Magnets & Copper Wire: A Pickup Building Thread

[Freekmagnet]

Crazy! I love it... thanks for making my morning...I have to admit I've been waiting for the moment this thing came to life whether it worked as expected or not.

Cool - I’m glad to have made your morning!

I’ve been waiting for this for a while as well. I’m not sure if it was a success yet, but we’ll see!

 

[Bruce Johnson]

All very interesting...

Here's one thing to think about: You've created Cubic Humbucking, Humbucking to the third power! Each String Unit (SU) is, in itself, humbucking because it has two coils going opposite directions, mounted on either side of a magnet structure, so the field goes opposite directions through the coils. That's the first level of humbucking.

Next, your magnet structure is arranged so the fields at the upper ends of the blades are alternating NSNS. The second level is that each pair of SU's is also humbucking. You've wired them alternating, and their fields are alternating so, as a pair, they are humbucking.

And then, at the third level, the two pairs of SU's may also be humbucking. You've wired them alternating, although it's not clear if the pair of pairs of blades are operating as opposites.

It gets a little mind-boggling. And explains why you are seeing some funny things with phasing!

One experiment you might try is to take the humbucking out of each SU. That is, on each SU, wire the coils going the same direction. That may significantly change the sound, for better or worse. Within each SU, it's not clear just what the shape of the magnetic field is, and how it's passing through the coils. Since they are round and turned sideways. Going around the coil, are the signals being generated by the moving magnetic field lines adding together, subtracting from each other, or what???

That's the biggest question of this design. It has the potential for an enormous amount of power (generating signal) and fine detail. But you've got to figure out how to get it all working together, not partially cancelling each other out. This is complicated pickup design.

 

[Freekmagnet]

One experiment you might try is to take the humbucking out of each SU. That is, on each SU, wire the coils going the same direction.

That’s a really good idea that hadn’t occurred to me. If I wired both coils in each SU in the same direction, that SU would be out of phase with itself, but it might be in phase with adjacent SU. And considering the behavior of this design, it would make perfect sense if it does work that way.

That’s some thinking outside of the box there, Mr. Bruce. I’ll try that tonight when I get home. Thx.

Yeah, lack of fine detail doesn’t seem to be a problem with this pickup so far. More like, too much fine detail.

 

[NKBassman]

This might be a dumb question, but could you get away with just a single coil between each string? ie one magnet/coil shared by two adjacent strings?

 

[Freekmagnet]

This might be a dumb question, but could you get away with just a single coil between each string? ie one magnet/coil shared by two adjacent strings?

It’s not a dumb question. I’m sure it would work - but then you’d have only 4 coils running. Might be easier to execute, but there’d be less output. I’m not sure what effect the asymmetric design would have on the sound.

 

[Will_White]

It’s not a dumb question. I’m sure it would work - but then you’d have only 4 coils running. Might be easier to execute, but there’d be less output. I’m not sure what effect the asymmetric design would have on the sound.

You could do it with 6 coils probably,

Edit: nope it would be 5.

 

[NKBassman]

It’s not a dumb question. I’m sure it would work - but then you’d have only 4 coils running. Might be easier to execute, but there’d be less output. I’m not sure what effect the asymmetric design would have on the sound.

The way I was thinking about it, it’d be 5 coils as you’d likely have one on either end as well?

 

[Freekmagnet]

The way I was thinking about it, it’d be 5 coils as you’d likely have one on either end as well?

View attachment 3478904

It’s a good idea - I’m going to try to get the “pairs” approach to work before I start thinking of a redesign. I’m pretty sure it’s a matter of the order in which the coils are wound.

And incidentally, I did one experiment where all of the coils except the outer coils were wired together, so your idea might actually work. It’s just that the inside poles were a little louder than the outside poles.

 

[Bruce Johnson]

I don't think the 5 coils idea would work. One of the SU's would be twice as powerful as the others, and you couldn't get a balanced humbucking.

You could do 4 coils; one coil for each SU, fastened on one side of each blade. The blades are alternating fields, so you'd wire the coils alternating and in series, and the whole thing would be humbucking. I'm sure that would work okay, and could be an interesting experiment in itself. And of course, it would be a lot cheaper to build. But it would only have about 1/3 - 1/2 of the potential power and definition of the 8 coil version.

Again, the central thing to this design is the shape of the magnetic field around each of the blades, because of the interesting magnetic structure you've built. What's the best way to position the coils to get the maximum saturation of the field through them? Probably about like you've got them, fairly narrow coils mounted tight on either side of each blade. But, what's the direction of signal being created within each coil? Are they adding together, or partially cancelling each other? The phasing of the signals coming from each coil is complicated. You may just have to find the answers by trial and error; connecting them up in different ways.

You may want to bring little extension wires out from both terminals of each coil; 16 of them. Then play around with multi-plexing them. You may find that you can get several distinct sounds out of this pickup, just in the way you connect up the wires. That by itself could be a valuable design feature.

I see other questions and variations, too. What would happen if you made the coils long ovals, where the ends that turn around hang out beyond the ends of the blades. That puts the turn-around part of the coil outside of the magnetic field, leaving just the straight sections going crosswise through the field. Would that increase the signal level and definition, or not?

 

[Freekmagnet]

@Bruce Johnson - well, you were right again! I wired the coils like this:

B=Beginning E=End

B>E>B>E > E>B>E>B > B>E>B>E > E>B>E>B

Thus removing 1 layer of humbucking and it definitely improved the sound. I have a considerable amount more output and bass and mids. I still have tons of treble, but it's more manageable. It's more proportionate to the amount of lows.

The pickup still hears EVERYTHING for better or for worse. If I'm playing and I shift my body position, the bass picks it up. I kid you not.

I wonder if I could remove another layer of humbucking? Maybe wire up the north-facing SU pairs to each other non-humbucking and vis-versa with the south-facing SU pairs.

 

[ardgedee]

I'm kind of surprised there's not modeling software for pickups like there is for speaker cabinets. Electromagnetics isn't straightforward but it's understood by its relevant experts. Probably moreso than acoustic physics are by its relevant experts, since humanity's been designing and studying electric motors and dynamos for over a century.

 

[Freekmagnet]

Alrighty, I just wired it like so:


N S N S
B>E>B>E | E>B>E>B | B>E>B>E | E>B>E>B

with the ends of each N pole together and the beginnings of each south pole together. The two NN and SS poles are wired in series. I hope that makes sense.

The increased output, bass and mids are noticeable. Again, still lots of treble, but it sounds more proportionally correct.

The pickup is still very sensitive and picks up EVERYTHING. That part hasn't changed a bit.

I'm thinking that this may have enough lows to run as a sole bridge pickup without a preamp. We'll have to see - I can't really say for sure about that yet.

I wouldn't mind if I could fine-tune it a bit and tighten up the sound a little bit. The highs I can knock down with a capacitor. The mids are a little wild, too. There's not too much further I can do that with wire. I might be able to use a smaller core like I mentioned earlier. Then I can either add more 44AWG or bump it up to 43AWG. The 43 would open it up more, but the bass would be nice and clear. Or I could use a smaller core and narrow the width of the coil and keep the turn count the same. That might make it a little more focused.

I think I can probably move on to the next phase. I'll marinate on this for a couple of days.

 

[MPU]

Can you swap the magnets for ceramic instead of neodymium? I've found out that ceramics have more pleasant tone in my ears when trying different magnets.

 

[Freekmagnet]

Can you swap the magnets for ceramic instead of neodymium? I've found out that ceramics have more pleasant tone in my ears when trying different magnets.

I gotta agree with you there.

I don’t have any ceramics that size readily available. I could cut some down. Although I’ve found that the big neos put out more bass than the ceramics.

I suppose I could try it. I hate cutting ceramic, though. Magnetic dust everywhere...

 

[Freekmagnet]

Ah! I’m so dumb!

I think it’s microphonic. I potted the coils, but it could be the steel slugs inside the coils. I never had much problem with microphonic pickups. I’ll have to figure this out...

 

[Jon Clegg]

I suspect that what you created is a very low inductance pickup, hence the brightness. Inductance is rather difficult to measure without a specialized meter. Which brings me to this gizmo, a pickup driver.

These are actually magnetic buzzers that I cut open, removed the metal discs and re-assembled. I ordered these originally to build a feedback sustainer - ala the Sustainiac, for one of my guitars. I had a few extra and decided to use two to excite pickups with a frequency sweep. The goal being to tailor passive tone control components to the pickup.

I use the Generate function in Audacity to create a 40-4000Hz 20 second 'Chirp' that I play through the headphone output of my audio interface while recording the response of the pickup via the input. I then use the Plot spectrum feature to see the frequency response.
Here's a plot of a mid-scoop filter in the tone control:

It's not perfect; I need to work out a good buffer with a flat frequency response for the input, but it does allow me to get an idea of how to tune things.
I've also used this setup to measure the inductance of pickups by putting a 10nF cap across the outputs of the pickup and looking at the resonant peak of the response, I can determine the inductance of the pickup by

L = 1 / (C(2 Pi F0)^2)

Typical pickups like the T-bird humbucker is 3.3H , the mini-hum on this bass measured out at 0.9H.

 

[Novarocker]

@Jon Clegg wow, genius idea!

Alrighty, I just wired it like so:


N S N S
B>E>B>E | E>B>E>B | B>E>B>E | E>B>E>B

with the ends of each N pole together and the beginnings of each south pole together. The two NN and SS poles are wired in series. I hope that makes sense.

The increased output, bass and mids are noticeable. Again, still lots of treble, but it sounds more proportionally correct.

The pickup is still very sensitive and picks up EVERYTHING. That part hasn't changed a bit.

I'm thinking that this may have enough lows to run as a sole bridge pickup without a preamp. We'll have to see - I can't really say for sure about that yet.

I wouldn't mind if I could fine-tune it a bit and tighten up the sound a little bit. The highs I can knock down with a capacitor. The mids are a little wild, too. There's not too much further I can do that with wire. I might be able to use a smaller core like I mentioned earlier. Then I can either add more 44AWG or bump it up to 43AWG. The 43 would open it up more, but the bass would be nice and clear. Or I could use a smaller core and narrow the width of the coil and keep the turn count the same. That might make it a little more focused.

I think I can probably move on to the next phase. I'll marinate on this for a couple of days.

Glad you picked up on em probably being microphonic. Pretty wild huh? I have a Tbird knock off that's the same. It's so sensitive it pickups up the impact of my ring finger all the way down the neck!

 

[Jon Clegg]

@Jon Clegg wow, genius idea!

Thanks. I've been using this to tune your mids-control circuit to the steam-punk bass.

Glad you picked up on em probably being microphonic. Pretty wild huh? I have a Tbird knock off that's the same. It's so sensitive it pickups up the impact of my ring finger all the way down the neck!

Yeah, some of the low-Z pickups I built were very microphonic even though I thought I had everything well potted etc. Kind of the down-side of building a pickup that will detect the tiniest string movement is that it will also detect the tiniest relative movement between the coil(s) and the magnet.

 

[Freekmagnet]

I suspect that what you created is a very low inductance pickup, hence the brightness. Inductance is rather difficult to measure without a specialized meter. Which brings me to this gizmo, a pickup driver. View attachment 3479367

These are actually magnetic buzzers that I cut open, removed the metal discs and re-assembled. I ordered these originally to build a feedback sustainer - ala the Sustainiac, for one of my guitars. I had a few extra and decided to use two to excite pickups with a frequency sweep. The goal being to tailor passive tone control components to the pickup.
View attachment 3479368
I use the Generate function in Audacity to create a 40-4000Hz 20 second 'Chirp' that I play through the headphone output of my audio interface while recording the response of the pickup via the input. I then use the Plot spectrum feature to see the frequency response.
Here's a plot of a mid-scoop filter in the tone control:
View attachment 3479369
It's not perfect; I need to work out a good buffer with a flat frequency response for the input, but it does allow me to get an idea of how to tune things.
I've also used this setup to measure the inductance of pickups by putting a 10nF cap across the outputs of the pickup and looking at the resonant peak of the response, I can determine the inductance of the pickup by

L = 1 / (C(2 Pi F0)^2)

Typical pickups like the T-bird humbucker is 3.3H , the mini-hum on this bass measured out at 0.9H.

That’s really cool! Thanks for posting that.

 

[Bruce Johnson]

Yes, microphonics is a bigger problem when you get into hi-fi pickups like this. Like Jon said, as you build a pickup to detect really fine movements of the string, then it also can detect really fine movement of itself in relation to the string. This can be internal movement of the coils in relation to the magnetic field, and it can also be movement of the whole pickup in relation to the bass. Its sensitivity is increased overall.

Is it locked down tight in your test bass rig? What was good enough for other pickups may not be good enough for this definition-monster.