As a long-time pro audio guy, I’ve been reading ProAudio Review for years. They do a good job of keeping me informed of the latest gear and spotting new trends. I received the September issue yesterday with the cover tease “Technically Speaking, Snake Oil Vs. Reality”. This was the topic of Editor Frank Wells’ column and I assumed that it was going to refer to a take-down of some dubious claims made by the “tweak audiophile” community from the perspective of audio professionals. What sorts of claims? Well, the audibility of $5000 loudspeaker cables, for example. What I discovered was pretty much the opposite.
Editor Wells was referring to his own review of the ESP MusicCord-PRO that appears latter in the issue (and which you can read on-line, here ). The product is an AC power cord that replaces the modular power cord found on most pro and semi-pro audio gear (versus consumer gear which normally uses a cord permanently affixed to the back panel). That’s right, an AC power cord, not loudspeaker cord. They’re asking $165 for a 1.5 meter length. This is an extremely high price for a power cord, although in the tweak audio community I have seen AC cords in the $1000 realm. I imagine it’s all part of the “bigger fool” theory of marketing. In any case, Editor Wells concludes the review by saying “What I heard was consistent and repeatable; significant, bordering on profound…” Mind you, the testing consisted of replacing the AC cords on some small powered monitors with the ESP cord and listening to the results (for the uninitiated, powered monitors are basically loudspeakers which contain their own power amp, so in effect he swapped out the power cord on the amplifier).
It may be best to hear what Michael Griffin, President of ESP says first:
“For dynamic peaks in particular, sometimes the amplifying equipment is just not getting enough current to reproduce the waveform accurately.”
And from the “Theory” section of the review:
“The wiring in between our breaker boxes and our outlets is typically flat with the ground lead in between the hot and neutral conductors; low magnetic/inductive interaction is in play. The typical Edison to IEC power cable is round and tightly bundled resulting in more interaction. Also, intuition would suggest that a move from a common power cord using 18 AWG wire to a larger gauge, say, 14 AWG, would yield more current capability, and it does, but Griffin says this actually compounds the problem as larger conductors have slower time constants, resulting in a sluggish response to demands for rapid current change.
The patented MusicCord-PRO approach uses an oversized 12 AWG ground wire as a core, spiral-wrapped with eight 20 AWG wires for a 14 AWG equivalent current capacity. Griffin elaborates: “As you go to smaller conductors, they have faster time constants. Of course, they don’t handle as much current. By using multiple 20 AWG conductors (that have no audible phase distortion) in parallel, it’s like taking these small conductors, with very fast time constants, and stacking them on top of each other. So you’ve got a very fast ramp up and drop off; AC current can flow faster. That’s how you improve the transient performance.” He adds that the ESP cables also have a braided copper RFI and EMI shield, offset by an inner jacket for improved EMI performance, particularly at 120 Hz.”
Outside of some religious dogma, it is difficult to pack more crap into two paragraphs that what we find above. You see, it is a hard job to sell someone a $165 1.5 meter long power cord once they realize that on the other side of the wall socket may be 50 feet of plain old Romex. It is true that Romex is laid out in a “flat” manner (the cross-section is a rounded rectangle), but the conductors themselves are round. This is the precise layout of common power cords. Heck, even the average lamp cord is two round wires run side-by-side. The major difference is that most power cords use stranded wire instead of Romex’s solid wire in order to be more flexible. The complaint about standard cables being “round and tightly bundled” isn’t just shit, it’s skunk shit. You see, if “tight bundling” is a problem for power cords, doesn’t ESP’s design of spiral-wrapping a series of conductors around the core make things worse in that regard? And to continue on this line, sure, the basic law of electro-magnetic induction states that the effect increases with proximity, but what sort of signal are you trying to avoid here? In studios, people are very careful about minimizing induced signals into mic and instrument lines because it degrades the audio. What sort of signals? Why AC power signals, of course. You don’t want 60 Hz hum in your mic line, right? But this is nonsense when applied to another power cable for a variety of reasons (not the least being the orders of magnitude difference in signal size).
OK, so what’s on the gear-end of the cable? Normally, the AC power signal feeds a transformer which scales the voltage to the required level. Following this are a series of electronic devices which turn the AC into pulsating DC, filter and smooth it, and finally, regulate it to a constant DC value. If you look at the pulsating DC waveform coming into the filter and realize what the job of the filter is, surely, the idea that some modest signal induced into the power line would have any real affect is nothing but a joke, and a joke squared once you look at the regulation circuit that follows.
Finally, there’s a lot of talk about “time constants” and “ramping up” the current to get it to “flow faster”. Yes, cables do have time constants and they do give a measure of how quickly a signal can change. The problem is that the time constants of even everyday cheap power cables are orders of magnitude faster than the 60 Hz power line frequency that you’re trying to pass. Don’t believe me? You can hook up some 18 or 16 gauge zip cable between your amp and loudspeakers (as most people do) and you’ll be able to hear plenty of signal to the upper limits of human hearing (20 kHz). Indeed, the fastest rate of change you should ever see on a 120 VAC 60 Hz line is found at the zero-crossings, and which just a little differential calculus reveals is only .064 volts per microsecond, pitifully pokey compared to the abilities of even very modest audio power amplifiers. In fact, you don’t want your AC cable to have an extremely fast time constant in the first place. Why? Because if it passes an extremely wide range of frequencies, then it could serve as an antenna, and if the power supply it feeds is not well designed, some of that interference could leak into the DC line. It is interesting to note that the reviewed product has added EMI shielding to prevent this scenario.
So, what’s the bottom line? I think there are two possibilities here. First, Editor Wells may be in possession of monitors that have an under performing AC/DC power supply. Given that he was using monitors from Event, JBL, and Genelec (all respected names), I’m doubtful of that. The second possibility is that Editor Wells is suffering from the well-known observer effect. At one point in his column he mentions that he could hear differences only after “training (his) ears”, and in spite of the fact that his audio analyzer could not discern any differences. (To be fair, he did not detail precisely what sorts of measurements he took, and the analyzer in question, while no doubt serviceable, is certainly not of the highest resolution available.)
I would never throw $165 at an AC power cord, but it could be much worse.
Updated October 7, 2009: Regarding the comment about the pokey slew rate of the AC power signal, I did not mean to imply that the line current rate would have to be the same, it’s just that when engineers talk about cable time constants, they’re usually referring to the voltage. It is possible for a power supply to demand current in very narrow high peaks, however this would not be an optimal design. And again, we’re not talking about signal current, we’re talking about the current drawn by the power supply which will in turn feed the amplifier. A properly designed power supply shouldn’t have a problem storing sufficient energy to handle musical transients, and I see no evidence that replacing five feet of power cord between the amp and tens of feet of Romex will somehow improve the situation further.