Does a voice coil have to be round?

[MAB]

I'm quite sure the voice coils of those drivers are still circles.

Perhaps I didn't understand OP's question!
And, due to the nature of inductors, the VC is going to have to form a loop!

Electrostatics are then different.

Again, perhaps I'm not thinking clearly.

 

[Timcognito]

We already have good options on midrange drivers.

But not the with wide dispersion pattern.

Philharmonic BMR Mini-Monitor (v2) Review

Philharmonic BMR Review

www.erinsaudiocorner.com

https://tectonicweb.wpenginepowered.com/wp-content/uploads/2019/01/TEL-DS-TEBM36S12-4A_Revision1.6.pdf

https://tectonicproaudio.com/wp-content/uploads/2022/03/DML500R-Manual.pdf

 

[Tim Link]

Perhaps I didn't understand OP's question!
And, due to the nature of inductors, the VC is going to have to form a loop!

Electrostatics are then different.

Again, perhaps I'm not thinking clearly.

I think you just need a magnetic gap and wires running through it with AC current. If the wires are straight instead of in a coil then it does seem there will be something different happening in terms of inductance. Less inductance?

 

[voodooless]

the VC is going to have to form a loop!

While a circle is a very efficient loop, it’s not the only way to make a loop

 

[Tim Link]

While a circle is a very efficient loop, it’s not the only way to make a loop

A single gap could have all wires running through in one direction. The return wires could be much bigger and heavier, not in the gap, and not attached to the diaphragm. I think it might be better to have two gaps

 

[Prana Ferox]

I think others alluded to this but there's a difference between the shape of the voice coil and the shape of the speaker diaphragm. There have been many non-round drivers built over the years, including those Genelecs, a lot of high end Sonys and Pioneers, a zillion car 6x9's etc. AFAICT their actual voice coils are always round (although some have multiple coils driving the same pistonic diaphragm.) However getting that pistonic diaphragm to behave properly and not have weird unpleasant breakup seems to be the issue. Note this doesn't apply to passive radiators and it's one reason you see more non-round ones of those.

 

[MAB]

While a circle is a very efficient loop, it’s not the only way to make a loop

Yeah. That's my point! Loop doesn't have to be wound into a circle of constant radius.

I'm probably really missing the point here!!!

 

[MCH]

you can use these calculators below to compare a circular loop to an elongated one (assuming a rectangle with very dissimilar sides). You will see that close to the plane defined by the loop, the magnetic field is much smaller in the center of the rectangle (=eliptical loop) assuming the same current.
I have no idea and i guess there are many other reasons, but the fact that many circular loops of the same area would be more efficient on the plane and the fact that the magnetic field within a non circular loop would be less homogeneous (weaker in the center vs near the shorter sides) could be the reason why they are not used (?)

Current loop [Encyclopedia Magnetica™]

www.e-magnetica.pl

 

[restorer-john]

Go look up Sawafuji Dynapleats.

I had a pair of JVC speakers with I forget how many Dynapleats in each, perhaps 12 or 13.

 

[Cbdb2]

you can use these calculators below to compare a circular loop to an elongated one (assuming a rectangle with very dissimilar sides). You will see that close to the plane defined by the loop, the magnetic field is much smaller in the center of the rectangle (=eliptical loop) assuming the same current.
I have no idea and i guess there are many other reasons, but the fact that many circular loops of the same area would be more efficient on the plane and the fact that the magnetic field within a non circular loop would be less homogeneous (weaker in the center vs near the shorter sides) could be the reason why they are not used (?)

Current loop [Encyclopedia Magnetica™]

www.e-magnetica.pl

Did you use the same wire length (proportional to vc mass) for both calculations? R=2L/π
Than the rectangle has a larger mag field than the circle because the flux density is flux/area (the total flux in side depends only on the current) and the circle has the largest area for a given perimeter. Squash the circle, increase flux density. Try it with that calculator by making the circle smaller.
The problem with square loops is the Flux density is higher in the corners than in the middle of the square. Flux density decreases with distance from the wire (1/r). In the circle if your a mm away from the wire your distance to the rest of the wire is the same all around the circle. In a square your much closer to wire in the corners so more Flux. So the push on the VC is harder in the corners

 

[egellings]

Isn't there a driver called Ellipticore (sp?) that has an oval-shaped voice coil?

 

[MCH]

Did you use the same wire length (proportional to vc mass) for both calculations? R=2L/π
Than the rectangle has a larger mag field than the circle because the flux density is flux/area (the total flux in side depends only on the current) and the circle has the largest area for a given perimeter. Squash the circle, increase flux density. Try it with that calculator by making the circle smaller.
The problem with square loops is the Flux density is higher in the corners than in the middle of the square. Flux density decreases with distance from the wire (1/r). In the circle if your a mm away from the wire your distance to the rest of the wire is the same all around the circle. In a square your much closer to wire in the corners so more Flux. So the push on the VC is harder in the corners

No, I compared the field of a circular coil (i.e. 10 mm diameter) vs a rectangle with a small side of 10 mm x a long side of >10 mm. Aiming to compare "several tweeters in line" vs one single long coil as per OP:
"I see line source speakers often made of a whole bunch of little tweeters. Why not just stretch the voice coil into a really long oval?"
And yes, agree with you. I have no idea about speakers but I can imagine having a much stronger field as you get closer to the shorter side/corner sounds like a bad idea.

 

[Cbdb2]

Not a driver guy but a long narrow driver is probably quit different than a row of tweeters. Cone resonance and break up is probably different. If that worked someone would probably already be using it.
I know, lots of probablies.

 

[Tim Link]

No, I compared the field of a circular coil (i.e. 10 mm diameter) vs a rectangle with a small side of 10 mm x a long side of >10 mm. Aiming to compare "several tweeters in line" vs one single long coil as per OP:
"I see line source speakers often made of a whole bunch of little tweeters. Why not just stretch the voice coil into a really long oval?"
And yes, agree with you. I have no idea about speakers but I can imagine having a much stronger field as you get closer to the shorter side/corner sounds like a bad idea.

Maybe no magnets at the very ends.

 

[Tim Link]

Not a driver guy but a long narrow driver is probably quit different than a row of tweeters. Cone resonance and break up is probably different. If that worked someone would probably already be using it.
I know, lots of probablies.

There are a lot of potential problems for sure. A straight former to hold the wires in the gap is not inherently stiff like a tube. Somehow that whole length has to be suspended and kept aligned in the long gaps.

 

[Randyman...]

I thought this was an interesting deviation from the standard circular designs:

Home

Dayton Audio RC180-55 2" x 7" Res-Core™ Neodymium Woofer 4 OhmDayton Audio RC180-55 2" x 7" Res-Core™ Neodymium Woofer 4OhmDayton Audio is proud to bring the RC180-55 2" x 7" Res-Core™ Neodymium Woofer to DIY speaker enthusiasts worldwide. This driver literally turns the voice coil on its side...

www.parts-express.com

 

[Tim Link]

I thought this was an interesting deviation from the standard circular designs:

Home

Dayton Audio RC180-55 2" x 7" Res-Core™ Neodymium Woofer 4 OhmDayton Audio RC180-55 2" x 7" Res-Core™ Neodymium Woofer 4OhmDayton Audio is proud to bring the RC180-55 2" x 7" Res-Core™ Neodymium Woofer to DIY speaker enthusiasts worldwide. This driver literally turns the voice coil on its side...

www.parts-express.com

Wow, that's kind of it! I wonder how long they could make something like that.

 

[Multicore]

There are a lot of potential problems for sure. A straight former to hold the wires in the gap is not inherently stiff like a tube. Somehow that whole length has to be suspended and kept aligned in the long gaps.

Owing to the combination of two factors that I mentioned in my first reply (1. electromagnetic efficiency is optimized by a circle, 2. circular speaker motors are easy to handle in design and production at a wide range of tolerances and are mechanically stable over a long service life) almost every engineer will choose a circle. It's not that you can't use different shapes. It's that you'll get worse results and it will be harder to do so why bother?

 

[Tim Link]

Owing to the combination of two factors that I mentioned in my first reply (1. electromagnetic efficiency is optimized by a circle, 2. circular speaker motors are easy to handle in design and production at a wide range of tolerances and are mechanically stable over a long service life) almost every engineer will choose a circle. It's not that you can't use different shapes. It's that you'll get worse results and it will be harder to do so why bother?

The reason to bother would be to get a wider bandwidth driver with constant dispersion and maybe retain a good portion of the benefits of dynamic drivers. It might have reasonable motor force to horn load well enough. A 1" wide by 36" long radiating surface should be able to cover 20k down to 200Hz easily with lots of power handling. You could use 36 tiny full range drivers but that's a major pain too, and you'll still get comb filtering issues.
But I'm with you. It probably just doesn't pan out or someone would be making it already.

 

[DonH56]

A planer voice coil, e.g Magnepan, is a good example of a non-round voice coil. But as others have highlighted, a rectangular or other design leads to a lot of complications in design and construction and usually degraded performance. Many, many years ago I asked a B&W (I think) designer about a rectangular woofer they had tried and rejected. One thing I remember was that corner stress and flex led to much higher distortion and worse long-term reliability, with additional undesired modes forming as driver excursion and frequency increased.

A (one) drawback of planer magnetic designs is that, with a one-sided magnet structure, linear excursion is fairly limited so as output increases distortion rises quickly. Using magnets front and back helps, but rather than the voice coil being "immersed" in the magnetic field by wrapping around the magnet, you still have a rapidly-falling magnetic field with distance that makes large excursion impractical (IMO). Perhaps when we have superconductors and much stronger magnets...

I had an idea many years ago of creating a magnet in the form of a tube or rectangle, pierced by holes so more of a screen, then suspending a conductive membrane inside so you had a planer speaker enclosed by the magnet. The idea was two magnets in the form of channels with the membrane between: [ | ] so magnet - diaphragm - magnet. A little thought shows it is not significantly better than a conventional approach though should offer higher field strength. Should have patented it just for fun.