Voice coil physics

[BugEZ]

I have followed the voice coil rail thread http://www.photomacrography.net/forum/v ... hp?t=33439
with great interest. I have contemplated a mineature stacking system that could be easily transported by air in a small suitcase. My intended subject would be a fly (wings,body eyes) and I only need 4-5mm of travel to allow good 2X-10X photos of the small flies I document. So, the simplicity of a voice coil actuator makes it worth exploring.

To this end I went to a thrift store and purchased a relatively small speaker box and opened it up. I removed the larger of the two speakers inside and dissected it. Prior to cutting the flexible diaphragms that support the coil and the speaker cone I applied 1.5 volts across the coil. The cone moved about 2mm.

As a mechanical engineer I found it quite enjoyable to examine a well engineered product. The coil was wound around a thin but rigid paper tube. The mass of the tube and cone (also thin rigid paper) is kept low so that the frequency response can be high. A relatively short length of the tube has windings on it. Quite astonishing to think that the coil is rated for ~10W. I would have expected it to overheat. The wires between the moving coil and the fixed terminal posts are fine braided brush wires to make them soft and to resist fatigue. The rear diaphragm (centers the coil in the magnetic gap) was surprisingly stiff both axially and radially. The front diaphragm (aligns voice coil) was more compliant.

I found a nice diagram of a voice coil showing the typical voice coil configuration. This link captures the physics of the magnetic flux path.

http://www.sageofathens.com/Documents/L ... cecoil.pdf


When hacking a speaker such as this one to create a bug actuator, I will need to extend the axial travel. Since I don't need high frequency response, I can wind a longer coil (the added reluctance of additional windings and the mass won't matter) I can get longer travel. If I can engineer radial supports that produce a lower axial spring rate the power needed to hold a fully stroked actuator will be low.

I'll post photos of the speaker distraction tomorrow.

Edit for spelling. K

[ray_parkhurst]

That is an excellent mini-tutorial on speaker design and modeling.

Should be an interesting project!

A few notes of my own:

The "spider" is all that is really needed for return force, but it only gives a small amount of stability to the coil. The coil must stay centered within the magnetic gap without ever touching the sides, or it will cause a spurious "scratching" sound. Not pleasant at all! It is the cone that offers triangulated support to keep the coil from skewing off-axis and rubbing against the pole pieces. So when you re-assemble the system, you need to ensure a similar structure is in place or you will lose stability.

The power dissipated in the coil is transferred to the pole piece through a combination of radiation and convection. There is almost zero conduction through the paper cone the coil is wound on, so the heat must exit elsewhere. Motion of the coil generates a fair amount of air movement, and this hot air then transfers heat to the metals of the magnet structure, which conduct to the frame.

Some speakers that don't need much displacement (tweeters and upper midranges) can use an iron-loaded thixotropic "ferro-fluid" material to fill the gap. This helps with many aspects: improved magnetics by effectively reducing the gap; improved thermal response by adding conduction to the thermal path; and dampening through friction. Unfortunately, this won't help us with the voice coil stacking rail, but it is useful to understand the principles.

[BugEZ]

Interesting comments Ray!

Here are the dissection/destruction photos.
First the speaker from the top with the top diaphragm partially sliced...

After slicing trough the upper and lower diaphragms the lower diaphragm is much easier to see...

Prior to cutting the brush wires that pass current to the coil...

The winding actually has two layers. Why a photo is worth 1000 words...


~ 30 turns per layer...


The air gap between the pole pieces is exactly as shown in the linked diagram. It is most narrow in a small axial section near the exposed face. I checked this with a small bent wire that could be inserted then rotated to check the gap.



The frame that supports the speaker cone is riveted and glued to the voice coil steel armature. The heads of the rivets are visible here. I will have to grind them off to separate the voice coil. More fun!

[mawyatt]

Looks like another fun project, good luck and keep us posted.

Thanks for the design site, very informative...even to an EE

Best,

Mike

[enricosavazzi]

(...) Quite astonishing to think that the coil is rated for ~10W. I would have expected it to overheat. (...)

Loudspeakers are often rated in "music power", which is quite different than either continuous (=DC if used for stacking) or RMS power. See e.g. https://en.wikipedia.org/wiki/Audio_pow ... eal_issues . Many loudspeakers are often also over-rated with respect to their actual power, however measured.