Product: Phono 2 Ci
Manufacturer: AQVOX - Germany
Price: 700€
Reviewers: Mark Wheeler & Werner Ogiers
Reviewed: April, 2006

Mark Wheeler writes

I have already wired my Hadcock GH242SE in balanced mode. It is balanced even when fitted with those truly horrible RCA connectors so contrarily loved by the domestic audio business. The Hadcock, thus wired, keeps the true earth separate from the signal return. Thus, playing with the Aqvox 2Ci is a doddle. At first I try it with my favoured rebuilt Decca fitted with a Len Gregory extended contact diamond. First impression is of a very lean sound, the White Stripes Digital is Evil sounding like the excesses of early CD. I worry that 2 days was inadequate burning-in with nothing more elaborate than a bit of wire to generate hum (I do not have an inverse RIAA filter to feed a RIAA input properly). Days of playing later I am realising that this is not the whole story.

Musical qualities of pace, pitch, rhythm & timing are all intact, indeed among the best I have heard from a moving magnet input. I am reminded of my modified Naim phonostage. The music bounces along and the traditional Naim forwardness is also present but there is much more to this. While I struggle to decide whether it is forward upper mid-range I am hearing, or a bass-light balance to exaggerate pace and speed, I begin to notice other qualities as I acclimatise to the balance.

The higher registers are spectacularly clear without ever becoming too much. Aqvox designer Carlos Candeias has chosen to adopt the Neumann variation on the usual RIAA equalisation. This has an extra pole, usually at 3.18uS to equalise for the 50kHz roll-off introduced by Neumann to reduce cutting head failure rates. This is well described by Allen Wright. The effect of this deviation from the playback standard is to extend the high frequency response of the cartridge-headamp system but also to reduce the out of band phase-shift, resulting in improved high-frequency timing. Even though the action is an octave beyond middle-aged steady-state audibility, the effect is similar to adding a supertweeter.

The Aqvox 2CI looks like a pro-sector product, sized very much like a 1U rack item, and their website lists a rackmount option. The XLR connectors reinforce that impression. Lifting the cover reveals a well constructed board populated with good quality parts and a modest power supply 40mm from the edge of the audio board. The gain blocks are (presumably) regulated by LEDs making this look as good as a valve pre-amp with the lid off. The sound opens up slightly with the lid off, as is a common experience with aluminium (but better than steel)casework. If I were selling this product to the domestic audiophile I'd fit a tinted perspex top to show off the innards.

The professional audio look inspired me to begin listening with balanced high-quality shielded multistrand cable output, which generally suits pro sources better than audiophile single strand. Substituting a single ended (shield connected at only the Aqvox end) phono-cable immediately improved the frequency balance. Typically the cable choice is dictated by the send-end.

Fortunately, while I had the Aqvox 2CI in my system, Bill Dyer arrived with his Digital Audio Systems Harefield active monitors. These have a more pro-sector balance and match the 2Ci perfectly, and will be reviewed on TNT-audio soon. This symbiosis in my listening room may seem fortuitous, but C G Jung might have suggested it is the synchronicity of the audio world collective unconscious at work. Bill Dyer is also on a mission to introduce accurate high-frequency phase to the domestic audio world, and this attribute of his speakers works wonders with the clear extended treble of the Aqvox 2Ci.

Sadly the various input impedance combinations available for the RCA inputs of the Aqvox 2Ci do not include the 30kohm that suits the Decca better than 47k. I try the Cartridge Man MusicMaker II but it too has the apparently lean balance, or forward upper-midrange I experienced with the Decca. This balance is mitigated by Bill Dyer's Harefields, but would not suit the majority of typical small modern domestic speakers. This is not a criticism of the Aqvox, more a comment on the lean balance of recent loudspeakers ever since the ultra-rich sounding Linn Sondek defined front-end expectations in the '80s.

Using my variable transformer as a passive volume control feeding the active DAS Harefields via 7m of balanced cable, the only single ended connection is the output of the Aqvox 2Ci to the transformer primary. Graham Nalty of Black rhodium was good enough to loan me 700€ worth of balanced Opera DCT interconnect fitted with high quality XLR plugs. I have no idea what the construction is but this combination takes another step closer to the music. The balanced output works better than single ended in my system, once the right wire type had been identified.

Rewiring my arm leads to XLR connectors I can now try the Aqvox 2CI trump card. The XLR input is a fully balanced current amplifier. Because it uses the cartridge current swings to modulate the gain stage its impedance should be irrelevant. I dig out one of my most awkward old cartridges to test this theory. This Linn Asak T-DC2100K (dc to 100k?!? dates from Ivor's finest hour of hype, bless 'im) is based around Supex innards and I recall I spent many hours in the 80s trying different input resistors and found 470 ohm worked best. The Aqvox RCA input can switch between 100 & 1k but nothing in between. What will the balance be like I wonder as I crimp the armlead terminals over the Linn's non-standard undersized four-in-a-row pins?

As good as the Naim 'S' boards designed for this cartridge is how the Aqvox sounds. Digging out the Nimbus Supercut of Little Feat's Feats don't fail me now that spun many times on my old Linn under this cartridge the sound is right there. The Asak was notorious for a spitty quality with the wrong moving-coil input; it could upset transformers and transistors alike. It doesn't upset the Aqvox, even when I hurl weird live recordings of Captain Beefheart's Magneticism. The final Asak test is Latin Quarter Modern Times, with its nasty sibilant vocal microphone EQ, and the Aqvox tames the Asak's wilder tendencies without masking any of the musical substance.

A more typical selection of old Ortofon, Audio Technica and something whose identity I've forgotten all give of their best into the Aqvox 2Ci. The balanced current sensing input obviously works for all my low output moving coils. I do not own any high output (above 2mV for 5cm/S) moving coil cartridges since my Sumiko Blue point Special broke, but I'm curious what might happen with a balanced output cartridge of moving iron persuasion. My Decca gets refitted to the Hadcock GH242SE, where it tends to reside most often. The Decca generator is unlike either conventional moving-magnet nor moving coil; it generates its output in lateral and vertical coils whose outputs are sum-&-difference wired to create left & right channel outputs. It is fully floating with respect to earth so is as 'balanced' as a moving-coil.

The Decca London does not give of its best into this input. It sounds slightly similar to the RCA input even though the configuration is completely different. The Decca is famous for its weirdness and so untypical that its mismatch is no criticism, but illustrates that no generator input (cartridge or microphone) can be truly universal, and for acoustic blues I'd choose the Decca feeding the Concordant Excelsior for scary delusions of realism.

Speaking of acoustic blues, the final curved ball thrown at the Aqvox 2Ci phonostage is the MusicMaker II into the XLR input. My hunch is that this shouldn't work. VTA just gets set cartridge top level with vinyl surface on thinnest 70s disc (dynaflex style) so tail-down on the Anadisc 200g Original Master Recording that next lands on the Michell Orbe SE platter: Muddy Waters' Folk Singer. The frequency balance is as one might expect from a moving magnet fed into a low Z input: deeply scooped midrange (the antithesis of the correct adjusted load RCA input) and muted highs...but despite this, the ambience retrieval is spectacular, suggesting good HF phase response despite the wrong resistance and capacitance to match the cartridge. The ambience is very well portrayed, surrounding a soundstage of average dimensions. Return to the low Z mc; the balanced input doesn't suit the mm MusicMaker.

This superb ambience is even present with mono recordings (as it should be) and the illusion of space was uncannily similar to the all-valve Concordant Excelsior. The accurate frequency response and phase performance are especially obvious with mono recordings; the central image remains accurate and in proportion, but the sense of space is present in the reverberation and flutter echoes.

Mark's Conclusion

The Aqvox 2Ci is a remarkable accurate phono amplifier, even more remarkable at a shade under 700€. The Dyer DAS Harefield active loudspeakers were a fortunate inclusion in my system while the Aqvox was performing head-amp duties. The Harefields allowed the Aqvox 2Ci strengths to blossom, where more typical modern domestic speakers might have masked the strengths while emphasising the lean balance. The Aqvox heavily regulated power-supply and modest power-supply capacitors keep the pace up with the quickest, but might contribute to the lean effect. The bass response sounds accurate but is qualitatively different from the familiar audiophile presentation, which can be offputting at first.

Musically, rhythms are flowing well and timing is impeccable right up to the limits of high-frequency audibility. Ambience is very well portrayed and full-bandwidth phase accuracy maintains soundstage proportions from low to high frequencies without inappropriate instrumental wanderings.

The Aqvox 2Ci doesn't pretend to the full gloss of high-end couture, and occupies a price point in the middle of the sector. It is much more accurate than many budget (sub 200€) products, and almost as transparent as more expensive products. This is a very accurate and versatile product and one I would recommend to anyone with a drawer-full of cartridges.

Mark's listening system turntable: Michell Orbe SE + Hadcock GH242SE cartridges: Decca London with Cartridgeman stylus, Cartridgeman Musicmaker II, Linn Asak, unspecified Audio Technica and Ortofon low-output moving coils phono preamps: Naim 'S' boards, Concordant Excelsior line preamp: transformer volume control active loudspeakers: Digital Audio Systems Harefield cables: Black Rhodium Opera DCT interconnects

Get a LEF! The Phono 2 Ci is built on two circuit boards. One contains a heavily-filtered switching power supply, the other the amplifier proper. Not too much is advertised about the circuit, so what follows is based mostly on my guesses. I feel quite confident about them, though. This phono preamplifiers seems to consist of three stages: two LEF amplifier modules and then an output buffer. Inter-stage coupling is with capacitors; there is always some subsonic filtering present, but an additional relay-controlled subsonic filter can be switched into circuit. (An older version of the 2 Ci used an opamp-based circuit for this filtering; this circuit was early 2006 abandoned for simple C-coupling, as this was found to sound better. At the same time the phonostage's gain was increased. The lower-gain opamp-filtered Phono 2 Ci was reviewed in Hifi World and LP Magazin.) The internal build quality and component selection is nice, but nothing special. The signal capacitors are said to be of the polypropylene film variety, but the MM-section cartridge loading caps are cheap and nasty ceramic disc types. (Apparently less of a problem than one might think: the good-sounding Trigon stage also uses these.) The LEF amplifier modules are Candeias Engineering CC80s (see here for a datasheet), resembling Marantz's HDAMs in shape and size, but then of red plastic instead of copper. The LEFs are plugged into the main board, and the area beneath the modules is itself filled with passive components, possibly for the RIAA correction. Each CC80 constitutes a balanced-in-balanced-out gain cell and is fed from, I think, a +22/-22V supply (I measured this, but forgot to write down the actual value). The modules employ some tens of surface mount components, including bipolar transistors, resistors, and capacitors. Each CC80 contains the actual amplifier, in addition to two voltage regulators and one or more DC servo units, making these modules really self-contained. The amplifier circuit is dual single-ended (dual, since balanced), without loop feedback, but with programmable local feedback, and using folded cascodes in the gain stage(s), presumably with emitter followers as output buffer. The folded cascode technique is relatively little used in circuits lacking loop feedback, but related art can be found in John Curl's Vendetta, the Pink Triangle PIP II, and my own JFET phonostage prototype. It is a circuit topology that, to me, has two major strengths: first, its current consumption is constant and entirely independent of the signal amplitude, which makes it very easy on power supply regulation, and second, as the gain is realised in a transconductance stage feeding an impedance, the RIAA de-emphasis can be built easily into that impedance. This makes it an intrinsic part of the amplifier circuit (as opposed to an add-on), and yet without any problems of input stage overload (the bane of traditional passive RIAA solutions). The first CC80 of each channel interfaces with the cartridge, with user options of connecting to the input transistors' bases, for standard voltage mode, or to their emitters, for current mode. These emitters each carry a -500mV potential, which is directly connected to the cartridge in current mode. This is not a problem, as the in-built servo mechanisms guarantee that all cartridge pins 'see' the same voltage, so that no current flows and hence no damage is possible. Even during power-on or power-off does the cartridge interface remain clean and safe. Hats off to AQVOX! (Still, should there ever be made accidental connection from a cartridge tag or tonearm hot/cold terminal to ground or to the unit's housing, then a - limited - current will flow where it should not and there is a slight chance that the cartridge might be damaged in the process. So beware, don't take any unnecessary risk, and be careful when making connections, especially with RCA-to-XLR cheater cables.) The second LEF is a gain stage with the input transistors' emitters linked to the front panel gain potentiometers. This allows gain control by modifying the stage's local feedback, instead of the usual attenuation (good!). Even so, these potentiometers are still in the signal path, and I don't like the long traces linking these pots to their circuits! The RIAA correction is passive, but it is not clear to me whether it uses one or two stages, and if it is interwoven in the folded cascodes (preferrable) or just inter-stage. (I suspect two-stage, interwoven.) With all those operating modes and alternative inputs there is, understandably, a lot of housekeeping in this phonostage, most of it neatly executed with relays. And yet, one wonders how much of a compromise this is, and how much better still a pure-CI pure-balanced Phono 2 Ci would sound. I would like AQVOX to consider 1) a more purist version of the same circuit, and 2) perhaps an even more elaborate version for the pro sector, including variable equalization for older pre-RIAA standards and for tweaking the sound during transcription to digital. As the accompanying documentation and manual are not entirely clear on this (I urge AQVOX to update their documentation!), I measured the Phono 2 Ci's gain in several modes. Unbalanced RCA-in to RCA-out, voltage mode, and the front panel gain controls turned to Min, the gain in "0dB", MM mode is 36.3dB. Using the "+6dB" switch brings this to 41.8dB (a 5.5dB difference). Engaging the "+20dB" switch yields 51.7dB, or 15.4dB difference. Turning the gain controls to Max then adds another 13.7dB. So the maximum gain available in unbalanced voltage mode is about 65dB. Using balanced-in, unbalanced-out, current mode, and testing with a DL-103 cartridge the gain of this particular combination, knobs set to Min, proved to be 57dB. From the listening tests emerged that channel balance was excellent, so I did not bother measuring it. Comparing to some other phonostages (Dino, a heavily silenced Rotel RQ-970, and a homebuilt Naim clone) indicated that the Phono 2 Ci is somewhat more sensitive to radiated disturbances than usual: I found it picking up more 50Hz and HF garbage than the others in the all-too-noisy lab corner of my home office. This happened both with the voltage input and the current input, and seemed to depend somewhat on the tonearm cable quality or geometry, an Incognito-wired RB-300 clearly bettering the SL-1200's stock wiring. I did not run a frequency response test, but an older version of the Phono 2 Ci was measured in the German magazine "LP", the review being available on AQVOX's site. The plot in that article reveals a RIAA curve doctored for a sweet sound, with a hump between 60 and 800Hz, and a depression from 1 to 15kHz. The Phono 2 Ci deviates from standard RIAA in the treble, having +0.5dB at 20kHz. This is done to compensate for the frequency response of real-world cutting machines, which can't keep boosting treble out to infinity and hence have to roll off somewhere above 20kHz, violating the RIAA 'standard' in the process. It seems wise to compensate for this deviation during replay, as Vacuum State (Allen Wright), Graham Slee, Whest Audio, and an army of DIYers found a long time ago. The silly thing now is that often the proposed replay compensation entails no more than a single zero at 50kHz, while e.g. Neumann's cutting curve happens to be treble-limited with a second-order Sallen-Key filter (-3dB at 33kHz in the old SX-66, 50kHz in the ubiquitous SX-74 cutter). Replay with a single-order zero at 50kHz then redresses the phase response somewhat, but leads to excess treble between 10 and 20kHz! Have a look at the above figures, which are simulations of a Neumann SAL-74B cutter amplifier followed with standard RIAA de-emphasis (red) and single-zero-50kHz-compensated de-emphasis (blue). The standard equalisation has a near-flat amplitude response out to 20kHz, but suffers over 25 degrees of phase shift at this frequency. The 'compensated' replay equalisation shows a slowly rising treble, but has less phase shift. (The green curves are for a single-zero 100kHz compensation, something AFAIK no-one uses.) In the end neither solution is right or wrong, although it must be remarked that the treble peaking in the compensated version might offset some of the treble losses inherent to vinyl's crude manufacturing process. Nevertheless, the correct replay of LPs remains an illusive target.

[Back to Part 1]