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
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.Continue on to Page 2 for more about the EAR Paravicini M100A/312 tube amp.