What do you understand about this article? what is it they did ?
https://hal.science/hal-02504314/document
https://hal.science/hal-02504314/document
Magnet-Only Loudspeaker Magnetic Circuits
- Thread starter arivel
- Start date
What I understand is that -basically- the iron centre piece of original the magnet assembly has been replaced by two ring-magnets and a plexiglass piece to hold it all in place. Also, the overhung VC has been replaced by an underhung VC. To my best of knowledge, the Purify motors also have a centrally placed magnet instead of an the classic all iron centre piece.
Maybe member Snicker-ish, very versed in motor simulation and design, will chime in.
Maybe member Snicker-ish, very versed in motor simulation and design, will chime in.
Same of some experiments appeared in the Blowtorch thread. The currents/voltage drive still Is discussed in the sequel
Is there anything in common with this other article?
https://audioxpress.com/article/voice-coil-spotlight-some-speaker-problems-that-needed-solving
https://audioxpress.com/article/voice-coil-spotlight-some-speaker-problems-that-needed-solving
The whole concept is to basically replace ALL iron (or as much as possible) in the entire motor structure.
See summary, conclusion and Figure 1 vs Figure 2.
The idea behind this, is because iron has quite some hysteresis and nonlinearities in it (Eddy currents).
Depending on the quality and type of iron obviously. (and costs)
The main reason why using demodulation rings is a necessity for best performance.
(for frequencies roughly > 200-400Hz)
Since magnetic material and processing techniques are becoming a lot more affordable, it's an interesting thought to make the magnetic structure completely out of magnetic material.
My take on this, is that this was also the starting point of this paper.
Constructing a loudspeaker motor is always a play of compromises, so it's very interesting to see what kind of compromises one can aspect. Probably also from a patent point of view.
Another solution (or a combination of), is just adding a huge magnet, creating a lot of BL.
Which helps with getting those pesky eddy currents at bay.
It's just not very price effective and can mess up some other parameters as well.
I absolutely love the work and papers from Antonin Novak btw, they are always extremely interesting to read! 🙂
Last edited:
They are not inventing a new wheel, at all.
Main problem in any speaker (or motor) design is horrible permeability of air (or vacuum) so magnetic path through it must be reduced to absolute minimum, a very narrow gap, hence the need to use iron or ome other high permeability material to, as they say, "canalize" it:
So what?
They imply that new material will be "more linear" than iron, I very much doubt that; to boot permanent magnet materials (as opposed to temporary magnet ones, such as iron) are precisely used because of their VERY NONLINEAR curves, which allow them to retain magnetism even after the magnetizing force is removed.
Main problem in any speaker (or motor) design is horrible permeability of air (or vacuum) so magnetic path through it must be reduced to absolute minimum, a very narrow gap, hence the need to use iron or ome other high permeability material to, as they say, "canalize" it:
So they remove iron BUT fill that space with some more magnet material.
So what?
They imply that new material will be "more linear" than iron, I very much doubt that; to boot permanent magnet materials (as opposed to temporary magnet ones, such as iron) are precisely used because of their VERY NONLINEAR curves, which allow them to retain magnetism even after the magnetizing force is removed.
In addition, the magnetic field generated by the voice coil will easily penetrate the low-conductivity ferrite, compounding the linearity issues.
With iron pole pieces, this effect is reduced, especially if the conductivity is enhanced by a shorting ring made of aluminum or copper
With iron pole pieces, this effect is reduced, especially if the conductivity is enhanced by a shorting ring made of aluminum or copper
Looking at the paper I don't see the overall advantage to their plan. A little bit of steel somewhere still gets the job done best. Maybe when there is a material yet stronger than neo magnets than can be charged through complex angles other than the pressing direction things will dramatically change.
why are you talking about hysteresis? in a loudspeaker the magnetic field is static. Are you referring to the moving coil?
Interesting paper. Thanks, OP!
Original circuit vs "no-iron" circuit:
No, everything remained the same save for the magnetic circuit; no-iron magnet assembly was designed in such a way that original coil of 18 mm height became underhang. Thus VC had more area to collect much more sparse magnetic field and Bl (=sensitivity) remained the same, albeit with much smaller (1.5 vs ~6 mm) Xmax.
Original circuit vs "no-iron" circuit:
You are reading the paper wrong (well, too small illustrations doesn't help). They don't fill the space around the coil at all - from magnetic point of view, it is suspended in thin air, so nonlinearities from magnet material (NdFeB discs in this case) are irrelevant. Merit & Novak threw away any "canalization" attempt.
Distortion elimination. There is no high order distortions left, 2nd order is reduced by 10 to 20 dB, 3rd by 40 dB. Without any complex arrangements, half a kilo of copper in the motor or peculiar VC winding they produced possibly the most linear motor I have ever seen.
Go look up what eddy currents are.
In general this doesn't seem to be understood well in this thread.
A rough explanation has been even offered in the paper we're talking about.
Air gap contains DC magnetic field from permanent magnet an AC magnetic field generated by the voice coil (i.e. AC field modulates DC field). The Voice coil with a metal core (a pole piece) behaves as nonlinear inductor whose inductance depends on both the voice coil current and the voice coil coordinate (i.e. Le(I,x) ).
Going a little bit deeper into physics, dependence of the magnetic polarisation J of the metal core on external field H is essentialy nonlinear process in the simplest case described in the frame of phenomenological Mean Field Theory, particularly using hyperbolic Langevin function for J(H) dependence. For this reason, the voice coil inductance also will be nonlinear and responsible for nonlinear distortion production in the mid frequency range. (because the J(H) dependence is described by hyperbolic Langevin function, which is odd function, the use of the metal core will cause nonlinear Le dependence on the current in the voice coil and results in the rise of odd order harmonics in the mid range. Moreover, additional source of the second order distortion is dependence of inductance Le on the voice coil coordinate).
In sum, the use of metal core led to an increase of odd and even order distortion in the mid range, the reduction of which is not an easy task. So, the magnet-only ironless construction looks quite reasonable.
Yes, there are a bunch of other solutions as well. Some of them are already mentioned in this paper. Each have there pros and cons. Besides there is in some cases an additional story that involves patents.
This paper just investigated one of the solutions. Nothing more, nothing less.
This paper just investigated one of the solutions. Nothing more, nothing less.