Humbucker base plates and eddy currents

[antigua]

A few months ago I compared a Donlis humbucker, a low cost PAF style pickup from China, with a DiMarzio 36th Anniversary PAF pickup.

The Donlis did no contain a metal cover, which is probably a good thing, considering that it would likely have been made of brass, and performed unfavorably compared to the nickel silver cover of the DiMarzio, but every PAF knock off has to have a base plate, and it happens to be that most costly domestic humbuckers have nickel silver baseplates, while import humbuckers have brass, and sometimes even aluminum, base plates. So the question becomes; how much is of a difference is made, on a technical level, when the base plate is nickel silver, brass or aluminum?

In the Donlis v. DiMarzio analysis, it appeared that the difference was about 0.4dB and 150Hz at the unloaded resonant peak. Since the implications apply to nearly all PAF knockoffs and clones, I decided to do this test again with a greater amount of care.

I have four pickups with base plates to be compared:

top: DiMarzio PAF Master neck (nickel silver), "HMV" Chinese humbucker (aluminum)
bottom: Seymour Duncan PG+(nickel silver), Donlis humbucker (brass)




The backs... I removed the black paint from a leg of the PG+ in order to verify that it was nickel silver. The PAF Master and the Seymour Duncan PG+ are nickel silver, while the Donlis at top right is brass, and the no-name at bottom right is aluminum. (I accidentally reversed the Donlis and the HMV in the two photo, my apologies)




Since it's impractical to solder to aluminum, the MVH humbucker has an eyelet at the corner for the ground connection:


I've removed the base plates (DiMarzio PAF Master, Seymour Duncan PG+, Donlis, HMV):




For the test, the Seymour Duncan Pearly Gates Plus is the common electronics. The other base plates are attached to it, and then measured.

First up, this is how the compare with no additional load, with common amplitudes up to 400Hz, which roughly correspond to the fundamental frequency of a "G4", or the 3rd fret of the high E.

Each division is one 1 dBV. In audio, it's said that 1dB is the smallest change in pressure that the human ear can detect. You can hear the difference in dB levels for yourself here www.audiocheck.net/blindtests_level.php?lvl=1

The four lines represent the frequency response of the four humbuckers. It can be seen that, while there is evidence off higher eddy current losses beyond 1kHz (the plot lines dip down), the deviation between all of them is less than 1dB:



However, then a pickup is in a guitar, it's "loaded". For test loading I've used 470pF parallel capacitance to represent a guitar cable, and 200k ohms in parallel to stand in for guitar pots. The plot with these values becomes:



The wider deviation seen between 2kHz and 3kHz in the "unloaded" plot carry down almost exactly to the "loaded" plot, where the resonance becomes ~2.8kHz. The widest spread is no more than 0.6dB.

This data suggests that the eddy current losses caused by both the brass and aluminum base plates have no audible effect on the pickup. They will sound too similar for a person to distinguish one from the other.

So if you're in the market for humbuckers and are considering import models, the base plate metal ought not weight heavily in your purchase decision, though if the humbucker as a metal cover, there's a good chance it's brass, and it will cause losses of 3db to 4dB at the resonance, destroying the Q factor.

Another interesting thing to note is that the nickel silver yields a slightly lower resonant peak than the brass and alumnimum, meaning the latter slightly decrease the inductance of the pickup.

Another data point, just for fun, this is what happens when the base plate is removed all together:



The unloaded peak shows a modest drop of 1.2dB, but the loaded peak only drops 0.2dB. This echoes the well known fact that nickel silver is superior to brass and aluminum with respect to eddy currents, but given that none of the differences exceed 1dB, it's all academic in the case of a base plate.



Here is a picture of the test setup. The driver coil that feeds a changing magnetic field to the pickup is taped to the top of the PG+ securely so that it's sure not to move when the different base plates are attached and detached. The pickup is connected to an integrator designed and built by Ken Willmott that removes the +6dB/oct inductive voltage, and the it's fed into a VEllemen PSCGU250 which has a built in function generator to feed the driver coil and produce a bode plot. 

[stratotarts]

Maybe the slot modification would help.

[sumitagarwal]

I know this is a few years old now, but I had some thoughts and questions around this.

The results from this study found differences that were measurable, but below the accepted threshold of human hearing. However, when playing a guitar through a tube amp aren't many of the more subtle differences magnified by applying gain, especially as the signal chain response varies widely across frequency ranges? So that even if some changes are inaudible listened to as a neutral signal, a typical guitar rig would make these differences audible?

As a tangent, I thought it would be interesting to post a pic of some of PRS's most recent pickups. I may be wrong, but to me this baseplate actually looks plastic! Is that likely or possible?

Thanks for actually executing these systematic studies and publishing them when nobody else seems to be doing so!

[mikecg]

Hello antigua,

Yet another 'tour de force' in pickup analysis - your detailed, careful work and clear presentations are much appreciated, hereabouts.

[antigua]

Feb 1, 2024 21:40:52 GMT -5 sumitagarwal said:

I know this is a few years old now, but I had some thoughts and questions around this.

The results from this study found differences that were measurable, but below the accepted threshold of human hearing. However, when playing a guitar through a tube amp aren't many of the more subtle differences magnified by applying gain, especially as the signal chain response varies widely across frequency ranges? So that even if some changes are inaudible listened to as a neutral signal, a typical guitar rig would make these differences audible?

As a tangent, I thought it would be interesting to post a pic of some of PRS's most recent pickups. I may be wrong, but to me this baseplate actually looks plastic! Is that likely or possible?

Thanks for actually executing these systematic studies and publishing them when nobody else seems to be doing so!

If we're talking about microphonics, then there would be new content added to the signal, acoustic sound, and maybe that acoustic sound adds in a treble that is brought out by clipping of a tube amp. But if you think about how to isolate a single voice on a recording of a crowded room, you get the best results by using an equalizer to raise up the frequencies of the voice you want to hear. I can't imagine that distorting a signal would make any one voice easier to hear, I think it would just make all of the voices harder to hear, and I think that's true in this case. 

But when talking about eddy currents, this is not like microphonics, there is no new content added to the signal. The eddy currents function as an unwanted electrical filtering. You could do the same thing with a graphic EQ. This raises the question, if a graphic EQ can bring out parts of a signal, and eddy currents are like an EQ, then can they bring out part of a signal? Probably not, because the eddy currents to not emphasize any part of the signal, rather eddy currents bring all the treble down, called a "high end roll off". Eddy currents roll off treble in two ways, 1) by putting inductive resistance across the pickup's coils, and 2) by cancelling out the high frequency magnetic content of the moving guitar string itself (and that's important because it means a bass plates and a covers and pole pieces will all cause these effects to happen in different proportions). If eddy currents were able to push a particular frequency up, then there would be potential for them to help bring out sounds you might not hear otherwise. 

It so happens that there is a feature of vintage style guitar pickups that does bring out a particular frequency though, the resonant peak of the pickup. Guitar pickups have a resonant peak somewhere between 2kHz and 5 or 6kHz, and the reason a lot of pickup on the market sound different is for no other reason than they have difference inductance values, and therefore different resonant peaks. If a guitar pickup pushes up content around 2.5kHz, it will sound different than if the pickup pushed up content at 4kHz. Eddy currents only serve to lower the Q factor at resonance, so if anything, eddy currents rob a pickup of whatever special voice that is associated with it, and makes it sound more flat and bland.  It's because of the resonant peak that electric guitars sound like electric guitars. Without the resonant peak, they would sound a lot more like an amplified acoustic guitar. 

[sumitagarwal]

This is a great explanation, thank you antigua!

Also, I never got an email notification about your reply... is there any way to turn emails on on this forum? I can't find it

EDIT: Nevermind, I think I found the settings!