Cable gauge and ohm load

Taking the last first, yes you can use a smaller cable for the high end if you biwire; whether you should is the subject of much debate. As a matter of fact there is little agreement about most of the questions you pose. My own view is that thin single strand wire does not do well in the bass. I think that this would be magnified by low efficiency speakers. The impedance of the SPEAKERS is an important factor in how much power is going to be transferred over the cable. The efficiency of my Spendor S 100s and my GamuT L5s are not wildly different but when I run the 100s with my CJ 350 the heat sinks get mildly warm while with the L5s they get quite hot. The Spendor is about a 6 ohm load while the GamuT gross below three ohms. This is not as relevant if you use tubes. I am sure Al could give you a good technical discussion of the factors involved; I won't try but confine myself to generalizations. First, heavy usually doesn't hurt; you will probably loose less power with bigger cable. Two, some thin cables work quite well, I have used OCOS and hear that Audience 24 is also good. It really boils down to your own system and the particular cable you have in mind.

There is a simple chart at www.roger-russell.com that provides table as a guide for 2 conductor copper wire. Here's the worst case information:

For a 2 ohm load:

22 AWG max length is 3 feet
20 AWG max length is 5 feet
18 AWG max length is 8 feet
16 AWG max length is 12 feet
14 AWG max length is 20 feet
12 AWG max length is 30 feet
10 AWG max length is 50 feet

The difference between silver and copper isn't a significant difference in my opinion. I would also think that this information could be used as a guide for multiple conductor cables using the aggregate wire size.

I second the comments by Stan and Elizabeth, and also the gauge tabulation that Mceljo quoted.

Anti-Cables, btw, are not narrow gauge. They appear that way because the insulation is thin. The technical info page at their site indicates that they are the equivalent of 12 gauge, which is consistent with their specified resistance of 0.00318 ohms/foot run ("foot run" would mean that the specified resistance is the total resistance of both conductors, or twice the resistance of a single conductor of that length).

As you stated, capacitance is not ordinarily a concern in a speaker cable, unless the particular cable has extremely high capacitance per unit length, and is particularly long, and the amplifier is particularly sensitive to load capacitance. Inductance, though, can be a concern, especially if the impedance of the speaker drops to low values at high frequencies (which is the case for many electrostatic speakers).

Regarding your first two questions, about length and gauge, the criteria that should be satisfied are as follows IMO:

(a)The total resistance of both conductors should be a small fraction of the impedance of the speaker at all frequencies.

(b)The inductance of the particular length of the particular cable should be such that the corresponding inductive reactance at 20kHz is a small fraction of the impedance of the speaker at 20kHz. Inductive reactance, which is measured in ohms, equals 6.28 x F x L, where F is frequency in Hertz, and L in inductance in Henries.

The obvious question, then, is how small a fraction is "small." In the case of resistance, that will depend on how critical damping and woofer control are for the particular speaker, and what the damping factor of the amplifier is. But in general my opinion would be that something in the area of 1% or 2% of the minimum speaker impedance would be comfortable, or somewhat more if damping/woofer control is not critical for the particular speaker.

Comparing the numbers shown in the tabulation Mceljo quoted with the data in this wire gauge table shows that what those guidelines are trying to accomplish is to keep the total resistance of both conductors in the cable at 0.1 ohms. 0.1 ohms is 5% of the the 2 ohm load the table is based on.

For inductance a considerably looser tolerance is acceptable, because its effects are only significant at very high frequencies, and also because of some complexities involving phase angles that I won't go into. IMO a comfortable value for inductive reactance would be in the vicinity of 20% or so of the impedance of the speaker at 20kHz. If cable length is short, say 8 feet or less, most cables will meet that criterion with most dynamic speakers.

I have no particular knowledge when it comes to metallurgy, so I won't comment on that question except to say that to the extent that silver may sound different than copper, the reason is most likely NOT that it is a better conductor. The higher conductivity/lower resistance of silver can be easily duplicated in copper by just going to a slightly heavier gauge.

Regards,
-- Al