
There is a particular kind of audiophile smugness that comes from owning an amplifier nobody at the show has heard of. I had it for months, I wasn’t proud of it, and I wasn’t planning on giving it up. Then I lifted the lid. More on that shortly. The object in question is the Audio by Van Alstine FET Valve 400R, a 200Wpc hybrid power amplifier built in a suburb of Minneapolis and descended directly from the work of a man who spent some sixty years arguing with the high-end audio industry. Frank Van Alstine died on January 29, 2025, at the age of 87. I want to say that near the top rather than bury it, because it changes how you read an amplifier like this one: the 400R is no longer just a product, it is a surviving argument. The company continues in Minnesota under new leadership with longtime lead engineer Mithat still designing, still building, still answering the email. But the man whose name is on the faceplate is gone, and reviewing his amplifier now feels less like reporting and more like reading someone’s handwriting.
Let’s dispense with the looks first, because you’re going to ask. The 400R is a black box. It is a handsome black box, in the way a good hand tool is handsome, but there is no billet faceplate machined from a single ingot of Swiss cheese, no VU meters swinging like a 1974 mastering suite, no 90-pound chassis requiring a friend and a hernia belt. Thirty-three pounds. Rack handles. There is exactly one concession to visual drama on the entire front plate: a single blue LED, which tells you the amplifier is on and then declines to tell you anything else for the rest of its service life. I have mixed feelings about blue LEDs as a category—the color is fashionable in the way that brushed aluminum was fashionable, and at two in the morning during a quiet passage a bright blue one can read like a landing beacon—but a lone indicator doing an indicator’s job is hard to object to. If it bothers you, a small piece of electrical tape has resolved this particular high-end audio controversy for decades. If your listening room is also your showroom, and your visitors judge with their eyes, this amplifier will disappoint them enormously. That is, until you play something.
The topology, because that’s where this gets interesting. Most “hybrid” amplifiers are hybrid the way a diner is Italian—there’s a tube in there somewhere, glowing through a window, doing roughly what a photograph of a tube would do. The 400R is not that. The voltage gain stage is a genuine 12AT7 triode circuit operating at real plate voltages, and it hands off to a fully complementary power MOSFET output stage built around Exicon double-die lateral devices. Van Alstine calls the arrangement a forward-transimpedance design, and he holds a patent on it, and I’m not going to pretend I can draw the schematic from memory. Here’s what matters about it:
- The tube does what tubes are actually good at—developing voltage gain with low, benign, mostly even-order distortion and no transistor grain—and it is not asked to drive anything difficult.
- The MOSFETs do what they’re good at—delivering current on demand (within the limits of the output fusing, of which more shortly). They’re lateral MOSFETs, which matters: their transconductance has a negative temperature coefficient at operating bias, so they don’t run away thermally, don’t need elaborate bias servos, and don’t drift as the chassis warms. They also clip gracefully rather than hurling square waves at your tweeters.
- There is no output transformer. None. That’s the whole trick: you get the tube’s character in the signal path without the bandwidth compression, core saturation, and impedance-fussiness that iron drags along with it. What sits in the output path instead is another matter, and we’ll get to it.
- Feedback is applied sparingly and locally rather than slathered on globally to make a spec sheet look pretty.
The result is an amplifier with 3dB-down points of roughly 4Hz and 350kHz. Three hundred and fifty kilohertz. Your bat can’t hear it, my cat can’t hear it, and it doesn’t matter directly—what matters is what that bandwidth implies about phase behavior and settling time inside the audible range. Elsewhere on the spec sheet: 200 watts per channel into 8 ohms, about half that on a sustained basis into 4 (the fuses have opinions, as you’ll see), single-ended RCA inputs only (no XLRs—Frank had firm opinions about balanced connections in domestic systems, and anyone who ever called the shop got to hear them at length, which by all accounts was among the great pleasures of owning AVA gear), and a power supply with far more reserve than the audio circuits will ever ask for. That last part is not marketing. You can hear it.
The spec sheet, in brief:
- Type: Two-channel hybrid (vacuum tube / solid-state) power amplifier
- Topology: 12AT7 triode voltage gain stage direct-coupled to a fully complementary power MOSFET output stage; AVA’s patented forward-transimpedance circuit; no output transformer; minimal global feedback
- Output devices: Exicon double-die lateral MOSFETs, complementary pairs per channel
- Tube complement: 12AT7 dual triodes (JJ Electronic supplied as stock)
- Rated power: 200 watts per channel into 8 ohms, both channels driven; approximately 100 watts per channel into 4 ohms (current-limited by the output fusing, see below)
- Output protection: 5-amp fuse in series with each speaker output, which caps deliverable current at 5A RMS regardless of load impedance
- Output coupling: Large electrolytic capacitors in series with the speaker outputs, fitted for speaker protection
- Intended load: 8-ohm nominal loudspeakers; 4-ohm loads retain full transient headroom but are limited on sustained high-level output
- Frequency response: -3dB at approximately 4Hz and 350kHz
- Inputs: One stereo pair, single-ended RCA (no balanced XLR)
- Outputs: Five-way binding posts, one pair per channel
- Chassis: Standard 19-inch rack width with handles, black finish; single blue power-on LED on the front plate, no meters or displays
- Weight: Approximately 33 lbs
- Distribution: Sold factory-direct from Audio by Van Alstine, New Brighton, Minnesota, with an in-home trial period
- Price: Approximately $3,000 direct (check current pricing, as it has moved over the model’s life)
Now for the part that the marketing copy tends to whisper: this amplifier is designed around 8-ohm loudspeakers, and it means it. There is a 5-amp fuse in series with each speaker output, and a fuse does not care about your fantasies of unlimited current. Power equals current squared times resistance: five amps squared is 25, times eight ohms is 200 watts—exactly the rated output, so into 8 ohms the fuse and the amplifier cross the finish line hand in hand, precisely as intended. Run it again into four ohms and you get 100 watts. Halve the impedance and you halve the power, which is the exact opposite of what every audiophile reflex in your body expects.
But that is a continuous number, and music is not a continuous signal. Music has crest factor—the gap between the brief peak of a snare crack and the far lower average underneath—and on most program material the average power is only 10 to 20 percent of the peak. A fuse, meanwhile, ignores instantaneous peaks entirely: it is a wire that must be heated until it melts, so it integrates I²t and answers to RMS current. It is also worth establishing what this amplifier can actually ask for, rather than inventing frightening numbers. Two hundred watts into 8 ohms means 40 volts RMS, so the rails sit near ±60 volts and the available swing is roughly 57 volts peak. Into 4 ohms that caps peak current at about 14 amps. Fourteen is the ceiling.
Against that, the fuse’s real behavior—not the IEC gauge table, whose maximum times are loose acceptance bounds, but the manufacturer’s average curve. A 5-amp fast-acting element opens in roughly 50 to 100 milliseconds at 15 amps, and in about a third to a half of a second at 10 amps. That second figure matters most: ten amps continuous does not give this fuse pause, it destroys it, promptly. But a kick drum peak lasts two or three milliseconds, so a 14-amp transient is over dozens of times before the element is in any danger. The drum on the first bar of “Cissy Strut” goes straight through and the full unclipped peak power reaches your speakers.
Where it comes apart is sustained demand: heavily limited pop masters, synth bass that just sits there at 40Hz, organ pedal notes, a party at three in the afternoon. A word on crest factor first, since it is the hinge of this whole argument. It is simply the gap between a recording’s peaks and its average level, expressed in dB, and it tells you how much of the time the amplifier is actually working hard. A pure sine wave—a test tone, the most thermally brutal signal there is—has a crest factor of 3dB. A well-recorded acoustic album runs 16 to 20dB. A typical modern pop master lands around 10dB. A master squashed to 6dB is one where the mastering engineer has limited and compressed the life out of it to win the loudness war: the average level has been shoved up so close to the peaks that only 6dB separates them, meaning average power is a full quarter of peak power. It is, thermally speaking, barely more forgiving than a test tone.
Now watch what that does at the fuse. Take the amplifier at full output into 4 ohms—peaks just touching that 57-volt ceiling—and compute the RMS current for each case: a 14dB well-recorded track draws 2.8 amps, a 10dB pop master 4.5 amps, and that 6dB brickwalled master 7.1 amps. The first two pass under the 5-amp rating, which is precisely why normal listening at normal levels never troubles the fuse. The third is 140 percent of rating, with the element cooking continuously and nothing to cool it. Push on into sustained clipping and you are at 10 amps RMS—gone in half a second. Note what the fuse is really doing: capping continuous output into 4 ohms at 5 amps RMS, which is exactly the 100 watts we started with. And before it opens, the element’s resistance climbs, sitting in series with your woofer and mildly compressing the very passages that heated it—the basis for every fuse complaint you have ever rolled your eyes at. I am rolling mine slightly less now. The guidance, then, is more nuanced than “8 ohms only”: a 4-ohm speaker of reasonable sensitivity at sane levels will never trouble the fuse, while low sensitivity in a big room, or a taste for sustained loud bass, means buy the 8-ohm speaker—or treat fuses as consumables, like tubes or strings. It is a deliberate choice, not an oversight: cheap insurance for your tweeters and the MOSFETs.
The fuse is not the only thing standing between this amplifier and your speakers. Lift the lid and you find the 400R’s real disappointment: large electrolytic coupling capacitors sitting right in the signal path, in series with the speaker outputs. Every note that the 12AT7 and those Exicon MOSFETs work so hard to get right has to pass through a can of aluminum foil and chemical paste on its way to your woofer. All that effort on a clean tube stage and no output transformer, and the last stop before the speaker is a part you would normally find in a power supply, which is where it belongs.
Why does it matter? An electrolytic is built to store a lot of charge cheaply, not to pass music cleanly. It has internal resistance that adds to the amplifier’s output impedance, loosening the amp’s hold on the woofer. It soaks up a little of each note and lets it back out a moment later, which smears the fine detail arriving right behind it. Its behavior shifts slightly with the voltage across it, adding distortion that grows as the frequency falls, right where the amplifier is working hardest. And tubes aside, it’s the one part of the signal path with a built-in expiration date: the electrolyte slowly dries out, so the capacitor never sounds better than it does when new.
None of this is dramatic. Nobody will walk into the room and shout “electrolytics!” It’s more like a thin coat of gray paint over the music. Low-level detail softens, the very top loses some of its air, the leading edges of notes lose a little snap, decays fade early, and the soundstage gives up some depth. Bass ends up a touch rounder than this output stage should allow. You notice it less on a quick demo than after months of living with it, which is exactly how it played out here.
The maddening part is that this amplifier doesn’t need them. An amp running from a single supply rail has to have an output capacitor, because its output sits at half the supply voltage. The 400R isn’t that kind of amp. It runs from a bipolar supply, with positive and negative rails, so its output naturally sits at zero, and it even has a DC offset adjustment to trim it there. Tellingly, you can’t check that offset at the speaker terminals, because the capacitors block it. You have to measure it at the top MOSFETs, ahead of the capacitors. In normal operation there is no DC to block. The capacitors are there for one reason only: speaker protection. If an output device fails and slams the output to a rail, roughly 60 volts of DC tries to cook your woofer, and the capacitor stops it.
That’s a fair concern, but the industry solved it decades ago with a small speaker protection circuit and a pair of relays. A little board watches the output, and the moment it sees DC, it opens the relay and disconnects the speaker. As a bonus you get a turn-on delay and no thumps. It’s inexpensive, it’s well understood, and it would take up almost no room in this chassis. Yes, a relay contact is technically in the signal path too, but a good one is a sliver of metal, not a can of foil and paste. Swap the capacitors for that circuit and the 400R would sound significantly better: same tubes, same MOSFETs, just more of what they’re doing reaching your speakers. As it stands, the music pays an insurance premium on every single note to guard against a failure that may never happen.
So what does it sound like? Better than those capacitors have any right to allow. The first thing I noticed, within about ninety seconds, was how quiet the background is. Here’s the specific, falsifiable version: on the Classic Records reissue of Kind of Blue, the reverb tail on Miles’s horn in “Flamenco Sketches” decays into a room instead of into a wall of low-level fuzz. The tail goes somewhere. It has an address. It just doesn’t go quite as far as it should, and the last whisper of it, the part that tells you how big the room really is, slips away a little early.
The second thing was the bass, and this is where I had to recalibrate my prejudices. I walked in expecting the usual hybrid tradeoff—lovely mids, soft bottom, apologies all around. Not quite. The bass here is tight, extended, and well-controlled, with more grip than most hybrids can manage. Kick drum has an actual leading edge. Standup bass has pitch, not just weight—you can follow a walking line as melody instead of as rumble. It stops short of the stop-on-a-dime control you get from big brutish solid-state, and with a fuse and a capacitor in series with the woofer, that’s no great surprise. I ran the amp on a pair of well-behaved 8-ohm speakers—the load it was designed for, and the load you should give it—and it never once sounded like it was working.
The midrange is where the 12AT7 makes itself known, and it does so with tremendous tact. This is not a warm, syrupy, euphonic amplifier. If you’re shopping for the audio equivalent of a scented candle, keep walking. What you get instead is tonal density—voices and instruments have body and a believable harmonic envelope rather than the etched, skeletal outline that lesser solid-state designs mistake for detail. Cymbals shimmer and then stop shimmering, correctly. Massed brass gets loud without getting hard. What’s missing is the last layer of air and inner detail, the difference between a very good recording of a singer and the feeling that she’s standing in your room. Given what’s in the signal path, that’s no mystery.
Soundstaging is good without being the magic act the reputation promises. Images are dimensional and stable, they extend well outside the speaker boundaries, and—more importantly—they stay put when the music gets complicated. But I ration the word “holographic,” and I’m not spending it here: depth is where the stage comes up short. Transients are quick, with the very first edge of a note sanded down just slightly.
Smaller nitpicks, because I have a reputation to maintain: the stock JJ tubes are perfectly musical, but a decent pair of NOS or Gold Lion 12AT7s is a cheap and audible improvement, so budget for one more evening of tube-rolling and one more argument with your spouse about why there are small glass objects in the silverware drawer. The single-ended-only input will annoy anyone running long balanced cable. And there’s no remote, no display, no app, and no way to make it talk to your phone, which I consider a feature and you may not.
The final verdict. The FET Valve 400R is a good amplifier. It just doesn’t live up to its hype. It isn’t the giant-killer its reputation suggests, and it’s a long way from AVA’s claim that its hybrids deliver “better sound at home than you hear at the performance.” It’s sold factory-direct, so around three grand buys you parts, labor, and six decades of accumulated stubbornness rather than a dealer’s showroom rent or a distributor’s boat, and that’s still a fair deal. The tube’s harmonic truthfulness and the MOSFET’s grip are both real. So is the asterisk. A design this thoughtful should not send its music through electrolytics that exist only to do a job a small relay board would do better. Frank picked a lot of fights with the audio industry, and he won more than his share. This is one I would have liked to have with him.
- Pros: Real tube gain stage, no output transformer, good bass control for a hybrid, dense and grain-free midrange, stable imaging, plenty of power into 8 ohms, fair factory-direct value, and a company that still picks up the phone.
- Cons: Large electrolytic capacitors in the signal path that a small speaker protection board would have made unnecessary, with some lost air, depth, and low-level detail as a result; utilitarian cosmetics, a blue power LED that some will want to tape over, single-ended inputs only, stock tubes are merely good, zero snob appeal at the country club, and the 5-amp output fuse makes this an 8-ohm design first: transient headroom into 4 ohms is genuinely intact, but sustained high-level demand into 4 ohms will find the limit.
Rating: recommended, with reservations. As for the smugness, it now behaves like the amp into 4 ohms: fine in short bursts, but it won’t hold up under sustained demand.

Setup guide: Audio by Van Alstine FET Valve 400R Bias & DC offset setting
