to the face of the pole piece, and the variation of the strength of the permanent field with distance from the magnet. I do not see any indication that hysteresis enters into this. The "minor loops" that you describe are extremely close to straight lines, and so I do not see how you get an effect from them.

Certainly materials such as you describe with low conductivity and permeability make a pickup with a higher resonance frequency and Q, and I think this would be very similar to using neo magnets with no additional pole piece. Neos are of course available in a vast range of sizes and shapes, and so the field strength variations and spatial variations due to magnet shape are easily produced. However, you seem to be describing changes other than those I mentioned, and I do not see what they can be.

Mikes comments are all valid but one of his statements gets very little attention in these discussions "the frequency doubling resulting from string motion parallel to the face of the pole piece". Horizontal string motion relative the the pole piece face and coil face plane induces more second harmonics than a more vertical motion which emphasizes the fundamental frequency. This effect can be highly subjective as it typically only occurs in the first 30 to 50 milliseconds after the string pick or strum. Instruments having pickups with two pole pieces per string emphasize this effect even more.

To add even more variables, look at the plectrum thickness; picking/strumming location relative to the active pickup(s); environment where we are playing; and developed picking/strumming techniques for its effect on the perceived sound. The sound of the pickup is an accumulative effect with some commonly understood effects of winding capacitance, guitar chord capacitance, resonance loading, eddy currents near pickup coil, pickup inductance and string window width being induced into the pickup coil(s).

Discussing any one of these in isolation does not do justice to the collective interaction of all of these to the perceived outcome.

Joseph J. Rogowski

[Helmholtz]

I don’t see how an impedance measurement is going to provide much info on the nonlinear core and wire losses that are significant sources of harmonic generation in a pickup.

While I generally agree that nonlinear effects (and consequent cross-modulation of frequency components) in PUs tend to be underrated, I wonder what you mean by "nonlinear wire losses"?

[Helmholtz]

Mikes comments are all valid but one of his statements gets very little attention in these discussions "the frequency doubling resulting from string motion parallel to the face of the pole piece". Horizontal string motion relative the the pole piece face and coil face plane induces more second harmonics than a more vertical motion which emphasizes the fundamental frequency.

The vertical string motion alone produces significant second harmonic content (around 25% according to Zollner), as the vertical magnetic field above the PU is not homogeneous but decreases approximately according to 1/d², d being the distance between string and polepiece.

[bbsailor]

The vertical string motion alone produces significant second harmonic content (around 25% according to Zollner), as the vertical magnetic field above the PU is not homogeneous but decreases approximately according to 1/d², d being the distance between string and polepiece.

Yes, and this is why the observed oscilloscope image of a guitar string pluck is asymmetrical because as the string moves down toward toward the pickup magnet, it produces a higher output than the upper movement farther away from the pickup magnet with a lower output.

Joseph J. Rogowski

[eschertron]

An immobile fixed magnet cannot produce a damping effect. It just acts like an ideal spring attached that absorbs and releases the same amounts of energy. This follows from the law of conservation of energy. Things change when eddy current effects are involved.
(I do have a scientific background.)

Could position of the magnet along the vibrating string length (other than center that is) create a dissonant, and therefor damping vibrations along the strings length? I'm thinking of a problem that is typically called Stratitis. Said to be caused by excessive magnetism due to pickups being adjusted too close to the strings. It causes odd harmonics and damping of sustain is usually reported.

I know this exchange was from a long way back, but I've been lurking on this thread since the beginning, and have been trying to follow along. So bear with me. I just stumbled across this youtube video Copper's Surprising Reaction to Strong Magnets | Force Field Motion Dampening and wondered if this played a part in some of the observations we've made. This is what I'm thinking:

At one point in the video there's a demonstration that a conductor moving relative to a magnetic field produces an electric current (whew! glad that's still true). The surprising twist to me was that if the current is not drawing energy out of the interaction, ie, if the coil is shunted, there is a "back magnetic force" (magnetic impedance? please let me know the actual term) that resists the relative motion. In this video, about 2:24 in, the magnetic object slows down its relative motion when the coil generates current. My takeaway from this is that since the pickup magnet->string->coil->tone/volume controls circuit is not 100% efficient at transducing the energy, there is some resistance to magnetized string motion by the coil. Is this something that's just painfully obvious to everyone but me?

edit: after some thinking, I know the shunted winding thing is true of other magnetic drivers such as speakers and permanent magnet motors, so I guess I shouldn't be too surprised to see it here. I guess I never thought that it may affect string mechanics.

[Helmholtz]

I know this exchange was from a long way back, but I've been lurking on this thread since the beginning, and have been trying to follow along. So bear with me. I just stumbled across this youtube video Copper's Surprising Reaction to Strong Magnets | Force Field Motion Dampening and wondered if this played a part in some of the observations we've made. This is what I'm thinking:

At one point in the video there's a demonstration that a conductor moving relative to a magnetic field produces an electric current (whew! glad that's still true). The surprising twist to me was that if the current is not drawing energy out of the interaction, ie, if the coil is shunted, there is a "back magnetic force" (magnetic impedance? please let me know the actual term) that resists the relative motion. In this video, about 2:24 in, the magnetic object slows down its relative motion when the coil generates current. My takeaway from this is that since the pickup magnet->string->coil->tone/volume controls circuit is not 100% efficient at transducing the energy, there is some resistance to magnetized string motion by the coil. Is this something that's just painfully obvious to everyone but me?

edit: after some thinking, I know the shunted winding thing is true of other magnetic drivers such as speakers and permanent magnet motors, so I guess I shouldn't be too surprised to see it here. I guess I never thought that it may affect string mechanics.

The principle described is called the "Eddy current brake". You always have both, the "back magnetic force" as well as energy absorption in the conductor. In PUs the load is usually high impedance, not allowing for noteworthy current. So the string motion damping effect is practically non-existent. But e.g. Alumitone PUs have a closed, very low resistance massive aluminum loop close to the strings, so some string motion damping effect is conceivable. But don't forget that the magnetic fields involved with PUs are very small.

[bbsailor]

The principle described is called the "Eddy current brake". You always have both, the "back magnetic force" as well as energy absorption in the conductor. In PUs the load is usually high impedance, not allowing for noteworthy current. So the string motion damping effect is practically non-existent. But e.g. Alumitone PUs have a closed, very low resistance massive aluminum loop close to the strings, so some string motion damping effect is conceivable. But don't forget that the magnetic fields involved with PUs are very small.

Once you understand how the Alumitone pickup works, you may see how some trade-offs between what you said about "eddy currents" in the string loop directly below the strings and the lower DC resistance in the two coils below the frame affect the tonality of the audio output. The metal frame is the very low impedance single turn string loop with the transformer lamination intersecting the metal frame in the location where the currents from both coils are common (adding). Thus, the induced current in this aluminum frame are primary current of a current transformer with two approximately 15000 turn coils wired in parallel for an effective turns ratio of about 1 to 7500. This is about the amount of turns on a conventional high impedance guitar pickup but with one big exception.

This exception is that guitar pickups have a single turn loop length that is twice the string width plus the radius of the end turn. If a pickup has a 2 inch string width, the straight length of the wire is 4 inches. The end turns are the diameter of the magnet core times 3.14159 to account end turn length. The two coils under the Alumitone frame are round and thus do not need the straight wire length so they are 4 to 5 time shorter with the same amount turns but induced by the current in the frame converted by the transformer action of the two parallel wired coils under the frame.

This Alumitone design now allows the tone of the pickup to be determined by the relationship of the very low resistance primary string loop relative to the amount of the turns of the high turns secondary coils to form the effective turns ratio of this transformer based pickup (much like a low level current transformer). Search "low impedance Pickup research" on this forum to see more details.

This is a prime example where trade-offs in design by minimizing the DC resistance of the secondary coils wire; the magnetic coupling of the metal frame to the two secondary coils; and the magnetic coupling to the strings creates a new set of variables to consider when analyzing pickup response and quality control for manufacturing many products with consistent results. Since the primary string loop is directly under the strings generating the current in the transformer primary, the output for this transformer action is much more efficient than a traditional high impedance pickup where the voltage induced in the pickup turns is more in the upper coil turn location and less in the lower coil turn location along with about one fourth to one fifth less DC resistance.

The bottom line is: today's high impedance pickups represent a 1930's technology that has not evolved much until more modern active pickups and Alumitone transformer-based type pickups have become available. It is now time to seek to better understand these new designs and relate their understanding to what we still build today. Magnets and induction are still involved but the evolutionary changes requires a more comprehensive understanding of the interaction of all the variables to the sound we characterize as "good".

Joseph J. Rogowski

[Mike Sulzer]

(1) There are no significant nonlinear wire losses. Although the core does cause frequency dependent losses, there is no significant nonlineaerlity from the coil. The changes in field are tiny and hysteresis is not an issue; this is why you need so many turns on the coil.

(2) You have claimed to have made a material with low conductivity and permeability. Therefore it interacts very little with magnetic fields. You are claiming that very small changing feeds cause significant hysteresis effects in a material with little interaction with those fields. That does not happen.

(3) These things affect the inductance and resistance.

(4) The ear-brain system does not care how the ratios of harmonics in the signal were generated, and the filtering effect of the resonance has a large effect on the levels of harmonics in and/or near to the frequency range where this system has the greatest sensitivity.

Impedance measurements are easy to perform but provide limited info on some of the physical mechanisms that are important contributors to pickup tone. I don’t see how an impedance measurement is going to provide much info on the (1) nonlinear core and wire losses that are significant sources of harmonic generation in a pickup.

Impedance effects cannot explain the significant (2) tone-modifying effects of a ¼ inch wide x 1/32 inch thick, unmagnetized strip of insulator bound alnico granules that is taped to the outside surface of a humbucker coil. In this position it adds hysteresis loss to the magnetic circuit that is formed by the pickup and strings but does little else.

(3) Impedance measurements also provide very little info on the changes in pickup tone that are associated with variations in wire tension, insulation type and scatter in the coils.

I think it is an oversimplification to claim that the pickup-cable resonance is the 'most important factor' in determining pickup tone. (4) Pickup cable resonance does not generate harmonics - it filters them. Nonlinear loss processes, including hysteresis, eddy currents, anomolous ferromagnetic losses and proximity losses in the coils, are harmonic generators. The time dependent spectrum of the signal that gets to an amp is determined by the harmonics that are generated by the pickup and the filter characteristics of the circuit that includes the pickup, tone circuit, and cable.