Faraday rings

[Puggie]

I fancy having a play at fitting a faraday ring to a speaker (sub) I'm re-coning. My plan is to do a full height (full length of the pole piece) Faraday ring. Would it be ok to machine the pole piece down by 1mm, and electro-plate the pole piece with copper up to 1mm thick and re-assemble the sub (I see no problem in this)? the copper ring does not need to be electrically insulated from the steel pole piece does it?

I know I'll lose BL due to the bigger magnet gap but this is not really an issue for my purpose (or rather I hope the improvements will outweigh this)

what advantages / problems am I likely to encounter?

 

[Puggie]

Anyone know if the faraday ring has to be electrically insulated from the steel pole piece?

 

[LineSource]

Hi Puggie,

The Faraday ring should be electrically connected to the steel pole piece to provide the best total current conduction path.

On the topic of copper plating. Lambda Acoustics built some speakers with just a copper plating on the pole piece, their LE series. The plating did help to modestly extend the frequency range and flatten the impedance curve, but the thick copper sleeves used for Faraday rings in their TD series were substantially superior.

 

[Puggie]

Thanks for the reply, what are you calling thick, I was planning on plating the pole piece with between a 0.5 and 1mm thickness of copper. I know Lambda seem to be producing a lot of drivers with a lot of copper in the motor but I've had problems trying to find much good technical info on them.

 

[hancock]

I know I'll lose BL due to the bigger magnet gap but this is not really an issue for my purpose (or rather I hope the improvements will outweigh this)

Saturating the pole piece is also really important for reducing inductance (and therefore flux modulation). You may be actually increase the inductance by doing this rather than lowering it.

John

 

[John Sheerin]

Thanks for the reply, what are you calling thick, I was planning on plating the pole piece with between a 0.5 and 1mm thickness of copper. I know Lambda seem to be producing a lot of drivers with a lot of copper in the motor but I've had problems trying to find much good technical info on them.

In addition to what John said, reducing the variation in Le with voice coil position will also reduce distortion. This requires that the copper (or some other conductive material) be placed in a certain location, and depends on the rest of the geometry of the motor. In general, though, Le will rise as the coil moves into the motor (more steel in the middle of the coil), and go down as the coil moves out. So you want to start by putting conductive material on the inside of the motor, then through the gap, then above the gap. How thick the conductive material is also effects how effective it is. For a subwoofer, 1mm might not be enough to make much difference at low frequencies. For a ring only below the gap, 10mm or more thickness of copper may not be excessive.

 

[Puggie]

Thanks guys, this was looking nice and easy untill those last two posts

looks like I'll have to download that FEMM magnetic modeling package and have a play.

So ideally I need to put as much copper in the motor as possible, while retaining the BL.

I have a sub where the motor has come apart, the top plate and pole piece/back plate glue joins have failed so I have 3 pieces, top plate, pole piece and back plate and 3 magnets still bonded to each other.

The voice coil is 2" dia

The pole piece is 50.5mm outer dia with a 19.5mm dia vent through it and protrudes 19mm above the top of the top plate. so not much room for putting much of a copper ring around there, my plan was to turn the pole piece down by about 1mm and electro-plate this with copper to form the faraday ring, I was planning on doing this over the whole length of the pole piece.

The top plate is 16mm thick and the hole in the centre 56.25mm.

The magnet is made from 3 ceramic 'slugs' 21.5mm thick, 165mm outer dia and 61.5mm inner dia. I was planning on fitting a copper ring in the void between the top and back plates that would fit flush against the inside of the magnets with an inner diameter as close as I could get to the hole in the top plate so its going to be about 4.5mm thick.

I have a couple of motors of dead subs which are to fit 3" VCs and magnets are 8.5" outer dia (BIG) I could look at turning new top plates and pole pieces to use these, this would give me the ability to fit a lot more copper round the outside of the VC and I would expect have a higher BL to offset the increase VC gap.

The sub is a very high excursion design (XMAX of 2" peak to peak but the physical limits are a good bit more) so I could lose the pole vent and replace the dust cap with a phase plug to increase BL maybe.


Any ideas advice etc is most welcome, this is really an experimantal excercise for me to see the effects of faraday rings and see if I can improve on this sub I have.

 

[hancock]

If you really want to get experimental, instead of a shorting ring, try wrapping some magnet wire tightly around the pole piece and run the wire in series with the voice coil, but so that the current runs in the opposite direction to the voice coil--if voice coil runs clockwise from + to -, run the pole piece coil counter-clockwise from + to -. You'll have to experiment with the number of turns, but you should be able to get the inductance down to zero. There is an article in one of the AES Loudspeaker Anthologies on this.

John

 

[John Sheerin]

Hey John,
Yeah, I have seen that somewhere... Do you happen to remember the author or title of the paper? I just searched and could not find it (I don't have the anthologies but can get any of the papers in them). If I can figure it out I'll build a speaker with and without next week, take some measurements, and give them a listen.

 

[BassAwdyO]

If you really want to get experimental, instead of a shorting ring, try wrapping some magnet wire tightly around the pole piece and run the wire in series with the voice coil, but so that the current runs in the opposite direction to the voice coil--if voice coil runs clockwise from + to -, run the pole piece coil counter-clockwise from + to -. You'll have to experiment with the number of turns, but you should be able to get the inductance down to zero. There is an article in one of the AES Loudspeaker Anthologies on this.

And BL goes down the drain
And the difference between this and a DVC woofer with the coils wired reverse polarity is?????

 

[John Sheerin]

And BL goes down the drain
And the difference between this and a DVC woofer with the coils wired reverse polarity is?????

Uh, no... BL goes up (a tiny, insignificant amount) in the case I have just been modeling.

Re goes up, of course. I need to play with this some more. I imagine I could increase the wire size from where I have it to lower Re (currently same size as the VC).

The difference from a DVC is that the DVC case (with the coils the same) would produce zero force (the Lorentz forces would all cancel, depending on the matching of the coils). BL would effectively go down in that case...

 

[BassAwdyO]

I dont quite understand this concept fully.... BL is B(the magnetic flux density) times L(the inductance of the coil in the gap). If there is no inductance, then how can there be BL? I'm assuming if that were to work then there would be inductance in the gap, but the inductance cancelation would occur outside the gap possibly thus negating the total inductance?

 

[Feyz]

L of BL doesn't stand for inductance, it stands for the length of wire that is inside the air gap where there is supposed to be a constant uniform magnetic field of B.

I have written a lengthy paper on how the Faraday rings do what they do, which you can access from here:

http://members.fortunecity.com/pirimoglu/

Just click the first link in there, which will open a pdf document I wrote. The references section of the document also contains the name, title and year of the AES paper by Birt, where he had experimented with the stationary coil that is in series to the voice coil among other things.

The document I wrote is a long one, and the first half is a summary of the basic electromagnetics theory that is needed in order to understand how the rings work.

 

[Puggie]

Oh no, now that has really set the cat among the pidgeons!!!

so turn the pole piece down but instead of plating a faraday ring on in, wind a coil over it (I presume that is what you mean by magnetic wire) and I have access to a couple of kilos of 1mm enammeled wire which i have inductors wound from.

would this also work with a coil OUTSIDE the voice coil, wound against the inside of the magnet?

Oh how the plot thickens.

BassAwdyO, the main pont here is you have a dynamic coil attached to the cone (the voice coil) and a static coil wound round the pole piece and unable to move.

I'll have a read at the AES paper and think then post on this in the week.

 

[John Sheerin]

Thanks for the reference Feyz. On the AES website, the summary for this article was coming up as something else, so I was passing it by.

I modeled the driver I have with several different coil geometries. I found one that will completely eliminate any inductance, but it would almost triple the series resistance over what the voice alone is. Switching to larger diameter wire to reduce resistance enlarged the coil size and reduced the effectiveness of the coil. I finally settled on a geometry that will add minimal resistance (around 0.7 ohms on a 5.6 ohm voice coil), only lose me a dB on my sensitivity, and not move the Q above 0.7. It should keep inductance flat with excursion, though.

 

[hancock]

I finally settled on a geometry that will add minimal resistance...

Sounds good. Are you doing that in FEMM? How about sharing the model with us?

John

 

[John Sheerin]

Yeah, I'm using FEMM. So my AC predictions do have the innaccuracy of not including the DC flux in the circuit. It seems like as I predict a flatter Le vs x in FEMM, I see a flatter Le vs x in the real world, though.
I guess this makes some sense, as I'm designing in methods to counteract any change in flux in the circuit, so no (or less) change at zero DC flux could correlate to no change or less change at some higher flux level. What doesn't correlate to the real world is the absolute value of Le...

Sorry, I can't share the model as it is a geometry that's not mine to share. It's basically a center neo slug on a pedestal inside a straight-sided cylinder. There's a steel disk on top of the magnet to form the gap. The coil is overhung. My correction coil is around the pedestal, or pole piece, below the magnet, inside the voice coil. So I have to glue in the correction coil first, then the magnet and disk. Unfortunately I think it will take me longer to make the coil than I thought - it's not a size I have a mandrel for, so I'll have to find or make something the right size to wind it on.

Btw, not to thread-jack, but I just saw your idea about making two planar bass sources on opposite walls and cancelling out reflections from the opposite sides. Have you had a chance to try that yet?

 

[hancock]

Btw, not to thread-jack, but I just saw your idea about making two planar bass sources on opposite walls and cancelling out reflections from the opposite sides. Have you had a chance to try that yet?

I've got something even better, but that's going to be my little secret for now. I've always been very non-proprietary about my ideas, but this one is really good and much more practical. I'm thinking of applying for a patent.

John

 

[hancock]

Hey John,

Your account doesn't allow personal emails. Send me a quick note to johndhancock at gmail.

John

 

[Puggie]

All very interesting guys thanks.

I'm going to have to try and get the AES paper and have a read then put some piccies of ideas up to see if you think I'm going along the right lines.