I smell BS. Fibonacci??

[salvarsan]

Anyway, i cant give away my secrets, they cost to bloody much to acquire,...

Even if you DID give away your secrets, it wouldn't matter.

Trade secrets, while necessary, aren't sufficient for a business plan or a production mechanism.

Since this is a forum whose original charter was the demystification of guitar pickups, you may want to take a pass on it.

[tedmich]

simple double blinded listening tests in a statistically meaningful number of randomly chosen and demographically diverse subjects would suffice

If you can "grok" Fibonacci numbers, simple statistics should be a breeze!

(Unless the first syllable receives the most emphasis, FIB-onacci)

[Possum]

Howdy Deacci! I also built my own CNC winder, which uses G-code, the beauty of that is that is that you can write infinite lines of code, and the software itself can handle slow starts and stops and other things as well. I purposefully made my traverse low tech, built of surplus parts and is not interfaced with the winder itself, barely cost $150. This was probably 2008? The software is a simple stripped down CNC plasma cutting software I learned to manipulate for coil winding.

I did a ton of programming experiments for lots and lots of winding patterns I would dream up and many based on my handwinding experience, but more importantly copying vintage pickup patterns from old pickups I dissected. Basically, its takes a pretty drastic pattern to make much of a change in audio response you can use while playing a humbucker; there's just not much height to these coils and they're not very fat either, its a small mass and takes a significant change in patterns to push it one way or the other. So, among the things I tried was "graduated" turns per layer formulas, might even possibly similar to what you did, where say you had 200 turns per layer on one pass, then the next one would be 199 turns per layer and so on. Ultimately the conclusions I came to is that these complex patterns don't really do anything I would call "magical." I did some that were pretty diverse patterns yet in playing they werent much distinguishable from eachother. I learned a ton of useful knowledge in the experiments, but basically I abandoned anything but straight forward winding formulas, that are stupid simple to program and give me the best results for what I do.

My point to you was that if when you built your auto winder, if the only patterns you worked on were the Fibonacci pattern and thats all you've done on that machine, you coming to hasty conclusions. Magical thinking is a big problem when you're trying to master the craft of pickup making, I see it all the time and I experienced myself probably for the first five years. In order to prove that your winding pattern does something "special" you also have to prove the sounds you're getting are different than just using ordinary simple winding recipes that every classic vintage guitar pickup used, except for Leo era Fender pickups. I'm not trying to be mean here, either. In your audio on the video I don't hear anything unusual, sorry just being honest. All I hear is a machine wound humbucker. A machine wound bucker is going to stand out from the hand wound stuff, so if some are buying your stuff who have only played hand wound buckers, yeah they will think its great for the wrong reasons. Winding patterns aren't the be-all definition of tone in a humbucker, the metallurgy and design of the bobbins has a huge effect too. So, by putting all your eggs in the Fibonacci basket, its a risk to use that in marketing, because that kind of marketing tends to backfire over a length of time, if customers of other machine winding pickup makers try yours out and don't find something you said was there.

But what the hell, if you're doing it just for fun, not trying to support a family with it, probably doesn't matter in the long run. Personally for me, my philosophy is to experiment all the time, try the most ridiculous out of the box ideas you can think of; over years of time you find out what doesn't work and its just as important to find out what doesn't work because it eventually leads you to what does. If you want to prove something then you also have to prove that something else won't give you the same results, and thats been an ongoing process for me in that things I have developed have become more and more simplified over the years as I discovered simpler ways to get the same results. My biggest teachers have been vintage pickups, those sounds are engrained in our heads, its what people expect to hear from a guitar, and the joy of this business for me has been figuring out why they sound that way, by deep research into materials and historical research into old technologies etc.

[David Schwab]

It states:

Each of the three pickup sets has its own sound. Wound to ohm ratings of 7.10k in the bridge and 6.51k in the neck, the LP59-Zero are the brightest, clearest set of PAF-style buckers in the line. Like the legendary originals, they feature low output with plenty of headroom. “They let you crank your amp without distortion,” Larkin observes.

The LP59-One set is wound at 7.96k ohm in the bridge, and 7.18k ohm in the neck to deliver a bright and articulate PAF sound with medium output. Larkin says rolling the volume back a little makes them sound almost identical to the LP59-Zero set, but when cranked, they have enough output to produce the classic rock tones of the 1950s.

The bridge pickup resistance of the LP59-Two set rates at 8.06k ohm and the neck at 7.30k ohm to produce a hotter, gutsier, and different PAF tonal “flavour” than either of its two siblings. “This set has a more noticeable difference between the neck and bridge pickups than the others,” Larkin observes. “It can roar and growl with the best of them.”

So it would be fairly easy to tell roughly how many turns he's using.

Those are some odd numbers... 7.18k? 6.51k?

[tedmich]

Fibonacci Calculator

(fun diversion but I still don't know how this addresses pickup manufacturing variability)

[rjb]

Those are some odd numbers... 7.18k? 6.51k?

Measured at 21°C (a Fibonacci number), no doubt.

[salvarsan]

The first 9 Fibonacci numbers are 1,2,3,5,8,13,21,34, 55, 89, 144.

I suppose its plausible (barely) that Fibonacci numbers are usable in a winding setup.
Using AWG 42, PAF bobbins fit at most ~89 winds vertically (winds/layer or wpl) while the scatter winds are ~55 wpl.

Whatever.

If you can control the diagonal lay of a winding, you can wind consecutive layers for maximal crossing instead of perfect lay alignment. With that level of control, you get high reproducibility. It should go without saying that a diagonal wind has fewer winds per layer than a straight lay (1:1 orthocyclic) wind.

The basket weave minimizes parasitic capacitance at RF.





Anecdotally, John Suhr was an early adopter of CNC winding and claimed to vary pitch and density on each of ~60 layers in his single coils.

[Mark Hammer]

They're not "odd", just unrounded. Plenty of pups out there that are 7.2k or 6.5k....or am I missing what you intended?