Cable Reviews

How can this be accomplished? The skin depth is inversely proportional to the square root of the product of electrical conductivity, relative magnetic permeability, and frequency. The relative magnetic permeability, or mu for short, is defined as 1 for air, copper, or any other nonmagnetic material. Thus, the way to reduce the skin depth at a given frequency involves increasing the product of permeability times conductivity. Even if mu remains at 1, one way to turn the trick is by greatly increasing the conductivity—which, practically speaking, means developing a room-temperature superconducting wire. For a superconductor, the signal will always propagate outside the surface of the conductor. A practical means of achieving this involves the use of highly magnetic materials.

Ordinary magnetic materials are well-known. These include iron, steel, magnetic stainless steel, nickel, and cobalt. As a class, these feature relatively low mu in the range from 100 to 1000. Specialty magnetic materials have been developed specifically for their magnetic properties and provide mus in the range from 10,000 to 100,000. Examples include grain-oriented nickel, chemically pure iron, and alloys of iron and nickel containing other trace elements going under such trade names as: Mu-metal, Molly-Permalloy, Supermendur, Permunder, and Hy-Mu 80.

L-G cable uses something similar to Mu-metal. The electrical conductivity of Mu-metal is much lower than that of copper, but this is more than compensated for by its extremely high mu in the equation for skin depth. Let's look at an example. Suppose the conductivity of Mu-metal is ¹/₂₀ of copper and that its mu equals 50,000. The product of these two is 2500, and because the skin depth is inversely proportional to the square root of this product at a given frequency, the skin depth is actually reduced by a factor of 50. Therefore, a 1mm diameter wire of Mu-metal would behave (in terms of phase dispersion) like a 50mm, or almost 2"-diameter, copper wire!

The high-magnetic permeability material must be selected carefully. It is highly desirable for the material to have low-hysteresis losses and not to retain a magnetic field. Materials such as Mu-metal, developed for use as magnetic shields, have these characteristics. On the other hand, magnetic steel would not be a wise choice and would sound bad as an audio conductor (footnote 2). Many people have also formed a low opinion of magnetic materials on the basis of bad experiences with magnetic chassis. Unquestionably, a magnetic or magnetizable chassis does adversely affect sound quality in amps and preamps (footnote 3).

Physical details
A total of four individually insulated, 40-mil strands of "Mu-metal" are used in the L-G Model 4-40 interconnect. One strand is used for the "hot" leg and three for ground. For a 20' run of interconnect between the preamp and amp, I measured a 4.6 ohm DC resistance for the hot and 1.5 ohms for the ground. Speaker cable is also available (Model 10-35), which uses 10 individually insulated 35-mil strands in a symmetrical geometry with 5 strands for hot and 5 for ground.

Because the audio signal propagates outside the conductor, by necessity it must propagate through the cable insulation. Therefore, the choice of insulation material is important. L-G has opted to use materials that do not polarize. Materials such as fabrics, which have a relative dielectric constant of 1.0, would be ideal. L-G uses cotton and silk for insulation. They claim that all plastics, including polypropylene, sound bad because their large molecules take too long to polarize under the action of electric fields varying at audio frequencies. Cotton is used to individually insulate each strand, and the outer jacket is made of silk. Neither is L-G happy with standard methods of shielding cable; they've opted for a tightly twisted geometry in which the hot strand is surrounded by the three ground strands. Because the ground strands are electrical conductors and highly magnetic, they act as both electric and magnetic shields. L-G claims that this configuration is just as isolated from electric fields as standard shielded cables, and much more isolated from magnetic fields. However, I discovered that the cable is sensitive to radiated hum, at least in long runs between the preamp and amp. I found it important to route the cable away from AC lines and transformers.

Each strand is individually insulated to eliminate non-linear contact resistance between strands, as you might expect to find in a multi-strand construction where strands touch loosely. Because the "Mu-metal" does not tarnish or oxidize, the longevity of the cable should be excellent, much better than typical copper cables. Copper oxidizes fairly quickly, which further exacerbates the performance of bare multi-strand designs because the oxide layers act as non-linear conductors.

The RCA plugs provided on my interconnect samples were far from visually impressive, looking every bit like cheap plastic and nickel-plate. But that apparently is by design. According to L-G, gold-plating is always dirty: first, because the layers of material under the gold (usually gold over nickel over brass) cause junction problems and partial reflections of the signal sinking into these layers; second (supposedly first noticed at Bell Labs during the Second World War), gold surfaces catalyze the growth of polymer chains, causing poor contacts that act a bit like diodes. A word to the wise: Be sure to clean gold surfaces on your RCA plugs and jacks frequently. L-G also feels that lots of metal in the plug body is detrimental to good sound. Consequently, they use plugs with a minimum of metal (ie, lots of plastic).

It's important to emphasize that my samples were pre-production, but I'm told that the only aspect of the cable that will change during production is the RCA plug. I was not very happy with the plugs I received. This was apparently the second type of plug to be tried, and one that was supposedly sonically superior to an earlier plug with a more metallic body. My problems had to do with reliability. I found it easy to break the solder joints at the connector, either by an inadvertent tug on the cable or by bending the cable near the connector. L-G has ordered sturdier connectors that will presumably be more reliable without affecting the sound.

Another peculiarity of this cable is its susceptibility to "ground loop" hum, at least in long runs between the preamp and amp. At first, this hum proved to be most puzzling, a gremlin if you will, because it seemed to appear and disappear unpredictably. There it was during one listening session, and nothing I could do short of switching back to the Cardas HexLink would get rid of it. Another time, there would be no hum. With a little bit of luck, I ultimately traced this anomaly to a dimmer-switch–operated floor-lamp plugged into the same AC circuit. During the day, the lamp was naturally off and the hum was gone. In the evenings, the lamp was on and with it the hum. This was not a case of radiated hum, so I'm postulating some sort of ground-loop problem as the culprit, possibly a consequence of the relatively large resistance of the cable ground. Of course, that spelled the end of the line for the floor-lamp, but I didn't really mind; candlelight listening is such an inspiring experience.

A final word of caution. The Model 4-40 interconnect is stiff, and consequently has a mind of its own. It is difficult to tuck neatly out of the way. I mention this because I presently share the listening room with Thomas J. Norton, whose aesthetic sense was offended by this cable. Tom and I are really an "Odd Couple" in this respect: he's Felix and I'm Oscar. I enjoy a controlled mess (eg, cables everywhere)—now that Tom's sorted and neatly tied down all the cables, I have a tough time finding things. A cable that asserts itself and does not blend into the carpet is a cable after my own heart.

But you do need to be careful when untying the L-G from the coiled mass it arrives as, inside its own plastic bag. It's just like a Slinky in this regard, and unravels with the lightening-fast strike of a cobra. It could easily bonk you in the face or, worse, zap you in the eye. Exercise caution; if you're a real chicken, you might even consider wearing safety goggles.

Sound
I first heard of the Lindsay-Geyer cable through a phone call from Irv Sherman, a Las Vegas audiophile. In his words, this interconnect was smoother in the treble, with fewer HF "glitches," than anything he'd ever tried. The few facts he was able to relay concerning the cable's composition not only had me dubious as to the viability of the concept, but at the same time piqued my curiosity. Subsequent discussions with David Lindsay alleviated much of my anxiety. Here was a man with substantial educational credentials who was able to eloquently explain the rationale behind the design. So it appeared that there was indeed a method to his "madness." However, all would have been for naught had the cable failed to perform as advertised.

The amazing thing about the L-G is just how convincingly it devastated the competition. It consistently displayed the ability to resolve spatial nuances at a level of precision I had not experienced before. Many cables are capable of preserving soundstage dimensions and of providing a realistic illusion of the depth and width of the original recording space. A few cables can even adequately delineate and focus instrumental outlines within the soundstage. But only the L-G was able to squeeze image outlines into the confines of a realistic parcel of space to a degree I had previously only associated with live music. And it was able to do this across the entire spectrum of instrumental size, from voice and violin to cello and piano. Take a piano, for example. What makes for a convincing illusion of a piano is not only timbral accuracy and proper dynamics, but also the image size projected by the system. Hearing a piano live, one cannot help but be impressed by the sheer size it projects. But no matter how big the wave launch, the piano's fundamental underpinnings, together with its harmonic overtones, can be clearly localized within the same unambiguous space. This rarely happens with reproduced music. The treble, midrange, and bass all too often fail to integrate.

A good example of this is Bruch's Kol Nidrei (from Collected Works for Cello and Orchestra, ebs 6060). With AudioQuest Lapis, Cardas, and Kimber interconnects in my system, the cello was right there during its lower registers; then, as more and more overtones crept into the music, the upper mids and treble would meander away from the cello's center of gravity. The end result was that the image became smeared across more real estate during certain passages. The effect is akin to an expanding balloon being pumped up, then deflated in concert with the harmonic content of the music. To a certain extent this is a natural effect, often referred to as "dynamic bloom." However, when the cello expands to width and height that threaten to engulf the soundstage, the effect looses its credibility. Like the science-fictional eggplant that ate Chicago: eggplants can get pretty big—but there are limits. It was indeed spectacular to have a cello engulf the soundstage, and I'm sure there are audiophiles out there who would enjoy this sort of effect, but for me it is a question of accuracy.

The upper registers were in general cleaner and smoother-sounding with the L-G in the system, compared with my previous reference interconnects. Background levels of treble grit, grain, edginess, sibilant sizzle, and transient smear all lapsed into inaudibility. Low-level upper-register detail was now much easier to resolve against a background of velvety smoothness.

Soundstage transparency increased as well. It was easier to see deeper into the soundstage and resolve layers of detail that were previously clouded over. For example, Lesley's vibrato (The Lesley Test: track 13, Stereophile Test CD) was clearly resolved to an extent I normally associate only with live music. Massed voices were naturally layered and spatially defined to the point where the depth perspective became much more palpable. It became easier to pinpoint individuals in a chorus, while their outlines remained rock-stable over the entire tonal range.

Every link in my system benefited from the introduction of L-G interconnect. I tried it between the turntable and preamp, preamp and amp, CD processor and preamp, and even as the digital link between a CD player and processor. In all cases, treble smoothness, image cohesiveness, and soundstage transparency improved. As each link was upgraded, I got another glimpse of heaven. When my system was completely wired with the stuff, it was time to celebrate: the musical experience had become that much more intense.

I even found a home for the L-G speaker cable. Clearly the realm of high currents is not entirely amenable to a highly-magnetic wire design. Even a 6' run of the Model 10-35 between the Classé DR-8s and the Apogee Stage was not entirely satisfactory. With the impedance rating of the Stage at around 3 ohms, this represented a worst-case scenario. The quality of the upper octaves greatly improved when I made the switch from SYMO cable, but the bass suffered. The impact and tightness of the deep- and midbass registers diminished. There was also a reduction of upper-midrange liveliness which the Stage could ill afford. However, paralleled with the SYMO, I was able to retain the best of both worlds: the treble purity of the L-G together with the bass drive and midrange clarity of the SYMO.

Summary
The Lindsay-Geyer interconnect is physically stiff and hard to diaper, but in the context of a high-end system I feel it pushes the art of music reproduction a mile forward. It will require careful routing to avoid radiated hum, and may still be susceptible to ground-loop hum in long runs. L-G offers an intensely musical experience, however, founded on three cornerstones: treble purity, harmonic integrity, and image cohesiveness.

Over a period of several months the L-G proceeded to trounce every cable I could throw at it, including my established references. It is the most musically convincing and accurate interconnect that I've heard to date. It is not cheap, but is much less costly than some, and as the best interconnect money can buy it has to be viewed as a bargain to boot—at least at the current asking prices.

It bears emphasizing that my samples represent pre-production prototypes. Full-scale production should commence in early 1991. I have been told that, except for the connectors, the cable will not change. Still, I would prefer to verify this, perhaps in a future "Follow-up."

After living with the L-G for a length of time, I discovered that nothing else would do. As always, it is difficult to fall back onto second bests. Here is an interconnect that, in a flash, sweeps clean a host of established cables. Everything that is not highly magnetic by design can be seen in retrospect as being either fuzzy or grainy. Long live the new king!—Dick Olsher

---

Footnote 2: I carried out some listening tests some years ago comparing resistors with steel leads against supposedly identical resistors using copper leads. The results were overwhelmingly in favor of the resistors with copper leads.—John Atkinson

Footnote 3: Engineers at the Japanese Kenwood company showed 10 years ago that the presence of a magnetic chassis in close proximity to a signal-carrying conductor induced distortion into that signal due to the non-linear magnetization of the chassis.—John Atkinson