The Best Tube Phono Preamps for a Turntable

A phono stage does two jobs. It raises a cartridge’s few thousandths of a volt up to something an amplifier can use, and it undoes the RIAA equalization curve that was applied when the record was cut — bass cut and treble boosted on the way in, so the playback stage has to put roughly 20 dB of bass back and take roughly 20 dB of treble out. Get the gain wrong and the record is quiet. Get the curve wrong and the record is a different record.

Which is why the specification that decides this page is the one almost nobody writing about tube phono stages mentions: published RIAA tolerance. Across the three boxes here it runs from a 0.8 dB window to a 6 dB window, and one of the three publishes no number at all. Six dB is a factor of two in amplitude. That is not a subtle valve character, it is a tone control soldered in place.

All three were confirmed as currently purchasable on 31 October 2026 and every figure below was read from the manufacturer’s own page the same day. We have not measured any of them and we do not claim to have listened to them side by side; this page ranks what the makers are willing to be held to, which in this category turns out to be plenty to go on.

What a tube does and does not do in a phono stage

Worth being blunt, because the category is sold on romance. A valve in a phono stage is a gain device, the same as a transistor, with different and usually worse measured behavior: higher distortion, higher noise, a warm-up period, parameters that drift with age, and a part you eventually replace. Our tube versus solid state comparison works through that honestly for power amplifiers and the same physics applies here.

There are two legitimate reasons to want one anyway. The first is that you like the result, which is a preference and needs no defense. The second is tube rolling: if the tubes are socketed and the maker names compatible types, the sound of the stage becomes something you can change for the price of a pair of valves rather than the price of a new box. Two of the three here support that explicitly. One of them will not even tell you which tube it shipped.

What a tube does not do is fix a mismatched cartridge, and it does not make loading or RIAA accuracy matter less. Everything on the cartridge-matching page still applies.

The criteria, in the order they decide it

1. Published RIAA tolerance.In dB, across a stated band. Without a number, “conforms to RIAA” is a claim about intent.
2. Gain, published as a number per setting. Cartridges publish their output in mV; a stage that publishes its gain in dB lets you check the pair before you buy. One of these publishes none.
3. Loading — resistance and capacitance.If your cartridge’s maker states a load, you want to be able to meet it. Audio-Technica states 47 kΩ and 100–200 pF for the AT-VM95E, for instance, and only one stage here can be set to anything.
4. Distortion and noise with a measurement condition. A THD figure with no frequency or input level attached cannot be compared with one that has both.
5. Tube disclosure. Named factory tube, socketed or not, compatible replacements listed. On a product bought for its valves this is not a footnote.

The three, on the numbers that separate them

Four figures, because these are the four that decide it. Loading, noise, cartridge support and the published dimensions are in each stage’s full spec table further down the page — and loading is the one where the gap is widest: the Pro-Ject publishes four resistive settings and six capacitance settings, and neither of the other two publishes any adjustment at all.

How much a ± 3 dB RIAA tolerance actually is

This is the arithmetic the category never shows you, and it is why the cheapest box on this page is third rather than first.

1. Fosi Audio Box X2 — the one to buy today

It is first on this page for a reason we would rather not have: it is the best-specified tube phono stage here that you can actually buy through a verified listing today. On published figures the Pro-Ject below it is better in almost every way. On availability, this is the one with a live offer, and a recommendation you cannot act on is not a recommendation.

Two socketed valves, three published gain settings (39 / 42 / 45 dB), THD at ≤ 0.1%, signal-to-noise at ≥ 98 dB, and 600 mV out at a 3 mV input. A 3.5 mm AUX input sits alongside the phono input, so one small box can feed a turntable and a phone into one amplifier channel — genuinely useful if your amplifier is short of inputs, which is the situation that sends most people to the do-I-need-one question in the first place.

The gain range is the right range. An ordinary moving-magnet cartridge wants somewhere around 40 dB, and 39/42/45 dB brackets it properly rather than offering one setting and hoping. Fosi also lists five compatible replacement tube types (6K4, 6J1, 6J4, GE5654, 6AK5), which means rolling is supported rather than merely possible.

Three real problems. First, there is no loading adjustment at all: no resistance switch, no capacitance switch, nothing. If your cartridge’s maker publishes a load — and the common Audio-Technica and Ortofon moving magnets do — you cannot meet it. Second, Fosi does not name the tube in the box anywhere on the product page, on a product whose entire premise is the tube. Third, the RIAA figure is the phrase “Standard RIAA” with no dB attached, and neither the THD nor the SNR figure carries a measurement condition — no frequency, no input level, no statement of whether the 98 dB is A-weighted.

And one oddity worth printing: Fosi’s own two gain-related figures do not agree.

2. Pro-Ject Tube Box S2 — the only one specified like a hi-fi component

Two ECC83 / 12AX7Avalves, named, in the signal path. Four MM resistive loading options and six input capacitance settings from 47 pF to 587 pF — more loading granularity than anything else in this registry, tubes or not. Gain settings at 40, 43, 50, 60 and 63 dB, covering moving magnet through high-output moving coil. A subsonic filter published as 20 Hz at 18 dB/octave, the only one on this page with a stated slope. And ± 0.4 dBRIAA accuracy from 20 Hz to 20 kHz — identical to the tolerance Pro-Ject publishes for its own solid-state Phono Box S2 Ultra.

It also publishes distortion and noise per gain setting, which is the detail that makes the figures usable: < 0.02% THD at 40 dB, < 0.05% at 50 and 63 dB; an 80 dB(A) noise floor at 40 dB falling to 75 dB(A) at the higher settings. You can see the cost of gain instead of guessing at it.

The honest cons are substantial and this site has published them before. That < 0.02% is roughly 38 times the 0.00052% Pro-Ject publishes for the Phono Box S2 Ultra, a cheaper box from the same maker with the same RIAA tolerance and the same adjustability. Pro-Ject publishes a noise floorrather than a signal-to-noise ratio, so its noise figure cannot be compared with the SNR numbers the other two publish — different quantities, not different values. Its page claims five MM loading settings and then lists four, and asterisks the 587 pF capacitance without explaining the asterisk. And it is by far the most expensive stage here.

It is second only because of the offer, not the engineering: on 31 October 2026 its Amazon listing returned an image and no price, and a fresh resolution found no other listing that cleared our brand-and-model guard. Pro-Ject still publishes the product, so this reads as Amazon stock rather than a discontinuation — the buy row below falls back to a tagged Amazon search, and a Pro-Ject dealer is the more reliable route while that lasts. If what you want is the best measured phono stage for the money, the answer is not a tube stage at all, and the under-$500 bracket page says so in its ranking.

3. Douk Audio T4 PRO — cheapest in, and the headphone one

The T4 PRO is the only tube phono stage here with a real headphone output and a published power figure for it: 150 mW at 32 Ω, with a stated supported range of 32–300 Ω, and Douk says the line and headphone outputs run at the same time. It publishes the highest signal-to-noise figure on this page (≥ 105 dB), it has a 3.5 mm AUX input as well as the phono input, and it costs a fraction of the Pro-Ject. If what you want is a valve in the phono path and a pair of headphones plugged into the same box, nothing else here does it.

It is third because of one number. ± 3 dB is the published phono response tolerance, the widest in this registry, and the arithmetic above shows what that permits. Its THD is 0.2% at 1 kHz— ten times the Pro-Ject’s figure at 40 dB gain. No gain figure is published at all, so you cannot check it against a 2.5 mV or a 5 mV cartridge before buying; the 0.775 V “input sensitivity” on the spec table is a line-level figure for the AUX input and says nothing about the phono path. There is no loading, no capacitance setting and no subsonic filter.

Douk also states that the included tube brand and model may vary by batch. That is honest of them and it is a genuine limitation: on a product bought for its valves, you do not know which valves you are buying, and two units may not match. The feature copy calls them “Dual GE Tubes”; the specification table names nothing. One last small thing, printed because it is printed: the net weight is published as “300 g / 0.73 lb”, and those two figures do not convert to each other (300 g is 0.66 lb).

Why the Bellari VP130 is not ranked here

It should be. It is the one classic American all-tube phono stage in this price territory, it is on Amazon today at a live price, and Rolls publishes a proper spec sheet for it: RIAA ± 1.5 dB from 20 Hz to 20 kHz, THD 0.02% at 1 kHz, S/N > 80 dB unweighted, a 20 Hz rumble filter, a 12AX7 tube, 8 Ω headphone and 100 Ω line output impedances. On RIAA accuracy that would place it second on this page.

Two things stop it, and both are about not misleading you:

• Rolls states the VP130 is discontinued and replaced by the VP130 MKII. The MKII has no Amazon listing that clears our brand-and-model guard, so the unit we could actually send you to is the superseded one. Recommending a reader into an end-of-line product because that is the version our network carries is the kind of thing this site exists not to do. If you want a Bellari, buy the MKII from a dealer that stocks it.
• Its published gain is 30 dBat 47 kΩ, which is low enough to be a practical problem with most moving-magnet cartridges. The arithmetic is below.

Four more candidates were resolved and dropped on the same day: the Pro-Ject Tube Box DS2 (every hit that cleared the model tokens was a power supply for it, so there is no verified listing), the Fosi Box X5 (its only hit was caught by the accessory guard, and a second Fosi box in a three-pick roundup would be padding rather than choice), and the Little Bear T11 and Nobsound tube phono stages, where no returned title cleared the model guard at all. A short roundup is better than a padded one.

Should you buy a tube phono stage at all?

On measured performance, no. Pro-Ject’s own numbers make the case against its own product: the solid-state Phono Box S2 Ultra publishes the same ± 0.4 dB RIAA accuracy, the same kind of adjustability and about 1/38th of the distortion, for less money. Every stage on this page is beaten on paper by a transistor box at or below its price, and the under-$200 bracket is full of them.

Buy a tube phono stage if you want a tube in the phono path and you want to be able to change it — that is a legitimate thing to want and nobody needs to justify a preference with a measurement. Just buy it knowing which number you are trading away, and do not expect the valve to make up for a loading mismatch or a loose playback curve. If the appeal is valves in general rather than valves in the phono stage specifically, the money usually goes further one box downstream: a tube amplifier for a turntable puts the valve where it is doing the most work.

And whatever you buy, the speakers see the result. If this is part of a larger upgrade, our bookshelf-for-vinyl roundup and the high-end bookshelf tier are the two ends of where that signal ends up.