[If there’s a name that’s synonymous with wire and cables for electrical connections, it’s that of Belden. Belden is a multi-billion dollar corporation based in St. Louis, founded 115 years ago. The standard audiophile view of Belden is that of a sleeping giant with enormous capabilities but little interest in the cutting edge. Just as General Motors has channelled their immense resources to produce the cutting-edge Corvette ZR1, Belden has applied all their design and manufacturing capabilities to high-performance audio cables for the audiophile market. Engineer Galen Gareis is the lead on the Iconoclast line of high-performance audio cables, and is happy to explain the science behind cable design. This is the first of several articles from Galen, with an assist from Gautam Raja. I hope you’ll read, discuss, and share!–-Ed.]
If you’ve spent money on high-quality interconnects or speaker cable, you may have wondered why a cable is designed a certain way, and how many of the supposed benefits are based on valid science—and not the snake oil that cable designers are often accused of. This series of articles is based on papers written by Galen Gareis to present one engineer’s view of audio cable design, discussing in detail every consideration of designing high-performance speaker and interconnect cables.
When a cable carries a pure tone, perhaps a sine or square wave, then frequency and time are interchangeable, meaning that the only distortion of the signal would be attenuation. But music is far from a pure tone, and is a complex flood of frequencies in the 20 Hz-20 kHz range. When you send multiple frequencies down a cable, you introduce the possibility of time-based distortion, as different frequencies are affected differently by reactive variables such capacitance and inductance. Our ears are quick to hear the deterioration in fidelity based on frequency-arrival time and phase coherence.
To compound the issues, audio frequencies lie in an awkward electromagnetic region for conductors. Don’t forget that audio frequencies are at the bottom end of the spectrum; these are among the slowest, longest wavelengths of electromagnetic energy we harness.
Electromagnetic wave propagation: what exactly is the “signal”?
To understand why cable design has an effect on a signal in the first place, it’s important to understand exactly what this “signal” is, and how it “travels” along the cable.
Visualize the wire as a tube that’s the diameter of a set of marbles which you can push down the tube; the marbles are the electrons. Electrons don’t move without also causing electromagnetic fields, so now imagine a donut with its hole centered around this marble tube. This is the magnetic wave (B field). Now, take a bunch of toothpicks and stick them around the outside of the donut—this is the electric field (E field) produced by the moving electron.
To send a signal down the cable, we apply an electromotive force to the wire to move an electron down it— or in our example, push a marble into the “send end” of the tube. Something funny happens with the donut, though—the electromagnetic B and E fields. When the marble is just halfway into the send end, the donut and toothpicks are already halfway down the entire tube. When the marble is fully inserted into the tube, the donut and toothpicks are already at the end of the cable.
What is effectively happening is that when you insert a marble into the tube, a marble at the opposite end pops out as quickly. So the “signal” we use travels at the VP (velocity of propagation) of the cable, and not the speed of the electrons at all. Those move very slowly compared to the electromagnetic B and E fields.
This should make it clear why the structure and dielectric, or insulation, of a cable is so important. The signal travels more around the cable than in it, and largely through the dielectric.
Meet The Distortions!
Cable is far from perfect at moving electromagnetic waves in the audio band. The accumulating time-based distortions in a cable carrying an audio signal are clearly measurable. Better designs minimize those distortions, but every cable places more or less emphasis on each one depending on the design engineer’s concept of audible influences.
The Iconoclast development program has documented many possible sources of time-based distortion—and there are undoubtedly many more yet to be discovered. The combination of these is much more significant than any individual source of time-based distortion on its own.
1) Varying velocity of propagation
To go back to our marble and donut example, a higher frequency would be represented by inserting a marble faster in the tube, with a correspondingly fast-moving donut. So with a multi-frequency signal such as music, the higher frequencies entering the cable reach the other end earlier than the lower frequencies.
Also as we saw, the “signal” moves down the wire’s outer circumference, and not in the wire. Therefore, the velocity of propagation of the signal (versus the velocity of the actual electrons) is determined by the dielectric or insulation material that the electromagnetic wave is predominantly traveling through. The slowing effect of the dielectric varies with frequency, throwing another variable into velocity of propagation—but giving us a way to play with it.
In one tested cable, the speed of a 20,000 Hz signal is about 110-million m/sec. The speed of a 20 Hz signal is about 5-million m/sec, or about 22 times slower. In other words, the impedance of a cable rises as we lower frequency due to the VP dropping, and capacitance value. Each is determined by the dielectric and the design spacing.
It is possible then, to tune the design of a cable to flatten the VP through the audio band, reducing the time errors across the frequency range. It is possible to measure the VP and show that a cheap spool cable has a much more drastic change in VP as frequency drops, than does a cable designed for audio.
2) Current and voltage phase relationship
Current and voltage are locked into a phase-shifted relationship, always. To send a signal down a wire, you apply a potential difference (voltage) and only then does current rise to meet this demand. When you have a capacitor in a circuit, you apply current, and only as current starts to reduce does a potential difference occur in the circuit, as the capacitor is now charged.
Thus, inductance and capacitance are responsible for a ninety degree time-based shift between current and voltage in all electronics, not just cable. And cables have measurable inductance and capacitance, so these locked-in relationships lead to all sorts of time-based issues in circuits and cables.
3) Impedance matching to the load
For a cable to transmit a signal with no distortion, it should connect to a load with the same impedance as the cable. But impedance is hard to define for audio signals, where the load is the loudspeaker at the end of a speaker cable, or the input section of a preamp at the end of an interconnect. After all, velocity of propagation varies with frequency, so impedance is not a constant. Also, speakers present a different impedance depending on frequency.
Another factor is that the wavelengths of audio signals are very long, even at 20 kHz. This means you can’t have a cable that’s long enough to propagate these signals, and this, plus an impedance mismatch, results in reflections off the load that cause distortion. Since the signal’s wavelength is many times too long to “fit” inside a given length of cable, the cable doesn’t yet have impedance. There can be reflections, but these are not to be confused with “return loss” reflections at RF.
[We will continue Time is of the Essence, with Galen Gareis’ explanation of time-based distortions found in audio cables, in the next issue of Copper.—Ed.]