There have been so many debates on newsgroups and in magazines about how different a wire can sound as opposed to another. Here we cover the psychological effects. I think you will find a recurring theme throughout these pages that seems to tie them all together. But I will leave this up to the discretion of the reader and see if their conclusion is similar to mine.
Wire. How many different kinds are there? How many different kinds can there be? Dozens, at least. I am being facetious here but for our purposes, we will look at three kinds of wire, coaxial, braided and solid core.
Coaxial wire is wire of a certain gauge that has a braid of wire or foil surrouding a center conductor, insulated from the braid by plastic. The outer braid is in turn usually covered with a plastic covering. The use of coaxial wire is for low level signals. The shielding is the ground conductor and the center is the main signal lead. The ground shield conductor acts as a protector of the very small (From microvolts up to tens of millivolts) signal from stray EM, which can induce unwanted noise.
One drawback of coaxial wire is that the capacitance of it can be quite high, from a few tens to hundreds of picofarads per foot. In tube amplifiers this can reduce the high end noticably. Not major significantly, but noticeably. This is because the inputs to tube amps, if not purposely designed for low input impedance, will have a high input impedance, from 100K to 2.2 meg. A 100 picofarad capacitor across a 470K (typical input resistor) resistor will start to roll off the signal at about 3368 hertz! Well within the audible range, in particular within the range of highest human hearing sensitivity (about 1000-4000 hertz). This is using the 470K value to obtain the -3dB amplitude point AKA the pole.
So, while the catch twenty-two here is that higher impedance yields higher sensitivity to stray noise, it also adversely affects the bandwidth using coaxial wire.
So use coaxial wire for applications where impedances are beow 100K.
There are many ideas about using solid wire versus stranded wire. We must consider the physics of it and the experimental results.
The physics of solid core wire for DC is the thicker the better because DC flows through the entire wire, therefore encountering less resistance..
AC on the other hand, for reasons yet unknown to me, tends to ride along the outside of the wire. That is, there is no current flow in the center of the wire. A couple of resources state that electrons being like charges, tend to repel each other from the center of the wire outward. But my question about this is that the same thing can be said of DC current flow. The negative charges there will repel also. So AC does something more to the electrons. (This is also true of a phenomenon in transistors where the electrons accumulate on the outside of the silicon leaving the center virtually devoid of electrons. This is known as current crowding. More on this soon in the solid state electronics pages.)
However, the fact of alternate direction flow tends to cause a stopping of current at a certain point in time. So is this enough time for charges to push each other out of the way? The more the flow has to stop and flow in the opposite direction, the more opportunities the electrons have to push each other out of the way from the center to the edges. Could this be the cause of the phenomena known as skin effect? Perhaps, because the effect increases in proportion to frequency. The higher the frequency, the more the skin effect. Or the less wire the electricity has to flow through.
What happens therefore is that the wire will begin to have more resistance (as opposed to impedance or reactance) to the flow of current because now the current has less wire to travel through. What are the options? Use thicker solid core wire, or stranded wire.
Stranded wire, by the laws of physics that seems to be forgotten by some engineers that tend to argue that there is no difference in audio applications, has more surface area than does solid core wire. In spite of the fact that the wire is bare and attached together, the areas that are not connected still provide more area for electrons at higher AC frequencies to flow through for a given gauge. If one has noticed, when stripping stranded wire, the wires lay there parallel to each other not twisted together. If one twists them together, then I can see there being less area. But it will still be more than a solid wire of the same gauge.
Besides looking at it from an electrons point of view, the distance between each strand may be large enough to allow for the indivuality of each strand as regard skin effect. I personally haven't done testing in this regard nor have I read anything about it, except for the diffrence I hear in my system when I use either. So this is my opinion.
What is the advantage? Higher frequencies get through better. More detail in the sound quality.
WHAT ABOUT THICKER GAUGE?
I have read quite a few arguments for and against using thicker gauge wire. Some claim that one gets more resistance with thicker wire. Hogwash. Some claim that one loses cartain "Maxwell effect" benefits using thicker wire. I haven't read the article yet, but my guess is that the in the thicker wire the lower frequencies cause the thicker wire to actually act as an inductor, actually hindering bass response.
I started to believe this at first. When I first modified my amplifier, I used a thicker (20 gauge) stranded wire rather than the 24 gauge solid wire that was in the amplifier at the time. I noticed cleaner high end, more midrange detail, but lower bass.
However, now I suspect that what happened is now the higher frequencies, which were once hindered by skin effect, are now coming through with less resistance. So the psychoacoustic effect is that now I get less bass, where I actually get more mids and highs.
On the other hand, there was at least one gentleman on another site, including some tests done by AUDIO magazine comparing frequency responses of 18 gauge, 16 gauge and monster cable for amplifier to speaker connections. They did find some differences. The gentleman on the web site found that there was more inductive reactance in the thinner wire.
The conclusion to both was that 16 gauge stranded wire was close enough to monster cable (14 to 10 gauge very thin stranded wire. Wire gets thicker [cross sectional area] with smaller number by a factor of about 1.5 for each two points. In other words, 14 gauge is 1.5 times as thick as 16 gauge) that it wasn't worth the expense. 16 gauge zip cord was good enough for most (read up to about 150 watts) applications.
Well, I would like to say that monster cable is worth it. But unless you have ultra-supersensitive hearing and a 500 watt per channel amp, I do not think that I could justify spending 2 to ten dollars a foot for wire, when I could not even hear a difference between 16 gauge and it. For a given power rating monster cable, generally a fine strand 12 or 14 gauge wire, will not have any different effect from a 16 gauge stranded wire. With many audiophiles today purchasing or building their own systems that put out from 2 to 100 watts of power, this wire would be overkill.
A decent 16 gauge stranded wire that costs about 25 cents a foot would be more practical and sufficient for such low power levels, especially since most of us run our speakers within six feet of our amplifiers. Now this is my opinion, but it is based on fact and personal experience.
However, when using wire as hookup, I seldom use anything less than 20 gauge for low level inputs. For hookup between the output of the preamp and input of the power amp section of an integrated amp, or the output jacks of a preamp, I use 18 gauge. I use 16 gauge from the output of the power amp to the speaker connections.
For power supply wiring, I use 16 gauge. I also make it a point to use the same length of wire from the B+ to each channel. I believe that it may be true that electrons actually flow much slower than it was once believed. Charge may move faster, but the actual current is slower. I say this because on one amplifier that I modified (solid state), the channels used to turn on individually and delayed (they had surge protection circuitry), I wired 16 gauge wire from the B+ and B- with equal length from each polarity and to each channel. The distance was only about 2 inches. When I did this the channels both turned on at the same time and instantly. I also (having bypassed the "protection" relay) did not get the characteristic turn on thump anymore.
So, how you wire your system is your choice, but I offer these
suggestions because they are the ones I have learned give the best results. As usual, I
also suggest that you experiment for yourselves and have fun!
Back to Papers Page
Copyright 1999, Gabe Velez