Gibson size humbuckers set up for single coil sound

[ms]

Here is part of a comment I made in guitarnuts2.proboards.com/thread/10690/humbucker-comb-filtering :

"...you should get close to a tele lead or strat sound without the hum by making humbuckers in this way:
1. Use fewer turns on each coil so that the inductance of two in series is what you want. Since inductance decreases faster than linearly with turns, you do not need to take off that many.
2. Use Alnico rod magnets for the right Q.
3. Do not use a tall Bobbin; the turns near the bottom do not contribute much to the level. So the turns you lose with respect to a strat pickup to get the right inductance cost you less level than you might think in terms of output if you are used to strat pickups with tall bobbins.
4. Lose the baseplate? Not sure; a nickel silver base plate kills some highs, but not that much. The cores are more important."

So a humbucker loses highs relative to a Fender type single coil as a result of the filtering effects of its electrical circuit (both higher inductance and higher losses, both largely from the permeability and conductivity of steel), and according to common belief, as a result of string filtering effects produced by the two spaced sampling regions. But it seems that the sampling issue really does not matter much, if at all. This is the result of the discussion referred to, and to some tests I did years ago.

So this discussion shows that you can make humbuckers with single coil properties.

First, why would you want to do this? A huge variety of guitars use original Gibson size humbuckers, but cannot be purchased with pickups that sound like Fender single coil, having both higher resonant frequency and lower loss (higher Q), and that reject magnetic hum. This discussion shows  an easy pickup change that does all of this. Note that split coil with a normal humbucker only goes part way because of the steel cores, and loses hum cancelation.

There are a few humbuckers available that use Alnico rods. For example, here is part of a description of a humbucker made by Dream Songs Pickups (www.dreamsongspickups.com/en/content/strat-rod-magnets-humbucker)
"...It is a beautiful sounding humbucker with rod magnets and it also has the sound of a true Stratocaster single coil in split mode. This pickup works well with 250k pots, but you may want to try it with 500k pots fot a clearer and grittier sound in humbucking mode.

DCR from 10k to 12k
... "

(I actually have some doubts about the claimed sound in split mode since this pickup is sold with a partial or complete metal cover, but I have never measured or heard one.)

This is not the kind of pickup discussed here. This discussion is about humbucker pickups that have a single coil sound when used as a humbucker, but they also can have a humbucker sound (with less output than a typical humbucker) by adding parallel capacitance and possibly resistance. That is, single coil sound does not require just a single coil, but can be achieved by designing for the proper electrical properties, and, almost as important, humbucker sound is retained by further modifying the electrical properties.

I have not found any such pickups for sale, but they are easy to make. Using Alnico rods results in less loss than steel, and less inductance, and so the required number of turns is similar to the minimum number used in available humbuckers. I have used two methods for making them:
1. Modify available humbucker bobbins, by drilling out the pole holes to fit available .195" Alnico rods, and wind with the required number of turns. (.187" rods might be easier to use; I have not tried these.)
2. Remove all the magnetic components and spacers from an available humbucker, modify the pole holes and install rods, using a custom made spacer so that standard rods can sit on top of the standard baseplate.

The first method offers more flexibility, and it is easy to drill out standard bobbins made from high quality strong plastic with a 5 mm drill (followed by hand clean up). For the second method, very inexpensive FLEOR pickups can be used, selecting the lower resistance two options (not the overwound ones). Things to watch out for:
1. Although the two coils in a pickup are well matched, the resistance can vary up to 1K from pickup to pickup.
2. DO NOT DRILL OUT WITH POWER. The bobbins are made from very soft plastic. Drill out by hand only, (if you must use power, use a special bit designed for soft plastic and be very slow and careful). A standard bit grabs, and you can destroy the coil and injure yourself.

It is not necessary to change the wiring at the pickup; in fact, it is not even necessary to disturb the tape around the two coils if you are careful.

Standard rods extend well below the bottom of the bobbin (both types of implementation, of course), and so you need to make a special spacer. I print mine, but cutting wood or plastic should work well also. Somebody might sell a shorter rod that would work with standard spacers.

So let's look at some measurements. These are impedance measurements, which are good at showing the location and width (losses or damping) of the peak and thus are good for comparing pickups and predicting the effects of a load. I display the magnitude of the impedance, which shows the peak, and the real and imaginary parts with the effect of pickup (and cable) capacitance removed. These days, I measure the capacitance with a meter at 100 KHz. This almost always gives a good enough value for the "unparalleling" computation.

The attached plot has six measurements. The three in the left hand column are standard pickups, and the three in the right column are humbuckers intended to sound like single coils. The one in the upper left is a Fender single coil type that I wound from a kit. The other two in the left column are Seymour Duncan humbuckers. The three in the right column have single coil properties, and they sound like them, too. They are not identical to typical strat pickups, but they work just fine in that role. I have not yet made measurements with an exciter coil. It seems that the output level is between a normal single coil and a humbucker, but it would be good to know for sure.

I have them in three guitars. Later posts will deal with wiring diagrams, etc., especially circuits to switch in typical humbucker sound. (In general, it is easier to lower a resonant frequency and increase damping, lower Q, than to go the other way, and so this is not so hard to do.)

[ms]

Here is a photo of a disassembled modified FLEOR showing the bottoms of the two coils.  The two black 3D printed spacers fit over the rods and keep them slightly spaced from the baseplate or just touching it in some cases.  I have not measured the electrical effect of the baseplate, but it must be small because the measurements in the previous post show low total loss, about what is expected from a single coil pickup using Alnico rods.  (The coils are secured using the same screws that came with the pickup, but they are barely long enough, and it would be better to use longer ones.)  The poles extend down different distances because I have staggered the pole pieces and have used two different lengths of rods (what I had).  This particular one is a prototype and the pole pieces are not necessarily set to the proper heights.  This is not the baseplate that came with the FLEOR, but those baseplates work fine, and I think I used the original one for something else.


Edit: I forgot to mention that these pickups use A5 rods, but I have not fully magnetized them, keeping the field measured flush against the face to about 600 Gauss or a bit less.

[ms]

The guitar in the photo uses the two pickups whose measurements are shown in the middle and bottom of the right hand column in the measurements shown in the first post. The neck pickup has 4250 turns of #42 per coil, while the bridge has 4934. (I think that strange number resulted from scaling up the inductance. It might as well be 5000.)


The wiring diagram is shown below. The volume pots are 250K, and I intended to use 250K for both of the tones pots, but I wanted a push-pull switch for paralleling additional capacitance across each pickup. So the neck pickup uses a so-called 500K pot that actually measures closer to 600K because that is all I had. I have not yet carefully listened to a set of capacitor values in order to select the optimum, but these are reasonable.

[aquin43]

This is really interesting work. The final result is like two strat pickups side by side but with the later half height coil to lower the inductance.

How does the output level compare with a strat pickup? I ask because a further modification that comes to mind is to parallel the virtually identical coils so as to lower the output impedance, use 50k pots and become virtually independent of the cable loading.

[ms]

May 15, 2024 9:20:41 GMT -5 aquin43 said:

This is really interesting work. The final result is like two strat pickups side by side but with the later half height coil to lower the inductance.

How does the output level compare with a strat pickup? I ask because a further modification that comes to mind is to parallel the virtually identical coils so as to lower the output impedance, use 50k pots and become virtually independent of the cable loading.

Thanks!  The output is significantly more than a strat pickup, but less than a normal humbucker, of course.  As we know, the turns further from the strings contribute little output in a strat pickup, and so in this design the average turn is located closer to the strings.  Also having two coils instead of one helps with the output/inductance ratio since inductance varies faster than linear with turn count.  I have not made accurate inter-comparisons yet.

It is probably not good enough for what you want to do,  that is, keeping the output equal to a standard strat pickup with the coils in these pickups in  parallel.  I am setting up a guitar to allow better comparisons, but that is still somewhat in the future.  That is, it is better to make comparisons of different pickups on the same guitar, but it would maybe still be useful to compare the output of a strat with a guitar with these modified humbuckers, but the with the coils in parallel.

[ms]

This black guitar uses two modified FLEOR pickups, measurements shown below. At the time I ordered these, the only bobbin choice with the correct resistance coils (neck, 7 t08 K, and bridge, 8 to none K, was zebra. In the impedance figure below, the differences between the impedances of single coil sound and humbucker sound with respect to damping (or Q) are indicated. (Check the first post for measurements of humbuckers.)













As described here (https://guitarnuts2.proboards.com/thread/10727/pickup-imp-apparent-fall-rise) the curved shapes with frequency of the impedances (rising real, falling imaginary) are the result of resistance in parallel with the coil, the result of magnetic coupling to metal parts, primarily the pole pieces.

There is a push-pull switch on each tone pot for the corresponding parallel capacitor. This guitar has 1000 pf on each switch. With both pickups on you can choose zero, one, or two capacitors.

[ms]

May 15, 2024 11:10:27 GMT -5 ms said:

May 15, 2024 9:20:41 GMT -5 aquin43 said:

This is really interesting work. The final result is like two strat pickups side by side but with the later half height coil to lower the inductance.

How does the output level compare with a strat pickup? I ask because a further modification that comes to mind is to parallel the virtually identical coils so as to lower the output impedance, use 50k pots and become virtually independent of the cable loading.

Thanks!  The output is significantly more than a strat pickup, but less than a normal humbucker, of course.  As we know, the turns further from the strings contribute little output in a strat pickup, and so in this design the average turn is located closer to the strings.  Also having two coils instead of one helps with the output/inductance ratio since inductance varies faster than linear with turn count.  I have not made accurate inter-comparisons yet.

It is probably not good enough for what you want to do,  that is, keeping the output equal to a standard strat pickup with the coils in these pickups in  parallel.  I am setting up a guitar to allow better comparisons, but that is still somewhat in the future.  That is, it is better to make comparisons of different pickups on the same guitar, but it would maybe still be useful to compare the output of a strat with a guitar with these modified humbuckers, but the with the coils in parallel.

I was wrong; your idea will work really well.  For one guitar in the comparison, I used the black guitar shown above.  Since the modified Fleor pickups are four conductor plus shield,  it was only necessary to remove the back control compartment cover and move a couple of wires to put the neck pickup coils in parallel.  I did not change pots, so this is not a listening comparison, just a look at the picked level on the scope.

The other guitar is this:




It uses a good quality roasted maple neck and the $97 JD Moon poplar body.  The pickups are Fleor kits.  The guitar has hum canceling ferrite core coils under the pick guard.  

The picked outputs viewed on a scope are almost the same.  I think the black guitar is possibly slightly higher in output, but it is too close to call for sure without a better comparison.  Changing the pots to 50K would load it a bit, but this idea should work. Even if you lost a bit of level, the signal to electric field hum ratio would improve because of the lower impedance, and of course the humbuckers lower the magnetic hum by at least 35 db if carefully made.  (Even the Fleor pickups do that well, well maybe not all of them.)  And of course the change in tone with volume setting would be gone.

[ms]

Let's compare the performance of a Fender type single coil pickup and an equivalent humbucker as described above. Actually I have changed to A3 magnets, both pickups made from rods from the same batch from Mojotone.  A3 can be fully magnetized without getting too strong, and it is easier to get consistent results when fully magnetizing the rods.  We expect the most differences from the open E6 string since the most string harmonics have frequencies in range of guitar electronics, but let's start with E1 and see if we can see any differences.  (Listening test will come later.)

Below are early time domain samples of the two (early time because the most high frequencies are present).  This is the neck pickup, maple topped guitar shown above, picked halfway between the two pickups.  The recording was made using instrument input 2 of an Apogee Boom, recorded by GarageBand into a mic/Line software input with all processing turned off.  A 24 bit stereo wave file (no compression) is written to the hard drive and then read into Python using the standard code for processing wave files.









They certainly do not look the same.  But do these differences matter for guitar playing?  First, in both cases, why is the first pulse so different from the other four? Explaining this requires some thinking about the timing and overlapping of the pulses that travel from the picking location towards the bridge and towards the neck, and recognizing that the polarity changes from the reflection.  Information about the differences between the appearances of the the two plots is contained in spectral plots, which will be shown in the next post.

[ms]

Here are the power spectra for the two pickups in question, E1 string.  The titles should read "32768 samples" not "reps".  The sampling rate is the usual 44.1 KHz.  The small differences below 5 Hz are probably just random picking variation, but there certainly are systematic differences above 8 KHz.  A guitar speaker falls fast by 5 KHz.  It looks as though these differences are not significant when played through an amp, but you should be able to hear them directly from the recording interface, especially with good headphones.  This depends upon how good your hearing is, of course!






Could the differences be the result of the spacing between the two HB coils?  To find out, the predicted Tillman spectrum is computed for the two closely spaced coils, assuming each is a point sample.  Here is the result:



The broad dip is the result of the spacing.  It appears to explain the systematic differences in the two spectra plots.

[aquin43]

I don't quite understand the "Tillman" factor. Does that include the effect of the picking and pickup positions? The twin coil filtering itself is much simpler: 3dB down at ~6 kHz with the notch at ~12 kHz and back to -3dB at ~18 kHz with a 25.5" scale and 0.7" pole spacing.

[gckelloch]

From what I figure, it makes sense there would be a reduction between 2.5 - 3kHz for the HB pickup considering the size of that harmonic on E1. Maybe it's not completely canceled because the string harmonics oscillate elliptically? Maybe the node at ~1.4kHz on the Tillman graph is from that harmonic node point being over just the outer portion of each coil, but it doesn't show on the string sample graph due to uneven ellipticity. Notice that both pickups have a reduction between 4-5kHz, which corresponds to the Tillman graph. I assume that would be from the partial cancelation of that harmonic on either side of each coil. Could all be coincidence, but it makes sense if I understand it correctly.

[ms]

Jul 16, 2024 11:13:30 GMT -5 aquin43 said:

I don't quite understand the "Tillman" factor. Does that include the effect of the picking and pickup positions? The twin coil filtering itself is much simpler: 3dB down at ~6 kHz with the notch at ~12 kHz and back to -3dB at ~18 kHz with a 25.5" scale and 0.7" pole spacing.

All I did was to use this article: Response Effects of Guitar Pickup Mixing (https://till.com/articles/PickupMixing/index.html#twopickupanalysis)

and follow this from the article:

A 50/50 mix of two pickups at two different locations is a simple sum:

V2pickups = 0.5sin((pi XpickupA Fstring)/(Lscale Fopen)) + 0.5sin((pi XpickupB Fstring)/(Lscale Fopen))


------
where pickups A and B are in this case 18 mm apart.  It seems to work well enough.

[aquin43]

I see, it does include the effect of the pickup placement.
Another approach is to calculate the response of just the pickup itself, which will modulate the response due to its placement and the picking position.

It is the modulus of the sum of two 0.5 long vectors at an angle phi(f) apart

where phi(f)=d*pi*f/(s*f0), if we ignore dispersion.

The response is sqrt((1+cos(phi(f)))/2)

f is frequency, s is scale length, d is pole spacing, f0 is open string frequency.

dB v Frequency, 25.5" scale, 0.7" spacing, 1st string (329.6 Hz)

[ms]

Jul 16, 2024 11:43:18 GMT -5 gckelloch said:

From what I figure, it makes sense there would be a reduction between 2.5 - 3kHz for the HB pickup considering the size of that harmonic on E1. Maybe it's not completely canceled because the string harmonics oscillate elliptically? Maybe the node at ~1.4kHz on the Tillman graph is from that harmonic node point being over just the outer portion of each coil, but it doesn't show on the string sample graph due to uneven ellipticity. Notice that both pickups have a reduction between 4-5kHz, which corresponds to the Tillman graph. I assume that would be from the partial cancelation of that harmonic on either side of each coil. Could all be coincidence, but it makes sense if I understand it correctly.

Let's put the Tillman curve and the HB pickup response (E1) on  the same plot and restrict the frequency range to 5 KHz for better visibility.  If the picking produced  a very short impulse rather than a pulse of measurable width, and if the pickup aperture were a point (and probably some other "ifs" not mentioned) then the harmonics in the pickup response would follow the Tillman curve.  It is pretty good for the  first cycle, but then it drops off in an irregular way, with the "deep dips" in the Tillman response probably still visible in the pickup response. (Remember, the locations of the "deep dips" is a function of pickup location.) Over some of this range, two coils might as well be one located between them, but we can see some drop off starting between 3 and 4 KHz in the Tillman plot.  This can be understood from aquin43's post just above.  For the lower harmonics the pickup spacing is small compared to the wavelength of the harmonic, but as the harmonic number increases, the relative phase shift becomes significant and there is a small degree of cancelation.

[ms]

Here are plots for the SC and HB, initial signal response, open E6 string.  The visible differences here are greater than for the E1 string as expected.




In addition to changes in the pulse shape, we see the spaces between the pulse fill in with "fuzz" progressively with time.  Much more "fuzz" is seen in the SC than the HB, indicating its frequency is in the range attenuated by the spacing between the two coils in the HB.  (Remember, this frequency range is four times lower than for the E1 string.)

The E6 string is much stiffer than the E1, and so we expect that the effect is caused by dispersion; that is, the higher frequencies travel significantly faster than the lower frequencies.

The spectra (to be in the next post) will show how much greater we expect the audible differences to be for the E6 than the E1.

[ms]

The plot below is for the E6 string; the blue is the SC, the orange the HB, shown in front so that harmonics where it is weaker can be easily seen and compared.  As expected, the spectra follow the Tillman curve well in the first cycle.  The broad dip associated with the close coil spacing is fairly well reproduced in the data, although not perfectly.  For the HB, some harmonics in the range of the dip are more than 10 db down from the SC, and this is within the frequency range of the guitar speaker.  So one might argue that this should be audible.  On the other hand, these harmonics near the center of the broad dip are typically more than 30 db down from the low number harmonics, and some might argue that harmonics with numbers in the high 20s and 30s should not matter much when they are so weak.  So the comparison is not complete without listening.

[ms]

The link to a test comparison wave file is below.  This is a picture of the single coil pickup used in the comparison in a 3D printed frame that mounts it into a humbucker rout.






Below is a low resolution plot of the 16 second test signal.  Each pair of signals is for a string (1 through 6).  There might be  a subtle problem with the A5 recording; that requires some further investigation, but this will have to do for now.

Only the picking transient and the signal just following it contain enough high frequencies to make good test signals, and so that is all I have included.



link to wave test file


Apple has provided a way to make a public link to a file stored in Apple iCloud.  It seems like the flakiest thing they have ever done.  If it does not work for you, let me know.

[frets]

MS
I may have missed this if you’ve mentioned it already; but, are you planning on doing both Humbuckers as single coils? I think your solution is brilliant. I have never seen it before.

[ms]

Jul 22, 2024 15:31:30 GMT -5 frets said:

MS
I may have missed this if you’ve mentioned it already; but, are you planning on doing both Humbuckers as single coils? I think your solution is brilliant. I have never seen it before.

Thank you!  And yes,  both pickups on the guitars I have put these pickups on have single coil sound humbuckers on the bridge pickup as well as the neck.  I have been using the neck pickups for comparison because  the lower harmonics are more balanced.  Not sure that matters, but it seems like a reasonable choice.

[frets]

MS

I think it would make for an interesting setup for a Les Paul. I could envision an HS with the single coil in the neck position.

I know you’re doing experimentation, do you plan on keeping the SS configuration on the guitar? I’m sorry, I’ve read the thread but could have missed some details😺.

[ms]

Jul 23, 2024 16:13:50 GMT -5 frets said:

MS

I think it would make for an interesting setup for a Les Paul. I could envision an HS with the single coil in the neck position.

I know you’re doing experimentation, do you plan on keeping the SS configuration on the guitar? I’m sorry, I’ve read the thread but could have missed some details😺.

That could make an interesting setup for a Les Paul.  My current preference (and that could change of course) is to put these single coil sound humbuckers in both positions, and then use a push pull pot to put capacitance in parallel in order to lower the resonant frequency  to move towards a normal humbucker sound.  Of course, you also need to adjust the tone control to get the right damping.  So it might be best to switch in a cap with some R already in parallel .  A more complicated circuit might be necessary.  The damping in a normal humbucker is the result of eddy currents in steel, and so the damping is a function of frequency, although it varies slowly.  In general it is easier to reduce high frequencies than to put them back in, and so I think this is the right method to try.