Pickup coil winder

Describe the coil you want, get the winding plan and the file the machine runs

Wire

Coil

Measure the core, not the finished pickup. On a Fender single coil the core is the row of magnets itself.

It changes none of the figures: the coil is the same either way. A humbucker only works with exactly one electrical and one magnetic reversal between its two coils, and winding the second one backwards is one of the two ways to get the electrical one, the other being to swap its leads.

Machine

Wire breaks at startup more than anywhere else. Start slow, then come up to speed over the first few layers. Resistance depends on temperature too: at 25 °C instead of 20, your meter reads 2 % more.

The defaults suit most coils. Only touch these if you know what you are looking for.

0 for turns laid against each other. Stacking says whether the layers sit flat (1) or nest into the hollows (0.866). Nobody knows which is right: your first coil will tell you, with a multimeter.
If the wire always turned back at the same spot it would build a bump at the edge, growing layer after layer. So that spot moves by a few wire diameters on every pass.
The even pattern goes back and forth at a constant speed, and it is the best of the three. Pulling back shortens the last layers a little, so the wire does not slide down the flange.

Result

Coil cross section

To scale, one line per layer. The saw tooth edges are wanted: that is what keeps the bumps away.

Machine file

Before you start Bring the wire guide against the flange, on the side you start from. Turn the coil by hand for one full revolution and check that nothing catches. Attach the wire, set the guide start on the machine screen, and go.

        

This page plans the winding of a pickup coil: how many layers, how long each one is, where the wire turns back, and how fast the spindle runs. It starts from the core dimensions, the wire and the number of turns you want, and works out the rest.

The gauge table and the turn length are the same as in the winding calculator, so the two pages cannot give two different wire lengths for the same coil. What this one adds is everything a machine needs and a hand does by feel, starting with the reversal points, which have to move or the edges pile up.

The file at the bottom is for my own two axis winder, an ESP32 driving two stepper motors. Nothing has been wound on it yet: every figure on this page is calculation, not measurement.

Frequently asked questions

Do I need the machine to use this?

No. The turns, the layers, the wire length and the resistance are just as true on a hand winder, and the cross section shows what the coil will look like. Only the file at the bottom needs the machine.

What is in the file?

One line per layer: cumulative turns, the wire guide position in millimetres, and the speed in rpm. A whole coil is about forty lines, which is short enough to read through before you run it. Turns are cumulative so that no error piles up along the file, and the speed is repeated on every line so that a line can be read on its own.

Why do the reversal points move from one layer to the next?

Because if the wire turned back at the same place every time, the extra wire left at that spot would stack up layer after layer and build a bump against the flange. Moving the reversal by a few wire diameters spreads that surplus. The page measures it: with no movement at all the figure is 100 %, well set it drops to about 38 %.

Does it agree with the pickup winding calculator?

On the resistance and the wire length, yes, exactly: the length is the resistance divided by the ohms per metre, and both pages use the same gauge table. On the turn count, no, and they cannot: the calculator spaces the turns 7 % wider than the wire, because that is what a scatter wind does, so its turns are longer. Here the guide feeds exactly what it is told. This page therefore asks for about three quarters of a percent more turns for the same coil, and that is why the button on the calculator sends the resistance rather than its turn count.

Has any of this actually been wound?

Not yet. The machine exists, its firmware is written, and none of it has run on hardware. Every number on this page is computed, including the winding time, and the one thing the model cannot guess is how the layers nest into each other: the first real coil will settle that with a multimeter.

See also the pickup winding calculator and the pickup specs chart.