Better understanding of speaker coils/magnets

[petemoore]

  I've looked at split-open side views and graphics on how speakers work, but what are the requirements for the magnetics ?
  The fixed magnet is pulling from exactly where ? ..outside and inside the coil cylinder ?
  What about top and bottom ? is the middle portion of the coil recieving the same push/pull messages as the outer portions all the time or does this ratio change when excursions occur ?
  When excursions occur does the coil recieve attenuated signals toward the ends of the swing as it's electromag relationship with the magnet changes because of the field orientation ?
  Has a mag/fixed coil been used for audio transduction which has the magnet only on the inside or the outside of the coil ?
  Anyone ever try winding a speaker coil...or see anyone else do it or and talk it over ?..how close must the relationship be between the coil and the magnet for effective magfield force ?

[Earthscum]

http://www.josepino.com/circuits/index.php?howto-speaker.jpc

http://www.instructables.com/id/Paper-Plate-Speaker/

That's just 2 quick hits searching for "paper plate speaker"... I remembered seeing a video someone had made one in, so I just searched for that.

Back in gradeschool my buddy showed me how to make them with foam cups.

[Earthscum]

Forgot to mention, you want your wire to be reasonably thin gauge, and we were finding it best to use a piece of cellophane tape wrapped backwards (sticky out) to wrap 1 layer as close as possible wrap to wrap (think inductors and high-Q techniques when wrapping).The cellophane, when you're done, slips really nicely up and down the magnet. Then you take another cellophane piece and use it like normal to stick the wraps in place. Light weight, and it's slippery... but, it's light weight. That's key, as well as getting the tightest wrap possible while still maintaining free slip.

If you do anything that is serious, you need air gaps for heating and shrinking tolerances, as well as other things I can't really remember right now. Basically, in a big speaker, you don't want your coil heating up and skidding across the surface of the magnet at 20-20,000 Hz.

[PRR]

> ..outside and inside the coil cylinder ?

"Right Hand Rule". Run a current through a wire, the induced magnetic field works at right angle to the run of the wire.

> does the coil recieve attenuated signals toward the ends of the swing

Ideally, no.

In real life, yes. This is bad, and we pick power and frequency to avoid it, to stay in the "linear" range.

Air motion is VERY small.

When cone is as large as an air wave, cone motion is SMALL. Not a problem.

When a 15" works below 500Hz (or a 4" below 2KHz), cone must move more than the air.

The magnetic flux leaks out of the gap, does not stop at the edge of the iron. Also a 20% reduction of flux isn't a terrible sound. So many old-time radio loudspeakers (the godfathers of classic guitar speakers) have a coil length similar to gap length. That's enough to do the job.

Some lightweight tweeters have coils shorter than the gap and never get out of the linear range.

Woofers: you can't get efficient deep bass from a small box. Instead you pick some bass frequency where efficiency is not too awful, and reduce midband efficiency so the result is "flat". To start, you add mass. You put more pulp and tar in the cone to damp the paper-flap sounds. You make the voice coil longer than the pole thickness. That adds mass, and also the overhung ends of the coil hurt midband efficiency more than bass resonance efficiency. And the long overhang means linear excursion much greater than pole thickness.

> the magnet only on the inside or the outside of the coil ?

It's a magnetic CIRCUIT. Just like a battery, both ends have to coome together at the load.

In principle, the magnet could be anywhere, connected by long iron bars. But magnetic circuits leak. The magnetic circuit must be very tight to get the most from costly magnetic-stuff.

You put the magnet inside or outside as appropriate. Alnico is dense and costly, you can usually fit enough inside the coil. Ceramic is bulky but much cheaper, you usually have to put it outside the coil. Neo-magic magnets are dense and expensive and usually put inside. But speaker design covers a long range, and designers are alert to economy. I know one big speaker with like 2 dozen small Neo magnets staggering around the outside of the coil.

[petemoore]

  Thanks PRR, this promises to make an excellent re-read for me.
  I know one big speaker with like 2 dozen small Neo magnets staggering around the outside of the coil.
  I would like to hear more about this one..

[Mark Hammer]

I recently rewound an earphone; the old school style mono units with the twisted white leads that used to come with televisions.  The structure is basically this: a piece of steel shaped roughly likely the male part of a snap fastener (i.e., looks like a teeny nipple) sits atop one end of a small cylindrical magnet, whose polarity runs from one "end" of the cylinder to the other.  The little nipple part has a few hundred turns of #40 or so wire.  On the other side of the magnet cylinder is a small circular piece of steel, thin enough to be flexible.  The coil pulls and pushes the metal disc.

I rubbed the disc on some of the ultra-fine plastic abrasive paper one uses for polishing frets to thin it up just a little more.  Improved the high frequency response a tad.  The rewind improved the efficiency of the earphone as well.

[PRR]

> circular piece of steel

Telephone receiver. Yes, used in a LOT of audio transduction.

NOT real closely related to the Voice Coil transducer that I assumed Pete was wondering about.

Problems: Sensitivity is strongly affected by diaphragm stiffness. A very sharppencil design can be unstable (diaphragm falls onto magnet). Electromagnetic audio flux must pass through the magnet and a lot of iron, lots of eddy-current loss. While there's no massy moving-copper, iron is not a lightweight diaphragm material either. And yes, upper frequency limit is the top resonance, which is raised with thinner diaphragm. Typically this resonance is "used" to balance eddy losses or to give a rising (to a point) response suitable for hearing loss.

Advantages: inexpensive, INCREDIBLY robust.

There's a classic paper on the design of the Bell 500 telephone with some theory on this transducer. (IIRC, they put the coil around the rim instead of the center tit.) I may have other material here. But if you don't have Bell Labs' machine-shop, I suspect sanding-off the thickness tolerance, and winding to suit your favorite impedance level, is the best hot-rodding.