Ridiculously expensive cable

[peter]

Taken from the aforementioned website:

 

My first impressions of The Silver PowerKordô were of an all pervading naturalness across the frequency range. The soundstage and presence of the music in the listening area was a lot stronger and more focussed. After the full burn-in period and depth of bass response increased in weight and tunefulness. The top end was wonderfully seamless and natural without any sense of brittleness.

 

So using a Silver PowerKord will actually improve the quality of the sound? And after its burn-in period, it will gain a greater depth of bass.

 

Hmm...you learn something every day.

[Matt]

well I thought it was true that a cleaner mains supply gives a better sound, to behonest, I cant tell the difference between using a normal 4 gang or a £150 power cleaning one. :blink:

[Joshwah]

Funny how this amazing mains cable makes so much difference when all thats supplying your socket is a bit of very cheap 2.5mm twin and earth! :blink:

[Andrew C]

OK Long post coming up. Credit to "Studio Sound" a Pro Audio mag who do not derive their income from selling add space to rip-off merchants!!

 

Sorry, no illustrations. Can't work out how to include those.

 

The misconceptions and arguments that surround the issue of cabling offer professional studios an ideal opportunity to distance themselves from hi-fi hype. John Watkinson gives a tutorial on the requirements of analogue and digital signals and speaker applications.

 

FOR SOME reason cables are always subjects of controversy and tend to accentuate the irrational in any debate. Despite the school story of the Emperor's new clothes, phenomenally expensive exotic audio cables are still touted even though their supporting arguments defy the laws of physics. I do get a lot of entertainment from these claims, but the truth has a habit of getting in the way of belief.

 

It's an understandable, if emotive, argument to suggest that having spent a lot of money on equipment that it is foolish to spoil the result with 'ordinary' cables. Unfortunately it is a fallacious argument because ordinary cables simply don't spoil the result. In fact, on a fixed budget, better results can be obtained in a typical system by using ordinary cables because it leaves more in the kitty for things which do matter like microphones and A/D converters.

 

Neglecting AC power leads, there are three basic types of cable used in audio: analogue signal cables, digital signal cables and speaker cables. The first two have to deliver information whereas the last also has to deliver power.

In analogue signal cables the goal is to replicate the original audio waveform at the destination. It is important to avoid any damage to the waveform which might take the form of additive noise or signal-dependent distortion. Noise pickup is avoided in professional analogue systems using screening and balancing. The balancing accuracy is largely a function of the associated input, but some cables have better screening than others. Noise can also be caused by triboelectrical effects in the insulation, which generate voltages when the cable is flexed. So called low-noise signal cables have an insulation material which reduces this effect.

 

TYPICAL cable lengths in audio installations are short compared to the electrical wavelength and this means that the distributed cable capacitance and inductance can be lumped as if they were real components having the same value. As modem professional signal outputs have a low output impedance (6OOΩ being seen only in museums), these stray components are not generally a problem.

 

Distortion can be linear or non-linear. Linear distortion is a change of the waveform which doesn't affect the spectrum, in other words lack of phase linearity. High capacitance cables can cause HF phase lag, but this is generally negligible except on the longest runs. Even the most humble cables are quite incapable of generating non-linear distortion, so forget it. If a system has harmonic distortion, it's not due to the cabling.

 

The oxygen-free debate rolls on, but really the only benefit of oxygen free copper is a marginal reduction in resistivity which has the same effect as using slightly thicker wire. However, oxygen free copper does at least have an effect that can be measured. In complete contrast, the practice of putting arrows on audio cables to show the signal direction is utterly without foundation. It is a tribute to the power of marketing that high-end hi-fi enthusiasts have managed to replace knowledge with belief to the extent that their behaviour has all the rationality of a Californian religious cult. I wonder how long it will be before the first audiophile mass suicide?

 

Skin effect actually does exist and it's worth knowing the basics.

 

Fig.1 shows an element of current in a solid conductor. A current produces a magnetic field, and if the current is alternating the surrounding metal forms shorted turns which cause power loss. Clearly at the surface of the conductor the effect is reduced because the shorted turns are broken. High-frequency currents then prefer to flow on the surface so the effective cross section appears to go down.

The surface area of a wire is proportional to the radius, but the section is proportional to the square of the radius. Thus as the wire gets bigger, the skin forms a smaller proportion of the section and thus skin effect is worse with larger diameter wire. Stranding the wire has no effect as the strands are in electrical contact with one another. The only solution where skin effect matters is to use stranded wire in which the strands are individually insulated.

In analogue audio signal wires, the conductor diameter is so small that the skin effect does not become significant until way beyond the audio band. Consequently silver or gold plating the wire to reduce its surface resistivity is a waste of time and money. Actually that also goes for connectors, although gold plated plugs do look nice.

In the digital domain, the goal is accurate transfer of data. In other words the numbers from the source should be transferred to the destination unchanged. When digital audio is serialised for transmission on an AES-EBU link, the frequencies are high enough for transmission line behaviour to be evident on longer runs. The signal amplitude falls with distance setting a length limit beyond which bit errors set in. It is important to match the source, cable and load impedances to prevent reflections. The type of dielectric in the cable affects the losses. Although PVC insulated analogue cable can be used, there are now lower loss dielectrics available for digital audio cabling which allow longer runs. The eye pattern should be checked at the receiver, also the receiver parity indicator. If these are both good, the interface is transparent. If changing a digital cable changes the sound quality, it's not the cable. There is a D/A converter somewhere which isn't rejecting jitter well enough.

 

My favourite quote on the speaker cable audibility debate is from Douglas Self: "Much of this has been raving nonsense". There's not much I can add to that view. Let's start as ever with the basics. At the highest frequency of interest, 20kHz, the electrical wavelength of a signal is around 15km. This corresponds to 1˚ of phase shift in 40m. Most practical speaker cables are far shorter than this and consequently the phase shift between the ends of the cable can be safely neglected. Consequently, speaker cables show none of the attributes of transmission lines. The easiest way to prove this is to test for the most fundamental of transmission line behaviours: a logarithmic signal loss per unit distance measured in dB/metre. Unfortunately speaker cables show a loss directly proportional to distance.

It's just as well that speaker cables aren't transmission lines because we deliberately don't do a very good job of matching. The calculated impedance of a speaker cable is high at audio frequencies, but it is swamped at one end by an amplifier which uses heavy feedback to give an output impedance as close to zero as possible. At the other end the impedance of a typical passive loudspeaker is hardly constant and highly frequency dependent.

As a speaker cable of typical length isnít a transmission line, it can only be a simple conductor and its important attributes are resistance and the ability of the insulation to resist abrasion in service.

 

UNTIL cryogenic superconducting speaker cables are adopted by the looney fringe, we have to face the fact that there will be an Ohmic loss in the cable. If the loudspeaker were to be a linear device which presents a constant and purely resistive load, the cable and the load would form a potential divider. The loss would take the form of a slightly diminished amplitude, with no change to the waveform.

However, a real loudspeaker doesn't have a constant impedance and so the voltage drop across the cable becomes greater at frequencies where the impedance is lower. This makes the cable appear to have a frequency response. Quality passive speakers have additional components in the crossover which attempt to keep the impedance more constant, but cheap units omit them.

Practical loudspeakers, particularly LF drivers, are not very linear and they generate harmonics. The amplifier output impedance is extremely low and controls the voltage so the harmonics appear as currents. When these currents flow in the speaker cable it then appears to be causing distortion. These distortion currents are within the passband of the tweeter and are audible.

The most effective way of cabling a passive speaker is to use bi-wiring or tri-wiring. Fig.2 shows that the loudspeaker crossover is divided so that separate wires can be used for MF and LF back to the amplifier terminals. This is a useful step because it removes the woofer distortion currents from the tweeter circuit.

In many woofers the coil inductance changes as the coil moves in and out. When the inductance is part of the crossover, the effect is that the tweeter signal is modulated by the LF cone position. Bi-wiring also removes this effect. However the advantage of bi-wiring will be obtained without exotic cables. Heavy-duty mains cable is perfect for the woofer except for kilowatt PA rigs. The damping factor is dominated by the coil resistance and provided the cable resistance is less than 5% of this the damping factor won't be affected. Tweeters actually work better when thinner wire is used because it keeps skin effect out of the audio band. They are mass controlled and don't need damping.

 

September 1997 Studio Sound

[Matt]

yeah ok, thanks for the advanced warnign at the start, I think I'll set by a sunday to read that...