Cones, coils and eddy currents

[Dogtemple]

I have been looking at the workings of drivers for a cutting head and am looking at fabricating a 'stereo' head entirely and skipping the usual step of using tweeters.

one thing I am not sure on is eddy currents when making a coil bobbin from aluminium and having a cone at the end, which would direct the movements to a rod and then the stylus.

typically with a bobbin using aluminium, there is gap to avoid a shorted coil. I am wondering, should a cone be placed directly on top/at the end of the bobbin, would this then short it out?


something I have in mind is to spin a disc of aluminium over a former, which will be a coil bobbin and cone in one. purpose here is for lightness, tidiness, accuracy and rigidity.

I was thinking of cutting a slit up the bobbin as normal and making a hole maybe at the top of this slit/at the base of the cone. I'm not sure if that will stop the coil from shorting or be completely futile.


anyone have any thoughts on this? I'm good on the mechanical side but not the electrical side, and the fields of magnetism is something I have no experience with, hoping someone more knowledgeable might be able to give some guidance please?

[dmills]

There are two 'eddy current' related loss mechanisms in play.

• One being what is essentially transformer action between the winding and the former, this is essentially independent of any movement in external magnetic fields, and produces no forces on anything, it depends on dI/dt, mutual coupling and resistance of the former only.

• The other is the eddy currents induced by the movement of the former in the field from the magnet assembly, and is essentially proportional to velocity and field strength and inversely proportional the former resistance.

Does it have to be ally? I mean the figure of merit here is that higher bulk resistance materials are likely better from a minimise the eddy current perspective, and I would point out that 0.1mm stainless steel stencils are cheap, readily available (SMT paste masks) and very, very precise, I would be very tempted to explore leveraging that technology.

Could you make it out of two layers overlap the gaps by 180 degrees, with a nice layer of anodising applied to both before bonding the two layers together? That way the eddy currents cannot circulate, and you have a solid surface on which to wind the coil?

Remember that for the losses from transformer action due to the drive coils, the coupling falls off fast as you move away from the coil, which hints that for this aspect a solid cone in front of the drive coil may not be that horrible.

Another thought is that if you are using a copper shorting ring on the stator pole pieces (and you probably should be) then the cone forward of that should have limited influence because current in the former will itself induce current in the shorting rings which will produce a field opposing the eddy current loss, needs a 3D field solver to see how that pans out.

Personally I would probably shy away from a spun metal former, square CCA wire and epoxy feels likely to make for a lighter assembly, with maybe flextures made by ordering suitably shaped stainless steel solder masks from an SMT stencils place doing the centring for me?

[Dogtemple]

thanks for that, interesting info. aluminium is not a prerequisite, its just strong and light, I'm open to anything that will do the job well. I had not considered the stainless stencils. the only thing with that that would concern me is the weight. I'm hell bent on getting it super light and minimal whilst retaining its function. would stainless have a huge affect on moving mass do you think?

I have also considered making the bobbin out of carbon fibre, this should do the trick? light and strong with none of the magnetic aspects...


I tried to spin a cone with coke can aluminium, but this didn't work. the metal wouldn't form at all and when annealed it would just fold up. so spinning the shape is a no go.

[dmills]

Interesting paper that may be relevant.

https://hal.archives-ouvertes.fr/hal-00436594/document

[Dogtemple]

Thanks, certainly food for thought