Statement amps are nothing new to Tim de Paravicini, his original EAR 509s and 549s hardly being minor projects. Neither were the first amps to bear the Yoshino names, a pair of designs showing that single-ended topology could be applied to both valves and transistors. But the latter, while Tim was able to voice it to ape the tubes, ran just as hot as the valves, and proved to be just as inefficient as all single-ended designs are known to be, so why bother? Damned if I know, because Tim's latest is a huge mutha of a single-ended amp...and it's solid state.

As 'single-ended' remains a buzz word that won't go away, and because there are those (such as I) who cannot live with, say, 8W, there's probably a ready-made audience/market for the M100A 100W monoblock. In this case, Tim says that the design is exactly as one would apply to valves, only using solid state devices, and incorporating an overkill power supply which contributes to a weight of 70kg per chassis, or 90kg in the crate. It is, unusually for a solid-state amplifier, transformer-coupled at the output, but this is necessary in true single-ended designs, especially as regards the inefficiency of pure Class-A, single-ended operation.

Tim states that the circuit design is 'modular' and can be scaled to any power requirement by redesigning the transformer. The M100A's output devices are individually biased and driven MOSFETs, the arrangement compensating for any MOSFET-to-MOSFET sample variation; moreover, the failure of any single MOSFET would be 'unlikely to precipitate the failure of the whole amplifier.' Only 15dB of feedback is used.

Each unit occupies 240x500x570mm (WDH), so the small footprint is little consolation. You can't miss this in the room. But forget the size: it's the styling. Looking exactly like something from one of those early 1960s steam-punk movies based on Jules Verne novels - think , with Vincent Price - the look of the M100A is best described as 'Victorian'. The meters - the top for output and the small one at the bottom monitoring bias - are simply gorgeous things of Nagra calibre, while the mix of natural metal and gilding is, er, quaint. My only tip would be to ditch the two 'balls' which split the wings on either side of the fascia, and make them of a single piece. Tim actually agreed, admitting that they'd cost less to produce, too...

The sides are pure heatsink consisting of thick aluminium sidepanels and 'chimneys', their sharp edges to be smoothed in future. At no point, regardless of the abuse during the sessions, did the M100As shut down or immolate, nor did I burn any digits. The top of the unit is made of mesh, to further aid cooling, and to provide a view of what looks like a transformer class reunion.

Operationally, the unit contains a front on/off switch and a volume control - set to maximum when used with the 312 pre-amp. At the back are both balanced (XLR) and single-ended (phono) inputs and a toggle to select between them, and an array of multi-way binding posts. The obvious connections are for 4, 8 and 16 ohm speakers, but Tim connected the 1 ohm Apogee Scintillas using an obscure combination of terminals ('something I chose to leave out of the owner's manual'). Clearly, he does not want to encourage experimentation.

Equally fascinating is the 312 preamp, already noted for its massive meter, which is actually a bit of a design conceit. This is not an 'active' meter, but a device mechanically connected to the level control to allow you to gauge the volume setting from across the room - far easier than a tiny tell-tale on the knob and, as Tim swears, 'more accurate than the big digital read-outs used by XXXX XXXXXXXX.' Across the panel, between the signature wings, are a rotary source select, two press buttons for tape and mute, the level control and a big orange on/off button. The operations, including source select, are accompanied by illumination below the arc of the meter.

Around the back of the 18kg, 432x305x140mm (WDH) unit, Tim has fitted sufficient single ended and balanced socketry to satisfy the arch audio lunatic, including three balanced line inputs and four single-ended line inputs plus phono, as well as three pairs each of balanced and single-ended outputs. Tim added these so that '...users could tri-amplify if they wish. You know the Japanese...' The 312 is transformer coupled throughout, each and every input and output included bar the tape monitor, to ensure freedom from hum and noise. And the 312 is quiet indeed, even for phono usage with the usually cranky London.

Yes, the 312 includes a full, pure Class-A phono section, with an m-c step stage equal to the Yoshino MC3 transformer. It is based on a three-transistor circuit with a 'unique RIAA equalisation network which gives excellent stability, high headroom, low noise and low distortion,' with RIAA accuracy stated as +/-0.3dB. Moving magnets connect directly to the input transistor, while moving-coils connect directly to the step-up transformer. The inputs are selected via relays, and Tim designed signal paths to be as short as possible. Volume is via a four-gang potentiometer, which has a terrific, Leica-esque 'feel', for those to whom tactility is crucial.

Other specifications of the 312 include a frequency response (at 1V output) of 3Hz-40kHz, -2dB, line sensitivity of 100mV with a maximum gain of 20dB and mm phono sensitivity of 3mV for 1V output. Channel balance is +/-0.2dB and residual noise (with volume at minimum) -90dB; signal-to-noise ratios for phono is 68dB (ref. 2.5mV).

We set up the system in fully-balanced mode, using the fixed level outputs from the Krell KRS25sc CD player, while the phono section was sampled with the London Gold/Decca International/Garrard 401 set-up for mm, and the Transfiguration/SME Series V/SME 10 rig for m-c. Balanced interconnects were Kimber's best, while speaker wire was courtesy of Transparent, for the variety of speakers listed in the sidebar "By The Numbers". I performed most of the listening with the Sonus Faber Guarneri and Apogee Scintilla.

Aside from listening for clipping - again, see the sidebar - the experience was, from the outset, overwhelmingly familiar. It was a pure EAR experience, Tim de Paravicini being foremost among those designers who seems incapable of stamping his identity on the sound. It's an experience which, for me at least, goes back to the TVA1 and the EAR 509, and must be admired for its consistency.

Before reading this, note that I am not ascribing these qualities to Tim as a person. I would hardly call him a 'control freak', even though the bass and drum sounds he favours are so stage-managed as to suggest a martinet with a whip standing over a battered and bleeding Jack Bruce. In my experience, the TdP sound is bold, up-front, dynamic and - to use modern parlance - kick ass. The sound is never shy, never lacking confidence and always close to the edge. If this combination were a car, it would be a TVR, not a Lexus. It is the kind of system which imparts excitement to almost everything - although that would be impossible to do to works by Andrew Lloyd Webber - and, as a result, calls for a red-blooded listener.

Everything about the sound of the combination is big and bold, a bottle of Brunello amidst a case of Evian, with sound so vivid that it can only be likened - sonically - to Decca cartridges, classic Lowther horns, dual-concentric Tannoys. While the M100A is capable of delicacy, if not quite the sublime finesse of the partnering 312, it almost begs to be fed a diet of huge works. It's as if to use it with chamber music would be to waste it, like pouring that Brunello into a casserole.

If the lower registers - the tight and dry bass, the bottomless extension - are as predictable as two-decades-plus of Tim's designs would suggest, then the rest is something of a surprise. While the top end is easily excited to near edginess, one suspects that it is the 312/M100A combination stretching the speakers and cables to their limits, rather than the amplifiers themselves misbehaving. Yes, there is a constant sense of a pending eruption, but it never arrives. I suppose it's as close as audio gets to a ride on EuroDisney's Space Mountain. You think you're gonna go off the rails, but you don't.

If ever a system begged to be fed the Kodo drummers, this is it. The Yoshinos portrayed them with all their power, all their mass, all of their impact. Conversely, a session with the Persuasions - unamplified voices - proved just as rewarding because the scale, the three-dimensional layering and therefore the sheer presence overcame a slight loss of textures in the deepest voices.

At no point was I bored using the 312/M100A combination, and I found myself actually looking forward to the listening periods, to the looks on friends' faces. And each face produced a rictus upon being told that the 312 cost £10,000 and a pair of M100As would set you back £20,000. This, of course, is loony tunes money, almost enough to pay for a decent tourbillon, and the competition is so fierce at this rarefied level (lots of products/few customers) that it's a buyer's market. There are a lot of dandy amps out there for twenty big ones.

Hand on heart, I have to say that the departure of the 312 - which I adored to pieces - caused me a greater sense of loss than the M100As, but their absence, too, causes a lingering wistfulness. Quite what has happened, I don't know. I suspect it's like being reminded of one's youth as one enters one's decline; the temptation is quite clearly the hi-fi equivalent of those sad fiftysomethings who run out and buy Harley-Davidsons or Ducatis the instant they realise that they can no longer keep it up all night without the aid of Viagra or scaffolding. So maybe the 312/M100A is a young man's plaything, because it is most surely an invitation to bang head, listen loud, and leave the session drenched in sweat.

This is, unashamedly, the most butch combination I've heard in a long, long time. Alanis Morissette fans need not apply.

EAR/Yoshino Ltd., Unit 9, Brook Road, Bicton Industrial Park, Kimbolton, Huntingdon Cambs PE29 0LR. Tel 01480 861834; FAX 01480 432006
earyoshino@aol.com
www.ear-yoshino.com

SIDEBAR: BY THE NUMBERS
It was with some concern that word reached me about the EAR M100A, prior to the amps even arriving Chez Kessler. Fair enough, Tim de Paravicini has antagonised the odd individual in the past, but to learn that some were attributing to the mighty M100A an output of only 10W before clipping struck me as particularly bizarre. Tim may be a lot of things, but a purveyor of gutless amps he is not. More to the point, I refuse to believe that any amplifier which can drive the Apogee Scintilla - in 1 ohm mode to some 102dB at 2m with no audible distortion - to be capable of yielding only 10W. And in normal listening with other speakers, the meters on the front of the amp were of little use, as I barely got them into the red, i.e. over an indicated 10W, because it was simply too loud for me.

Accuse me of lacking faith, but merely driving the 1 ohm/82dB Apogee, the 6 ohm/86dB Wharfedale Diamond 8.1, and the 8 ohm/85dB Spendor S3, plus original Quad ESLs, assorted LS3/5As and the Sonus Faber Guarneri to levels which nearly threatened their existence wasn't enough evidence. Even though a simple calculator and an SPL meter would prove the delivery of more than 10W, still I confronted Tim and had him bring along, when he collected the monsters, a slew of test equipment.

Tim duly arrived with an oscilloscope, signal generators, a 200W/8 ohm dummy load, a mouthful of names like 'Flume' and 'Neutrik' and fistfuls of cables, in full, mad-professor-from-Back-To-The-Future-Mode. Arms flailing, a dervish in full flight, he hooked it all up, twiddled with knobs, fed it some signal, cranked the loud button all the way around to Armageddon. The result? Before my very eyes, 107W before any TRACE of clipping. The lesson? Where there's smoke, there's not necessarily fire.