Help with split coil bass pickup design

[JayGunn]

Here's my proposed project. If you guys tell me it's hopeless or needs major revisions, I'll listen and learn.

The idea here is to make a split coil pickup to put in the neck position of a 4-string bass build I have started, roughly where a P-bass pickup would be.

I'd like it to sound similar to a 'modern' P-bass pickup, meaning a big fat fundamental and bright enough so you hear some attack. Instead of "moom-moom-moom" (thiink P-bass with the tone turned down) I want to hear "doom-doom-doom," meaning bright enough to bring out the attack. I always play with fingers (so far).

I am definitely going to use a preamp, so I am not too concerned about hot output or internal capacitance, since I'm taking the guitar cord out of the equation.

Here's the plan view I came up with after studying this site a lot. This is a see-through drawing:



It's 2 coils, epoxy potted in an MM shaped case. The case has a solid top and the coil is 1/16" below the surface. Maybe plastic case for the trial version, then a macassar ebony case for the production pickup if I love the results. The neck pickup will be a real MM clone pickup, not home made. By choosing these case dimensions, if I hate the sound I can just buy another MM clone for the neck.

Here's how the coils are constructed:



I'm using 4 layers of low carbon steel for the core, out of concern about loss of highs from eddy currents (forbidden topic?). I chose 4 layers with a total width of about 1/4" so that there's a reasonable amount of surface area to stick the neo magnet onto at the bottom. The blades/layers would be dipped in lacquer or something before assembly. And the whole bobbin lightly epoxy coated before I start to wind.

Here are the things I hope you guys can advise me on:

1) This is a pretty high mass core compared to a single blade. Is that a problem?

2) If I underwind this in comparison to a typical P-bass pickup will I sacrifice the beefy bottom end? I am thinking of trying 4000 turns of 42ga single poly insulated wire, not heavily scattered. I don't want to build up 10,000 turns on that thick core and make coils too fat to fit into the case as shown above.

3) The neo magnets I have seen on the 'bay come in 40mm X 10mm x 5mm or 40 x 10 x 2. Since I don't need high output, is the 2 mm thick one sufficient?

4) In the diagram I said 1018 steel, but I might end up using HR1 or CR1 commercial steel, which have a maximum of .12% to .15% carbon, so maybe a bit more permeable than 1018.

One final note--I have read here that if the resonant peak is very high (let's say over 6khz) the pickup may sound darker, because it's too high for the bass rig to reproduce. That would certainly be the case with my G-K MB115 (no mid, no tweeter). So I might end up adding a cap across the pickup to lower that point until I like it.

Thanks in advance for any design suggestions you make.

John

[CRU JONES]

Hi JayGunn

I think you are just going to have to wind one up and see! I would say it looks pretty good! I have not tried laminating blades. If you want me to take a stab in the dark, I think you should do less steel mass. I think you will have excessive eddy loss with 4 blades... but, then again, it might be perfect for Neo magnets!

When trying a new design, I like to be able to change as many variables as possible. If you can make a bobbin that allows for replacing the blades, if you can replace the magnet with magnets of different types and field strength, and if you build many bobbins to wind up for comparison; you will be doing well. When I try a new design, I am usually significantly far away from where I wish I was! But I feel I have gained an ear for how to modify it to get where I want. Building a terrible sounding pickup can be a better learning experience than getting lucky on your first shot.

That being said, building an active pickup gives you much room for adjusting the load to suit your pickup. By that I mean, when designing a passive pickup you are stuck optimizing your pickup for the 250kOhm or 500kOhm control values the user will choose, plus the 900pF or so cable capacitance. If your first shot doesn't work for the 250k controls you planned on, it still might work with 500k controls, otherwise it is a useless pickup... for the market. With an active design you don't have the cable to deal with, which automatically means your transient response will be very good over a broad range of resistive loads. I very much like your idea of experimenting with differing values of capacitance

I have only experimented with single blade designs. The difference between 22ga and 16ga is quite significant in my designs. I have yet to try 14ga, but I think I would like it. 10ga is a bit much for what I have been doing.

A trick I learned from Marku52 to figure out if you are saturating your permeable mass, is to put your magnet against your material and see if you can measure any magnetism on the other side. a paper clip would be a decent measuring device.

Good luck! Hope that helps Let us know how it comes. I personally would like my bass to go "dooom, dooom", so maybe you can pass some pointers in return

cheers,
Ethan

[JayGunn]

Ethan,

Thanks for your encouraging reply.

Since I am planning to epoxy the laminations to the bobbin top and bottom wood of fiber retainers (passing them through a rectangular hole or mortise), I won't be able to make the blade count changeable. BUT I can certainly make a pair of 4-blade and pair of single blade bobbins and I'll do that.

I had noted the formula for H, magnetic field strength as

H = NI / lower_case_L

where N is number of closely spaced turns, I is current and 'lower_case_L' is length of wire. And it occurred to me that making a fat core would somewhat lower the magnetic field strength since it takes more feet of wire to create a given number of turns around a bigger object. But maybe only a 5-15% increase in the denominator, so perhaps not a big deal from that POV.

But then I read the formula for B, flux density, as:

B = mu*H

where mu is the permeability of the core material. I am way over my head with this stuff, but it just seemed to me that the mu of a material is fixed, not dependent on the thickness. So it suggested to me that while a wider core would give me more intimate contact with the flat, somewhat wider magnet beneath, that the strength of the magnetic field from 1 or 4 layers of core would not be much different. This is purely superstition on my part, I admit.

So if I wind a single-blade core and a quad blade core with the same number of turns, I'll be surprised if they sound drastically different...but I'll have 2 sets of similar sounding pickups if I read these formulas right. No harm done unless that all sound lousy(!)

Actually on the single blade core I'll probably do the magnet(s) like this, so maybe it WILL sound different:



And the idea of using strips of steel in the cores rather than just a chunk of steel 1/4" thick is (superstitiously again) to prevent eddy current losses, even though, hey, they might sound great.

[CRU JONES]

Hi JayGunn,

Ok, I am going to have to let you in on a secret. I don't use formula's much in the beginning. I wind, measure , and play After awhile you start to get better at guessing.

The formula: H = N*I/L might be more useful for creating an electromagnet. Our pickups won't be generating much current through the coil. H will be a tiny number

Eddy currents are relevant though! They are very important and eliminating them is not necessarily better.

It is a balance. The main ingredients are: inductance, capacitance, resistance, eddy losses and Magnetic field strength. It is easier to know these quantities by measuring after winding

I think your design will be a success first try. It helps immensely to be able to choose the load the transducer will feed. I have a humbucker in one of my guitars that sounds really good into 100k controls. It uses Ceramic8 magnets as rails. C8 is saturated, so it has no permeability to increase inductance. Nor is C8 electrically conductive, so you get really low eddy damping. I put a steel plate on the bottom to boost the inductance and to get some eddy losses going on, but it still needed the 100k controls. With 500k controls it sounded like my cat coming down a chalk board. It made me shiver and want to give up winding forever! With 100k controls it is still pretty tough. I am thinking I will demagnetize the magnets a bit. By demagnetizing the C8, you are "un-saturating" the permeability and some inductance can be gained. It is harsh, but records distorted guitars really well. The sound is barely tolerable on its own, but sounds pretty amazing in a mix I could never sell the pickup, it is just too unlikely that anyone would obey the advice of using 100k controls. A cat would slide down a chalk board and I would have my reputation forever banished! ...I could make it active though

Now wind that thing up!!!

Cheers,
Ethan