Power amplifiers are perplexing beasts. From experience I know that a panel of listeners can listen to three different pairs of high-quality speakers and describe the differences between them as “significant” or even “huge.” But that same panel, when listening to three amplifiers back to back, will rarely describe the differences with the kind of force and magnitude they apply to descriptions of loudspeakers. I don’t think that is too surprising, given that the differences among speakers in frequency response, power response, and phase response are much greater than those one finds in amplifiers. The perplexing part is that, when you talk to members of the listening panel over a beer, you find that they often consider the differences between amplifiers to be as important as the differences between speakers, if not more so. Assuming that experienced listeners aren’t crazy, you have to ask: “What about amplifiers is so important, even if it is subtle?”
Recently, I rounded up a group of amplifiers to shed some light on this question. I wanted to work with amplifiers that are relatively high powered, mainly because my speakers—MBL 101Es—are pretty inefficient (82dB) and provide a 4-ohm load. With these speakers, I didn’t want clipping behavior to dominate my listening. I thought the test group of amplifiers should have relatively mainstream prices (for high-end audio that is). While it is interesting to find that some esoteric and very expensive technology provides unusual benefits, I wanted to think about amplifier differences in a way that would apply more generally. Finally, I chose amplifiers with obvious circuit differences, to maximize the chance that I would find those important sonic differences.

Representing the traditional class A/B transistor-amplifier camp for this session was the Classé CA-M400 monoblock. With 400 watts output into 8 ohms, and 800 watts into 4 ohms, the Classé easily met my definition of highpowered. The CA-M400 retails for $10,000 per pair, which from my research was toward the high side of average for this level of power. Next, I tested the McIntosh MC 501 monoblock, which is a transistor design with the unusual feature of having transformer- coupled outputs (as you would typically find on a tube amp). The MC 501 delivers 500 watts into a 2, 4, or 8 ohm load, and is priced at $9400/pair. With the burgeoning Class D market getting some buzz, the Audio Research 300.2 stereo amp seemed a natural. While ARC calls this a Class-T design (because it uses the Tri-Path module), in a broad sense it is a switching amplifier with similarities to Class D designs. It delivers 300 watts per channel into 8 ohms, and 500 watts into 4 ohms, and is comparatively inexpensive at $3995 for two channels. Finally, I included my reference Musical Fidelity kW 750, because I am familiar with it and because it easily fits into this power spectrum (750 watts per channel at 8 ohms, 1100 at 4 ohms).
With all this power capability on hand, a few of you will want to be assured that adequate AC supply was part of my test rig. To address this, I connected each amplifier to a dedicated 20-amp circuit. This is relevant only because many high-powered amplifiers will not meet their rated spec on a 15- amp circuit. Logically a 15-amp circuit tops out at about 300 watts per channel, continuous, for a traditional stereo amp. I think it unlikely that the continuous demand during my listening would ever have bumped up against this limit, but I did my best to take it out of the equation.

I could bore you with other aspects of my test setup, but I won’t.
I started by listening to each amplifier for about a week, and then rotated them in and out of my system in pairs. This took some effort, but the differences were easy to hear, particularly in longer listening sessions. That isn’t to imply that the differences were what I expected.
First up was the McIntosh MC 501. The basic character of the MC 501 revolves around smoothness. This isn’t achieved by rolling off the highs, which by the way are appealingly delicate and well delineated. Rather, the MC 501 has less grain than we are ordinarily accustomed to. To put this in a positive sense, the MC 501 sounds more continuous than the typical amplifier, in that each instrument seems whole and complete. I think continuousness is a better descriptor of what you hear from amplifiers like the MC 501, because other good amplifiers do not sound grainy. It is only by comparison that you realize that the continuousness of the MC 501 is on a higher plane.
On strings, for example, you hear what the bow is doing quite well, but the sound of the bow and the sound of the resonance from the body of the instrument seem to be completely integrated. Similarly, as electric guitar notes decay, you find that the sound just seems to be there as one unified thing, rather than a collection of elements.