[We began this series of articles by Belden engineer Galen Gareis in Copper #48 , and continued it in Copper #49. We now present the conclusion of Time is of the Essence, on time-based distortions in audio cable design,—Ed.]
7) Dielectric effects
Dielectrics have a disproportionate impact upon weak electromagnetic signals. Four-fifths or more of the current magnitude at audio frequencies is below 3,000 Hz, and this clumping is often called “spectral density”. This energy does not stop in plastic or air. It emanates out in an inverse-log power decay through all the materials it encounters along the way.
Electromagnetic fields are most influenced by dielectrics nearest the wire. Using too many small wires splits up the current, and starts to allow the dielectric to influence the sound more and more, negating the advantage of dielectric uniformity. The electromagnetic field is strongest nearest the wire, decreasing with the square of the distance moving out away from the wire. With smaller wire, the electromagnetic signal moves from being in the dielectrics to being around it, as well.
So weaker signals are affected more by the dielectric in general, as they decay the most as they move away from the wire (a great proportion of the signal is in the dielectric). Also, weaker signals are affected most by the dielectric closest to the wire, especially as we go up in frequency. This can be measured by how easily velocity of propagation is nearest to one (100%) for a given cable size. The better the dielectric, the closer the VP will be to one, and at the smallest size.
If we use two dielectrics in layers, say air and plastic, we also see the signal slow as it moves from air to plastic, and speed up as it moves from plastic to air. Some cables (including the Belden Iconoclast) put air nearest the wire, where the signal strength is highest. This reduces the outer plastic dielectric’s contribution to the group velocity, and translates to a lower capacitance number when we use the cable in audio applications. It also significantly reduces the cable’s SIZE.
8) DCR influences
This is mostly a speaker cable issue, as the amplifier’s negative feedback loop works best with ZERO cable resistance. To control the drivers— the woofers being the worst back-EMF devices the amplifier sees— we need as much CURRENT as possible to be delivered to offset their unwanted motion. The higher the cable resistance, the less current the amplifier can generate to manage the back EMF from the speaker. We need to manage DCR with many smaller wires, yet still arrive at low inductance and capacitance, too.
A second effect of too-high DCR is that it can vary the speaker’s frequency response. Voltage at a given frequency dropped across the cable is signal that doesn’t get to the speaker. This is frequency dependent in complex speaker loads.
On interconnect cables, the small to near-zero currents allow small wires to be used to improve current uniformity and lessen proximity. Here we want flat Rs electricals measured out above the audio band.
I will go out on a limb and say stranded wire doesn’t seem to “measure up” to its bad reputation in my studies. I see no need to use them as the cost goes up, and the advantages go down as I already design with small AWG wires, negating the need for the flexibility stranding provides. Bad sound with stranded wire has always been due to the dielectric design, and not the stranding.
9) Cable symmetry
How do you make a complex cable’s cross section look like one simple wire electrically, and have every wire sound the same?
Matching multiple wires into a complex structure isn’t easy to do well. The ideal cable has one wire that has exactly the same structure and length as the opposite-polarity wire. However, if we run two large conductors in parallel, the inductance, capacitance, proximity effects and skin effects between them will be less than ideal. Running several smaller conductors in a single polarity and weaving them around each other to increase the average distance between the center of any one conductor and another in an opposite polarity will reduce the inductance. A carefully designed weave pattern will help cancel out magnetic fields, further lowering inductance. Does it work? A single bonded pair made for Iconoclast has a 0.196 uH/foot inductance. One we complete the cable assembly inductance drops to 0.08 uH/foot.
The use of BONDED pairs in each polarity helps cancel the magnetic fields. No magnetic field means zero inductance! The two flat-polarity halves pulled tightly together with a textile wrap keeps loop are to a minimum, further keeping inductance in check. The proximity effect is lowered by using many, many (fourth eight!) small wires. Alternating current near one another lessens conductor efficiency, especially with larger wire. High current in speaker cables mean that proximity effect is a larger problem than in interconnects. The weave pattern keeps wires less-parallel to one another. The final results are plenty measurable, and are best in class in the designs evaluated.
However, more small wires with a dielectric between them can create a nice big capacitor which negatively affects the signal. Inductance and capacitance have an inverse relationship, so to get both intrinsically low, you can’t go “whole hog” on the opposite variable. This means a compromise, and yes, audio is a whole set of compromises. The end view explains why this is so in the speaker cable. The bonded wires “trade places” between the minima distance and the maxima distances. The weave insures wires end up at the same electrical length. This keeps every wire electrically the same, and lowers bulk capacitance while the bonded pairs lower inductance. The tight spacing and magnetic cancellation in “star quad” cross-over points are unique their ability to lower capacitance and inductance, both, and still address multiple cable issues even handedly and at the expense of nothing. Well, except being easy to make. We’re not an easy crowd, though, are we?
Wire direction showed no impact on the sound. But, the manufacturing process insures ALL the wires are “directional”, so flip the cable around at will if you like, it’s free. I fully refrain from setting critical design goals on immeasurable attributes. As proof to this, we sell the exact same designs in three copper grades, separated by the copper’s costs. The copper quality factor has zero influence on the design, and all measurements are identical. Are there differences in sound? Yes, there are, but the influencing factors do not derail the design between the copper draw sciences.
10) Attenuation at audio frequencies
The common assumption is that better cables are better simply because they cause less linear attenuation. However, when cables are designed for time-based issues, they clearly sound better. The effects of more- or less-linear attenuation are far harder to hear. With a non-optimized cable, attenuation is not linear, and resistance increases from 5,000 Hz onwards in the standard 1313A 10 AWG zip cord style cable, and is essentially a wall of totally lost energy above 20 KHz.
Taking the audio band from 100 Hz to 20 KHz, we see a 6.75% change in overall impedance with Iconoclast, versus a 71.8% change for the 1313A cable. But, this is NOT just attenuation per say, but skin effect and proximity effect variables that effect higher frequencies too.
The important thing to remember with all audio cables is that arrival times are more important than raw speed down the wire. We call it “sound quality” when we use the cable, but it really is the time-alignment of all the signals. The human brain hears superimposed time alignment and amplitude preservation first, everything else a distant second. Efforts to manage phase (lower inductance) and VP are important. Attributes (smaller wires, lower L and C) that are time-dependent seem to provide the most benefit.
By touching upon some of the considerations of a cable designer, I hope to convey how complex this issue is, and that a cable is not “just a cable”. I’ve often said, “Rest assured, if there is snake oil in these products, it sure looks like physics to me. All the data is measured and real.”
[We will be featuring more articles on cable design by Galen Gareis, with an assist from Gautam Raja, in the near future.—Ed.]