Does hand winding mean hand on the wire?

[R.G.]

If the pissing match has died down, there were some useful things to talk about.

The issue of identical coils is a difficult one. How can they be identical unless you wind the first entirely by hand with a computer mapping every movement and then replicating that on the entirely machine-wound bobbin. I can't give you a difinite answer to the question of would they sound the same. I'd love to do the experiment for real but I don't have the time or the computer systems to do it.

And that is one of the issues in question. Until recently, that wasn't even an option. But there are schemes now that makes this a possibility. We may be able to do that experiment soon.

The cachet attached to older hand-wound pickups is really more a reflection of the way in which hand-winding increases the variance across pickups such that some are true "tonal outliers", in a statistical sense; they weren't ALL winners - they couldn't be - but some were truly great and distinctive. At the other extreme, the need to run a business and have a product line, implies that one should respect the customer's expectations and provide a consistent product, and consistency implies mechanization and reduction of variance. If you could run a business doing ONLY hand-winds for wealthy customers willing to wait, and willing to accept the result, then you can afford to hand-wind in the purest sense. If your clientele expects to place an order and get a pickup with a known sound at a reasonable price within a reasonable period of time, AND your landlord likes the rent cheque delivered on time, then many aspects of hand-winding will need to be set aside in favour of mechanization.

The thing about low cost computers is that they enable you to do both. You can hand wind for variance, picking out the good ones. This scheme is in a sense what engineers now call genetic algorithms - using semi-randomized recombination of state vectors to get a sampling of the result space. This happens to turn up near optimal solutions which other directed algorithms don't.

But I'm reminded of the question about what to do when you find a gold nugget. When one is looking for gold, one searches until they either give up or find a nugget. Once you find a nugget, you have to choose - do you keep wandering around using the same search methods, or do you dig where you found the nugget?

It seems like once you find a nugget, you ought to dig there - that is, remember how you made the especially-great pickup, and replicate it. Or at least replicate it as closely as you can when that's the right pickup for the situation, playing style, player, guitar, phase of the moon, etc.

That's a good point about repeatability. As random as hand guided winding is (I still call it "hand winding" because not too many people ever actually hold the bobbin in their hand), I think we can agree that once you come up with a formula, they do all sound the same. And that's a good thing.

I know we've had this discussion in the past, but it seems that there is a difference between a neatly machine wound coil, and one with a scatter pattern. It doesn't seem to matter what the scatter pattern is, as long as it has some randomness to it.

That's is an interesting conclusion. Is there agreement on that? Exact scatter doesn't matter, as long as it's scattered?

Is there varying degrees of scatter? Or does it just affect the size and tension of the coil?

Also a good point. Has any testing been done on that?

If I had the opportunity to try winding pickups on a CNC winder, I would. There are a lot of great sounding pickups out there that are not hand wound, so I think, and this is my opinion, that if you can program the correct tension and scatter, you shouldn't be able to tell the difference.

I happen to be in the camp that Mother Nature doesn't care if someone's hand touched the wire as it went on the coil, and all that She cares about is the resulting coil, including all its minutae - wire positioning, tension, etc. I don't think She remembers who touched the wire after the coil is done.

It might even be that those settings must vary as the coil is wound. I'm sure I don't apply consistent tension as I wind, as I often change the speed of the winder to see what I'm doing, and I know the scatter is random.

What leads me to think this is hand carving guitar necks. If you have a nice neck shape and carve it on a CNC router, what's the difference? It's more precise, and you can see that. I have a Warmoth neck here that a customer supplied for a guitar I'm putting together from parts. I don't care for the feel of the neck, but it's nicely made. The Warmoth brothers don't play guitar, so how would they know how to shape a neck?

But I'd love a CNC setup. It would allow me to remove the drudgery of building, and still make the parts the way I want. I love carving necks, but some of the stuff is just tedious and a lot of work!

And that's a good statement for why one ought to test and think about how the winding is done. Knowing what's good and selecting the final results to match what's good is what matters, right?

There's a few people like Don Mare who swear by a precise, repeatable scatter patter. I've never been able to repeat a scater pattern so I tend to agree with you Dave that just so long as it's random it works.

If I thought a CNC could turn out consistently brilliant pickups that sound as good as mine I'd be tempted to use one. But whilst I am not convinced and no one has been able to show me proof I am going to stay with hand winding.

That's a fair enough statement.

Would you be willing to help (in a modest, non-time-interfering way) in trying to replicate a scatter pattern?

[David Schwab]

I happen to be in the camp that Mother Nature doesn't care if someone's hand touched the wire as it went on the coil, and all that She cares about is the resulting coil, including all its minutae - wire positioning, tension, etc. I don't think She remembers who touched the wire after the coil is done.

I really think as long as you have some kind of randomness to a pickup, it will sound the same, as long as all the other criteria are met.

I say this because all my pickups have a certain tone, and if I'm winding the same design more than once, without changing wire gauge, number or turns, coil geometry, etc., they sound the same. I've even wound them with pole pieces and then blades, and it would be hard to tell the difference.

So if you program a winder for your approximate hand tension and scatter, that should do it. And also the tension and scatter probably vary as the coil is wound, adding another layer of randomness.

I remember reading an interview with electronic music pioneer/composer Wendy Carlos. She had an early digital synthesizer, and it had a feature where you could hand draw a waveform with a light pen on a screen. She said surprisingly many of the really way out shapes she drew all sounded like triangle waves or something with a very uniform shape.

So we must be sensitive to some changes more than others.

I'm game to do tests with a hand wound pickup verses a CNC wound pickup of the same design if someone here who has a snazzy winder wants to try it out.

[kevinT]

I really think as long as you have some kind of randomness to a pickup, it will sound the same, as long as all the other criteria are met.

So if you program a winder for your approximate hand tension and scatter, that should do it. And also the tension and scatter probably vary as the coil is wound, adding another layer of randomness.

I don't think CNC coil winding machines are as precise as folks think. I actually see some randomness occuring even as the wire is being wrapped around the bobbin on my machine. In fact, i see some peculiar things happening as the wire is being wound...and when i program a very large pitch, here are some crazy patterns that are occuring naturally--its hard to explain . It could be my machine, or it just could be a characteristic of winding with a machine.

Just like hand wound pickups--each winder's pickup will sound different, it also is the case with regarding to machine winders----the pickups will also all sound different because of all the small quarks of each winder's machine and the many parameters that one can program into the machine.

[Spence]

Quote:
There's a few people like Don Mare who swear by a precise, repeatable scatter patter. I've never been able to repeat a scater pattern so I tend to agree with you Dave that just so long as it's random it works.

If I thought a CNC could turn out consistently brilliant pickups that sound as good as mine I'd be tempted to use one. But whilst I am not convinced and no one has been able to show me proof I am going to stay with hand winding.

That's a fair enough statement.

Would you be willing to help (in a modest, non-time-interfering way) in trying to replicate a scatter pattern?

Yes, why not.

BTW, Bare Knuckles are a big business and everyone of their pickups is handwound. I'm pretty sure that other than custom stuff, no one, rich or poor has to wait.

[mkat]

Can the same wire placement, side by side, that is produced by a machine be replicated with hand winding (ie. hand guided and tensioned by holding the wire in one or two hands) consistently over a few hundred layers? Maybe if winding at an unproductive speed.

It seems that there is no standard definition of "hand winding", subjective. Is there a term for a combination of both?

EDIT:
Ignore the above. What I thought I was getting at has been covered in another thread.

[David Schwab]

Can the same wire placement, side by side, that is produced by a machine be replicated with hand winding (ie. hand guided and tensioned by holding the wire in one or two hands) consistently over a few hundred layers? Maybe if winding at an unproductive speed.

I really doubt it. You can do many things on a machine winder that would be impossible by hand. You see it on various coils like inductors. I don't know if any of that style of coil would be good for pickups... probably not!

On the other hand, I do think you can program a winder to simulate hand winding, but first you have to figure out what hand winding does.

At the very least I bet you could get something that would sound very close, if not maybe different.

[Mark Hammer]

The issue of whether scatter is scatter is scatter is debatable. I am reminded of the way "noise" is conceptualized in the modular analog synth world. You would think that if a circuit produces what is identifiably white noise (all frequencies represented in no particular pattern) there would be no need for any other circuit to do the same function, barring adjustments to suit other supply voltages and other operating contexts. But you'd be wrong. One random does not always equal another "random", and when synth users/players do something like use noise source A vs B for sample & hold purposes, they find clearly audible differences in the nature of the randomness obtained. Some noise sources, while noisy, are not nearly as random and non-repeating as others. Same thing goes for so-called random number generators.

That's not to say that one version of scatter is more purely random than another, but the degrees of freedom afforded by something that has to be produced within specific physical parameters (THIS bobbin shape/size, THIS wire gauge and associated tension/winding-speed, THIS number of turns) implies that there WILL be different kinds of scatters that vary somewhere between something purely systematic and orderly (all turns line up like little soldiers, one beside the other until you reach the flatwork, at which point you simply change direction) and something that is "purely random" (any turn can be at any angle, with no particular location or circumference ever replicated).

In the case of my own efforts (and one should NOT consider them exemplary), the randomness and scatter I tend to introduce, generally results in a bit of a bulge in the middle, largely because one starts to worry about not having turns near the flatwork slip over the flatwork. In other words, as "random" as it may be overall, there is a form of regression towards the mean, whereby most turns, regardless of their angle or scatter, tend to be clustered somewhere near the geographical middle of the coil. It is hard for me to imagine that everyone's scatter turns out that way, or that scatter resulting in an asymmetrical profile on a coil (more turns clustered toward the top or bottom) behaves exactly like a beautifully parallel/flat profile coil with "scatter inserted", or a nicely symmetrical scattered coil bulging in the middle.

Again, I will emphasize, that does NOT mean that somebody has got the scatter down "right" and others have it wrong. Rather, there will continue to be tonal variations (many of which will be hard to detect unless under VERY favourable conditions) associated with different means of producing scatter, and different properties of coils obtained when all that scatter is added up at the end. The proposition that "scatter is good" because it seems to add something, does not automatically imply that it adds the same "thing" every single time, independent of how the peculiarities of that scatter are produced.

[mkat]

I really doubt it. You can do many things on a machine winder that would be impossible by hand. You see it on various coils like inductors. I don't know if any of that style of coil would be good for pickups... probably not!

But isn't this what some pickup makers are selling, some of it marketed as hand wound even though it clearly isn't (well, according to my definition...)?

On the other hand, I do think you can program a winder to simulate hand winding, but first you have to figure out what hand winding does.

At the very least I bet you could get something that would sound very close, if not maybe different.

It would seem that way. As a software engineer, I understand the possibilities of writing algorithms to program virtually any pattern into the software. I'd be very surprised if this isn't happening now already, ah Wolf's winder is one example that I think has the ability for a myriad of scatter patterns from memory but don't recall what the tensioning device was.

Again, I will emphasize, that does NOT mean that somebody has got the scatter down "right" and others have it wrong.

If a scatter pattern cannot be easily replicated, then wouldn't close enough would be good enough to a pattern that is known to work?

[David Schwab]

But isn't this what some pickup makers are selling, some of it marketed as hand wound even though it clearly isn't (well, according to my definition...)?

Well I don't know... But I do know that there are machine wound pickups that sound great. I've never heard a Duncan pickup I didn't like. I might not use some of them, but they have a good tone. I don't think they hand wind all of them... just the Antiquities? (I know Seymour signs those pickups).

Personally I'm interested in the end results. If I could do that on a machine and they sound the way I want them to, that's a good thing I think. But that's just me... I like using power tools and jigs and stuff to build things, even if I can do it by hand. I like the repeatability.

It would seem that way. As a software engineer, I understand the possibilities of writing algorithms to program virtually any pattern into the software. I'd be very surprised if this isn't happening now already, ah Wolf's winder is one example that I think has the ability for a myriad of scatter patterns from memory but don't recall what the tensioning device was.

Exactly. The thing it discovering an algorithm that works.

If a scatter pattern cannot be easily replicated, then wouldn't close enough would be good enough to a pattern that is known to work?

I think the fact that we can't humanly repeat an exact scatter from one pickup to another, and yet they will sound consistent if all other things are equal, has to be an indicator that as long as you replicate the hand winding process within some kind of tolerance, the pickups should sound like the ones you wound by hand.

You have to look at each parameter... you are applying tension with your fingers. Now we can do things that require minute feedback that is hard to do with machines, but how much does the control over the tension matter while you are winding? How many of you vary tension while winding? I think I try and keep it consistently tight, but maybe someone has a certain technique where that matters. But then some winders use a felt pad and some kind of clamp for tension. So that's a static tension being applied.

So, there's one parameter. We need control over tension. Maybe. Perhaps with some kind of feedback control based on an algorithm.

Then there's the scatter pattern. That shouldn't be too hard to replicate. And it can include a randomness factor, where it cross fades to a few algorithms. At the very least don't lay the wire in neat rows next to each other. Maybe skip a few rows and then vary that.

Then keep the speed keyed into the rest of the winding process.

I guess one could program coil shapes too... maybe you want a bulge in the middle?

I bet Possum has some insight in these things, since he has a spiffy programable winder he designed.

[kevinT]

I'm sticking to my guns on this one. IMO ....A true handwound pickup is one in which the winder hand guides the wire and applies tension to the wire simitainously as the wire fills the bobbin which turns on a machine (generally an electric driven machine, but not ruling out hand turned if the person has the time and patience to crank it 5000 times ).

Guiding and tensioning by the fingers is the key to the definition.

Hand guiding the wire using some other aparatus attached to a machine with some kind of machanical tensioner (ie. clamp with felt and knob) should be termed a hand guided pickup.

I think it's our obligation as pickup makers not to confuse folks even more.

I do both hand and machine, so I have no interest in scewing the definintion to my advantage. I just think it needs to be accurate and a generally accepted definition.

If one is programing scatter on a CNC machine, i don't think there is anything wrong with stating that they are simulated scatter wound, or simulated hand wound.

[JGundry]

After getting R.G.'s suggestions I figured out a "quick and easy" way to copy a hand guided or machine pattern in real time. First you attach a linear potentiometer to a wire guide shaft that you control with your hand. You want a sort of rack and pinion set up with the pot. You can simply put a rubber collar around the shaft of the pot and have a press fit against your wire guide shaft. Then hook up an old analog multimeter to the pot. Then take a video camera and videotape the movement of the meter while you wind. You can either read the resistance change from a linear taper pot or run 10 vdc across it and read the voltage change. Then play back the video through editing software like imovie and graph it out on paper so it can later be programmed into your winder. Since the video tape is time coded you will have 30 frames of resolution for each second which should be more than enough resolution unless you are doing a super fast scatter. It would take forever to plot out and program in but this should work. I am going to do this with my Leesona 102 so I can program in every quirk of the traverse into my computer controlled winder.

[Spence]

Interesting idea Jon. Does your computer winder have a 3 bobbin capability too. I'm trying to figure why you wouldn't just use the Leesona.. I suppose if you can replicate the action of the Leesona then you can sell it on.

[JGundry]

Interesting idea Jon. Does your computer winder have a 3 bobbin capability too. I'm trying to figure why you wouldn't just use the Leesona.. I suppose if you can replicate the action of the Leesona then you can sell it on.

My computerized winder will do 6 bobbins but I can't rig the Leesona tensioners to work with it. The original tensioners are meant for 40AWG and I think this fact is important to wire stretch etc. Part of the reason to program the winding pattern into the computerized winder is so I can easily visually analyze the pattern the Leesona 102 is doing. The benefit and problem with computerized winders is the linearity and perfection of the traverse. If I can take a look visually at the Leesona 102 pattern I will be able to see what the cam adds to the pattern and some other variables that the machine ads and evaluate them very precisely since it is time coded. I may have two price levels of pickups one done on the Leesona 102 with the original tensioners and one done on the computerized winder but I'm not sure. One other reason to profile the Leesona traverse is that the machine set-up for the Leesona is a pain but the set-up does make a differencee in how the coil is wound. So once I get the Leesona 102 where I want it I may want to leave it. But the computer can change set-up in an instant. So experimenting with the computerized winder will be much easier and would translate into easier set-up changes of the Leesona 102 if I have a profile of the Leesona traverse in my computerized winder. I hope that all makes sense.

[Spence]

The problem with mechanical things like cams is that they wear. If that machine was used for any length of time the profile of the cam lobes will have flattened and you'll be getting something different from the original spec. That's just one component in the chain. Are spares available?

[JGundry]

The problem with mechanical things like cams is that they wear. If that machine was used for any length of time the profile of the cam lobes will have flattened and you'll be getting something different from the original spec. That's just one component in the chain. Are spares available?

No spares available. But the Leesona 102 is way over engineered for what it does and it is built like a tank. It has several oil reservoirs and the manual gives a detailed oiling schedule for the machine. My machine although it looks filthy is mechanically meticulously maintained. I have taken my Leesona apart to reverse the winding direction and I can assure you that it has almost no visible wear to the cam. The cam itself sits in a pool of oil. The act of reversing the winding direction actually is a defacto clutch overhaul with the Leesona 102 as well. If wear were an issue then you would be correct but it is just not an issue with this machine.The cam is large and the lobes are perfectly in tact. My father restores Studebakers as a hobby and I have helped him since I was an early teen so I'm well aware of what a worn cam looks like. I also know that he often takes an engine apart to find that even with the number of miles the engine has seen the cam is still within factory tolerance. So maintenance is everything.