Hand-winding humbuckers?

[Steve Sawyer]

I'm looking forward to winding some of my own pickups. I've been collecting some parts to build a winder and have been looking at other folks' winder designs and techniques. One thing that I noticed is that the articles and demonstrations are always for single-coil pickups, and was curious about how humbuckers could be hand-wound.

I think I've found an answer - that hand-winding humbuckers is challenging because both coils have to be very close to identical - inductance, impedance and resonant frequency - otherwise the humbucking won't work because the two coils won't be producing identical but out-of-phase signals. My understanding is that there are so many variables at play in a pickup that there are slight variations even in production machine-wound pickups. This also adds to the fun of hand-winding as each pickup will be unique, but could make it extremely difficult to produce a set of matched coils to mount in a humbucker.

Is this the case?

Thanks!

[Jason Rodgers]

You don't have to get the two coils as exact as you think. Just make sure you have a counter on your machine. Even then, some folks have their own special recipes for scatter-winding (not perfect layers) and off-setting (couple hundred turns difference between coils). Now, if you want consistency in your winding, yes, you should figure out how to make your coils as similarly as possible from one coil to the next and one set to the next. Those parameters include wire gauge, wind count, turns per layer, wire tension, and even consistency of your parts suppliers.

[Steve Sawyer]

Thanks, Jason - I'm a long way from taking a crack at a humbucker, but it's good to know that it's not a fool's errand to think about...

[Jason Rodgers]

For your first attempt, just aim for some simple goals: for example, 5000-5500 turns per coil of 42awg for the bridge, 4500-5000 for the neck, and even tension and coil build. The StewMac kits will give you quality parts that all fit together.

BTW, here's my winding adventure http://www.mimf.com/phpbb/viewtopic.php?f=10&t=3037

[Steve Sawyer]

Thanks for the link to your winder build, Jason. That looks really nice.

I have some digital counters, an Arduino and a couple of stepper motors that I'm going to try using. I also have an old sewing machine on a shelf in the garage that might get cannibalized at some point if it looks like going the DC motor route will work better. Either way, it's going to be fun to build.

I have a couple of OEM Tele single-coils that I bought from the clearance bin at GF awhile back, so I have something to put in the Tele I'm building when it's done, but I also picked up a couple of the StewMac single-coil Tele kits you mentioned.

All this stuff unfortunately has to get put away for awhile as I'm still finishing up the shop remodel. The end of THAT is finally in sight - still have a bit of flooring to finish up (in the laundry room underneath the washer and dryer - ugh), but I started assembling the cyclone DC unit on Saturday as the DC system is the next thing I have to install. After that the ceiling needs to be done then I can finally rock 'n' roll again...

[adam johnson]

It depends...

We are going to do some practical simplifications ahead but the important stuff is right, or should be, I am abit rusty. Either way, it is practical and will get you were you need to go, corrections welcome.

A humbucker works by canceling out anything that is equal but opposite between the two coils, -1 + 1 = 0 right? The coils are out of phase and they both collect the same ambient noise assuming they are positioned reasonably close to each other, the same noise on each coil but out of phase, that noise cancels out, -1 + 1 = 0 after all.

So what happens when the coils are not matched? Lets use Jason's numbers of 5000 and 5500, we will also assume these are noisy coils and on their own have 100 mV of noise. 5000/5500=.90, the coils are off by 10%, we will just call that 90% efficiency and 90% of the noise is canceled so our once noisy coils are now a quiet 10 mV.

Lets throw in the magnet and see how this affects the strings.

The strings are not canceled thanks to the fact that the coils do not just have opposite polarity but opposite magnetic polarity, one is south the other is north, the out of phase coils each see an in phase string, so the signals reinforce each other when the coils are summed instead of canceling, but not completely. The two coils each see different parts of the string and different parts of the string vibrate in different ways. We can see this if we look at a vibrating string under a strobe light or we can hear it by playing our harmonics at the 5th, 7th, and 12th frets. These are vibrational nodes, between them we have the loops, a harmonic is the vibration in the loops of the previous harmonic.

Each harmonic is the harmonic + the fundamental, so our E is 82hz, the second harmonic is 164, the third is 246 and so on. The first (fundamental) is the entire string length vibrating, the second is half the string, the third is half that and so on, when you get to the higher harmonics it is just a small part of the string vibrating, a length which could be half the distance between the two coils, one coil sees that harmonic in phase and the other sees it out of phase, so that harmonic gets canceled, or more likely partially canceled since it will not be exactly in phase or out of phase, just abit off. The higher the harmonic the more this can happen. So if our two coils are a perfect pair they can cancel out a fair amount of higher harmonic content, we call this phase distortion.

Back to Jason's pickup with 5000 and 5500 turn windings. Not just the turns are mismatched, the inductance, capacitance, impedance, and distortion* are as well, and they are loaded abit differently by the volume and tone controls as a result. The outputs of the two coils are no longer quite the same, they see the same string but give us different information about the string, so less (different) harmonic content is canceled and our sound is said to be richer and/or brighter, or less pure and noisy if you are in that camp.

Reality check time, lets stand on the shoulders of giants for abit.

Back when Seth Lover was working out that holy grail of humbuckers for Gibson, the PAF, he wound to fill the bobbin. The folks working the winders just wound til it was full, wrapped it in tape and tossed it in a bin for the folks assembling the pickups. So any given old PAF likely will not even have coils wound by the same person, let alone match, but who complains?

Off to Fender and the strat. A modern strat may have 2k difference in DC resistance between coils and yet we get adequate humbucking in positions 2 and 4, especially when compared to the noisy 1, 3, and 5 positions. In Leo's day they wound the coils to be more or less identical, he was a penny pincher and practical, an acoustic is quieter when plucked at the bridge than the sound hole so why should and electric be any different? It was what the musician expects after all.

So to answer your question it depends on the sound you want and the amount of noise you will accept. In reality a well wound single coil is quieter than many amps and most rooms/houses in this noisy world so we have lots of room to play. To start with just do as the old timers did and fill the bobbins and call it good, work on your winding technique, it will have a greater affect on noise and sound quality than abit of mismatch. Get a tight evenly wound coil free of microphonics before you trouble yourself with turns counting, you may find you do not even need it by then. You can get those coils surprisingly close to each other just by filling the coil with a nice winding once you get the technique down, you can always unwind a few turns from the larger coil afterward if you want to.

Adam

*Distorion is the main player here I suspect and is primarily if not completely phase distortion, so more cancelation. This can get out of hand if it is a sloppy uneven coil leading to a thin sounding pickup. An even but randomly wound coil seems to give the ideal amount of distortion for most of us. An even, perfectly wound coil has very little distortion.

[Steve Sawyer]

Adam - a great & informative analysis!!

The more I learn about hand-winding, the more I'm looking forward to giving it a go. Obviously you aim for consistency, but the little variations that inevitably creep in make it a bit like a box of Forrest Gump's chocolates!