Each have their virtues, but I'd venture to say if both ferrite and alnico were made in the same manner (slug or other) with the same gauss ratios, voice coils and cones, that they'd both *sound* the same.
There is a lot of theory behind dynamic drivers, a heckuva lot more than I used to think, and there are parts of it that would disagree. The term for the day here is "Barkhausen noise".
The idea there is that the major differences in sound between AlNiCo and ceramic or ferrite magnets is that the former is electrically conductive compared to the latter. In the case of electrically non-conductive ceramic magnets, the theory, strongly forwarded by John Wilkinson (look his name up, he's not just another throw away "I know it exists because I can hear it" audio BSer) amongst others, is that eddy currents differentially form between the top and bottom plates that are both formed by the magnetic field created by the voice coil and interact with it again making for non-linearities in the overall magnetic field.
The basic idea is that this interaction causes all the materials involved to experience some jumps in magnetic domains that creates rapid changes in the magnetic field that works with the voice coil to create electromotive motion that in amount to a background noise. This noise is thus named for our Dr. Barkhausen from his first discovery of this effect back in 1919. The theory here is that this lowers the effective signal to noise ratio of the driver and the aforementioned Wilkinson argues that in effect, ceramic drivers are in capable of recreating the dynamic range found in modern digital media and may even have less effective SNR than even quality vinyl.
With drivers that have greater excursion, such as woofers, this has been measured as impedance variations during excursion. The resultant non-linearities result in a great deal of distortion as well. A fix has been implemented with the addition of copper rings to the magnet structure around or inside the voice coil to help break up the worse of these eddy currents and thus improve the linearity quite a bit. Sometimes multiple shorting rings are used such as in the XBL^2 technology that some subwoofer drivers use.
However, there are those who argue that shorting rings are only a bandaid that holds up improved impedance behavior as a quality while ignoring the low SNR due to the Barkhausen noise altogether. Proponents of the idea hold that only electrically conductive magnetic material such as AlNiCo and Neodymium can minimize this issue and allow maximum fidelity.
But hey, don't take my word for it. Let's go back to 1991 with US Patent #5,070,530:
Electroacoustic transducers with increased magnetic stability for distortion reduction
BACKGROUND OF THE INVENTION AND PRIOR ART
The above-referenced copending application discusses observed differences between loudspeaker constructions utilizing non-conductive ceramic magnets and loudspeaker constructions utilizing alnico-type magnets and points out the general inferiority of the ceramic magnet structures. In particular, the effects of distortion due to eddy currents, in the top and bottom plates, which generate local magnetic fields that are coupled back to the voice coil are noted. These eddy currents produce harmonic distortion effects due to the non-linear iron characteristics as well as frequency selective, i.e., frequency dependent, distortion effects because the amplitudes of the eddy currents are proportional to frequency.
In the prior art, the effects of energy loss due to eddy currents in the conductive parts of an electroacoustic magnetic transducer motor structure have been misunderstood. For example, extra conductive material, generally in the form of copper, has been added to the motor structure to flatten the loudspeaker impedance characteristic, i.e., make the characteristic more uniform with frequency. The fact that the energy transferred into the conductive material reduces the energy that is transformed into useful loudspeaker diaphragm motion, and that this effect, which is non-uniform with frequency, results in a reduction in the accuracy of reproduction of transients, has either not previously been recognized or has been ignored. This frequency selective loss in prior art transducer constructions results in transducers with reduced ability to track the rapid changes in audio signals. Indeed, a flat impedance characteristic in a loudspeaker driver has been found to be of secondary importance and is even undesirable when it is produced by non-linear or frequency selective losses in any of the parts of the magnet structure.
It has been discovered that in addition to eddy current effects in ferromagnetic structure parts, ceramic magnets in contrast to alnico magnets, introduce another distortion component. Magnetic fields are introduced into the magnet material by the motion of the signal-carrying coil, whether in a loudspeaker or a microphone embodiment. This energy, which is effectively subtracted from the available useful energy, is proportional to coil travel and is thus inversely proportional to frequency. There are undesirable consequences associated with the phenomenon. For example, the signal-related AC magnetic energy that is induced into the magnet causes distortion. While the exact mechanism has not yet been proven, it is believed that the induced AC magnetic field modulates the DC field in the magnet.
I recommend doing your own research and reading on the subject, because I'm really just trying to figure it out. Well, actually, I kind of gave up trying to figure it all out once I kept running across all these little things that add up to the total driver behavior. It's beyond me, simply put. Can we hear the end result? Dunno, but if there's such a bias for the sound of AlNiCo and Neodymium magnets without people having previously read about this effect, then maybe there's something there. Definitely some interesting stuff, though.
Oh, contrary to what the name Barkhausen may imply, the theory does not apply merely to
woofers. (Yeah, yeah, lame, I know.)
- JP
![JimJ[VT]](/forums/data/avatars/m/20/20031.jpg?1443894878){kind=avatar}
