A peer-reviewed research paper Comparative Measurements on Loudspeaker Distortion: Current vs. Voltage Control has been recently published in Archives of Acoustics.
Have you heard the postulation that low-impedance voltage or big "damping factor" "controls cone motion"? In the engineering sense, however, there is really no such thing as control of the cone motion. It is therefore quite the greatest and most regrettable fallacy that the misinformed audio community has unsuspectingly adopted and that is being used to justify the way things stand and discourage efforts to reconsider. See the article published on the EDN network: Part 1 & Part 2.
(Note: Like in the book, existing faults are brought out at times even quite vigorously. However, the issues treated are universal, and nobody has reason to take them personally although recognizing one's own misconceptions may sometimes produce inconvenience. Yet of course there are always those who do, reacting with personal smear instead of technical facts. Examples can be found on Diyaudio.com. There is nothing surprising; it has always been so. That's also why Copernicus postponed the publication of his work De revolutionibus orbium coelestium until the last days of his life.)
An interview by the keeper of a Polish tube amp site, Hiend Audio.
An interview article about the book by Kari Nevalainen has appeared on Inner magazine together with an outline of each chapter. Many key points have been understood and represented here quite accurately.
Hifi & Musik, a Swedish print magazine, published an article about current-drive in issue 10/2013 based on the above interview. They have added some own images and also dropped some offered, like the book's cover. The technically demanding text contains a number of errors, already starting from the book's year, but is nevertheless, as a whole, rather informative and shaking and something unprecedented in the history of hi-fi.
A review of the book by Hans Polak has appeared in Linear Audio vol. 12. (The article is orderable in PDF format if one is not willing to buy the whole volume.)
The presentation is a kind of outline about the operation of current-drive and the benefits achievable. The bottom line says: With this book, the author succeeded in convincing me that for loudspeakers, current drive is, in theory, superior to voltage drive. But the industry took another road and has most likely passed the point of no return long ago. The reviewer seems to believe, however, that drivers would yet evolve better and thereby improve things for voltage drive (e.g. ironless magnet circuits, moving-magnet principle).
For the most part, things have been grasped and rendered quite accurately, and the stuff serves as an introduction to the subject but unfortunately not without fallacies. Some remarks:
p. 185: It is said that on current drive the influence of back EMF on the current is roughly Re/Ro. -Not comprehensible and not from the book.
p. 186: Spring constant (K) confused with compliance.
p. 187: The bump in impedance is the same irrespective of the driving mode.
p. 187: The microphonic EMF was not measured around the resonance frequency but in the lower midrange showing a standing-wave peak centered at 380 Hz.
p. 188: The fluctuation of inductance Le with displacement causes distortion also when the dependency is linear.
p. 189: Here, the BL nonlinearity has somehow weirdly been mingled with the current-dependency of inductance though the book discusses only the latter. Also, eddy currents in the iron parts do not constitute a distortion mechanism, nor is such notion from the book.
A review of the book by Joseph DeMarinis has appeared on the AudioXpress site (originally published in the July 2010 issues of AudioXpress and Voice Coil printed magazines). While it was to a large part rather accurate and relevant in what it discussed, unfortunately though not so surprisingly, the most important findings of the book, dealing with the flaws of voltage drive, were not really considered. Also, there appeared some notions and interpretations, reflecting the accustomed ways of thought, that deserve amendment.
My comments on the review were eventually published in the November 2010 issue of AudioXpress but not in Voice Coil. Main remarks are in the following:
"According to conventional wisdom, there are two very good reasons why voltage drive is used universally: loudspeaker resonance and damping..."
The damping of a loudspeaker's resonance is a single well-defined task to be performed - not two; and once done in the frequency domain, the time domain will automatically settle in place since the time behavior of a linear system is always only a reflection of its frequency behavior, as determined by the Fourier transform relationship. As explained in the FAQ:s, electrical damping is nonexistent beyond an octave or two from the resonance; and in the resonance region it can be substituted in all aspects by mechanical damping with the same outcome. Thus, there are no valid reasons, let alone very good ones, to perform against the clear directive of the governing law F= Bli, especially in the middle and treble regions. It is also fully feasible to use some amount of passive or active electrical damping for bass and increase the impedance level for other frequencies, where the benefits of current-drive mostly appear.
"In other words, implementation of these ideas requires a system approach, as can be done effectively with powered loudspeaker products."
In principle, current-drive doesn't necessitate powered loudspeakers more than voltage drive does, and passive speakers with flat-response amplifiers can also be made to work. With dedicated amplification for each driver, the source impedance can be kept more ideal, but just the same kind of restrictions apply when one is striving for voltage drive.
"Most of his objections to voltage drive stem from the back-EMF..."
Here, we come to the main omission of the review. The objections are only halfway due to the motional EMF. The effects of the inductance EMF are even graver, but this major issue was not even given a mention. (The voltage across any inductance, in this case that of the voice coil, is also an EMF, as it is induced by a fluctuating magnetic field.) The most important chapter is #4, where the flaws of voltage drive are demonstrated; and without an idea of these phenomena, readers are not able to make reasoned judgments on the subject. Trying to reproduce some of these effects would have been more enlightening than just focusing all attention on the bass region damping, as the loudspeakers could not be modified.
The pulse tests
(Im)pulse responses, in general, do not render useful information as to the qualities of the driving modes since the impulse response of any linear system contains only the same information as the frequency response but in a different form. This is because the frequency behavior and time behavior are not separate issues but one and the same characteristic only viewed differently. Therefore, if we have a current-drive speaker that exhibits the same frequency response (including phase) as some voltage drive speaker, their time responses (for a given input) are also inevitably identical, as dictated by the Fourier theorem; so in this regard one cannot be better than the other.
When interpreting oscillations, one should always be mindful of what frequencies are dealt with. In Fig. 7 (KLH pulse), one can observe a periodicity of some 0.5 ms, which corresponds to the 2 kHz response peak appearing in Figs. 5 and 6. In Fig. 8 (SEAS pulse), there occurs quite pure 5 kHz oscillation that stems directly from the sharp cone break-up peak seen in Figs. 3 and 4. In Fig. 9, the oscillations are above 10 kHz, the corresponding response peak being left beyond the scale of Figs. 1 and 2.
In each case, the oscillations reflect the high-frequency prominences occurring in the respective frequency plots but are unrelated to the Q-value or any bass damping properties of the systems.
The step tests
The same principles also hold for the step responses shown in Figs. 10 and 11; the time domain properties are a consequence of the frequency domain properties rather than the technology used. In Fig. 11, you can discern a period of about 30 ms, which is attributable to the 36 Hz peaking found in Fig. 4. The current-drive amp, with its over 1 kΩ impedance, cannot be responsible for any damping in any of the tests.
The acoustic phase
In a 2nd-order high-pass system, like a woofer in a sealed enclosure, the phase shift approaches zero at mid frequencies, rises to 90º at the resonance frequency, and levels off to 180º below the resonance. In Fig. 12, the excess phase lead in addition to that is likely due to the room effect, as the wavelength becomes large compared to room dimensions and the pressure field becomes static. Anyway, the result also shows that low frequencies are reproduced in advance of higher ones, at least in terms of phase, and not after, which contradicts common thought.
The bass distortion
Concerning Fig. 13, lowering the mechanical Q is indeed key to control this kind of distortion; but it is also noteworthy that the driver was mounted in a very large enclosure (I was told 95 ltr) and without proper stuffing, so the system is operating at the mercy of the driver's own nonlinear spring force and damping force. Both of these can be linearized considerably with a proper size cabinet and effective use of damping material (and driver optimization).
"It seems that the reverse current generated by spurious cone movement "distorted" the voltage drive net current in such a way as to correct some of the distortion."
This is the partial velocity feedback effect that occurs within an octave or so from the resonant point but is ineffective elsewhere, where the EMF current is smaller and perpendicular to the signal applied.
The constancy and also the level of the distortion (THD I suppose) from 100 Hz up suggest that here the distortion is coming more from the microphone itself than the speaker. This is rather common in close-up measurements.