|Elliott Sound Products||White Paper on Speaker Cables and Interconnects|
By Roger Sanders
Reprinted By ESP with kind permission of Roger Sanders (Sanders Sound Systems)
This article is a republished version of that originally written by Roger Sanders (Sanders Sound Systems), one of the 'fathers' of the electrostatic loudspeaker. Although the bulk of the material is from one of Roger's previous companies, the references have been changed at Roger's request. While it may appear that this is an advertisement for Sanders Sound Systems, this is not the case at all, but to remove the name would be to unfairly remove the appropriate references to the source of this material. I have no affiliation with Sanders Sound Systems or Roger Sanders, and references should not be seen as an endorsement or criticism of the products offered - I have not heard the ESLs, preamps or amplifiers, so any further comment on this is not appropriate.
This material is not published without considerable thought and soul-searching on my part. It reflects many of the things I (and many others before me) have already written on the subject. It is presented here as a public service - the hi-fi fraternity needs to be aware that a great deal (the vast majority - OK, all) of the hype about 'cable sound' is pure, simple, and unadulterated horse feathers !
There are a very few things that make a difference, and these are explainable, repeatable and measurable.
There is no basis in reality for most of the claims made, and a search of the ABX site will confirm this. It is to be hoped that this article (which has apparently already caused some flak) will help to clear up some of the gross exaggerations and dis-information that abounds on the Net and in magazines. Most of us may think that the dissemination of information should be factual, but unfortunately there are many people who publish 'information' purely for their own benefit.
Note: Apart from substituting Australian spellings, the article is untouched in terms of the original content. The comments on directionality have been adapted from the latest version of the paper.
Electrostatic loudspeakers (ESLs) are different. The load they present to an amplifier and speaker cables is quite unlike that of conventional magnetic speakers. To a speaker cable, they appear as a capacitor, while magnetic speakers appear as a combination of a resistor and inductor. It therefore is not surprising that cables for ESLs have different requirements from those for magnetic speakers.
Cables have inductance, capacitance, and impedance. Cable manufacturers juggle these parameters to get the cables to sound the way they want. Let's look at these elements more closely and see how they should be optimised for ESLs.
An ESL is driven by a high-voltage, step-up transformer. This transformer is inside the speaker and converts the relatively low voltage of an amplifier to the several thousand volts needed to drive an ESL. Unfortunately, all transformers have leakage inductance. This inductance interacts with the capacitance of an ESL to form an L/C (inductance/capacitance) resonant circuit. This produces an undesirable, high-frequency peak in the frequency response of the ESL.
It is essential that this resonance be kept well above the audio spectrum to prevent the sound from being excessively 'bright'. Since the capacitance of the ESL is fixed, the only way to get the resonance high is to build a transformer with very low leakage inductance.
Designing and building very low leakage inductance transformers that will operate over a wide frequency range and at high voltages is extremely difficult. One of the reasons that some ESLs sound better than others is the design and quality of their transformers.
Inductance is a big problem with ESLs due to the L/C resonance described above. ESL manufacturers expend great effort to obtain transformers with low inductance. So it is vitally important that the cables have low inductance too. If the cables add a lot of inductance to the circuit, they can undo the transformer designer's best efforts.
In a speaker cable, inductance is largely determined by the area between the conductors. Most speaker cables have conductors that run side by side ('twin-lead') and that are separated by a small distance, so have moderate inductance. They do not have the low inductance desired for the best performance when driving ESLs. Some cables use many small wires that are woven together. This reduces inductance greatly, but at the cost of increased capacitance.
The capacitance also should be low. This is not as critical as inductance, but it is important. Remember that an ESL is a capacitor, and amplifiers find capacitors very hard to drive. If the cable adds more capacitance, it only makes things that much worse for the amplifier. Capacitance is highly affected by how close the conductors are to each other. So to keep the capacitance low, the conductors must be widely spaced. Note that this is just the opposite of what we need for low inductance.
Many cable manufacturers deliberately add a lot of capacitance to their cables. For example, you will find a box at the end of MIT cables, which contains capacitors. Alpha Core (Goertz) cables are made as a sandwich with two ribbon conductors very close together, and this type of construction produces high capacitance and often, amplifier instability. Woven wires are close together so have high capacitance. These types of high-capacitance cables are best avoided when operating ESLs.
Impedance is the resistance to the flow of current in a cable. Most cables are designed to have low impedance so that they don't significantly reduce the damping factor of the amplifier. But some manufacturers deliberately use high impedance cables to alter the sound of the speaker by both interacting with the speaker's crossover and reducing the damping factor. When the damping factor is reduced, the amplifier cannot keep the woofer under good, tight control. The result is that the bass becomes 'loose'.
In the case of an ESL, it is best to use a medium impedance cable as this will damp the L/C resonance and reduce its magnitude. Since the L/C resonance should be supersonic, this damping effect may not be audible. But reducing the resonance will make life much easier for the amplifier. Of course, if the ESL's transformer is poor, the L/C resonance will be in the audio range and damping it with a medium impedance cable will help smooth out the high frequencies.
Sanders Sound Systems cables are uniquely designed to meet the needs of ESLs in three ways. They have low inductance, low capacitance, and moderate impedance. How is this done?
Because the conductors need to be close together for low inductance, but wide apart for low capacitance, simultaneously obtaining low inductance and low capacitance seems impossible. But surprisingly, there is a solution to this problem. Coaxial cable construction runs one conductor inside the other. So electricity 'sees' the conductors in the same place. This results in very low inductance.
Sanders Sound Systems' coaxial, low-inductance design is enhanced by spiral-winding the conductors in opposite directions. This further cancels inductance.
But what about capacitance? Doesn't a coaxial design place the conductors close together forming a high-capacitance cable?
Not necessarily. The conductors can be physically separated by a significant distance using a thick, high-value dielectric to produce very low capacitance while maintaining ultra-low inductance.
The impedance is determined by the size and length of the conductor. Sanders Sound Systems sizes the conductors to obtain medium impedance in the typical range of cable lengths used by most audiophiles.
For driving the conventional, magnetic woofers used in hybrid ESL/woofer systems, the demands for low capacitance and low inductance are relaxed, although maintaining these parameters at low levels is still desirable. At the same time, the impedance needs to be low to maintain a high amplifier damping factor to achieve tight control of the woofer.
Sanders Sound Systems bass cables meet these criteria by using dual pairs of coaxial cables. This technique drops the impedance to very low levels while maintaining low inductance and capacitance.
All interconnects are NOT equal. There are some very specific features that interconnects should have. Sanders Sound Systems offers excellent interconnects with all the finest features. But all the hype surrounding interconnects makes it very confusing to know what is important. The purpose of this paper is to explain the facts so you can make intelligent decisions. And the facts can be quite surprising as you will soon see.
There is no doubt that speaker cables can exert a small influence on the sound of your audio system. But interestingly, all well designed interconnects sound identical.
The above statement sounds absurd, since interconnect manufacturers all claim that their products will make your system sound better. They also claim that different types of wire (copper, silver, oxygen free copper, etc.) sound different, how skin effect causes transient smearing, and how dielectrics change the sound. So the idea that all interconnects sound identical is outrageous.
Or is it? Have you actually done a well controlled test to verify their claims? I strongly urge you to do your own testing rather than taking my word for it. It is very simple and easy to evaluate interconnects. Let me show you how.
The idea behind the test is to make it possible for you to switch back and forth between interconnects instantly and repeatedly while all other components in your stereo system remain the same. You can then listen very critically for any difference in sound between the interconnects you wish to test.
You cannot accurately test interconnects by listening to one for awhile, then unplugging them, connecting another set, and listening again. Our 'audio memory' for subtle details is too short to accurately remember any differences in sound in such a test, and we cannot check repeatedly to be sure of what we hear -- so we are easily deceived. You must be able to switch instantly and repeatedly to hear any real differences between interconnects.
You do not need any test equipment. You can use your preamplifier to do the switching. You will need a Y connector so you can connect the two interconnects under test (let's call them 'A' and 'B') to the same component -- probably your CD player.
Note that the Y connector is the same for both interconnects, so even if you believe that the Y connector somehow corrupts the sound (they don't), the same corrupted signal will pass through both interconnects so the test will still be valid. Remember that we are only listening for any difference between the interconnects, and you can hear that difference (if present) on any signal, even a corrupted and distorted one. Inexpensive Y connectors can be obtained from Radio Shack. If you want audiophile grade Y connectors, Sound Connections International (phone 813-948-2907) sells beautifully built, gold plated units at reasonable prices.
Connect one end of interconnects 'A' and 'B' to the Y connector. Do so for both channels.
Connect the other end of interconnect 'A' to one of your preamp line level inputs (such as 'CD'). Connect the other end of interconnect 'B' to your tape monitor input. Do so for both channels. Be sure you don't reverse the channels. All line level inputs on a preamp are identical, so it doesn't matter which ones you use.
You could connect the interconnects to any other line level input on your preamp instead of Tape. But the tape monitor inputs allow to switch back and forth between interconnects by toggling the tape monitor switch instead of having to press different input switches, or rotating a knob. Toggling a single switch is more convenient and makes it easy to do the test 'blind' so you don't know which interconnect you are listening to. Doing the test blind is desirable so your personal prejudices don't influence the test results.
If your preamp doesn't have a tape monitor function, then use any two line level inputs. If you have to use a rotary selector switch, use two inputs that are next to each other on the rotary switch so you can easily move back and forth between them.
The test is done by simply listening to music while switching back and forth between the two sets of interconnects as much as you wish. The idea is to try to hear any difference between the interconnects. There is no time limit, you may switch whenever you wish and take as long as you want.
The test is easiest to do if you have a remote control preamp so you can sit in your listening chair and simply push the Tape Monitor button on the remote whenever you want to switch to the other interconnect. If you don't have a remote control preamp, then you may need an assistant to switch for you whenever you signal them to do so.
To do the test blind, press the tape button several times quickly so you get confused and don't know which interconnect you are listening to. If your preamp has an indicator light showing what you are listening to, then either put a piece of black electrical tape over the light or close your eyes while you do the test.
After doing this test, you will discover that all the hype surrounding interconnects is just that. The fact is that all well designed interconnects sound identical.
But please carefully note that I said all well designed interconnects sound identical. Some interconnects are badly designed and do indeed sound different. So just what is a 'well designed' interconnect?
First, the interconnect must be shielded. Shielding prevents RFI (Radio Frequency Interference) and EMI (Electromagnetic Interference) from corrupting the sound. RFI can take several forms with the simplest being a buzzing sound (usually caused from light dimmers), to actually hearing radio or TV program transmissions faintly in the background of your music.
EMI is caused by magnetic flux lines cutting across the interconnect and inducing currents in it. This can take the form of hum if the interconnect is near an electrical transformer or motor, or will be crosstalk if the interconnect is near another interconnect that is active with a different signal.
Shielding is usually done by braiding a fine wire mesh around an internal conductor(s), making the interconnect coaxial in design. Although this mesh is usually adequate, there are small spaces between the wires in the mesh so that there is not 100% coverage. To obtain the greatest shielding, some interconnects are designed with a solid foil shield. This foil is prone to cracking and breaking if it is flexed, so the foil (usually aluminium) is often deposited on Mylar film that is wrapped around the wire to improve flexibility. But still, foil-shielded cables should only be used in stationary applications since frequent flexing will eventually crack the shield. Braided-mesh shielding should be used for interconnects in home audio systems.
The second requirement is that the interconnect have low impedance. High impedance can cause loss of output at both high and low frequencies depending on the loads presented by the components connected to the interconnect. And when the frequency response is restricted in this way, the effects are indeed audible. Buy why would you want to limit your system's frequency response?
The third requirement is that the connectors at the ends of the wire be practical and trouble free. This encompasses several factors:
Amazingly, many very expensive interconnects fail to meet these basic criteria. In particular, many have no shielding at all! This is inexcusable in an expensive interconnect. The manufacturers of such poor interconnects only get away with this because most home environments have little RFI and EMI. But this isn't always the case and there are many systems that are plagued with buzzing and other noises due to the lack of shielding. The owner is very frustrated that he can't get the noise out and never suspects that his exotic interconnects are the cause.
Some interconnects have very high impedance. This is because the interconnect uses extremely tiny wire. The manufacturers of such interconnects claim that very small wire prevents 'transient smearing' due to 'skin effect' or some other arcane reason. But the fact is, wire size and type does not affect the sound (unless the impedance is too high). There is no such thing as 'transient smearing' in interconnects and 'skin effect' does not alter the sound at audio frequencies. You discovered this in your listening tests. But some of these interconnects have several thousand ohms of impedance and can adversely effect the frequency response of your system.
Very few interconnects have connectors that meet the 'practical and trouble free' criteria outlined above. There are too many connector types to discuss here, but if you will examine them, you will see that few meet the criteria outlined above.
Sanders' cables and interconnects do not have any 'signal flow' arrows on them. This is because wire is not directional. It has no magnetic polarity and has no rectifier properties like diodes. It behaves the same regardless of current flow direction. It has identical resistance, resistance, capacitance, and inductance regardless of the direction of current flow. If you doubt this, then I encourage you to actually measure wire and see for yourself.
Even if wire did have some sort of directional quality to it, it wouldn't matter in an audio cable because the audio signal is AC (alternating current), not DC (direct current). This means that the current reverses direction at the frequency defined by the music -- it does not 'flow' from your source components to your speakers.
So even if you could find wire that actually did flow current better in one direction than the other, its orientation would always be wrong half the time and right half the time, no matter which way you connected it because the signal is constantly changing directions. So orientation wouldn't matter. Cable manufacturers who claim otherwise are operating outside the realm of science.
At Sanders Sound Systems, we make no extravagant or false statements about our interconnects. We don't claim that they sound better than any other well designed interconnect. What we DO claim is that they are the very finest quality and are superbly engineered. And we sell them at a reasonable price.
Specifically, the cable itself is coaxial, low impedance design, with a braided mesh shield. The mesh is of unusually high quality and has a very tight weave. The shielding is so well done that we use a transparent covering over it so that you can actually see the quality. The bright copper braid is also looks very elegant. The cable is quite flexible and a medium size (5.5mm). Although it doesn't really matter (as proven in your listening tests), the metal used in the wire is oxygen free copper.
Our RCA connectors have precision machined, parallel jaws, in the shape of a cylinder. They grip firmly and so perfectly that you can actually feel a suction and 'pop' as you remove them from the jack. All contact surfaces are gold plated over brass. Insulation is Teflon. The jaws are protected by a strong outer cylinder that is separated from the actual contact jaws. This prevents any damage to the precision contact jaws. The outer surface of the connector is deeply knurled for a good grip and hard chrome plated for superior wear resistance. It has a superb strain relief with tapered jaws that clamp down on the outer coating of the cable as a ring clamp is tightened at assembly.
We also supply balanced interconnects that use professional studio cable with a black covering. It is 7mm in diameter, heavily shielded, and flexible. XLR connectors have anodised aluminium housings and gold plated, brass connector pins. Excellent strain relief clamps are used.
Much as I would like to hope otherwise, I know that this will not make a great deal of difference to the believers of the 'cable gods', who postulate that the use of unshielded pure silver interconnects make a difference (they do, because they pick up noise), or that exotic mains or speaker leads will change the character of their systems (they won't). Actually, everything makes a difference, but what is important is whether the difference is audible. The tiniest variation can be measured, but this doesn't mean that you will hear any difference, let alone an 'improvement'.
If you have vast amounts of money and want to impress your friends with your $5,000 speaker leads, then far be it from me to deny you this (dubious) pleasure. However, if you are like most of us, and don't have that sort of money to throw around frivolously, then don't for an instant think that you are missing out on musical 'Nirvana', because it just isn't true. As I have suggested before, make your own leads, and use the money to buy more music! This is infinitely more satisfying in the long run.
For the original article, see the Sanders Sound Systems website (the link to various 'white papers' is on the home page). Also, look at Cables (etc.) - The Truth.
I strongly suggest that the disbelievers visit the ABX site, and look at some of the tests that have been performed on all manner of equipment.
Further information can also be found in some of my own articles, in particular Cables, Interconnects and Other Stuff - the Truth. Also see how you can make your own AB switch box, which can be used to test amplifiers, cables, capacitors and most other audio components.
|Copyright Notice. This article, including but not limited to all text and diagrams, is the intellectual property of Roger Sanders (Sanders Sound Systems), and is Copyright © 2000. Reproduction or re-publication by any means whatsoever, whether electronic, mechanical or electro- mechanical, is strictly prohibited under International Copyright laws. The author (Roger Sanders) and editor (Rod Elliott) grant the reader the right to use this information for personal use only, and further allow that one (1) copy may be made for reference. Commercial use of this published material is prohibited without express written authorisation from Roger Sanders and Rod Elliott.|