Paul McGowan: High-End Survivor Page 2

Atkinson: The British term for "vacuum tube" directly reflects that reality because it behaves as a valve . . .

McGowan: It's a valve that controls the power supply. So it's logical to assume that the power supply is critical. Absolutely critical. If you can't get it right there . . . as Ivor Tiefenbrun of Linn used to say, if you can't get it right at the very beginning, the rest of it doesn't mean squat. And he's right. I can't argue with the man.

So yes, I've been thinking about this for years. I always had this Rube Goldberg idea of a product of taking a motor, plugging it into the wall, and connecting it up to a generator. Because then it wouldn't matter what the wall did (as long as the motor had a governor on it). But I could see the warranty claims—"My bearing's going out," "My wife hates the noise," "I can smell ozone." It would have been like the Hill Plasmatronics [ionic loudspeaker], you know: a great idea, but a little goofy in the execution.

But the Power Plant is basically an electronic version of that motor/generator idea. It receives AC energy from your home's wall socket, converts it to DC, then regenerates regulated, balanced, 115V AC power.

Atkinson: Paul, earlier today you showed me—using a Fluke storage oscilloscope intended for power-line analysis—the dramatic improvement the Power Plant has on the 120V/60Hz AC waveform out of the wall. Just why should there be anything wrong with the AC in the first place? Why isn't it a perfect AC sinewave?

McGowan: As you saw, John, it's not a well-done sinewave; it would never pass muster out of an audio generator. There are a number of problems with the AC sinewave. One of them is that it's at 60Hz—which is something that perhaps we'll go over in a moment—or perhaps 50Hz, which is even worse.

But in virtually every home and office and listening environment I've been to, which is perhaps 25 or 30, the sinewave has been flattened to some degree, lopped off on the top [see fig.1]. When the sinewave gets lopped off, higher frequencies that are known as harmonics are created. And they are the dreaded odd harmonics, the 3rd, 5th, 7th, 9th. So you have several problems: First, you have an increase in harmonic energy, the average being perhaps 3% [see fig.2]. Second, you have a lowering of the overall energy level, because the sinewave is no longer able to deliver the appropriate amount of power, and that gets worse depending on where you are.

Fig.1 Typical wall-socket 60Hz AC waveform (red), and as output by the Power Plant (blue).

Fig.2 Spectrum of typical wall-socket 60Hz AC waveform (red), and as output by the Power Plant (blue).

The reason this happens is because increasingly there are products in use in our homes, from computers to televisions to stereos, that use a rectified power supply. This has a diode bridge that converts the 60-cycle AC into DC with a 120-cycle AC ripple superimposed, which is followed by the reservoir capacitors.

Atkinson: And those power supplies suck current from the wall only as required to top up the reservoir capacitors, which presumably will be at the peaks of the sinewave. So while the AC voltage should be a sinewave, the current waveform out of the wall could be anything.

McGowan: Yes. The current waveform feeding one of these rectified devices ranges from about 50% distortion up to about 180% or 190% distortion. Typically, what'll happen—depending on the device and how well it was designed with this in mind—is that you can be drawing no current at all, and then, right at the crest of the sinewave, there will be a huge "sucking sound," as Mr. Perot would say, as the supply draws current from the peak of the sinewave. Since everyone on the power grid to which we're all connected is synced up to the exact same 60-cycle frequency, that means everyone is drawing current at exactly the same point on the waveform. This [synchronized current draw] will both flatten the crests of the sinewave and produce some pretty large harmonic content.

Atkinson: The question I have to ask is: doesn't the preamp or the CD player or the power amplifier's power supply filter out all that garbage?

McGowan: Yes and no. In its simplest form, if you were to look at a power amplifier's [voltage rails], you would see that these harmonics are pretty much gone. They're gone because the capacitors in the power supply have shunted them off to ground. Most people would stop at that point and say, "That's fine, they're gone." Unfortunately, in the way that power amps are designed—and preamps and DACs, for that matter—the point where the noise is shunted off, our power-supply ground, is directly wired to the signal ground.

Atkinson: All the signal voltages are referenced to that ground point . . . ideally, the signal ground should have an infinitely low impedance to the reference ground outside, but it doesn't.