[ETM]

The technical issues seem to mushroom with this, but still, it's not rocket science.  Do you have some plots you can share?

Indeed, there is nothing high-tech involved. These are things that can be demonstrated with basic measuremet equipment.

Below is a plot of how the impedance magnitude of a typical hi-fi driver changes at 2 kHz frquency as a function of cone displacement. The drivers tested were A: Vifa M13SG-09-08, B: Seas P14RC4Y/DC, and C: Peerless 833429. In each driver, the measurement has been extended to the rated Xmax limit. The displacement was actuated mechanically and measured with a scale. This is only one out of the many deleterious distortion and interference effects voltage drive is beset by and not necessarily the most serious.



Due to changing inductance, or inductive EMF, the impedance typically varies +/- 10% within a driver's linear excursion range. This means that on voltage drive a 2 kHz tone reproduced by the same driver exhibits +/- 10% amplitude modulation with accompanied phase modulation also. According to basic theory of amplitude modulation, the distortion components thus generated amount to 5% of the original amplitude, giving rise to a full 7% of total intermodulation distortion, all this eliminable by current-drive operation.

...infinite output impedance...

Can you explain the phrase above a little more?

An ideal current source is characterized by infinite internal impedance just as an ideal voltage source is characterized by zero internal impedance. Neither of these extremes can ever be achieved in practice, but for loudspeaker current-drive, a few hundred ohms is well enough.

[ETM]

The price is now 21% off ($27.76) at Amazon, so if you have any honest interest in sound quality matters, do not hesitate any more!

Here is a link to the news release at mi2n.com: http://mi2n.com/press.php3?press_nb=126137

[Tyson]

So, the idea of matching amp output impedance with speaker impedance is possibly one reason why going to an active crossover works so well with tube amps?  For example, I recently bought a DEQX, which does preamp functions, crossover functions (I set my at 96db/octave slopes), phase correction, group delay correction, time-alignment corrections, and frequency response corrections (both inside of each drivers passband, as well as between the different drivers).  I have a 4 ohm tweeter, which I use the 4ohm output tap from my dedicated tweeter tube amp, and I have an 8ohm midrange driver, which I use the 8ohm output tap from my dedicated midrange tube amps. 

Is this essentially the approach you are advocating here?  If so, it would explain a lot of the dramatic jump in sound quality when I went active and was able to do phase and frequency response corrections. 

[ETM]

Tyson,
In current-drive design, impedances are not matched, but the output impedance and the impedance seen by the drivers is kept as high as possible. With infinite output impedance, there is not necessarily need to "match" amps with speakers any more than there is such need in voltage drive. Note also that the nominal impedance of the output taps is a different thing than the output impedance.