[SoundStage!]The Y-Files
Back Issue Article
January 2000

Sound Applications

James Weil of Sound Application makes power-line conditioners. He has for the last 15 years, starting well before the latter-day craze of Bybee filters, Chang Lightspeeds, MITs, VansEvers and countless others. And while he'd rather not talk about his competition, he has opened up their various boxes and knows exactly what's inside. He hands me an Allen key and minutes later I peer inside his now-bottomless CF-X conditioner. The first thing that catches my eye are the mirror-polished, solid copper bus bars to which the flying leads from the high-grade Hubbell plugs are soldered. As Weil explains, these 1/8" bus bars are no ordinary copper but OFE grade. Contrary to popular opinion, OHFC copper is only 99.95% pure. Only OFE is 99.99% pure, true 4N copper and comes with authentication papers. The "six nines" or "nine nines" purity designations are often and misleadingly bandied about without, in fact, referring to the very expensive, ultra-pure copper that comes either out of Japan or South Africa. While 5/100th of a percent of impurity differential seem negligible, Weil assures me it's anything but. The immaculate surface gleam radiating off the bars stems from gradual-progression hand polishing, from 600 to 1000 to 1200 to 1500 grit. To preserve the surface from corrosion, it is sealed off with a German polishing compound that leaves a mono-molecular protective layer so the copper will never oxidize.

It's clear right out of the box (pun intended) that James Weil takes his work very seriously. Seemingly insignificant details are attended to, like the choice and placement of the rubber feet underneath or the actual material of screws, all hardware being brass. As a long-practicing Buddhist and martial artist, Weil believes in the old and sage advice that one's livelihood should be made in the most honorable of ways. To him it means giving his customers the very best he can produce. That forcefully eliminates compromises. I'm soon to find out to what extent.

For protection, the Sound Application unit uses a custom-made, very expensive high-speed magnetic circuit breaker with an interrupt time of less than 1 millisecond. This is far beyond the ubiquitous thermal breakers that are industry standard and can take up to three seconds to respond. Should voltage spikes get past the breaker, a bank of eight varistors, value-staggered from 2000 volts on down, act as protection followers. Then the RF filtering begins. A similar organ-pipe array of staggered custom capacitors addresses various frequency bands until the caps decrease in size to tiny silver mica units that filter in the gigahertz range -- 2.5 gigs to be precise. That's RF filtering bandwidth a 1000 times greater than the industry norm. We're talking far-out cell-phone territory.

When I shoot Weil a quizzical look, he wryly counters that, amongst others, the importer of the excellent and ultra-wide-bandwidth Swiss Goldmund gear uses four of his units in parallel in his own shop. Considering that Goldmund makes its own conditioner, selling at the same retail price, this quietly but succinctly speaks volumes. When Weil adds casually that every single component of his unit is rated for 100 amps of continuous current, I'm beginning to get the picture. Overkill is the name of his game. Or, as marketing would phrase it, when only the best will do. However, James Weil lives on the South Pole of marketing and would be hard-pressed to talk anything other than facts. He is, in fact, outright disgusted by a lot of unethical marketing practices that he finds in alarming evidence in his particular sector of the industry and in some instances feels compelled to call outright frauds.

To protect his design, Weil can't disclose the actual vendors of his varistors or custom capacitors. He openly admits that varistors, while ultra-fast, sound terrible as a genre. It took him years of evaluation, to the boring tune of listening to more than 5000 individual components, to find the right ones that afforded the necessary protection without sacrificing audio and video performance. He now has custom capacitors manufactured to Sound Application specifications. What he can talk about are the capacitor sleeves. They contain an interesting damping material used extensively in computers. He gifts me with some of this stuff for my own experimentation. It looks like a thin, 2"-wide aluminum film with glue backing. He has applied, in small 1/4" dots, this very material to windows and doors in his Berkeley Hills home. It magically eliminates buzz and rattles and has great applications for car damping, too.

Jim adds that less is more with this material. When he first experimented with it, he applied bandage-size strips with typical American bigger-is-better gusto. He was soon to learn that less went much further and purchased a few hole-punches to produce Marigo-type dots of various diameters. He eventually eliminated the larger sizes in favor of a 1/4" punch. A dot of that precise size sits proudly atop the magnetic circuit breaker whose side is covered with a different type of damping material. Jim volunteers that he knows people who have successfully used the shiny metalized stuff to treat resonances in load-bearing roofing beams.

While there are no ridgepoles in the CF-X, every one of its bigger caps is fully encapsulated in this material. It acts to cancel microphonics to which larger capacitors in particular are very susceptible. A cleverly implemented RC network prevents further ringing. A top-of-the-line Caddock UltraPrecision TF020 resistor acts as a bleeder resistor. The Caddock Precision resistor Weil tried first proved too noisy. When you consider that competitors use variations of 1% metal film resistors for this circuit juncture, the implications are rather self-evident.

What we have here then in the CF-X Sound Application 12 outlet power-line conditioner is an ultra-wide-bandwidth, RF capacitive filter with no series elements, no transformers, no inductors and no current-limiting devices of any nature; all components are rated 100 amps. Component quality and implementation are as good as it gets. The ground wire used is 11 gauge van den Hul induction-canceling, monocrystal OFHC copper while the power distribution wire is by MAC Wire. Weil calls the MAC wire by far the best wire he's ever worked with in all his years. He is also very familiar with the Harmonic Technology wire and calls it a fabulous product for the money. His stand-offs for the bus bars are custom too and were specified with resonance control in mind. Thus the raw parts cost of his Sound Application unit exceeds what other makers retail their finished goods for. And then there are the details: 100% point-to-point wiring, 3/16" leads on miniature polypropylene capacitors with leads hand wound 720 degrees around 1/8" OFE bus bars. The hand winding is crucial. The same circuit does not work on a circuit board. All hardware is brass, not inductive steel. Although I'm not permitted to say how, the left and right banks of dual Hubbell AC outlets each are separated to provide essentially isolated analog and digital filter circuits.

Flipping the unit over, I notice that all the outlets are mounted with nylon screws to the top plate. All except one just off-center that uses metal screws and is thus grounded to the chassis. The other five dual-gang outlets are floating while, of course, remaining AC grounded. This de-coupling from the chassis eliminates potential ground loops between components plugged into the unit.

Weil points to his solder joints next, contrasting them to standard crimp connections for power-distribution products. He explains that a true cold-weld connection for a 10 or 12 AWG wire requires thousands of pounds of pressure per square inch. I understand why soldering is his weapon of choice. "Most of the manufacturers claiming to do cold welding are, in fact, doing nothing more than standard industrial crimping -- the same techniques used in your Ford automobile."

He explains the basic differences between signal path and power distribution soldering. The techniques for either are diametrically opposed. Signal soldering requires minimal soldering while power distribution demands larger contact areas. "You need to use about 50% more solder and run the joints hotter to ensure full secondary wetting over a larger surface area, thus avoiding any possibility of high-current hot spots due to lack of surface contact area of two dissimilarly shaped materials." Weil adds that he has never been able to find a technician capable of soldering to his specifications, and he's been refining his own technique for 35 years. To this day, every single unit is hand-soldered by the designer himself, and the standard of soldering exceeds NASA Zero Fault requirements. While this sounds like a lofty claim, James Weil readily opens his unit to anyone foolish enough to challenge him to in-depth scrutiny. Weil explains further that when solder joints cool down from their melting point, surface oxidation begins to develop. That is, unless you're committed enough to apply inert gas during the setting period. Weil only knows one guy crazy enough to do that.

Why this entire obsession with removing surface imperfections? Weil’s relaxed answer is monosyllabic: "Noise." Well, of course, we already knew that, right? After all, noise and its artifacts, or more precisely the elimination of either, is the primary design objective for a RF filter. Anybody mirror-polishing his solid OFE copper bus bars and gas-cooling his solder joints must know a thing or two about noise to go to such trouble avoiding it. By implication, this requires a product quiet enough to be transparent to the removal of such noise.

To understand the product's evolution into its present-day incarnation, a cursory glance at its inception will prove useful. As Weil recollects, he began his slow, gradual but continuous descent into the minutiae of power-line conditioning details with a casual, somewhat flippant remark of acquaintance-turned-eventual-mentor John Curl who became a legend in his own time as one of America's most gifted, unconventional audio designers.

Some 15 years ago, Jim had John over at his house for dinner and noted out loud how his TV picture's quality decreased the moment he powered up his ionizer. John's "should be an easy problem to fix" retort causes Weil to crack up in the retelling. "Yeah, 15 years later, I got that fixed all right." That's how long he's been at it. While he is very modest about revealing who his current customers are, he does let out that famous reviewer Dick Olsher has used a Sound Application unit to power his reference gear for a long time. Mobile Fidelity Sound Lab employed two of his units after a massive 30-second audition.

For those like me who had never before heard of Sound Application, this cursory name-dropping validates that while Jim Weil's products may not have the visibility that large advertising budgets afford, some well-known listeners in the industry own his pieces and seem rather reluctant to let them go. Jim reckons he's sold about 100 so far, and the CF-X is his first and only production piece. Everything else he's sold over the years has been purely custom, so the CF-X stands at the current tail end of a long evolution of one-up products. Jim calls the protracted refinement phase "cut'n'paste," suggesting that modeling software is absolutely useless to predict outcomes in this sector. Rather, tried-and-true parts swapping and testing are the only ways to arrive at satisfactory results. As with many other audio designs, part of what one pays for in a design such as the CF-X are many years worth of accumulated experience that lead up to and make possible its eventual inception.

When the infamous Berkeley Hills fire destroyed his entire record collection, Weil abandoned serious listening to music for most intents and purposes. He went video instead. To this day, he finds a TV screen a much more useful barometer to qualify and monitor his ongoing research into the effects of power-line conditioning than listening to speakers. Our visual sense is much more accurate because a larger percentage of the human brain is devoted to visual processing rather than auditory processing. Relying on eyes rather than ears eliminates a lot of subjectivity and is much easier duplicated and verified by test subjects as well. He claims that substituting internal parts has an immediate and noticeable effect on picture quality and, unlike with audio, usually generates uniform responses from by-standing observers.

He also notes that video and digital audio are very similar with regard to the problems they cause and the solutions they require. "The picture on my Sony XBR with 35,000 hours use, although not as bright of course, looks better than current-generation Sony XBRs; the color and resolution of fine detail is better," he quips. Of course, it’s plugged into a CF-X, not the wall.

I ask him about break-in, one of the audiophile religion's ten commandments. Weil reckons that 2000 hours of use should do it, the harder the better. How do you drive a power-line conditioner the hard way? Play rap music at deafening levels around the clock? That was a dumb answer, as I'm sure you know. For maximum effect, you'd want inductive loads -- in other words, large motors. A table saw would do nicely. In the absence of such humdingers, Weil suggests something like an 1800-watt electric heater. Set it at a temperature where the internal thermostat will cause the heater to go through repeated auto on/off cycles. Each time it turns back on, the initial current surge is way in excess of its steady-state current draw, thus speeding up break-in of the power-conditioning device. Vacuum cleaners work well, too.

Such practical advice from the designer suggests once more that he is not of the snake-oil mentality but would rather chop off a few fork-tongued heads in the field if business etiquette didn't demand otherwise. Off the record, we do cover some competitors. From my laundry list of products, all of which he is familiar with down to the exact amount and origin of internal components, Weil singles out PS Audio’s Power Plant devices as honest and value-oriented engineering. He does have a simple and cheap solution to deal with domestic-power-factor issues that McGowan feels need addressing (and Weil isn't talking marital advice or canine alpha dog rituals), but improper implementation of his fix could prove fatal to the operator, so I won't discuss it.

To summarize, the Sound Application CF-X device is an extraordinarily broadband RF filter fully effective into the gigahertz range. It’s purely capacitive and employs no series parts whatsoever. Its primary application is for audio and video performance enhancements and works great with your Mac or PC. More specifically, it’s a product designed to allow reference systems to do what they’re were designed to do -- produce reference-caliber performance. Built entirely by hand and customizable, the standard CF-X with Harmonic Technology power cord retails for $4200, and with the MAC Wire power cord for $4800.

How does it sound? I thought you'd ask that. I'm certain that an in-depth SoundStage! review should be forthcoming, so stay tuned. In the meantime, visit www.soundapplication.com and arrange to review one for yourself. But to tie you over until then or our review, here's a tip from James Weil that you may not have been hip to: AGS Line Dielectric Silicon Compound, catalog number DS-1, for use on GM modules, Ford rotors, spark plug boots, marine light sockets and other connections. Read: all contacts in your audio system. This stuff is cheap and sells in automotive stores (try NAPA). Does it work? Would I be writing about it if it didn't? You can, of course, also buy something that is exactly the same in an audiophile-approved wooden box with gold engraving and pay a premium instead. In either case, you get mo’ betta sound. Cheers.

...Srajan Ebaen


[SoundStage!]All Contents
Copyright 1999 SoundStage!
All Rights Reserved