Now for the relevant data regarding how much the drivers in a multi-way speaker system affect one another.

First, the details.

Data was taken on a two-way 8 ohm nominal, 8" woofer, 1" soft dome tweeter system, with a second order crossover network.
Drive level to the speakers was approx.  1W, 2.83V  Curves were taken in an anechoic chamber.

The graphs have 0 dB at 0.316 Volts RMS.

In order to correlate the drive level with the output level, subtract 19 dB.

FIGURE A
Output of Woofer when acoustically driven by Tweeter in Red.
Output of Tweeter when acoustically driven by woofer in Blue.
Direct connection to driver not electrically driven (crossover bypassed).

FIGURE B
Output of Woofer when acoustically driven by Tweeter, through 2nd order crossover in Light Blue.
Output of Tweeter when acoustically driven by woofer, through 2nd order crossover in Green.
Connection through crossover (crossover not bypassed).

Analysis

Figure A indicates that the voltage coming out of the woofer when driven by the tweeter is approximately 69 dB down from the input drive level.  The voltage coming out of the tweeter when driven by the woofer is approximately 47 dB down from the input drive level.

Figure B indicates that the voltage coming out of the woofer through the crossover when driven by the tweeter is approximately 70 dB down from the input drive level.  The voltage coming out of the tweeter through the crossover when driven by the woofer is approximately 64 dB down from the input drive level.

These levels might seem to put the amount of inter driver IM down to levels that are below audibility, and for some that believe that distortion has to be higher than -40 or -50 dB,  this would seem to be the case.  These levels would then be attenuated by the figures in the driver to driver attenuation charts.

However, those familiar with my dither algorithm argument will realize that I believe that if signal aberrations are above -90 dB, they might be audible.

Figure C is the telling one, as the levels are NOT that low.  Using the -19 dB to equate input level to output level, and subtracting another 6 dB for a typical hi-fi subwoofer/woofer, we can see that peak levels reach - 34 dB.  This is much higher than one might have expected.  This means that any system with a subwoofer will be modulating the woofer to act as a significant signal generator, and this is what could feed into the tweeter through the crossover network.  Note that the curves were taken in an anechoic chamber, and so, room modes are not included, these would tend to cause peaks and dips in the overall bass levels that existed within the room, and what levels the various drivers would experience.

This got me to thinking, there is one other source of external acoustic output:  THE OTHER SPEAKER/WOOFER IN A STEREO PAIR.  This output would not be rolled off at higher frequencies that are still within the range of the woofer output.
If we allow for the distance from the other woofer to arrive at the levels that would occur, we would subtract another 10-12 dB for a 10-12 foot distance from the other woofer in a stereo pair.  This has the peak level now at approx. -45 dB due to acoustical output from the other woofer!  Due to the acoustic travel delay, this other signal would NOT be coincident with the original bass signals, and so, would constitute a definitely undesirable signal aberration.  Delay would be on the order of approx. 3-4 mS.

So now we return to the tables of crossover attenuation with the various wiring schemes, and see that if we single wire, the levels of these distortions might be high enough to hear, and that by reducing them through bi-wiring, they would become less audible, or even fall below audibility.

So the action of the various speakers in a multi-way, multi-channel system WILL likely have an influence on the sound, and bi-wiring can be shown to have a potentially positive effect in some cases.

As seen in the tables Fig. 5 and 6, the amount of inter-driver attenuation will vary depending on the amplifier source impedance, as well as the effective wire gauge used for the single and bi-wiring.  This explains why some folks have very positive results with bi-wiring, and some hear no changes at all, it is very much system dependent.

Further research will continue into this aspect of speaker system behavior, and I will post the results as they are obtained and verified.
 

All applicable copyright laws apply, all rights reserved, except transmission by USENET and like facilities granted.  Any use or inclusion in print or other media are specifically prohibited.  The informational content is not warrantied in any way or form, and any use of said content are at the reader's own risk, the author shall not be held responsible in any way for any damages or injuries arising from the content of this web site.  Common safety practices are encouraged at all times.  Do not fold, spindle or mutilate.