Ayre Acoustics V-3 power amplifier:
Page 3
Don't read this as an indictment of MIT's AC-conditioning products. Though I haven't used them for long, or in many systems yet, I have heard them improve the sound of other gear; so I can't really can't generalize. What I think I was hearing with the Ayre/Krell combo was the payoff for all of the attention paid to the power supplies.

Just to show you what a fun job this can be sometimes, I had two other AC conditioners around the house—Transparent Audio's Powerlink HPAV and API's Power Wedge 112—and neither of them afforded any improvement in the sound, either. Nor did high-end IEC power cords make much difference. In short, you're going to have to really work to improve the Ayre's power-supply performance with audiophile Band-Aids. As far as isolation from line noise and the usual line-carried garbage is concerned, the V-3 works.

Filled to capacitance
The Ayre was an easy amp to like: It swung like a good 'un. From the minute I put it in the system until I came into my office to write this paragraph (I just had to go out and listen to it one more time), its lithe, swaggering way with a rhythm never failed to impress me. Part of this was no doubt due to the way the amp reproduced the leading edges of transients. I was just listening to Medeski, Martin and Wood's "The Lover," from friday afternoon in the universe CD (Gramavision GCD-79503), and was particularly struck by how the song was carried along by Billy Martin's loosely swinging polyrythmic drumming. The cymbals cut through the Hammond B-3 murk like sunlight piercing fog, and the trapset pushed the performance forward like the funkiest tugboat in the harbor.

Of equal note was the way Chris Wood's acoustic bass was reproduced: deep and woody, but not partaking of that too-tightly-controlled signature so common in solid-state amps. I know a lot of people like that kind of sound, but real bass—deep bass—has a lot of slop in it. The Ayre, like most tube amps, probably got a little more slop in there than is realistic, but that's preferable to sterile aridity. This doesn't mean the V-3 lost all differentiation on the bottom, because it didn't. It had enough control to sort out the Hammond and the stand-up bass in the same frequencies; it just warmed up the bottom end ever so slightly.

I surely couldn't quibble with the Ayre's response through the rest of the frequency range, either. The midrange sounded natural and relaxed and about as far from a solid-state signature as you can go without becoming euphonically colored. On top, it had gobs of detail and an easy, unforced sound that I could—and have—listen to for hours.

It also turned out to be a grand amp for the human voice, especially women's voices. I couldn't get enough of Emmylou Harris, Enya, or Janet Baker through the V-3/10-T combo.

It didn't lose its cool under pressure, either. One of my most consistent references for complex, full-tilt orchestral boogie has been Corigliano's First Symphony (Erato 45601-2). I've seldom heard its equal for dense timbral textures and dynamic gradations, so I just had to throw it at the Ayre. Even at its loudest and most congested, the symphony didn't faze the V-3 a bit. Offstage piano "heard as from a distance" was as ethereal as a memory, while the brass tuttis and crashing gongs remained distinctly articulated even at volumes approaching 100dB.

Restrictions?
There's been some talk concerning Hansen's choice to limit bandwidth in the V-3. Ayre doesn't print many specs on the amp, so most people weren't even aware of this aspect of the design until Audio published its finding that the 3dB-down point was 25kHz. Now people are talking about how this results in a closed-down top end. I remain dubious about this piece of common knowledge—certainly one could as easily attribute the sound of the Ayre to its use of MOSFETs as to the deliberate limitation of bandwidth. I asked Charlie Hansen about it.

"We can't really run more current into the input stage through the cascode stage for practical reasons. That means that when we're driving that output stage—the input capacitance of the output transistors—we run into the bandwidth limiting.