The points being brought up about speaker cable impedance altering crossover points does not really come into play in the real world. If one were using cables with drastically different electrical characteristics for each individual driver / frequency range, i could see this becoming a factor. Otherwise, all of the drivers / crossover points are subjected to the same equivalent load introduced by the speaker cable and will act accordingly.
If one were to change individual runs of speaker cable BETWEEN the passive crossover components within the speaker and the individual drivers, one could measurably alter the hinge points and cause a lack of "synchronicity" within the speaker itself. The end result would be drivers that no longer crossed and blended at the factory specified rates, resulting in a lack of continuity from driver to driver with increased irregularities in frequency response. This could also drastically effect both vertical and horizontal radiation patterns due to cancellation between the drivers, etc...
When changing speaker cable, what does change is the total Z ( impedance ) that the amp sees. Some amps are very sensitive to terminating impedances, various levels of reactance, etc... This means EVERYTHING from the binding posts back. In some cases, using 6 ft of Brand X wire and then switching to 18 ft of identical wire MIGHT make both a measurable and audible difference. After all, any electrical characteristics that the cable might introduce into the system would have been tripled in this specific case. Obviously, this would be most likely to occur on amps that were not extremely load stable AND if using cables that were highly reactive ( either very inductive cables like zip cord or highly capacitive cables like Goertz, Polk, various braided designs, etc... ). The more stable the amp and less reactive the cables, the less likely you are to notice anything other than increased series resistance. Since most speaker cables are typically of relatively heavy gauge to begin with, it is not a big deal.
As far as Ernie's comments go about running long balanced lines, most of it is true, but maybe not for the reasons that he's thinking. Running balanced ( aka "XLR" ) type cables does have some very specific benefits. Before going over those, let's look at how a balanced design works.
In a "normal" ( rca or "single-ended" ) cable, one has a "hot" or positive wire and a ground wire. In a balanced line, you have one positive wire, one negative wire and a wire that is ground. By using this approach, you can have a voltage ( or level of "gain" ) that is twice that of a single ended circuit. This is because we can swing voltage both above ( positive ) and below ( negative ) zero volts ( ground ).
In a balanced installation, a signal is fed onto the positive wire and the inverse of that is fed into the negative at the source end of the chain. At the load end, a circuit is used to compare the two signals. If there are any variances that cause the positive signal to diverge from what the negative shows ( or vice-versa ), the "distorted" or "non-linear" part of that signal is not reproduced. These "distortions" can show up as being differences in signal amplitude between positive and negative, differences in phase, etc... The bottom line is that, if the signals don't jive EXACTLY ( or as close as is humanly possible to design into a circuit ), that part of the signal is not reproduced.
Obviously, this gives us two noticeable advantages over a single ended design. First of all, we've increased the drive levels from the source to the load. Just as i stated with amps & speaker cables vs components & interconnects, the more signal that you have, the more you can afford to lose without noticeable degradation. As such, they can afford to "lose" a little bit of signal using phenomenally long runs ( with the resultant higher series resistance ) as they've drastically increased the amount of signal that they had to work with.
Besides the above, most professional installations make use of compressors, expanders, effects units, equalizers, noise gates, mixing boards, etc... Since most every one of these mentioned has the potential to add gain to the signal if adjusted in that manner, it is possible to take a signal of microscopic amplitude and boost it enough to blow your eardrums out. Signal loss is NOT a big deal unless one is worried about the "ultimate in audio reproduction". As most of us know and have heard, the majority of the professional recording and sound reinforcement field do not consider this a priority. Keeping the noise out of the system and capturing "most" of the musical event is good enough.
The second advantage is that this approach is obviously a lot less susceptable to outside interference. Since most RFI / EMI is introduced via some type of varying magnetic field, the interference does not ( typically ) show up in the same amplitude and phase on both the positive and negative runs. As such, it is simply cancelled and you are left with a lower noise floor and extremely minimal outside interference.
If you noticed, i said that the second advantage was from OUTSIDE interference. The cables themselves can introduce errors into the system. If the cables make use of lossy materials with non linear levels of dielectric absorption or have specific levels of reactance that cause frequency response abberations, the sonic traits of this cable WILL be passed down the chain. This happens due to the fact that the use of identical materials used on both the positive and negative runs of wire would present a "match" that was still balanced between the two, even though they are distorting the signal. As such, a longer set of cables will still introduce more of their own sonic characteristics into a system than an identical yet shorter cable would. As such, one still needs to keep cables as short as possible and purchase cables that make use of high grade conductors and insulating materials IF one is looking for the most "transparent" system possible.
In terms of potential problems with the amplifier / speaker interface due to driver control and damping, this should not be a problem in the least. That is, so long as one selects cables that are MORE than capable of handling the power levels being generated within the system and are of low enough series ( "loop" ) resistance so as not to impede the flow of current in any way, shape or form. As such, i personally prefer an "overkill" approach when it comes to speaker wire gauge.
Selecting an amp that is both well designed and suitable to drive the specific load that they will see is also desirable and highly recommended. Otherwise, one might have to result to "band-aids" such as shorter speaker cables to reduce series resistance for "under-muscled" amps. In some cases, going to very small gauge cables to introduce series resistance / minimize communications between the amp / speaker interface also works. If the latter sounds crazy, believe me, it is true. Such may be the case when one has VERY reactive loads and / or amps that have a hard time with that specific type of load.
Some cable manufacturers take all of these factors into account and even provide charts as to what speaker cables they recommend for specific power levels. The smart ones also consider that the length of the run ( series resistance ) can come into play in such a situation and recommend heavier gauge cables for longer runs. This is not to say that you will personally like the products that these manufacturers produce, but to say that they are at least applying common sense, science and engineering to their designs. Companies that can provide no reasoning as to why or how their products work the way that they do are probably selling snake oil at exhorbitant prices. Sean
>
