magnet design 101 - saturate iron with a ferrite magnet?

[Bill F.]

Here's that model I cooked up of the ceramic-mag motor with over 2T in the gap--the outer limit of what would be achievable with ceramic. The majority of the return circuit is 1006 steel, but the piggy-back top plate and the pole cap are permandure--an unavoidable requirement to achieve 2T+.

Notice how much of the flux projects from the surface and is wasted. That is one of the downfalls of trying to achieve high flux densities with ceramic magnets. Rare-earth would be much better suited to this application.

Bill

[capslock]

Hi Bill,

I guess that software will keep me happy at some point. Just need to find somebody to build the magnet systems for me...

In your new example, I am once more surprised at the low flux in the magnet, around 0.2 T this time where I would expect more than 1 T. I guess Permadure is the right choice if maximum flux is your aim. If you're shooting for saturation, sticking to steel might be a better idea.

Cheers,

Eric

[capslock]

Double and even triple magnets are becoming increasingly popular. However, the field in the gap should be primarily determined by the cross-section of the magnet, not it's hight. So in first approximation, the added hight has no effect except to allow a longer VC stroke.

Secondary benefits might be:
- the energy stored in the magnets is bigger compared to the changing energy of the VC, so there is less modulation
- it is easier to concentrate the field onto the gap with a longer magnet assembly??
- ?????

[Bill F.]

I believe you're correct that maximizing magnet cross-section is the most efficient way to harness its flux. I've heard an an approximation that it takes close to 10x the stack height to double the flux density of a single magnet.

I don't know about field stiffness benefits of stacked magnets, but I believe stiffness is more directly linked to how much surplus flux is available at the gap to maintain saturation under power, independant of the geometry of the magnet that supplies the flux. I suppose you could think of this surplus flux as "transient headroom."

Another interesting thing is that the level of field modulation is inversely proportional to the efficiency of the voice coil's coupling to that field. For example, think of a ribbon driver in which there is a wide air gap flanked by poles saturated by very large magnets (necessary to achieve sufficient flux across the gap). The ribbon couples so inefficiently to the B field that it will cause negligible field modulation.

In a similar sence, driver efficiency is your friend. To achieve a given SPL, an efficient driver will generally have a stronger B field and a lower current in the VC--a recipe for reduced field modulation.

it is easier to concentrate the field onto the gap with a longer magnet assembly??

Yes, to a point. However, the thinner the magnet the more flux short-circuits around the edges. I think there's an optimum balance. You'll always lose some. Notice in the HE ceramic model I posted above, a large portion of the magnet flux flies off into space or short-circuits and is wasted, even though the return circuit doesn't approach saturation anywere but near the gap.

[capslock]

Originally posted by Bill F.
Another interesting thing is that the level of field modulation is inversely proportional to the efficiency of the voice coil's coupling to that field. For example, think of a ribbon driver in which there is a wide air gap flanked by poles saturated by very large magnets (necessary to achieve sufficient flux across the gap). The ribbon couples so inefficiently to the B field that it will cause negligible field modulation.

In a similar sence, driver efficiency is your friend. To achieve a given SPL, an efficient driver will generally have a stronger B field and a lower current in the VC--a recipe for reduced field modulation.

Aren't these two examples contradicting? The ribbon has inefficient coupling and the the efficient driver has efficient coupling.
The efficient driver is less effected because there is less AC current and because the magnet is bigger compared to a more inefficient driver.

Greetings,

Eric

[capslock]

I have read that modern ferrite magnets are magnetized with the iron system installed. Apparently, the assemblies are inserted into the machines with their back plate first, i.e. there is no adaptor that attaches do the pole piece and top plate.

So it would appear that they use a coil that is essentially wrapped around the circumference of the ferrite magnet. It's field will magnetize the ferrite ring accordingly. However, it also extend into the interior of the magnet ring. This space is filled by the pole piece. The field inside the magnet ring will be conducted through the plates into the pole piece, but inside the pole piece, the orientation is exactly opposite to the coil's field. So it will act to oppose the external field. No harm is done because the iron is not permanently magetizable, but it seems a waste of energy compared to coupling the field into top plate and pole piece.

How are field coil speakers built? It would seem the only sensible way is to build them similar to old AlNiCos with an outside return, with the magnet being the pole piece. If they were built like conventional ferrite designs with a center return, the flux there would be opposing the external field.

[Bill F.]

Aren't these two examples contradicting? The ribbon has inefficient coupling and the the efficient driver has efficient coupling.

Not contradicting, but perhaps not necessarily related. I guess what I'm trying to say is (assuming good design and saturated poles) the greater the ratio of B flux to coil current, the smaller the field modulation--obvioiusly.

In the case of the ribbon driver, there is a great deal of gap flux that the ribbon doesn't harness efficiently. It doesn't need to because it is so light.

In the case of the high-efficiency cone, the voice coil does not necessarily have to make efficient use of the B field. Consider an underhung topology.

[destroyer X]

A friend, from Hungary, will build one of my amplifiers...but a special underpowered one..output may be 12 volts...so...1.5 amperes will cross the output coil.

The coil is very small, for instance, if he intend to use it with a 22 turns coil (air core is around 2uH.... what gonna be the resultant inductor value and i would like to know if you had that experience..if this saturates and produces an audible "tac" sound when a peak of audio crosses and saturates the core.

So.... three questions in the reality:

Will this saturate with only 12 volts and 1.5 amperes crossing?
What gonna be the resultant inductance of that coil, now over a ferrite core?
Had you experienced the audible "tac" noise of saturation?


If you have a chart, coil turns/coil wire diameter/Inductance.... please... send me or post it here:

carlos.eugenio1951@yahoo.com

regards,

Carlos

[speakerman19422]

What driver companies use are for profit not just performance. Feroba 2 is only used by Lowther, once it was common. The choice of voice coil formers are the same Royalties are paid on liscensed materials. Nomex is more neutral then aluminum and kapton, rarely used now . Pro musicians that play guitar like Celestion woofers due to the neutral sound the nomex formers give. Double stacked off centered magnets are the best. The cost is not profitable when off centering the magnets. Less leakage occurs in the magnetic field.. The same can be said of cone materials There have been some clasic designs that have not been recreated. I know of one company that had USA made stamped frame drivers that were superior to the European cast frame counter parts at a fraction of the cost that were superior in t-lines.