DIY opposing voice coil drivers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Ok so I did an internet search and a diyaudio search and couldn't find out much about this, there's probably a simple explanation but still, here goes:

Magnets:
Modern (magnetic) drivers have a fixed magnet that the voice coils magnetism drive against, to create movement from the current an amplifier delivers.
The disadvantage is that the magnet's non-linearities are a source of distortion, which is not such an issue for the earlier Alnico magnets.

Field coils:
Field coil drivers replace this magnet with a large DC electromagnet, field coils speakers were used in some old tube sets were there was a useful constant current feeding the amp and the electromagnet acted as a useful choke too.

Opposing driven coils?
So the obvious next solution is to simply replace the magnet/field coil assy with a slightly bigger voice coil from another driver.
If the field lines need concentrating I would think some soft iron tubes might work as cores between the two.

Then the amplifier current would create both attractive and repulsive forces between the mobile voice coil and the fixed one.

Has anyone ever tried making a driver like this, perhaps modding an old driver? Any ideas of sound and/or efficiency.
Why don't any driver makers ever do this?
 
Something to ponder is the "Left hand rule" of magnetism. The force generated by a conductor carrying a current is at 90 degrees to the magnetic field lines.

I don't quite follow your comments on the relationship between magnet types and non linearity/distortion. The voice coil encountering non uniform flux density throughout its travel is certainly a distortion mechanism; but is probably more related to the compromises that have been made in the design of the pole pieces, underhung versus overhung etc.

Where there is a distinction between magnet types is in magnetic materials that are electrically conductive versus non conductive such as ferrites. The latter types give rise to flux modulation, which is to say that the flux lines can "bend" or be displaced under load. Englishman John Watkinson has commented on this subject in relation to ceramic magnets. He suggests that the bending occurs in quantum steps to an extent that it is akin to going from 16 bit word depth to (say) 14 bit.

Keith
 
I assume with an energised coil the magnetic field forms along the axis of the coil, but whether an iron core rather helps that I'd not sure.

Assuming the force is axial like in a solenoid then I would think the force would pretty much be in-out as require. Maybe I need to find a couple of scrap drivers and experiment!

The distortion of ceramic magnets I think related to their influence in the inductance of the voice coil but I could be wrong, I just hear that Alnico gives a different sound to ceramic. Which presumably means that field coils would sound different too, as would opposing coils - if they work..
 
...there's probably a simple explanation but still...Why don't any driver makers ever do this?
You would like your driver to have a linear relationship between applied voltage and SPL output. In conventional dynamic drivers using permanent magnets or field coils, the SPL output is proportional to B x L x I, where B is the magnetic field strength in the gap the voice coil moves thru, L is the length of wire in the magnetic field, and I is the current thru the wire. If the impedance of the voice coil is Zvc, then I = Vapplied / Zvc. So assuming a stable B, SPL is directly proportional to Vapplied.

If you create the magnetic field B with a field coil hooked up to Vapplied rather than a constant DC voltage supply, B would be proportional to Vapplied. In this case, SPL would no longer be proportional to Vapplied, but to the square of Vapplied...a decidedly non-linear relationship.
 
The disadvantage is that the magnet's non-linearities are a source of distortion, which is not such an issue for the earlier Alnico magnets.
Linear or not, the fixed magnet supplies a constant magnetic field, so whether its own curve is linear or not it´s irrelevant, you always operate at the same point so no non linearities can develop.
To boot that magnetic field operates at 90 degrees from current in the coil.

As of having an audio created/modulated "mechanically fixed" field , and applying same audio to both coils, you create square Law response, what bolserst said.
 
As JMFahey posted, as long as you have lots of coil in the gap (AKA overhang coil design), makes no difference if magnetic field is somewhat irregular.

But perhaps the fertile question is: given a square-law activation, how can that be leveraged into a suitable sound output? Does it vary with frequency in a helpful EQ manner? Can it turn a tapped "horn" into something musical (AKA pig's ear to silk purse conversion)?

B.
 
Last edited:
I have performed this exact experiment, together with aluminum and copper VC material evaluations, and I can tell you that:

1. The comsol and matlab modeling (Biot savart law, etc) did not predict strong forces at DC or AC.

2. An 8ohm 3" VC needs around 3kw of power to generate a perceptible output from a driver (VC, former, cone, suspension, frame)

3. Inductance is a HUGE issue, and counterwinding, layering, etc all proves mostly useless.

From my limited experience, I would expect that the only way that you would get any decent reaction would be with a DC/AC pair, in the style of the field coil. Since the DC section usually is very heavy, the magnet is almost always replaced with the DC section.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.