I hauled the Ayre V-3 up my stairs the way I’ve hauled up every amplifier I’ve ever loved: badly, with one hand under the front lip, muttering things my mother wouldn’t approve of. Fifty-five pounds doesn’t sound like much until it’s fifty-five pounds of Colorado machine-shop aluminum with no handles and a center of gravity somewhere near the back panel, where the transformer lives. By the time I set it on the amp stand I had already decided I was going to be objective about this thing. That lasted until the second side of side one.
What Charles Hansen actually built
Most solid-state amps of this vintage were five-stage affairs — a differential input pair, a voltage amplifier, a driver stage, a pre-driver for the driver, and then a wall of bipolars strapped to a heatsink, the whole teetering pile held upright by a generous dose of global negative feedback. Hansen looked at that and asked the impolite question: how many times do you want to photocopy the photocopy? The V-3 answers with three stages. An input stage, a folded cascode, an output stage. That’s the tour. Please tip your guide.
The output devices are MOSFETs, and not because MOSFETs are fashionable — they were, briefly, in the way parachute pants were — but because they’re voltage-driven. A MOSFET gate doesn’t demand the current a bipolar base does, which means Hansen could delete the driver stages entirely instead of designing clever ones. Deleting a stage is the only truly distortionless way to improve it. There is no overall negative feedback loop wrapped around the whole thing, which is a design decision with consequences, and we’ll get to those consequences in a moment because they’re audible, and because pretending otherwise is how reviewers lose their souls.
Pop the lid and the circuit board looks lonely. Ninety percent of what you’re looking at is power supply: a transformer roughly the size of a bundt cake, a stack of chokes, and a small municipality of filter capacitors regulating the front end. The chokes are the tell. Nobody puts iron in a power supply in 1994 because it’s cheap or light or convenient. You do it because an LC filter attenuates rectifier hash in a way brute-force capacitance simply doesn’t, and because in an amplifier with no feedback the power supply is the signal path — the output devices are really just valves modulating the rails onto your woofers. Everything ugly in the supply arrives at your speakers with a stamped ticket.
Specs, briefly, because nobody ever fell in love with a spec sheet
100Wpc into 8 ohms, 200Wpc into 4 ohms
32dB of gain — generous; be careful with your volume knob the first time
10k ohms unbalanced, roughly 20k balanced
Bandwidth deliberately capped around 25kHz, a consequence of how much current the input stage can push through the cascode
Output impedance in the neighborhood of a third of an ohm through the midband, climbing toward an ohm-plus at 20kHz
55 lbs, 18″ wide, and it idles warm enough to be a cat magnet
Hold that output impedance figure. It’s going to explain the one thing this amplifier doesn’t do.
Class A first, then A/B, and why that matters
The V-3 is biased into class A for its first handful of watts — call it five to ten, depending on how the unit was set and how warm the room is — and then slides into class A/B for everything above that. Purists will sniff. They shouldn’t. In a normal room with normal speakers you spend the overwhelming majority of your listening life inside those first few watts; the big power is for the transient that lasts eleven milliseconds. The real trick is the handoff. A badly biased A/B amp announces the transition like a gearbox with a bad synchro — you hear it as a hardening, a faint grain that shows up on massed strings and vocal sibilants and never goes away once you’ve noticed it.
The V-3 doesn’t announce anything. Whatever Hansen did to set the bias — and lateral MOSFETs, with their forgiving thermal behavior, make this easier than it is with bipolars — the crossover region is inaudible to me. I listened for it. I put on the kind of material that exposes it, cued up the same passage at whisper level and at levels my downstairs neighbor still resents, and I could not catch it changing clothes.
The sound, which is the whole point
This is the smoothest all-transistor amplifier I have had in my system. Not the most transparent, not the most powerful, not the most impressive on a showroom floor at 100dB — the smoothest, by a margin that isn’t close.
The midrange is where it gets uncanny. Voices have body and a lack of electronic edge that I associate almost exclusively with good tubes. There’s no glassiness, no sense of a signal that’s been through a lot of transistors and is a little tired from the trip. The top end follows suit: cymbals have shimmer and decay instead of that white-noise sizzle, and they stop when the drummer stops. Detail is abundant but it never arrives with elbows out. You can play this amplifier loud for four hours and get up feeling fine, which is a specification nobody prints.
And then there’s the bass, where I have to be honest with you. It’s punchy. It’s tuneful, it has real weight, and on a lot of music it’s a delight — there’s a woody, unforced quality to double bass and kick drum that the tightly-corseted solid-state competition can’t do. But it is not the best transistor bass I’ve heard. It’s slightly loose in the bottom two octaves, a hair soft on grip, and on a demanding woofer load it can sound like the amplifier is politely suggesting where the cone should go rather than telling it.
My hypothesis, and it’s a hypothesis: it’s the feedback, or rather the absence of it. Global feedback is the cheap way to drive output impedance toward zero, and Hansen declined to use it. A third of an ohm at the speaker terminals is a damping factor in the low twenties — not the several hundred your typical big bipolar amp claims. Add a MOSFET output stage, which is inherently a lower-transconductance, higher-source-impedance device than a bipolar, and a supply whose chokes trade some instantaneous current-delivery stiffness for quiet — chokes love steady state and merely tolerate a 30Hz transient — and the woofer’s back-EMF gets a slightly softer argument in return. Everything that makes the midrange sound like a tube amp is exactly what makes the bass sound a little like one too. You don’t get to unbundle it.
Final verdict
The Ayre V-3 is a beautifully reasoned amplifier that made a specific, coherent set of choices and then had the nerve to live with all of them. If you want a transistor amp with a tube amp’s midrange and treble and none of the biasing, retubing, or 3am glow-in-the-dark anxiety, this is the one, and I don’t say that lightly. If your speakers are a 3-ohm dip with a 15″ woofer and you listen to a lot of synth bass, look elsewhere, or at least audition first with the material that matters to you.
Me? It’s still on the stand. My back has forgiven it. Mostly.
I hauled the Ayre V-3 up my stairs the way I’ve hauled up every amplifier I’ve ever loved: badly, with one hand under the front lip, muttering things my mother wouldn’t approve of. Fifty-five pounds doesn’t sound like much until it’s fifty-five pounds of Colorado machine-shop aluminum with no handles and a center of gravity somewhere near the back panel, where the transformer lives. By the time I set it on the amp stand I had already decided I was going to be objective about this thing. That lasted until the second side of side one.
What Charles Hansen actually built
Most solid-state amps of this vintage were five-stage affairs — a differential input pair, a voltage amplifier, a driver stage, a pre-driver for the driver, and then a wall of bipolars strapped to a heatsink, the whole teetering pile held upright by a generous dose of global negative feedback. Hansen looked at that and asked the impolite question: how many times do you want to photocopy the photocopy? The V-3 answers with three stages. An input stage, a folded cascode, an output stage. That’s the tour. Please tip your guide.
The output devices are MOSFETs, and not because MOSFETs are fashionable — they were, briefly, in the way parachute pants were — but because they’re voltage-driven. A MOSFET gate doesn’t demand the current a bipolar base does, which means Hansen could delete the driver stages entirely instead of designing clever ones. Deleting a stage is the only truly distortionless way to improve it. There is no overall negative feedback loop wrapped around the whole thing, which is a design decision with consequences, and we’ll get to those consequences in a moment because they’re audible, and because pretending otherwise is how reviewers lose their souls.
Pop the lid and the circuit board looks lonely. Ninety percent of what you’re looking at is power supply: a transformer roughly the size of a bundt cake, a stack of chokes, and a small municipality of filter capacitors regulating the front end. The chokes are the tell. Nobody puts iron in a power supply in 1994 because it’s cheap or light or convenient. You do it because an LC filter attenuates rectifier hash in a way brute-force capacitance simply doesn’t, and because in an amplifier with no feedback the power supply is the signal path — the output devices are really just valves modulating the rails onto your woofers. Everything ugly in the supply arrives at your speakers with a stamped ticket.
Specs, briefly, because nobody ever fell in love with a spec sheet
Hold that output impedance figure. It’s going to explain the one thing this amplifier doesn’t do.
Class A first, then A/B, and why that matters
The V-3 is biased into class A for its first handful of watts — call it five to ten, depending on how the unit was set and how warm the room is — and then slides into class A/B for everything above that. Purists will sniff. They shouldn’t. In a normal room with normal speakers you spend the overwhelming majority of your listening life inside those first few watts; the big power is for the transient that lasts eleven milliseconds. The real trick is the handoff. A badly biased A/B amp announces the transition like a gearbox with a bad synchro — you hear it as a hardening, a faint grain that shows up on massed strings and vocal sibilants and never goes away once you’ve noticed it.
The V-3 doesn’t announce anything. Whatever Hansen did to set the bias — and lateral MOSFETs, with their forgiving thermal behavior, make this easier than it is with bipolars — the crossover region is inaudible to me. I listened for it. I put on the kind of material that exposes it, cued up the same passage at whisper level and at levels my downstairs neighbor still resents, and I could not catch it changing clothes.
The sound, which is the whole point
This is the smoothest all-transistor amplifier I have had in my system. Not the most transparent, not the most powerful, not the most impressive on a showroom floor at 100dB — the smoothest, by a margin that isn’t close.
The midrange is where it gets uncanny. Voices have body and a lack of electronic edge that I associate almost exclusively with good tubes. There’s no glassiness, no sense of a signal that’s been through a lot of transistors and is a little tired from the trip. The top end follows suit: cymbals have shimmer and decay instead of that white-noise sizzle, and they stop when the drummer stops. Detail is abundant but it never arrives with elbows out. You can play this amplifier loud for four hours and get up feeling fine, which is a specification nobody prints.
And then there’s the bass, where I have to be honest with you. It’s punchy. It’s tuneful, it has real weight, and on a lot of music it’s a delight — there’s a woody, unforced quality to double bass and kick drum that the tightly-corseted solid-state competition can’t do. But it is not the best transistor bass I’ve heard. It’s slightly loose in the bottom two octaves, a hair soft on grip, and on a demanding woofer load it can sound like the amplifier is politely suggesting where the cone should go rather than telling it.
My hypothesis, and it’s a hypothesis: it’s the feedback, or rather the absence of it. Global feedback is the cheap way to drive output impedance toward zero, and Hansen declined to use it. A third of an ohm at the speaker terminals is a damping factor in the low twenties — not the several hundred your typical big bipolar amp claims. Add a MOSFET output stage, which is inherently a lower-transconductance, higher-source-impedance device than a bipolar, and a supply whose chokes trade some instantaneous current-delivery stiffness for quiet — chokes love steady state and merely tolerate a 30Hz transient — and the woofer’s back-EMF gets a slightly softer argument in return. Everything that makes the midrange sound like a tube amp is exactly what makes the bass sound a little like one too. You don’t get to unbundle it.
Final verdict
The Ayre V-3 is a beautifully reasoned amplifier that made a specific, coherent set of choices and then had the nerve to live with all of them. If you want a transistor amp with a tube amp’s midrange and treble and none of the biasing, retubing, or 3am glow-in-the-dark anxiety, this is the one, and I don’t say that lightly. If your speakers are a 3-ohm dip with a 15″ woofer and you listen to a lot of synth bass, look elsewhere, or at least audition first with the material that matters to you.
Me? It’s still on the stand. My back has forgiven it. Mostly.