Continuing to revert exclusively to vinyl usage for my 'pleasure', as opposed to 'reviewing' listening, I'm ecstatic about the flood of killer phono stages available today. My absolute references, but beyond my means, are the Manley Steelhead and the Audio Research Reference Phono, while I adore the EAR 324 for its unbelievable flexibility and quietness - a masterpiece. EAR's 834P remains my sub-£500 all-tube reference, while I can't recommend too highly either the Pro-Ject Tube or the NAD PP2 for sheer value for money.

But it's time to get selfish. I try to treat myself to a piece of hi-fi each year, having bought London and Koetsu cartridges, the Denon DVD-2900 universal player and other bits and pieces since the change of millennia. Now I'm after a phono stage with three specific requirements, beyond sublime sound. They are, in ascending order of importance, 1) a price I can manage, 2) valves inside and 3) ultimate suitability for the two cartridges I love best: the aforementioned Koetsu and London. Which makes it so tough, because one's an MC and the other's a high-output freak of science.

Enter a new phono stage from Audio Research, with trickle-down technology from the REF. ARC explained the arrival of the PH5: 'The PH3 series had been in our line some nine years and needed replacing. Discussions with our retailers and importers indicated that a well-performing product around US $2000 was what they and their customers were looking for. So, our task was to give them a phono stage that was a clear sonic advancement upon the previous products, with enhanced features, a little more gain (+3 dB), at a lower price (some $600 lower than the previous SE). Thus the PH5.' Not an easy task, as the PH3 and PH3 Special Edition have been major successes for ARC, and had remained unchanged, for nine and six years, respectively.

Their response? The new PH5, replacing both the PH3 and PH3SE, with a retail price of £1799. Its 19in front panel layout matches the SP16 and CD3 Mk II, with green LED indicators on the left side and four soft-touch buttons on the right. The buttons control Power, Mute, Mono and Loading for 47K, 1000, 500, 200 or 100 ohms through press-press scrolling. Here's the irresistible kicker for analogue addicts: the PH5 comes with a remote control for all functions, so you can listen to the adjustments from your hot seat. [But see the sidebar, 'To Load Or Not To Load'.] Handles are a £70 option.

PH3/PH3SE owners will recognise the layout of the rear panel, with one pair each of phono-socket inputs and outputs, an earthing post and an IEC three-pin mains socket. Alas, due to the prohibitive cost, balanced output is not available, but ARC will consider it for a dearer variant if there's enough demand.

Like the PH3, the PH5 is a hybrid, using a high-gain, 5-JFET per channel, non-inverting input stage with constant-current stabilisation and no overall feedback. A direct link to the flagship REF is new RIAA circuitry, patterned after the Reference Phono's RIAA stage, with passive high frequency and active low frequency equalisation. ARC believes that this results in an improved margin of stability under the most demanding conditions.

Four 6922 twin triodes make up the non-inverting gain and output stages, with high-stability power supplies and four separate regulators for all valve heaters, plate supply and microprocessor logic. Bandwidth is 0.7Hz-400kHz (-3dB), while gain has been increased to 57.5dB. The latter was a revelation: for the first time, I had to trim the phono stage output down to match the line sources. The PH5 pumps out plenty of signal, resulting in a lowered noise floor for most installations. And it was consistent with the McIntosh C2200/MC2102 and Musical Fidelity kW combinations, and the PrimaLuna Prologue One, Audio Analogue Maestro and Audion Lo Sfizio integrateds. I doubt that even an Audio Note Io or early Ortofon SPU could stump it.

Regardless of how hard I made the system work - Apogee Scintilla, Wilson WATT Puppy 7, LS3/5A, Sonus Faber Guarneri - and irrespective of my hunger for level, the PH5 never failed to deliver sufficient oomph. Main listening involved four wildly differing designs fitted to the SME V/SME 30/2 front end: the London SuperGold and Grado Prestige MM/MI/MF cartridges, and Koetsu and Transfiguration MCs.

No question: the PH5 met all three of my criteria, caressing the Koetsu and preserving its warmth and sheen, while opening the window for the peerless frequency extremes of the London, especially the snappy bass. Treble? Crystal clear and - above all, when you're dealing with Londons - scarily fast, the transients snapping without smearing, without hindrance. With so much gain, background noise was never an issue. With so much transparency, everything else profited.

It was a litany of superlatives: massive soundstage, wide and deep to the point where there was no doubt of the presence of specific sounds beyond the outer edges of the speakers. Front-to-back depth turned into an added benefit especially for mono listening - all those wonderful Capitols and RCA and Columbias I've been enjoying - so now I'm on the hunt for a mono cartridge to maximise the experience. And the vocals! Natural, detailed, with sibilance that sounded really rather than spitty. I implore you: audition this with some Peggy Lee!

But there is one area where the PH5 nears the REF Phono and the Steelhead at three or four times the price: retrieval of low level detail. I simply heard 'stuff' that I hadn't before, including the very artefacts that make a listening experience more real, regardless of the recordings' age or mono/stereo status: Keely Smith's breathing while Louis Prima was mugging about, greater separation of voices amongst the Hi-Los and the Crew Cuts, more subtlety in the harmonic overtones on acoustic guitar (try Doc Watson on Cisco), dazzling tastes of mallet and drumstick and pedal in well-recorded percussion tracks. Microphone fetishists will want to dig out test discs that tell whether they used Neumanns or AKGs.

Simply the most 'real-sounding' phono stage I've experienced below £2000, the PH5 is a natural upgrade from the delightful EAR 834P. My only gripe? I can't choose between the Koetsu and the London, and the PH5 only has one set of inputs....

Had I more patience, I might have waited before buying, for there are two phono stages due soon that should also tempt me: EAR's valve version of the 324, and the long-awaited all-valve Quad phono stage to match the QC24, which I've used in 'beta test' form - a stunner at under £1500, believe me. But I've always had a soft spot for Audio Research. But now I actually own a piece.

Absolute Sounds 020 8971 3909

SIDEBAR: To Load Or Not To Load
There's nothing as rewarding as learning something new, and the PH5 completely altered my understanding of moving-coil cartridge loading. Like most audiophiles who lived through the Great Moving Coil Renaissance of the 1970s, I thought of loading to be as important as driving sober, safe sex and paying VAT. Boy, was I wrong. What happened was this:

After carefully installing the PH5, with the Koetsu Urushi in place, I was not merely surprised but distressed to hear absolutely no level changes whatsoever when scrolling through the PH5's impedance settings. I freaked - was it a dud? So I entered a phone-and-e-mail exchange with Terry Dorn of Audio Research.

He told me, 'Ken, don't worry: so did I. Our techs explained that with low-impedance cartridges, the actual impedance differences between 100 and 47k ohms work out mathematically to be almost negligible. However, with a cartridge having an internal impedance of, say, 500 ohms, the difference becomes much more audible. "On a higher output cartridge," one of our techs explained, "the different loading options will act like a very subtle tone control, mainly affecting the high frequencies and reducing overall volume slightly as you load the cartridge down."

'At most we're looking at subtle differences, not radical ones, and it is possible that, depending how the record was recorded and cut, one setting may be preferred and on another record another loading option might be preferred. We have also found that some cartridge manufacturer's loading recommendations do not accurately reflect the actual internal impedance of the cartridge. But, again, they may feel that most users will find their recommended loading the best-sounding option for most systems and most recordings. So, as it is with so many things in this industry, cartridge loading is a combination of measurement, subjective preference and system synergy. Not a "black art", perhaps, but certainly as much art as science.'