Speaker Cable Face Off 1

Speaker cables have become one of the least quantified and objectively analyzed audio products in the audio marketplace. Many cable vendors and forum cult hobbyists insist on the abandonment of measurements and engineering truths in favor of subjective listening tests in hopes that the listener will think they hear a difference, even if none exists, to justify their unproven claims and sometimes extravagant prices. Let's refer to these people as "cable soothsayers" since they seem pre-dispositioned against applicable engineering truths and proven sciences despite reliance at times on the names of these truths (IE. Skin effect, diode rectification, etc) as a mechanism for the alleged problem(s) and their corresponding solution(s). The belief set forth by these "cable soothsayers" that "you can't measure it" leaves open the avenue of convincing prospective buyers of the differences in product by pseudo-science, clever marketing ploys, and sometimes just blind faith. Audioholics.com on the other hand is not a faith based website, nor are we cable "cable soothsayers". We believe and cherish objective measurements and feel they are the first step in developing and/or analyzing a high fidelity product. If a product measures well, and the measurements are applicable to its corresponding fidelity, it will usually sound good, however the converse is usually not true unless one seeks inaccurate audio reproduction of music. With that in mind we have analyzed six different brands/models of speaker cables ranging from as little as $0.46 / ft to $55 / ft. Don't let price fool you however, as in many cases, cables especially, price doesn't always dictate quality.

Cable Descriptions

The Original Monster and Monster Navajo, and Sound King are typical 12AWG stranded Zip Cords, while the Cobalt Cable is 10 AWG stranded. The Stealth Fine Ribbon is a flat ribbon silver cable with unpaired conductors individually encased in a quasi air dielectric with equivalent of 17-18 AWG (Note the manufacturer claims an equivalent of 14.5 AWG per cable pair, but we measured between 17-18 AWG as can be seen in the Rdc and Rs measurements in Table 1. I could only assume the manufacturer estimated their gauge spec based on a bi-wired configuration. Since their website doesn't explicitly state this, it is open for interpretation. The Stealth Premier II is basically a Litz construction with individually paralleled insulated strands for each unpaired conductor to yield an equivalence of about 7 AWG (Note: The manufacturer claims 4 AWG per pair which is about what we measured considering the inaccuracy of measuring small cable resistance of cables so low in DCR and associated length). According to the manufacturer, the Stealth Premier II has recently been revamped with more conductors per cable resulting in a slight lowering in gauge to 6AWG per conductor, or 3AWG per pair.

Cable Measurements Comparison

Tabulated below are the cable metrics of six different speaker cables.

Table 1: Measured RLC Parameters of Various Speaker Cables

Cable Metric Definitions

• Rdc - Commonly referred to DCR which is the series resistance of a cable at zero frequency.
• Rac - The resistive portion of the cables series resistance as a function of frequency due to skin effect.
• Rs - Total Series Resistance (mohms) measured tip to tip at one end of the cable while the other end is shorted. Note: Rs = Rac + Rdc (minus instrumentation inaccuracies identified below)
• Ls - Series Inductance (uH) measured tip to tip at one end of the cable while the other end is shorted.
• Cp - Parallel Capacitance (pF) measured tip to tip at one end of the cable while the other end is open circuited.

Cable Measurement Test Set-Up Notes
All of the above measurements were completed on a fully calibrated and certified, WAYNE KERR Model# 3260B Magnetics Analyzer. The 3260B was calibrated for two different bandwidths for greater accuracy (100KHz or less - for audio frequencies, All Frequencies - for greater than 100KHz measurements). The measurements and calibration process was repeated twice for consistency. All cable lengths were between 6 to 12 feet and divided by their length for a normalized per foot measurement. The Monster Cable and Sound King cables were unterminated (bare wire), while the Stealth cables were terminated with spades and the Cobalt cables were terminated with banana connectors. Special care was taken to ensure solid contact was made with these cables when all Rs, and Ls measurements were taken.

At low frequencies the results illustrate Rs being lower than Rdc, which is inaccurate, as Rs tends towards Rdc as frequency approaches zero or DC. The LCR measurement derives Rs from signal phase and amplitude, while a DC meter measures exactly what it is looking for, thus this discrepancy is likely due to a meter resolution issue, as the meter in AC mode does not sport the high accuracy it would in DC mode. The cable should ideally be modeled as multiple parallel resistors, and those resistors treated as a lumped element in series with an ideal inductor. Each resistor is a frequency dependent element, and the inner ones fall out as the frequency increases. It is important to note the difference in measuring techniques, and caution the reader not to attempt to derive any relationships with the two numbers, as the absolute accuracy between the methods has not been established. However, the rising trend of Rs vs frequency is indicative of increased Rac due to skin effect and should also be noted.

Figure 1: Inductance vs Frequency Comparison

Summary

As we can see by the graph above, one of the most expensive cables (Stealth Fine Ribbon) in this comparison has the highest measured inductance. In fact, the manufacturer of this cable ships these cables and the Premier II two as separate loose conductors. The measurements above are based on keeping the two conductors as closely together as possible, using electrical tape every foot to ensure no open loop areas. When the conductors are left dangling on the floor like the manufacturer recommends, inductance measurements of these cables more than doubled! See tables 3 and 4 for the resultant impact this had on frequency and phase response.

The Cobalt Cables measured the lowest inductance out of all the cables in this comparison, despite the commonly held myth that the lower the gauge, the higher the inductance. Conductor spacing plays a significant role in the overall loop inductance as clearly evident in this example.

Notice how the inductance of these cables began to decrease slightly between 50 KHz to 100 KHz and continued to decrease linearly with rising frequency. This is a clear result of skin effect pushing the current density profile to the outer surface of the conductors and thus minimizing internal inductance.

The Original Monster Cable seemed to prematurely lower in inductance at around 20-30 KHz. "This premature breakpoint is most likely due to inter-strand copper chloride build up from to the PVC insulation, causing a non skin-effect based current profile variation as a result of magnetic field variation within the conductor bundle." (John Escallier)

Unfortunately my test sample had clear evidence of copper oxide and copper chloride corrosion as indicated by its browning green conductor color. Since most PVC insulation's outgas chlorides with time, and the time any cable spends on the shelf prior to use is uncontrolled, it was decided to use the existing, old sample. Characterizing the shift in electrical parameters being discussed here as a result of inter-strand chemistry changes falls outside the scope of this article, and will be considered for further evaluation. In any event, this did not appear to cause any significant measurable effects within the audio band, and is unlikely to result in any deleterious audible effects.

The bottom line about internal inductance vs frequency is "at infinite frequency, the internal portion of the wire has no field, hence, no energy stored, and no inductance." (John Escallier) For more information on cable inductance relationships, review our article on Calculating Cable Inductance of Twin Feeder Speaker Cables.

Speaker Cable Face Off 1 - Measurements

Figure 2: Resistance vs Frequency Comparison

Summary

Once again the Stealth Fine Ribbon Cable had the highest Resistance out of all the cables in this face off. Its resistance was nearly 11 times greater than the lowest resistive cable in this face off. Incidentally the lowest resistance cable was the Stealth Premier II, followed by a very close second, the less expensive Cobalt Cables.

Notice again how Rs begins to increase significantly above 50 KHz to 100 KHz (depending on the wire gauge) and increases linearly with increasing frequency. Originally we predicted a 34% increase is AC Resistance (Rac) at 20 KHz due to skin effect for 12AWG wire (see Skin Effect Relevance in Speaker Cables), but this was assuming solid core wire as explicitly stated in the article. I suspect that in reality, employing stranded un-insulated wire further reduces skin effect in the audio band to result in a negligible change of resistance at 20kHz and merely a (4.14 / 3.1) = 33.5% at 50kHz with respect to 100Hz, which is well beyond the audible range. Although this is not true Litz wire (individually insulated stranded wire), this may indicate subtle, but apparent benefits of un-insulated stranded wire over solid core wire of similar gauge. John Escallier suspects " It is possible that oxide and chloride formation between strands interfere with eddy current formation, producing less of a skin effect ". These conclusions must be considered, with repeatable and verifiable measurements and analysis between un-insulated stranded and solid core wires. However for now we will defer this overall issue to further research as it goes beyond the scope of this article. Perhaps a future article may also consider the effects of PVC jacketed stranded wire, and how aging affects skin depth, and therefore, inductance between solid and stranded wires of similar gauge, including tracking of parameters after aging of the cables in an oven and/or cryogenically freezing.

Note, the analysis used for skin effect in the previously mentioned article only applied to round conductors utilized in all of these cables except the Stealth Fine Ribbon. The flat ribbon cable employed in the Stealth Fine Ribbon utilizes a different Bessel function. However these calculations are beyond the scope of this article, and since there is no significantly measurable skin effect problem with these cables in the audio bandwidth, the exercise is mostly academic.

Figure 3: Capacitance vs Frequency Comparison

Summary

The Stealth Fine Ribbon cables had the lowest measured capacitance out of all the cables in this face off. This was due to the conductor spacing of the cable.

Note: It is likely that the manufacturer choose to minimize capacitance at the expense of a dramatic increase in cable inductance.

All of the other cables in this face off had appreciably higher capacitance, which should not represent any real world problems of amplifier stability or frequency peaking for moderately long runs (50ft or so).

Frequency Response Comparison

The measured RLC parameters of all the cables in this face off comparison must now be analyzed in the frequency domain to determine if any appreciable resultant attenuation or peaking is likely to occur in a typical high fidelity audio system. Before doing so, we must make a few basic assumptions for consistency in our comparisons.

Assumptions:

• Speaker Load (RL) = 4 ohms
• Ls - use typical Ls measurement within audio band measurements from Table 1.
• Cp - use worst case measurement within audio band measurements from Table 1.

Using an ideal source and resistive load, we will now analyze the following cables based on the RLC metrics in Table 2.

Table 2: Measured Cable Metrics Comparison (typical)

Figure 4: Graphical Frequency Response Comparison of 10ft Cable Lengths

Table 3: Frequency Response Comparison of 10 ft cable lengths

Table 3 illustrates comparative insertion loss, total loss at 20 kHz, and associated change in group delay within the audio band for 10ft lengths of cables based on their measured RLC parameters all terminated into 4 ohm loads. The yellow highlights represent the best performing cables while the red highlights represent the worst. As we can see by the results, the Stealth Fine Ribbon cables performance was dependent on conductor spacing. As conductor spacing increased, its performance decreased. Given the short cable lengths in this comparison, it is unlikely any of these cables would yield audible degraded performance with perhaps the exception to the Stealth Fine Ribbon cables. The excessive attenuation characteristics of the Stealth Fine Ribbon cables may yield a perceivable softening of the highs resulting in a commonly referred by audiophiles "warm" sonic signature.

Table 4: Frequency Response Comparison of 50 ft cable lengths

Table 4 illustrates comparative insertion loss, total loss at 20 kHz, and associated change in group delay within the audio band for 50ft lengths of cables based on their measured RLC parameters all terminated into 4 ohm loads. The yellow highlights represent the best performing cables while the red highlights represent the worst.

Note that at 20 kHz, a phase shift of 36 degrees represents 5 microseconds, this delay being considered as close to the limit of human directionality perception. The worst performing cable in this respect (Stealth Fine Ribbon) phase shifts within the audio band close or at the threshold of humany hearing. With the exception of the Stealth Fine Ribbon cables, it is unlikely any effects due purely to phase shifting within the audio band would be audible for any of these cables.

As we can see by the results, the Stealth Fine Ribbon cables performance was dependent on conductor spacing. As conductor spacing increased, its performance decreased. Given the long cable lengths in this comparison, it is highly likely that audible degraded performance would be perceived if the Stealth Fine Ribbon cables are utilized. The excessive attenuation characteristics of the Stealth Fine Ribbon cables may likely yield a perceivable and excessive softening of the highs not dissimilar to attenuating a high frequency tone control in an audio system. It is interesting to note that that 50ft of the Stealth Premier 2 (closely spaced conductors) and Cobalt cables performed similarly to 10ft of the Stealth Fine Ribbon cables when loosely dangling.

Speaker Cable Face Off 1 - Conclusions

With the exception to the Stealth Fine Ribbon cables, all of these cables performed admirably well within the context of electrical test performance. Based on the poor electrical test results of the Stealth Fine Ribbon cable, we discourage its usage for serious audiophile applications, unless the intent is to lower the high frequency response of the speaker system, using the wire as a "tone control". More consistent results would be obtained by the insertion of a lumped element at the speaker of the correct inductance, for the desired result, or using room treatment to tame an overly bright room or loudspeaker system. Note that simply spacing the conductors to insert a lumped element such as an inductor or ferrite will add inductance by geometry. The Stealth Premier 2 and Cobalt cables had the highest performance in this comparison. However, special care must be taken to ensure closely spacing of the (+) and (-) conductors of the Stealth Premier 2 to achieve comparable performance to the Cobalt Cables. Since this could prove quite challenging, we recommend considering the cheaper and consistently excellent performing Cobalt Cables. From a build quality and cosmetics stand point, the Cobalt cables were our top preference. On a budget, we recommend either the Monster Cable Navajo or Sound King 12AWG along with quality spades or banana terminations of your choice. The Original Monster Cable appeared to have oxidizing problems over time due to the employed insulation and may result in added contact resistance over time.

We feel choosing a speaker cable should be a function of electrical performance, build quality, durability, longevity and system compatibility. There are many aftermarket banana plug and spade type terminations that may compliment the un-terminated Monster series cables and Sound King Cables. I really liked the overall construction of the Monster Navajo Cable. It had nicely crafted insulation that was off white in color making it blend somewhat more easily when installed near a similarly colored wall or ceiling. The Cobalt Cables build quality was quite impressive. The jacket covering the conductors was esthetically pleasing and the banana plug terminations fit very snug against the binding posts of my amplifiers and loudspeakers. The spade terminations Stealth Audio Cables employed on the other hand were not wide enough to interface with the binding posts of any of my hardware. In addition, the stiffener at the end of the cables where it was terminated made it very difficult to interface with my amp as it was hitting my amps supporting shelf.

With that in mind, our top three choices in order of preference were:

Stay tuned for Speaker Cable Face Off II where we will be analytically comparing a larger variety of exotic speaker cables to determine if they measure up to their claims.

Acknowledgements
I would like to thank the following people for their peer review and useful feedback.

• John Escallier
• Dan Banquer ( http://www.redesignsaudio.com )
• Rod Elliot ( http://www.sound.au.com )

I would also like to thank:

• Cobalt Cables ( http://www.cobaltcable.com )
• Monster Cable ( http://www.monstercable.com )
• Stealth Audio Cables ( http://www.stealthaudiocables.com )
• Parts Express ( http://www.partsexpress.com ) seller of Sound King Cables

for providing samples of their products which made this face off comparison possible.