Single coil size '51 P bass humbucker

[okabass]

Testing a bass PU. Try to get single coil sound and size. And humbuckers no-noise.  

www.dropbox.com/s/kltt7ddgmsd3ymq/2018-05-11%2009.10.56-1.jpg?dl=0

The frame is a cheap China rail PU (series, 2x 5,5 kΩ, around 5 H). Magnet is now Alnico bar, 300 Gs. I have stronger caramic too, 800 Gs. Self made soft iron rails. Idea for asymmetric rails is from a Yamaha patent.
 

Not bad, but needs trimming.Perhaps parallel connected coils, but that could be too weak (around 2,8 kΩ).
Any Tips?

[straylight]

Are you using the winds already on the bobbin or rewinding it?

I would go parallel to get the single coil sound. The original winds on the bobbin will be 44 or 45 AWG and really tightly packed so high capacitance and dark sounding. Don't worry for output.

If rewinding perhaps try 42 or 43 AWG in series. Those rail bobbins are really awkward to wind though.

I would get some 19mm alnico rod magnets, cut some flatwork from 1.5mm plywood and wind up a stacked single. I don't know if I could get enough 42AWG wire on, but that's where I'd start. You need to wind bass pickups pretty hot or they sound really thin. P-Bass split-pickups need 8500 turns a side minimum for a bright sound, and I run out of space on the bobbin before I get to 12000 turns of 42AWG unless I make the bobbin taller.

[okabass]

May 30, 2018 22:52:32 GMT -5 straylight said:

Are you using the winds already on the bobbin or rewinding it?

I would go parallel to get the single coil sound. The original winds on the bobbin will be 44 or 45 AWG and really tightly packed so high capacitance and dark sounding. Don't worry for output.

If rewinding perhaps try 42 or 43 AWG in series. Those rail bobbins are really awkward to wind though.

I would get some 19mm alnico rod magnets, cut some flatwork from 1.5mm plywood and wind up a stacked single. I don't know if I could get enough 42AWG wire on, but that's where I'd start. You need to wind bass pickups pretty hot or they sound really thin. P-Bass split-pickups need 8500 turns a side minimum for a bright sound, and I run out of space on the bobbin before I get to 12000 turns of 42AWG unless I make the bobbin taller.

Good advice.

Actually I ended similar idea two weeks ago. I made a stack frame with 20 mm Alnico 5 rods. I wound (≈AWG 43,  0,055 mm ) the lower coil to 8,6 kΩ and the upper to 15,5 kΩ (ca. 8100 and 14600 turns). 24,1 kΩ in series and 7 H.
It sounds surprisingly good, the single coil character is there. Bass end is a bit thick but that's no problem. 
Here's a pic of a proto with AWG 42 which is too thick to wind enough turns. 
www.dropbox.com/s/hlk2qoffwp0kbsy/20180521_215827-1.jpg?dl=0

[straylight]

Ok, in terms of your sounds design, here's something to consider, and I'm sorry this is maths heavy but I'll try to explain as I go.

Your output voltage is given by

e = N dT/dt

where
e is the output voltage
N is the number of turns in your coil
T (should be greek theta but...) is magnetic flux
t is time

dT/dt is rate of change of magnetic flux with respect to time, which in our case is the signal from the string moving and the background EM noise. 

Notably, output is directly proportional to your number of turns of wire. But DC resistance? Also directly proportional to the number of turns of wire.

The second thing going on is inductance, which at the risk of sounding patronising is the property that opposes changes in current flow. 

L = N^2 u A / l

where
L is the inductance 
N is the number of turns in the coil
u is the absolute permeability of the core material
A is the Area of the coil
l is the lenght of the coil

this means a long thin coil will have less inductance than a short fat coil and there's a big inductance penalty for more winds. This is why 16kOhms (using DC resistance as a measure of wire length) of 44AWG is thereabouts the limit.

Why does inductance matter? this brings in our third equation

f = 1/(2 pi SQRT( L C ))

where
f is the resonant frequency
pi is the constant begining 3.14..... 
L is the inductance as calculated above
C is the capacitance of the pickup

the higher L and C get, the muddier things get.

To get your noise cancelling single coil sounding single coil ish, and hum cancel, you need to a) keep the number of turns in both halves of the coil close to balanced and b) keep your inductance under control. Capacitance is harder to calculate as it's dependant on your winding density, but hand winding gives you a headstart as you're likely to get a fairly good scatter and reduced capacitance.

I would try equal sized coils as tall as you can fit in the space you've got to work with, and keeping the wind count as close to that of the single coil you're trying to emulate as possible.

Another approach might be to make a wooden spacer that matches core shape made by the two outer magnets of your pickup and wind your lower coil around that so only your upper coil goes around the magnets.

I'm quite tempted to try a strat stacked single made out of two stacked humbucker bobbins wound at 4000 turns each with short magnets, I might have something on my desk that will just work for that.

[okabass]

Thank you. That kind of info I like and need. Math is no big problem. 

" I would try equal sized coils as tall as you can fit in the space you've got to work with, and keeping the wind count as close to that of the single coil you're trying to emulate as possible."

To make it sure. If the original SC pickup is around 6,5 kΩ, do you mean 2 x 6,5 kΩ  or 2 x 3,25 kΩ?  I've made a split coil PU with equal coils 2 x 3,25 kΩ.  

I made a proto with 20 mm magnets cause I already have them. My first idea was to made equal coils like Seymour Duncan and some others. But I read that it is possible to wind unequal coils, only the humbucking effect is not as good than equal coils (more noise). But in places I play there have not been problem with the PU noise. Noise is less than upper single coil alone.

The PU's capacitance is 83 pF, with Cu-foil shield 150 pF (measured with DE-5000 meter). Seems not to take too much hi end away. With a cable and an amp the res frequency goes about 6% lower than without the shield. 

My intent is to get longer magnets / or try thinner wire to make ca. 15 kΩ + 15 kΩ pickup. Or perhaps with the 20 mm magnets equal 12 kΩ +12 kΩ coils.
But as I say the PU sounds very good now. Thicker bass is no problem: if I cut 3-6 dB to the lo end (ca. 40 Hz, shelf eq), I get quite good original P single coil type sound. To me it's important that the '51-type P bass character is right.

[straylight]

EDIT: SCREW-UP, SEE MY NEXT POST

From the bench testing i'd be inclined to say try going with about the same number of winds in total as your target single coil. However I need to throw this creation in a guitar and see what happens, and it's 1am here, so it shall have to wait, I was just messing around with it to see what different sets of magnets and polepieces did to the numbers, but it was a convenient set of windings.

However I note the Seymour Duncan STK-1S set I just sold hat an astronomicly high DC resistance and sounded strat-ish so I may have missed something. They were stupid-hot though. I suspect I might be short on output here.

[newey]

Straylight-

T (should be greek theta but...) is magnetic flux

Θ or θ can both be bbc coded.  (i.e., ampersandTheta; or ampersandtheta;)

[okabass]

Jun 4, 2018 19:17:59 GMT -5 straylight said:

From the bench testing i'd be inclined to say try going with about the same number of winds in total as your target single coil.  

However I note the Seymour Duncan STK-1S set I just sold hat an astronomicly high DC resistance and sounded strat-ish so I may have missed something. They were stupid-hot though.  

Ok. I'm not good to read those graphs, but to me it seems the PU is thin sounding. But it probably should be so. Quite same graph, despite 3 different type PUs. 

There's good info about stack PUs. 
music-electronics-forum.com/t6893/

As I understand " David Schwab" writes that a stack PU with equal coils tends to cancel lo end and that high DCR (many turns) tries to compensate this. I think that my PU's 8,6 kΩ and 15,5 kΩ coils produce quite good lo end. But I'm going to try an equal turns stack with AWG 44 and 20 mm magnets.

[straylight]

I've screwed up, several times. For a start what I bench tested was 4000 turns and then another 4000 turns in the same direction, not hum cancelling. Secondly I tried to post saying so yesterday and failed to do check i'd actually posted correctly before I dashed out.

Lower coil needs to be out of phase magnetically and signal-wise to work, at which point having a long top coil and short lower coil may seem like a sound design decision (see below first) but same number of turns or close to that is still optimum for hum cancelling regardless.

Wind hot.

Perhaps separate coils with a brass or copper plate much like a tele base plate. Forget about keeping induction low. I was wrong. Resonant frequency goes higher. Below is a correct graph. My first guess after seeing this would be to try 6000 turns on each coil to match an 8000 turn single coil. I'm just messing around with the parts that happen to be on my bench because this problem interests me, and it fits my wider investigation of compairing inductances of materials from my local industrial bolt supplier using the 2x4000 turn PAF* style humbucker.



Note that the PAF actually matches a strat single coil pretty well in terms of resonant frequency but the single coil has a higher quality factor (Q). It's more of an accident of numbers than anything else as the paf design has a much greater inductance and more output. When considering pickup output the behaviour is a bit counter-intuitive, the pickup magnets (and iron poles if separate components) magnetise the strings which then behave as vibrating magnets moving their magnetic field through the coils. Magnetic field strength thereabouts follows an inverse cube rule so winds nearer the strings are significantly more efficient than winds further away. 


*should that be patent long expired?

[okabass]

Well I managed to fit ca. 14 kΩ per coil using  AWG 43 and 22 mm magnets. 
www.dropbox.com/s/g5ymscshkcc42kj/20180615_013354.jpg?dl=0

Sounds quite good. I think it is similar to Seymour Duncan stack PU.