Guitar Pickup Analysis Web Link

[salvarsan]

A few years back, some of us created FEMM magnetic models for pickup magnet geometries.
Steve Kersting saved a few of them.

http://www.skguitar.com/SKGS/sk/Imag.../Magnetics.htm

-drh

[Mal88]

Yes, you are right, there is a discernable difference in level as the string is pulled laterally over the magnet poles! Now that is interesting because this implies that the level change is about 3 dB; which implies that the average effect on the magnetic field potential changes by as much as 1.414 times - but the magnetic modelling does not show that.

There is definitely a ripple in the magnetic field strength near the adjacent poles, so it could well follow that the energy transferred from the relative instanteanous movement of the string as it is laterally dragged over/between the poles, sits over a 'saucer' of the magnetic potential between the poles, and consequently vertical and horizontal movements are somewhat self-cancelling; thus causing the discernable drop in induced signal level.

My modelling to date shows that magnetic 'rails' would not suffer this 'ripple/fade' fate and that although magnetic rods look asthetically pleasing, rods/screws may be a rather poor engineering choice!

[David Schwab]

My modelling to date shows that magnetic 'rails' would not suffer this 'ripple/fade' fate and that although magnetic rods look asthetically pleasing, rods/screws may be a rather poor engineering choice!

We are used to the look of rod poles, but if pickups never had them we wouldn't care!

I'm convinced that blades are a better way to go.

[Mal88]

As your political combatants keep saying - "Lipstick"!!!

[Mal88]

Excellent work there! I did the same totally unaware that this had been worked on, and Steve had gone the one step further with the plot across the tops of the pickups (which I alluded to but never seriously followed up)!

Remember that this is a 'slice' of the magnetic field right through the centre of the magnetic structure and it appears bigger on the 'sides' - but - if you took several 'slices' (parallel to this one and normal to the strings) I believe that the overall magnetic field is 'fatter/wider' in the middle, but you are not seeing this in this case (because you are rignt in the centre); and like Bragg diffraction, the fatter/wider field intensity will considerably cancel these magnetic intensity ripples seen on this plane.

[salvarsan]

caveats

The other thing we noticed is that a pickup has a magnetic aperture
with a width and strength, a quality factor perhaps.
P90's and PAF's have a wide aperture, Fender SC types are more focussed.
Dual blade pickups have the best-focussed field of any of them.

The other thing the simulations let us do is estimate how much steel to put
in the magnet path if using rare earth magnets.

BTW, do _not_ believe anything FEMM has to say about Alnico 8.
It lists permeability as ~6 when it should be closer to 1.1.

-drh

[Mal88]

Yes I agree with the wider sensitive pickup area for the PAF and I measured that against the standard Strats. But the Strats do have a fairly consistent pickup width that is somewhat wider than the poles.

Remember that the FEMM analysis is an infinitely thin slice of 2D right along the centre axis and that is rather like a 'null' point - so you will get 'misleading' results if you don't think beyond the null situation. Case in point when you model a bar or cylinder magnet - you will see that the vast majority of the graduating magnetic potential is close to the side surfaces (because the magnetic lines oppose each other and there is nowhere else to go).

Unfortunately I have not been able to get my hands on a working P90, but I have seen the structure of the magnets / poles and that is a real worry. I have heard people say they are noisy, and the reasoning is there because the magnetic field is everywhere but via the strings - and that is why they are noisy!

The difference between the PAF and the Strat is that the PAF has a slightly improved magnetic circuit - but it really has a very long way to go to get it to be anywhere towards 'good'. It (PAF) really is not a hum bucker as the second coil does not 'buck hum' it merely partially closes some of the magnetic circuit nearer the strings. If this was really to buck hum, then it would have a very much smaller third winding under the mounting plate to cancel far-field excitations - but it can't - because the magnetic is so leaky and the common magnet is ill positioned (and there is virtually nil magnetic shielding)!

[David Schwab]

The difference between the PAF and the Strat is that the PAF has a slightly improved magnetic circuit - but it really has a very long way to go to get it to be anywhere towards 'good'. It (PAF) really is not a hum bucker as the second coil does not 'buck hum' it merely partially closes some of the magnetic circuit nearer the strings. If this was really to buck hum, then it would have a very much smaller third winding under the mounting plate to cancel far-field excitations - but it can't - because the magnetic is so leaky and the common magnet is ill positioned (and there is virtually nil magnetic shielding)!

Yeah, but for the hum they intended on bucking, the 60 (120) cycle stuff, the humbucker works, as the magnets are not needed to pick up hum, or buck it. It's the out-of-phase coils that do the trick. It doesn't mater which coil is bucking the hum... they both pickup the strings and the nose. The magnets are only needed to pick up the string's vibrations. Audio inductors are often wound as humbuckers and have no magnets at all.

The higher frequency electrostatic noise slips in between the coils as it were, so you need good shielding for that... something guitar humbuckers generally never have, unless they are covered.

Some of the newer stacked pickup designs are interesting as they use the top coil only to pick up the strings, and the bottom coil to induce noise into the system. The bottom coil is often wound less and has higher inductance due to more metal included.

See patent #7166793 by Kevin Beller with Duncan. It's similar to the Kinman and DiMarzio designs, but has a few twists.

[SK66]

Remember that this is a 'slice' of the magnetic field right through the centre of the magnetic structure and it appears bigger on the 'sides' - but - if you took several 'slices' (parallel to this one and normal to the strings) I believe that the overall magnetic field is 'fatter/wider' in the middle,.

Yes, the pickup's aperature is a bit "wider" in the middle. Also note that some of that extra strength from the end poles is cast off into useless space (off the end of the pickup).
The one thing that I find interesting is that magnet "strength" doesn't seem to be a huge factor in a pickups "volume"....

[Mal88]

Yes I believe that you are right in that the magnet strength is not a major factor in sensitivity (but don't tell the musos)! The sensitivity is really locked up in the incremental change in overall inductance caused by the relative movement of the string. Just as dV = L di/dt, so also dV = R dL/dt (where R is the magnetic potential (or flux) of the field where the movement of the string is taking place). Plus / minus signs neglected for simplicity in these equations.

The magnetic flux is not linear with distance from the magnetic centre so if the pickup is slightly closer then it will be far more sensitive to string movement than if the magnet was proportionally weaker or stronger.

[Mike Sulzer]

Yes I believe that you are right in that the magnet strength is not a major factor in sensitivity (but don't tell the musos)!

I do not agree. If the magnet has zero strength you get no output from the pickup because the there is no changing flux through the coil from the vibrating string. A weak magnet weakly magnetizes the string and gives a small output, and a strong magnet gives more. That is an observed fact. Of course, you cannot make the magnet too strong or you affect the vibration of the string.

[Mal88]

Right Mike! Look at the equation again - if there is no magnetic strength then there will be no output. The output is proportional to the change in magnetic field. If the field is twice as strong, then the change for the same relative movement wil be 6.02 dB louder (agreed), but move the string to 0.7071 times the current distance position and the magnetic field will be double the potential and also be 6.02 dB louder. (compre') Alternatively use a more effective magnetic circuit - (most manufacturers neglect the inefficiency of the distant end pole and leave it floating in the breeze).

[David Schwab]

I do not agree. If the magnet has zero strength you get no output from the pickup because the there is no changing flux through the coil from the vibrating string. A weak magnet weakly magnetizes the string and gives a small output, and a strong magnet gives more. That is an observed fact. Of course, you cannot make the magnet too strong or you affect the vibration of the string.

Right, but getting past using a weak magnet, if you take a fully charged alnico V, and a fully charged Ceramic 8, the ceramic is quite a bit stronger, but the pickup will not be louder by the same degree, if at all. It might be brighter though.

Also the shape of the pole piece makes a difference... alnico rods will pull the strings a lot more than a blade charged by even a strong neo magnet. I'm working with N42 grade neo bar magnets which are spec'd at: Pull Force: 22.40 lbs; Surface Field: 4480 Gauss; Brmax: 13,200 Gauss; and I have no string pulling problems when used in the context of a pickup. I would also imagine my cores are quite saturated by the magnets, but that's another subject.

Interestingly I get a lot more low end with the neos than with a C8 in the same pickup.

[Mike Sulzer]

Right Mike! Look at the equation again - if there is no magnetic strength then there will be no output. The output is proportional to the change in magnetic field. If the field is twice as strong, then the change for the same relative movement wil be 6.02 dB louder (agreed), but move the string to 0.7071 times the current distance position and the magnetic field will be double the potential and also be 6.02 dB louder. (compre') Alternatively use a more effective magnetic circuit - (most manufacturers neglect the inefficiency of the distant end pole and leave it floating in the breeze).

This is how I see the B field vs. distance thing: If you weaken the field, you can move the pickup closer to the string without getting string pull. Then you actually get more output from the pickup because: 1. The magnetic field at the string is the same in the two situations. 2. The change in magnetic flux through the coil when the string vibrates is greater because the magnetized string is closer to the pickup. There are limits as to how close you can go, of course.

I do not see where the .707 factor comes from. Far away, the field is like a dipole, varying with inverse distance cubed. Close up it hardly varies at all. In between is complicated.

Do you have an example of how a more efficient magnetic circuit can help?

[salvarsan]

Do you have an example of how a more efficient magnetic circuit can help?

Reluctantly, variably, no.

-drh