Homebrew guitar pickup winder

[defspace]

I'm a bit on the fence if this actually counts as a handmade item. If nothing else, I figure it fits right in with Vsalon's over the top, no stone unturned ethos.

Inspired by @VertigoCycles' guitar and amp builds, I'm sharing my recent guitar hobby project. I designed, printed, and assembled a semi-automatic guitar pickup winder.

For those unfamiliar with the construction of guitar pickups, they're pretty simple. Many (5,000+) turns of very thin (42awg - 44awg) copper wire are coiled around a plastic bobbin. In the case of Gibson style pickups, a bar magnet is placed below the coils. When steel guitar strings move in the magnetic field a current is induced in the coil and can be sent to an amp.

I drew up the design in Fusion 360:


The bobbin will be mounted to the disc in the rear. The off axis cylinder on the left is a cam that will rotate and push the traverse bar (near horizontal bar) back and forth to guide the wire back and forth across the bobbin. The limit bar (distant horizontal bar) will allow me to use collars to limit the wire travel so I'm only winding on the bobbin. The three grey blocks on the right are displays for wind plate RPM, wind count, and traverse cam RPM.

I designed it to print in three sections, the winder enclosure on the right, the traverse motor enclosure on the left, and the middle section in the middle.

After catching a few errors and redesigning, the final product looks something like this:




I wasn't clever enough to figure out how to do the math to design a cam that moved exactly like I wanted right off the bat. I ended up just printing and evolving the design bit by bit. After a while I stopped including the part of the design that held the magnets for the traverse cam speed sensor, just so I could print cam designs faster. Wouldn't you know it, it was lucky cam number 13 that was good enough to keep.


The cam design isn't perfect, but the coil comes out reasonably even. At this point, I'd rather spend the time to design and build version 2.0 with stepper motors and Arduino control than continue to spend time tweaking the cam shape. The old Gibson factory mechanical winders didn't wind perfectly either, so I figure I'm in good company.

I'm tensioning the wire by hand. You can see that the edges of the bobbin are slightly warped. That shows that I was putting too much tension on there. I have some ideas for a mechanically controlled tensioner. I've left a notch cut out in the middle section of the winder to allow me to clip it in place.

Here's a coil all wound up, waiting for it's partner to be wound:


Here's a short video showing an early iteration of the machine in action:


By logging the wind plate speed and cam speed, I can estimate the number of turns per layer. If something sounds good, I can check my notes and rebuild the same pickup again in the future. When winding entirely by hand, I wasn't able to be very consistent. Imagine winding up a ball of yarn; you'll never do it the same way twice.

[VertigoCycles]

Super cool! Thanks for sharing that. Doing it imperfectly, like the Gibson winders, can't be all bad right? The Throbak guy seems to be making a good go of it.

[Mark Kelly]

By logging the wind plate speed and cam speed, I can estimate the number of turns per layer.

I used to have a coil / transformer winder, a commercial unit from the salad days of V/T electronics. It was somewhat like a bigger, motorised version of this one:

winder1.jpg

(pic courtesy University of Surrey)

The thing was wonderful to watch operating but more importantly it had a simple resettable counter to give turns per layer and total turns. It seems to me that would be a simple addition to your machine.




FWIW I gave the winder to the Melbourne Audio Club when I got out of the audio electronic business, they probably still have it.

[ColonelJLloyd]

Very cool. I think Lollar wrote a book on building your own winder. Have you seen it?

[defspace]

Very cool. I think Lollar wrote a book on building your own winder. Have you seen it?

For some reason I thought it was out of print, but sure enough, it's up for sale on Lollar's page. I just placed my order.

The thing was wonderful to watch operating but more importantly it had a simple resettable counter to give turns per layer and total turns. It seems to me that would be a simple addition to your machine.

I've got a wind counter on there, but the TPL is a bit trickier. I'm tensioning by hand, and the tension applied slows the winding motor a bit. Without any resistance, at full power, the winder will run at about 1210 RPM. With a bit of tension to produce a nice coil, the winder bogs down to about 1150 RPM (I can hold it steady +/- 5rpm or so). The traverse motor doesn't speed up or slow down due to tension. I figure I get two layers per full revolution of the cam. If the winder sits around 1150 RPM and I'm shooting for 50TPL, I'm looking for 11.5 RPM on the traverse cam. I've got the traverse cam with four magnets on it to increase my sampling rate, so if I set the traverse readout to 46, and wind around 1150 RPM, I should be close to the mark. I made a reference sheet comparing various RPMs and the resultant TPL. I can track TPL and design to it, but it ends up being the product of other variables rather than a parameter I'm influencing directly. Seems like some pickup winders (the people, not the machines) really care about TPL, others don't. I cant imagine anyone winding by hand has a particularly precise TPL.

[Mark Kelly]

Is it a DC motor? If so I can show you how to stop it slowing with increased torque. You need to know the DC resistance of the motor winding.

[defspace]

Is it a DC motor? If so I can show you how to stop it slowing with increased torque. You need to know the DC resistance of the motor winding.

Indeed it is! 12v DC. I'm reading anywhere from 13 to 30 ohm (mostly around 25ohm) depending on the rotation of the spindle.

[Mark Kelly]

OK what is probably happening is you have a winding resistance around 26 ohm and there a re some points where your comutation is in contact with two windings in parallel.

The trick is to construct a motor drive which has an output impedance that is negative by just less than the winding resistance, eg about -24 ohms.

Assuming you have a normal DC motor supply with a speed control that puts out a set voltage, what you then need to do is increase this voltage as the load increases.

When the torque demand on the motor rises it increases the current draw, which increases the voltage drop across the windings. Since the forward speed is set by the difference between the supply voltage and this voltage drop, the motor slows down.

If the supply voltage moves with the current draw, this voltage difference stays constant so the motor doesn't slow.

You need to measure the current: a simple Widlar current mirror is useful, see 11.8 here

If you add a resistance to the collector output of Q2 you get a voltage that is proportional to the motor current (the motor is the load shown as R1 above). If the resistance is variable you can tune this voltage.

Find the voltage control signal on the DC drive and insert an op amp adding buffer into the loop, connect the current mirror voltage as the second input and you have a motor supply with negative output impedance. If the sum of the impedance and the motor winding resistance is zero the output speed will stay constant with load. If it goes even slightly negative the thing will become unstable, so it is best to tune to get a very small positive sum.

[defspace]

Mark,
Thanks for typing that up and sharing your knowledge. I have to admit that it's over my head right now. I may come back to it later after I brush up a bit.

I've been using the drop in motor speed as an indirect measurement of tension, so I don't mind the motor slowing. However, I recognize it would be best to be able to measure and set those two variables independently.

[Mark Kelly]

If you use the method I've outlined* you can put an indicator on the current feedback which will give you a direct measurement of torque: DC motors have linear current v torque characteristics.


* Which is called I x R compensation by the way. You can buy DC drives with I x R built in from Maxon etc but expect to pay Maxon prices.

[Ignignokt]

Super cool. I need to build the plans I bought from Highline Guitars, I have that same wind counter in my build. What RPM are you using for the pseudo scatter wind motor?

[defspace]

Super cool. I need to build the plans I bought from Highline Guitars, I have that same wind counter in my build. What RPM are you using for the pseudo scatter wind motor?

The Highline setup provided a bit of inspiration for this build. I've been shooting around 50 turns per layer, so at 1150 wind RPM, that works out to 11.5rpm on the traverse cam (two layers per 360 degree rotation of the cam). Some people go higher, some go lower. I'll experiment some more when I get some more time.

[Ignignokt]

Nice. Chris at Highline just made a new winder model that is CNC, that looks really cool because you can program the scatter.

I'm working on making some pickups to fit these pickup covers I made. Gonna try to make them like Jazzmaster pickups.

DoOzGlc.jpg

To go in this, a Wandre Spazial style aluminum neck guitar I'm building.


https://i.imgur.com/uElM0ny.jpg

[defspace]

Well, you're definitely going to need to post pics of that project coming together.

What's the plan for mounting the pickup? I really dig the cover. Can you share any information on how you made it?

[Ignignokt]

It's pretty simple to make a die using the method I used-
XX1IjBp.jpg

I made this press die out of some aluminum, then made a dummy of the pickup cover. I put the dummy pickup cover in the female die, poured some aluminum epoxy in, let that dry, then put a sheet of wax (that matches the thickness of the sheets to be pressed) into the female die, closed it up, poured in the male die epoxy. I'm using one of those cheap 12 ton harbor freight presses to press the sheet.
The covers aren't perfect, the edges can be sort of ugly and I'm trying to figure out how to get them a bit more presentable and how to attach them to the flatwork but it works out pretty cool. Would be an easy way to make bicycle headtube badges.

Heres the kind of guitar I'm trying to recreate, they were really strange guitars made in Italy in the 50s-60s that had aluminum necks, fiberglass over foam bodies, and really crazy designs.

https://cdn.shopify.com/s/files/1/21...g?v=1505599953

here is a 12 string thingy I built last
https://i.imgur.com/qWnC1Rv.png

https://imgur.com/a/TSqk856

[defspace]

It's pretty simple to make a die using the method I used-

Only on Vsalon would this be simple! Super cool.

I'd really love to see any build photos you have on that 12 string, it looks like there's some really interesting construction there.