In the last episode of this ongoing saga, I (to the undoubted delight of the troll brigade) compared copper to cookies, and said that it could either be an alloy (like chocolate cookies) or a conglomerate (like chocolate chip cookies), or both. I said that most copper alloys were done "on purpose," and mentioned a couple of them, and said that some of the natural alloying elements -- silver and gold, for example -- have such little negative effect that IACS (the International Annealed Copper Standard) doesn't even bother to list them separately, but just considers them as part of the copper content of IACS-rated coppers.
What I didn't say was that the "on purpose" alloys, like brass, bronze, phosphor-bronze and even Tellurium copper, are typically of lower or even much lower conductivity than "unalloyed" copper, and are always used for some purpose other than just their ability to pass signal. You'll see what I mean if you'll remember that the conductivity of even the "worst" copper, ETP, is 100-101% IACS and then consider that the IACS conductivity of the best cartridge brass is 28%, silicon bronze is 7%, phosphor bronze is 15%, and Tellurium copper, the very best of the bunch, is still only 90%.
The copper impurities that are more likely to cause problems (other than simply acting like a volume control, to resist overall signal flow), are, I said, the non-alloying ones like copper oxide and the "iron triad" elements (Iron [Fe], Nickel [Ni], and Cobalt [Co]), all of which can have a deleterious effect on signal transmission completely independent of their effect on wire conductivity. These, I said (although CuO will also be found on the surface of any piece of copper that has ever, even if only briefly, been exposed to the atmosphere) tend to be like the chips in a chocolate chip cookie, and to appear as distinct inclusions at the junctures of the copper crystals of which the wires in our cables are made-up.
The reasons that I gave for the copper oxide and the iron triad metals being problematic were, respectively, that CuO in very thin films (such as between the crystals of a copper wire) is a semiconductor, which makes every one of the tiny inter-crystalline inclusions of it into a mini-diode to filter out low-level signal information. The iron triad metal inclusions do much the same sort of thing, only by becoming magnetized and thereby affecting the actual electron flow of the signal.
All of these inclusions are the "chocolate chips" in the conglomerate that is the copper, and if they are potential sources of distortion, signal loss or other problems, then doesn't it stand to reason that copper with fewer of them would be better to use for making cables out of? I think -- and this is just my personal opinion -- that we've all been laboring under a misapprehension when we've accepted purer copper as being better just because of its greater purity. The way our thinking seems to work reminds me of the old joke about the perfect contraceptive being an aspirin tablet -- held firmly between the knees. Certainly it would work, but what would do the trick wouldn't be the aspirin, but the knees held tightly together, and in this case I think that it's the purity that's the "aspirin."
We've already seen (in the last part of this article) that the thing that makes better cables better isn't likely to be the higher conductivity of purer wire. The conductivity differences simply aren't that great, and can too easily be compensated for. So if the thing that makes the difference isn't the greater purity and there is a difference, isn't it possible that what's making it is the fewer impurities? In short, fewer "chocolate chips" in the cookies?
One way to lessen the number is to use "continuous cast" copper. If this "melt-produced" copper is allowed to cool slowly (as it normally is), longer crystals will form, naturally reducing the number of crystals per length and correspondingly reducing the number of crystal junctures for "chips" to form at. While I can't personally vouch for it, and have no idea at all of how they could possibly trace a single crystal through its full length, one of the manufacturers of OCC ("Ohno Continuous Casting") copper claims single crystal lengths of as much as 700 feet (213.36 meters)!
Obviously, how many crystals wind up in your set of interconnects or speaker cables will depend on not just the cables' length, but how many strands are used, and (thickness-wise) how many crystals go to make-up each single strand. Even so, no matter what the actual number is, the crystal count will certainly be less if the crystals are longer, and fewer crystals will certainly make for fewer crystal junctures. Voila! The "chip" count goes down and the sound quality, if I'm correct, goes correspondingly up -- even as compared to other copper of the same analyzed level of purity. (The point here is that it's the number and not the size of the "chips" that makes the difference.)
Right now, I can imagine you saying, "That's all well and good, but what if I already have all the cables I need, and don't know what kind of copper they were made out of?" Rejoice, there still might be something that you can do, especially if you are using inexpensive or "free" (the kind that some manufacturers still include free with their products) cables, that will either work to make your cables sound better or, if done properly, at least do no harm. All you have to do is to pre-heat your home oven to no more than 160 degrees Fahrenheit and bake your cables for three to four hours.
No this is NOT carrying the chocolate chip cookie thing to the point of ridiculous. The fact of it is that most less-expensive cables are made with "hard annealed" ETP copper, which is very cheap and is "hard" because it has very small, closely-packed crystals. Copper is, however, a very peculiar metal, with a very active molecular structure -- meaning that its molecules, even without high temperatures to energize them, move around a lot. It's so active that, for example, if it has a microscopic hairline crack, its molecules will "leap" across the crack to join with the other side, and the crack will eventually heal itself and go away entirely. It's also so active that, just left alone, its molecular motion will be sufficient to eventually bring it to a much softer state of anneal, with much longer individual crystals and correspondingly fewer crystal junctures per length.
What the baking process does is, by the application of heat, to speed up that process and accomplish the equivalent of three or four years of self-annealing in an equivalent number of hours. Be careful with the temperature, though. The reason for the 160-degree temperature maximum is that many of the insulating materials used on lower-priced cables have either a softening or an outright melt point at 180 degrees Fahrenheit, and you don't want to get your cables too close to it if you want them to survive their baking experience.
Incidentally, I mentioned, last time, that there was one weird thing about the metals for cables that has even me baffled. That's still true, but I'm out of space for now. Maybe I'll tell you about it next time.