So the next thing was the pickup. Basically, a magnetic pickup is just a coil of wire with a magnet below it. The magnet slightly magnetizes the string, which then vibrate around the coil, which changes the magnetic field slightly. That magnetic field then induces a voltage in the coil, like in a generator, transformer, or dynamic microphone.
Usually, there are also iron (or steel) rods under the strings that go through the coil (pole pieces), which get magnetized by the magnet and then magnetize the string above. If the pole pieces can be adjusted up and down, you can adjust how loud an individual string is.
There also is a magnetic pickup type that consists of two coils, the humbucker, which is the type I want to use. They work on the same principle, but instead of just having one coil, they have two next to each other (or stacked, but I won't get into that now) which are connected in series, but one in reverse polarity (so connecting both inner or both outer wires together if they're wound in the same direction). This means that in theory, if both coils have the same amount of windings, the signals cancel each other out. This is great for reducing noise, as it should be canceled out, but the actual signal shouldn't get canceled. To achieve that, on one of the coils, the magnet faces the other way, so there is one with the north pole and one with the south pole facing the strings. That way, the polarity of the voltage induced in one of the coils is opposite to the one of the other coil, and since one of the coils is also connected in reverse, the voltages add up again, while the noise that's received by the coils still cancels out, so the signal to noise ratio should be better.
So that's the theory on how they work. Now from searching around a bit, I found that apparently, you need about 30-40g or wire for one pickup. I had a 450m roll of 0.1mm wire, so that wasn't very helpful (also, that site was using 0.0635mm wire, so it would be even longer). With some math (ignoring the density of the enamel on the wire), I figured out that I would need 580-1500m of wire, more than I had. The DC resistance of pickups is also apparently in about 6-12kOhm, and measuring the resistance of my spool of wire, it was at 870Ohm. Because it's fairly quick to try out though, I just soldered some wires to both ends of the spool and connected it to the line-in port of my PC. And with the coil placed maybe 1.5cm under the strings, with a magnet inside the coil, the signal already clipped, and with the magnet removed, it still clipped a little bit.
With it looking like I may actually have enough wire, I decided to just try it (worst case, I have to order some wire and wind new pickups), so I modeled a bobbin to wind the wire around and printed two.
View attachment 316809
Now I made a bit of a mistake here: it would've been better to not just connect the upper and lower plate on the sides, but throughout the whole length of the coil, as one of the sides separated on both coils, which caused some issues.
However, I did not know that it would cause some more annoying issues later, so I started winding.
Now obviously, I'm not going to wind 450m of wire by hand, that would take way to long. So instead, I did it on my lathe (though it would be perfectly possible with a drill, but I have a lathe, so that was the easier way for me). First, I have to mount the bobbin somehow. To do that, I took a piece of wood and cut it to the width and length of the bobbins base plate, and that makes alignment easier. That piece of wood was then screwed to a dowel, which was mounted in the lathe. The bobbin was then attached to the wood with double sided tape in the middle and some painters tape to hold down the ends, as they liked to bend up a little bit.
View attachment 316810
The rod in the tool holder is just there to guide the wire and to give me a reference of how much to move it side to side.
So after securing the start of the wire with some tape and winding a few windings by hand, I wound the rest at 100-250rpm. Now a bit of an issue was judging how much wire to put on each coil, as I had to split what I had had rather evenly (I didn't get it too even).
As you also might be able to see, I put a few pieces of threaded rod into the holes for the pole pieces. I didn't do this the first time, which was a bit of a mistake, as the wire squeezed the bobbin a bit and shifted the top plate, so the holes didn't line up perfectly anymore.
However, after winding both coils, I had one coil at 540Ohm and one at 330Ohm, so the balance isn't that great. This should just affect the noise canceling though, which won't be quite as good as it would be otherwise.
Now, two rather loose coils aren't very good for mounting them to a guitar, and I also didn't have a way to adjust the pole pieces yet (and I haven't even put them in), so the next step was a base plate.
I made that out of some 4mm acrylic, as I thought it might be a bit easier then making it out of aluminium, and using aluminium wasn't really needed here. Steel might have messed with the magnetic field a bit, so I didn't want to deal with that.
The base plate basically is just a rectangle with a few holes in it to screw down the bobbins, and 16 tapped holes for the pole pieces, M5 for the second coil and M4 for the first one (because the holes were out of alignment too much to reasonable fit an M5 thread through there).
View attachment 316811
One of the M5 pieces is still missing here as I had to drill out the holes in the bobbin a bit to fit it in, and one of them doesn't have a slot to adjust it. That's because when I glued the bobbin back together after it broke, I used that piece to align it, and I accidentally glued it in. It's at a decent height though, so I just have to adjust everything else around that one.
For the magnets, instead of using one long magnet, I have a small 5mmx2mm round magnet for each pole piece (mostly just because I could get those locally).
Now that's the pickup pretty much done, I also made the neck a bit thinner. At 32mm thickness, it was a bit uncomfortable, and it got thicker too close to the first fret, which made it even more uncomfortable. I thinned it down to about 27mm, and moved the spot where it got thicker again further towards the headstock, which made it a lot more comfortable.
View attachment 316812
(towards the left it's already thinner, but on the right it's still at the previous thickness)
View attachment 316813
(And now with everything thinner)
With the neck being thinner, it now also bends more under the string tension, but that's what the truss rod is for. However, the neck is supposed to be slightly concave, so it works pretty well.
View attachment 316814
So here's the current state of the guitar. I still need the clamp to hold the nut in place, as I didn't fill the slot enough, but that's something that's fairly easy to fix. The strings also don't line up perfectly with the pole pieces, but it works, so that's not an issue. It's also fredless again now, sorry.