Magnets & Copper Wire: A Pickup Building Thread

[dwizum]

Thank you for the specific feedback @Bruce Johnson, I was hoping I'd get that kind of guidance!

I've run the motor up to it's full speed (4,000 rpm at 12v, more than I will ever run it at in use) and the platen spun true. At least, far more true than the sloppy plywood platen I had on my last setup. So for me it's an improvement. I'm going to round off the corners and edges at least so they're less grabby. If that's not enough to keep wire from hanging up I'll change it to a round platen. And if the whole arrangement is too wobbly I'll put it on a better-supported shaft and use a belt to drive it, although I was hoping for a simple direct drive. I did get the "high torque" version of this motor which has significantly bigger bearings that most of these cheap can motors so I'm hoping that helps in stability, but we'll see...

With respect to the distance from the feed on the traverse to the bobbin - I was conflicted - I felt that closer would be better for the reasons you stated, but further would be better in terms of dealing with the up/down change in feed angle as the bobbin turns through each rotation. I can understand what you're saying though and in hindsight I'll probably move it closer. Right now its about 3 3/4" inches away from the traverse. My old winder had it more than 7 inches away and I did find what you're saying to be very true.

I tried to measure tension on my velcro tension block with a simple spring scale and it was a little hard to get a steady reading, but it was in the low 20's. I think I'm set there. I may swap the spring clamp to a thumb screw so it's adjustable.

What are you all using for a guide/eyelet on the traverse? I've seen videos of people using a simple plastic block with a vee shaped groove cut in it. I was going to use a plastic block with a small hole drilled in it, since having the wire captive like that strikes me as more reliable.

 

[dwizum]

Well it works!

I don't have the gear motor for the auto-traverse hooked up with it's limit switches yet but I couldn't resist winding a coil, just moving the traverse back and forth by hand.

I drilled a tiny hole in a piece of scrap PVC to use for the eyelet for the traverse and it worked fine. For now, winding "by hand," I just put collars on the steel rod to act as stops.

That coil is on one of the humbucker bobbins I'm going to be using. They're pretty tight, but they need to be since I need to squeeze two of them into a Bartolini BC shaped cover, which is made from wood with 1/8" thick sides so there isn't much room.

For an experiment, I just wound it until it was quite full (probably more full than I'd want in practice). It's 4,500 turns of 42 awg wire. It reads 3.8k ohms. That feels a little low to me, but I'm not an expert. I need to get the wire tension higher, the coil is pretty spongy and loose. I'm hoping that getting the tension up will let me wind a tighter coil and fit at least 5,000 turns on without the bobbin being over-full like this. If not, I'll switch to 43 awg wire so I can fit enough on.

The only other real issue is that the reset button on my counter doesn't seem to work. Which is odd, since it worked fine a few days ago when I was testing things on my kitchen counter. It has a reset wire for a remote reset, so for now I can just touch that to ground to reset it. I'll probably just run that wire to a switch to make it more legitimate.

Otherwise I'm pretty happy! I'm pretty much out of time today I think, but hopefully I'll get some time to fiddle with the tension and try a couple more coils this weekend. And I need to buy more 42 awg wire, this spool is gonna run out real soon, especially if it takes me a few coils of experimentation before I get a good one.

 

[rudy4444]

Well it works!

View attachment 4130725

I don't have the gear motor for the auto-traverse hooked up with it's limit switches yet but I couldn't resist winding a coil, just moving the traverse back and forth by hand.

I drilled a tiny hole in a piece of scrap PVC to use for the eyelet for the traverse and it worked fine. For now, winding "by hand," I just put collars on the steel rod to act as stops.

That coil is on one of the humbucker bobbins I'm going to be using. They're pretty tight, but they need to be since I need to squeeze two of them into a Bartolini BC shaped cover, which is made from wood with 1/8" thick sides so there isn't much room.

For an experiment, I just wound it until it was quite full (probably more full than I'd want in practice). It's 4,500 turns of 42 awg wire. It reads 3.8k ohms. That feels a little low to me, but I'm not an expert. I need to get the wire tension higher, the coil is pretty spongy and loose. I'm hoping that getting the tension up will let me wind a tighter coil and fit at least 5,000 turns on without the bobbin being over-full like this. If not, I'll switch to 43 awg wire so I can fit enough on.

The only other real issue is that the reset button on my counter doesn't seem to work. Which is odd, since it worked fine a few days ago when I was testing things on my kitchen counter. It has a reset wire for a remote reset, so for now I can just touch that to ground to reset it. I'll probably just run that wire to a switch to make it more legitimate.

Otherwise I'm pretty happy! I'm pretty much out of time today I think, but hopefully I'll get some time to fiddle with the tension and try a couple more coils this weekend. And I need to buy more 42 awg wire, this spool is gonna run out real soon, especially if it takes me a few coils of experimentation before I get a good one.

My tensioning / guides are spring loaded wood disks fined with felt. They are shown and detailed on my winder shown in post #120 on page 6.

Most of my coils are wound with 9,000 to 10,000 turns of #42. Your number of winds seem pretty low for a conventional bobbin. Automating your transverse feed might not be ideal, scatter winding has a lot to offer from what I've read about the resulting sound of various pickup wind layer configuarations.

 

[dwizum]

Most of my coils are wound with 9,000 to 10,000 turns of #42.

What size/style of bobbin? humbucker or single coil? What kind and size of magnet?

I'm trying to gather all the bass pickup specs I can find, especially non-fender layouts. There's just so little out there about Bass pickups other than P's and J's...

I like the spring in your tensioner. Somehow I missed that when I read through the thread. I was playing with a new tensioner after my above post and I made one with felt between two scraps of plywood, held down by a wing nut. The problem is, any time you need to re-thread the wire, you have to loosen it up and tension from scratch. Using a spring would avoid that!

 

[dwizum]

I made a new tensioner and checked it by hanging 25 grams of fishing weights on the wire and adjusting until it just stopped slipping. That seemed more consistent than my spring scale. And it made a big difference. My coils are nice and tight. I was still feeling iffy about getting enough turns on these small bobbins so I switched to 43 gauge and would two at 6,000 turns each. (Well, to be honest, one was 6,000 and the other was 6,007 ).

I assembled them into one of the wood housings I have for the bass I'm building right now. But I don't have the hardware for that bass on hand yet so I stuck it into last year's build off bass. Well, I tried to, but I made a critical error with the install. Let's see if anyone can spot it:

After sorting that out I plugged it in to give it a try. It makes noise! Two issues though. One, I think I have the start/finish on one of the coils backwards, because parallel mode sounds like series and series sounds like parallel. Not a big deal. Luckily this bass is set up with a series parallel switch so I could spot that problem easily. Second problem, although the output level is pretty good, they're a little harsh.

If these were guitar pickups, I'd rewind them with more turns in the hopes of taming the treble. I probably have room for another thousand turns. For all you guys who have experience with bass pickups, does that sound about right? Any other suggestions? I'm stuck with this magnet thanks to the specific bobbin I've built around...

 

[Freekmagnet]

I made a new tensioner and checked it by hanging 25 grams of fishing weights on the wire and adjusting until it just stopped slipping. That seemed more consistent than my spring scale. And it made a big difference. My coils are nice and tight. I was still feeling iffy about getting enough turns on these small bobbins so I switched to 43 gauge and would two at 6,000 turns each. (Well, to be honest, one was 6,000 and the other was 6,007 ).

I assembled them into one of the wood housings I have for the bass I'm building right now. But I don't have the hardware for that bass on hand yet so I stuck it into last year's build off bass. Well, I tried to, but I made a critical error with the install. Let's see if anyone can spot it:

View attachment 4132167

After sorting that out I plugged it in to give it a try. It makes noise! Two issues though. One, I think I have the start/finish on one of the coils backwards, because parallel mode sounds like series and series sounds like parallel. Not a big deal. Luckily this bass is set up with a series parallel switch so I could spot that problem easily. Second problem, although the output level is pretty good, they're a little harsh.

If these were guitar pickups, I'd rewind them with more turns in the hopes of taming the treble. I probably have room for another thousand turns. For all you guys who have experience with bass pickups, does that sound about right? Any other suggestions? I'm stuck with this magnet thanks to the specific bobbin I've built around...

I spotted the problem - that’s pretty funny.

If I remember correctly, you’re making humbuckers and winding directly around ceramic bar magnets(?). If that’s the case, I have had similar problems with excessive brightness using that particular structure - in fact, my efforts had a really unstable sound to them. I didn’t have much luck with adding more turns - the bass started sounding really flabby above 7000 each. But, YMMV.

I’ve had better luck using steel cores with the ceramic magnets powering them from underneath the pickup. You could try putting a 1018 steel bar underneath the pickup with both of the magnets making contact with the metal. That’ll increase your inductance, maybe give you some more lows while taming that brightness a bit. The more metal, the more inductance (I think). Again, YMMV.

 

[rudy4444]

What size/style of bobbin? humbucker or single coil? What kind and size of magnet?

I'm trying to gather all the bass pickup specs I can find, especially non-fender layouts. There's just so little out there about Bass pickups other than P's and J's...

I like the spring in your tensioner. Somehow I missed that when I read through the thread. I was playing with a new tensioner after my above post and I made one with felt between two scraps of plywood, held down by a wing nut. The problem is, any time you need to re-thread the wire, you have to loosen it up and tension from scratch. Using a spring would avoid that!

My pickups were occasionally single coils but I wound about 50 sets of "split coil buckers" as 6, 7, or 8 pole pickups. The odd pole configuration is because these were fro custom lap steels. I use Garolite XX in various thicknesses for flatwork and 3/16" diameter magnets. Here's a composite photo of my winder, which despite it's "unusual look" works really well. There's a 8 pole single coil mounted in one of the shots and one of the photos is a six pole split coil bucker.

I haven't wound any of my own bass pickups, so I can't help out with specifics for that.

 

[dwizum]

I spotted the problem - that’s pretty funny.

If I remember correctly, you’re making humbuckers and winding directly around ceramic bar magnets(?). If that’s the case, I have had similar problems with excessive brightness using that particular structure - in fact, my efforts had a really unstable sound to them. I didn’t have much luck with adding more turns - the bass started sounding really flabby above 7000 each. But, YMMV.

I’ve had better luck using steel cores with the ceramic magnets powering them from underneath the pickup. You could try putting a 1018 steel bar underneath the pickup with both of the magnets making contact with the metal. That’ll increase your inductance, maybe give you some more lows while taming that brightness a bit. The more metal, the more inductance (I think). Again, YMMV.

Thanks for the feedback. Good idea on the bar across the bottom, I'll try that. If I can find some scrap that is the right size at the local hardware store...

The other thought I had was to just cut some steel bar to fit these bobbins (they're about 3/16" x 1/2") and then I can put whatever bar magnet I want across the bottom, PAF-style, with a pole touching each steel blade.

 

[dwizum]

Ok I feel totally stupid.

This bass I'm testing in has push pull puts for parallel/series and coul split. The pickup wires for the bridge pickup connect first to the push pull on the treble pot. When I was soldering the new pickup in, I squished the wiring to the pot and made a partial short across the pot. The bass sounded harsh because of that short, not the pickup. I fixed the short and fixed the series/parallel wiring and it sounds GREAT. Pickup is going to stay as-is!

Thank you all for playing along in the meantime. Sometimes when you're deep in this stuff you miss the obvious and go on wild goose chases I guess.

 

[dwizum]

I made a matching bridge pickup this morning, but before I did that I made some test pickups at 5,000 and 6,600 turns to compare to the 6,000 I had on the working one last night. I like all three but 6,000 sounded the most balanced so I'm sticking with that - the 5,000 turn version didn't really have enough oomph. I kept the 6,600 turn version for the bridge position. The 6,000 turn version for the neck is 7.08k per bobbin, 14.16 in total. And the 6,600 turn version for the bridge is 7.8k for each bobbin, 15.6k total.

I took a few photos of the assembly process for the bridge pickup. These shells Bartolini BC size, which are 3.95" by 1.25". The sides are 1/8" thick and the end blocks are just under 1/2". The walnut tops are 1/16" thick. To put the coils in, I made little base plates out of 1/16" acrylic. I put copper tape on the base plate, then two little hardwood spacers with a trough down the middle. The pickup cable sits in a notch in the end of the plate and runs up through the trough. I filled the trough with hot glue and then covered it with another piece of copper tape, with the cable's shield wire captured in the tape.

All the parts:

Base plate put together:

Plate with wire in the trough:

Then I put the bobbins on top of the spacers with a tiny dot CA, connected the wiring and tucked it between the coils, then put a strip of tape across the top and a wrap around the coils to protect the wiring. Makes for a nice compact little assembly that's a friction fit in the shell.

If I were to start from scratch with another bass pickup, and I wanted to stick with a common soapbar shape, I'd probably choose a slightly bigger shape to make the layout easier. It would be nice to have larger bobbins or baseplates on the bottoms of the bobbins so I could thread the coil wire through eyelets or wrap it around little pins to help facilitate soldering to the leads. I basically did these like a PAF guitar humbucker, partly out of familiarity and partly because there wasn't really enough room in these shells to allow for eyelets or posts.

But I'm very happy with these. This pair in these walnut shells are going in the bass I have a build thread for currently, but as soon as that bass is done I'm going to pull the Bartolini soabpars out of my main bass and replace them with a pair of these as well.

Then I want to move on to something a little more out there. Maybe a sidewinder in an MM shape? Or a stacked coil in a jazz shape...

 

[dwizum]

A question as I think about finishing the auto-traverse, or leaving it as-is and continuing to "hand feed" the wire.

How are you all choosing a traverse speed? Do you just move back and forth at a rate that gives an even-looking layer? Are you doing math? Do any of you actually keep track of the speed or think about it? Do any of you know what your fill percentage is, per layer or for the whole coil?

I'm going slower than planned on this new winder, at least for now. Around 600 - 700 rpm. I've been traversing at a rate that seems to give a consistent layer of wire, but I can't say that I know how "full" the layer is. I would a bunch of PAF bobbins today, and I counted off 6 seconds to go from one side to the other as I traversed. At 600 rpm, that's 60 turns per layer. The wire I was winding is .0025 in diameter, and the bobbin is .25" tall. So in theory I'm wrapping layers that are 60% full. I don't know how that translates to an actual fill percentage for the whole coil though, but I've got to imagine it would be much higher, since a given layer is going to naturally sit down in the "gaps" in the prior layer to an extent.

 

[Jeff Siddall]

A question as I think about finishing the auto-traverse, or leaving it as-is and continuing to "hand feed" the wire.

How are you all choosing a traverse speed? Do you just move back and forth at a rate that gives an even-looking layer? Are you doing math? Do any of you actually keep track of the speed or think about it? Do any of you know what your fill percentage is, per layer or for the whole coil?

I went full on math. The winder I built is computer controlled and uses only stepper motors so I know exactly where everything is for the compete wind, down to the nearest 1.6 degrees for the bobbin motor and 20 microns (about 1 mil) on the traverse motor.

You just enter the thickness of the wire and the width of the bobbin and it automatically figures out how to lay perfectly flat layers.

The wind motor speed is limited in my implementation to 500 RPM but I can make it, quite literally, any speed up to that.

Of course that is all overkill but it was cheap and easy for me to do it so I figured, why not?

Now, one of these days I hope to actually get the wire, bobbins magnets and pole pieces so I can start making some pickups!

 

[Freekmagnet]

A question as I think about finishing the auto-traverse, or leaving it as-is and continuing to "hand feed" the wire.

How are you all choosing a traverse speed? Do you just move back and forth at a rate that gives an even-looking layer? Are you doing math? Do any of you actually keep track of the speed or think about it? Do any of you know what your fill percentage is, per layer or for the whole coil?

I'm going slower than planned on this new winder, at least for now. Around 600 - 700 rpm. I've been traversing at a rate that seems to give a consistent layer of wire, but I can't say that I know how "full" the layer is. I would a bunch of PAF bobbins today, and I counted off 6 seconds to go from one side to the other as I traversed. At 600 rpm, that's 60 turns per layer. The wire I was winding is .0025 in diameter, and the bobbin is .25" tall. So in theory I'm wrapping layers that are 60% full. I don't know how that translates to an actual fill percentage for the whole coil though, but I've got to imagine it would be much higher, since a given layer is going to naturally sit down in the "gaps" in the prior layer to an extent.

I’m glad you got your pickup figured out!

My winder has an optical trigger on the platen that triggers a stepper motor by 3-5 increments each time it turns one full revolution. That 3-5 number is controlled by printed patterns on the back of the platen. The increments on the motor are defined by stepper driver (200, 400, 800, 1600). The stepper then turns a guide screw that feeds the wire on to the bobbin. I have a couple of threaded rods with different thread pitches. Between the three of these adjustments, I can get really really close to spot on spacing with 42, 43 and 44 AWG wire. The X-Y limit is a little tricky to set up, but I’ve got it down enough so that I get consistent results. I can run it up to 900-1000 RPM before the sensor starts getting confused.

To answer your question, the traverse rate is determined by the speed of the motor driving the platen. The spacing always stays the same.

 

[Bruce Johnson]

Feed rate:

My winder is mechanical feed, using a double ended leadscrew. The leadscrew is mechanically driven from the spindle, so the feed head moves sideways an exact amount with every rotation of the spindle. It doesn't matter what speed the spindle is turning, the feed movement is the same amount per spindle revolution.

I did the math and made my feed rate 0.0025" per revolution. This is a combination of the thread pitch of the leadscrew and the (1:16) reduction of the toothed belt drive to the leadscrew. With every revolution of the spindle, the head moves 0.0025" to the right or left. That's fixed, unless I wanted to change the belt drive pulley ratio.

I chose 0.0025" because most magnet wire used in making pickups is in the range of 0.0020"-0.0023" diameter. The 41 ga wire that I use is nominally 0.0023". Making the feed rate 0.0025" makes the winds about as tight and compact as you can get. When I look close at my coils with a magnifier, my winder will actually make a single layer like a spring, with all the turns tight in against each other, not overlapping.

That was a decision on my part to make my machine this way, to make all my coils very tight and dense. I'm really not interested in experimenting with lower TPL rates and making looser sloppy coils. If I want to some day, I could rework the belt drive to be changeable.

 

[rudy4444]

A question as I think about finishing the auto-traverse, or leaving it as-is and continuing to "hand feed" the wire.

How are you all choosing a traverse speed? Do you just move back and forth at a rate that gives an even-looking layer? Are you doing math? Do any of you actually keep track of the speed or think about it? Do any of you know what your fill percentage is, per layer or for the whole coil?

I'm going slower than planned on this new winder, at least for now. Around 600 - 700 rpm. I've been traversing at a rate that seems to give a consistent layer of wire, but I can't say that I know how "full" the layer is. I would a bunch of PAF bobbins today, and I counted off 6 seconds to go from one side to the other as I traversed. At 600 rpm, that's 60 turns per layer. The wire I was winding is .0025 in diameter, and the bobbin is .25" tall. So in theory I'm wrapping layers that are 60% full. I don't know how that translates to an actual fill percentage for the whole coil though, but I've got to imagine it would be much higher, since a given layer is going to naturally sit down in the "gaps" in the prior layer to an extent.

All my coils are formed by simply watching the wire and slowly traversing the path to get as flat of a layer as I can. The photos of the women who were winding those highly sought after Fender pickups certainly weren't using any sort of precision measuring equipment. Your coils look great as they are.

Have you read the Lollar winder book? Since it has a wealth of information about winding I picked up a (expensive) one on Ebay, extracted what information I found useful and then re-sold it on Ebay. That made it almost like a library check out, but with no time limit.

I stop winding every thousand turns and paint about an inch layer of polyurethane finish on each flat side of the coil with a small artist's brush. That works really well to prevent any tendency for your pickup to be microphonic.

Since my split coils are visable I capped them off with black electrician's cloth friction tape and top-coated the tape. I capped off a few with heavy black cotton string, but it's more work than the simpler friction tape. Enclosed coils obviously don't need that attention.

 

[dwizum]

I haven't read the Lollar book but some day I probably should. I see it on Amazon and eBay used for north of $200. But you can get it on the Lollar website for $60!

 

[Freekmagnet]

I haven't read the Lollar book but some day I probably should. I see it on Amazon and eBay used for north of $200. But you can get it on the Lollar website for $60!

I had the Lollar book. It was alright. It’s basically a book of plans to build his winder, and a little bit of extraneous pickup information you can find with a quick search on the internet. His winder is pretty good for what it is, but functionally, it’s a mechanical version of what you just built. It uses sewing machine motor for a spindle and an oscillating fan motor to run the auto-traverse. Both motors run independently, so the only way to control your spacing is by reducing your winding speed. It’s really good for making P90s.

The Lollar book got me started with pickup making. I ended up giving the machine and the book away to some kid who said he wanted to make pickups.

 

[dwizum]

I'm just starting to wonder. If people are paying $250 for the book on eBay, and you can order it brand new on Lollar's website for $60, that sounds like a better business to be in than actually reading it and making pickups. Buy a box full, wait for it to go out of print (which apparently it has many times), and cash in...

 

[ardgedee]

Heck, buy them for $60 and sell them for $100 on Amazon. You'll stand the chance to make a sale or two before everyone else's pricing scripts update to undercut you.

 

[Freekmagnet]

I'm just starting to wonder. If people are paying $250 for the book on eBay, and you can order it brand new on Lollar's website for $60, that sounds like a better business to be in than actually reading it and making pickups. Buy a box full, wait for it to go out of print (which apparently it has many times), and cash in...

Yeah, I think Jason Lollar himself was irked about that very thing. It seems that at one time there was a cottage industry of selling bootleg copies of his book for hundreds of dollars. He posts, or at least used to post, over on the pickup maker's forum.

Lollar Pickup Winding book - Music Electronics Forum