Can cables matter?

I set out looking for whether it was physically possible for a cable to make a difference to sound, if so, how it could be proven, and if not, what evidence to the contrary exists.   Mind you, this has been debated for as long as audio has been around and I am not likely to be the authority that puts an end to it.

My uncle worked as a physicist at ORNL for many years and so I had access to ask him and a few of his colleagues some questions.  I figured who better to ask about the physics of electrons moving on a wire than a group of internationally recognized experts in the field who happen to work for the Department of Energy.   In asking questions, the first thing I quickly learned is that deciphering their answers to my questions was going to take some effort on my part.  Think of asking the Secretary of the Treasury what a dollar is worth, and you begin to get the picture.  I learned a lot more than just the answer to my initial questions.

For instance, I asked about the speed of electricity on a wire and very quickly learned that at least 3 different answers are possible depending on how you look at it.  We have electron velocity, drift velocity, and signal velocity.

Electron velocity is the speed of movement of individual free electrons within the wire and is the only one of the three that can be measured without a current on the wire.  Free electrons move in a straight line until they collide with something, then move off again until the next collision.  Movement is random and thus no measureable current is produced, and what energy is measureable is is measured as heat. Electron velocity is measured in nano-meters per second.   This doesn’t help us as it has no direction and thus cannot carry our signal from one end of the wire to the other.

Drift Velocity is the movement of an electron away from an applied force.  When a battery is attached to a wire, an external electromagnetic field is applied to the wire and free electrons react by moving away from the source (since like charges repel).   Electrons still bounce around and exhibit the same style of collsion to collision motion, but now instead of  entirely random movement, we see a generalized direction of movement.   If you can imagine trying to measure the progress of a bumper car from one end of the arena to the other, it is not an easy task, and not particularly fast or efficient so speed of progress is fairly slow.  Although this can transmit a signal, it would be very slow (measured in meters per second) and thus is not useful for real-time transmission of signals.

Our third factor is signal velocity.   Instead of being the movement of electrons, this is the movement of the electromagnetic fields around electrons.  Electrons repel each other, but not by colliding with each other, instead, as the two fields around electrons begin to overlap,  the one furthest from the battery will move away.    The field around an electron can be 50-100 times the size of the electron itself, making movement of the field much more rapid than drift itself.   Much as a skunk can clear an area without ever actually seeing it, electrons will move away from each other at the first hint that another is approaching.  Fluctuations in the magnetic field are directly proportional to the applied force so changes are measurable and repeatable.  Now we are talking, this is fast enough to be useful, directional, and directly influenced by changes to applied force so we can control it.

 

So, perhaps unsurprisingly, what we are interested in is signal velocity, and while it has been argued that it is a purely electromagnetic wave effect, that isn’t the case.   Were signal velocity a pure wave effect, we would expect it to travel at the speed of light (in a vacuum), and the argument that cable material makes any difference would be moot since wave behavior would be unaffected by materials changes.  In reality, measured signal velocities are considerably below the speed of light, so some other factor must be involved.    Signal velocity involves both the electromagnetic wave fluctuations and movement of the electrons themselves (drift velocity).

When asked for hard numbers, I was told “Somewhat near the speed of light” and also told that it was not possible to calculate the speed of a wire using simple equations and that measurements to find the exact speed were very difficult as few instruments were capable of measuring differences that small at velocities that high.    When asked to define “somewhat” I was told somewhere between half the speed of light, and 9/10ths of the speed of light.   While this seems like quite a difference,  the reality is even half the speed of light is so blazingly fast as to be nearly impossible to measure accurately.

When asked if they believed it was possible for cable materials to alter the signal being carried, the answer was “yes it could”.  Two reasons were given, the first was that differing materials would impact the speed of transmission, and the second was that different materials could cause the transients to behave differently.  As an example of this second phenomena, they pointed to the fact that when flipping a switch to double the current applied to a system, changes measured immediately ahead of the switch were very abrupt while changes measured miles down the wire show a much more gradual rise.  I asked if that rise was always the same or if different materials showed measurable differences and while the response was that all expected it would cause differences, none had ever done an experiment expressly to test that hypothesis.

I asked if there was a difference in transmission speed between copper and silver (assuming equal diameter and 100% purity of both) and was told that because of the differences in the two metals that “differences in signal speed could be expected”.

When asked how much distance would have to be involved to measure a difference in signal transmission speed between copper and silver wire, the general consensus was that somewhere between 10 and 50 Kilometers could provide measurable performance differences with the most sensitive equipment available and that longer distances would make it easier to accurately measure.     One of the things that was repeatedly mentioned was the need for the best possible instrumentation because of the need to see the tiniest leading edge of the wave.  (Due to the differences in rise time previously mentioned).

When asked point blank, do you think these differences could make an audible difference when differing materials were used between the source and transducers, the answer was “could it, yes, is it likely, no”.  While they were not willing to rule out the possibility, both suggested that the differences in question in a system no larger than a few meters would be measured in femto-seconds and that any difference was probably well below the threshold of other components used in the system.

 

I came away with several thoughts but not a lot of answers.

Materials do make a difference at least at a technical level,  thus, in theory, it is possible that a cable can cause audible differences.

The fact that the observable differences are very difficult to measure in wires that are kilometers in length using the most precise instruments available, make me skeptical that those same differences are large enough to play a key role in audio over distances measured in centimeters.

Measured differences in cable transmission speed are below the currently accepted limits of human hearing.  femtosecond differences shouldn’t be recognizable if our hearing is incapable of distinguishing tones shorter than a 100 milliseconds.

The human brain is an amazing system that has proven time and again to outclass some of the worlds greatest instruments, so it is possible that our understanding of the limits of human hearing is flawed or incomplete.   (How probable it is that our understanding is flawed to a relevant degree, I leave to my readers discretion).

I heard repeatedly about the limits of instrumentation and the difficulty with measuring phenomena this fast and differences this small.

Testing the materials in question would require a very substantial investment as the instrumentation is largely custom built and designed to perform a single measurement.   In order to both measure speed and rise time, you would need 2 different instruments either of which costs hundreds of thousands of dollars to produce.

The ROI simply is not there to do the testing.  Even if it were possible to prove a difference (regardless how small), it would not be economically prudent to do so.