Srajan Ebaen
Financial interests: click here
Sources: 27" iMac with 5K Retina display, 4GHz quad-core engine with 4.4GHz turbo boost, 3TB Fusion Drive, 16GB SDRAM, OSX Yosemite, PureMusic 3.01, Tidal & Qobuz lossless streaming, COS Engineering D1, AURALiC Vega, Aqua Hifi Formula, Fore Audio DAISy 1, Apple iPod Classic 160GB (AIFF), Astell& Kern AK100 modified by Red Wine Audio, Cambridge Audio iD100, Pro-Ject Dock Box S Digital, Pure i20, Questyle QP1R
Preamplifier: Nagra Jazz, Esoteric C-03, Vinnie Rossi LIO with DHT module, COS Engineering D1, Wyred4Sound STP-SE Stage 2
Power & integrated amplifiers: Pass Labs XA30.8; FirstWatt SIT1, F5, F6, F7; S.A.Lab Blackbird SE; Crayon Audio CFA-1.2; Goldmund Job 225; ; Aura Note Premier; Wyred4Sound mINT; Nord Acoustics One SE UP NC500MB; AURALiC Merak [on loan]
Loudspeakers: Albedo Audio Aptica; EnigmAcoustics Mythology 1; Sounddeco Sigma 2; soundkaos Wave 40; Boenicke Audio W5se; Zu Audio Druid V & Submission; German Physiks HRS-120; Eversound Essence, Rethm Bhaava [on loan]
Cables: Complete loom of Zu Event; KingRex uArt, Zu and LightHarmonic LightSpeed double-header USB cables; Tombo Trøn S/PDIF; van den Hul AES/EBU; AudioQuest Diamond glass-fibre Toslink; Arkana Research XLR/RCA and speaker cables [on loan]; Sablon Audio Petit Corona power cords [on loan], Black Cat Cable Lupo
Power delivery: Vibex Granada/Alhambra on all components, 5m cords to amp/s + sub
Equipment rack: Artesania Audio Exoteryc double-wide 3-tier with optional glass shelves, Exoteryc Krion and glass amp stands [on loan]
Sundry accessories: Acoustic System resonators
Room: Rectangular 5.5 x 15m open floor plan with two-storey gabled ceiling, wood-sleeved steel trusses and stone-over-concrete flooring
Review component retail: €4'990/pr incl. 19% VAT

If one had desire for an ultra-wide bandwidth DC-coupled amplifier, until now it could have meant very serious Swiss franks for a CH Precision, Goldmund or Soulution model. Or it could have meant a mere fistful of watts from a Bakoon AMP-12R. But what if 55 watts into 8Ω were your personal G spot, even more so if they doubled down to 2Ω, with 4Ω in-between? Say hello to the LinnenberG Allegro monos from Germany. How ultra is their bandwidth? 1MHz, with a rise time of 700ns. To avoid stability problems with partnering gear, the i/o filters on either side of this circuit set a full-power limit at 350kHz.

With his former pro-audio background, Ivo LinnenberG's credo is dual differential. Whilst the Allegros have switchable RCA inputs, their XLR are the preferred entry. His output devices of choice are the legendary Hitachi lateral Mosfets of 1977, the only transistor type to ever be made exclusively for audio purposes. In purist mode, that means a single complementary pair; what some clever marketeers elsewhere call single-ended push/pull because two transistors/tubes per channel are the absolute minimum for such a circuit. To achieve an honest 20-bit analog resolution (anything much beyond that disappears in the Johnson/thermal noise of resistors), the Allegro's A-weighted noise is 25μV, thus the equivalent of 120dB of S/N ratio. Attendant harmonic distortion is just 0.004%. The power supply runs off 60.000μF of storage capacitance and a 230VA mains transformer. Input sensitivity is a usefully low 1.5Vrms. Your preamplifier's volume won't max out at barely there. DC offset, thermal and over-current protection are via non-intrusive microprocessor. The astonishing thing? All this packs into a case too small to accommodate the 63A terminal posts, input sockets, power IEC and heat sink on its rear. The cooling ribs thus relocate to the front. As a class A/B circuit, nothing bigger was required. To me, the whole Allegro concept spoke loudly enough to request a review loaner pair. Let's reiterate: petit very practical form factor; reasonable not overkill power with low-impedance stability; top resolution (low noise + low distortion); no signal-path capacitors; no phase shift from insufficient extension (at least 10 x the audible bandwidth, here 0.3Hz-350kHz); high transient speed; high drive current. As Ivo puts it to those who'd correlate small size with small ambitions, "clever construction enabled us to build an amplifier with high packing density, making its signal and power paths extremely short. Creating circuits with unrivalled speed and response time is, amongst other things, a matter of physical closeness of electronic parts. The higher the frequency, the more important it is to reduce every millimeter of board trace." For those of the visual persuasion, Ivo has these graphs.

20kHz square wave

We'll get busy for those who'd rather have listening impressions. On paper alone, it's clear that for half the money, one could easily procure 5 times the power with class D. Just as clear is that on bandwidth, one wouldn't come close. One interesting question will be, how would bandwidth versus power manifest sonically on minimum-phase speakers which sing on 50wpc? Here we remember that a doubling of power brings just 3dB of extra gain whilst a perceived doubling of loudness needs a spare 10 decibels. Hitting marketing's magic 100-watt/8Ω figure without higher THD from running the Mosfets harder would have meant four output transistors, ergo a bigger power supply, larger chassis and more radiator surface for higher dissipation. Clearly the Allegro's compact size, power and tight packing are intrinsic to its behavioural specs. Small as not simply chic but essential? How weird that should seem to the bigger-is-better crowd. Here are some design notes from Ivo:

distortion with frequency

"My circuit does not aim at reinventing the wheel. It is neither a historic design carried out with modern parts nor a modern design with NOS parts. Instead, it is vital to do a proper job and avoid mistakes every senior designer should know about. I am always interested in the internals of other gear. All too often I ask myself what the hell a designer was doing by violating simple rules of analog circuit design. For the Allegro, we see a balanced input buffer, giving both hot and cold XLR inputs exactly the same impedance. Different impedances ruin the real world of common mode rejection which after all is the aim of balanced transmission. These input buffers drive a single-stage transconductance amplifier in differential mode. The current output of this stage drives a closed-loop 3 x gain output buffer. That output buffer is an ultra-wide bandwidth design, making use of the high-speed capabilities of lateral mosfets.

distortion with power

"Generally, the design principle could be identified as low-impedance high-current class AB, enabling high speed, wide bandwidth and low noise. Low distortion—a fundamental discrepancy in ultra high-speed design—is achieved by distributed gain with only modest amplification factors per stage. A final word on class D and why I decided against it. The efficiency of switching amplifiers is unbeatable by linear circuits which always need some quiescent current in the output stage. That is a big advantage for any environmentally friendly design. On the other hand, we see residual high frequencies on the output terminals, often in the range of 100mV or so. That's not a big problem within active speakers but a serious problem when speaker cables act as perfect antennae, polluting the room with high-frequency noise. This distributed frequency is mostly in the 100kHz range, not harmful to your health (by contrast to micro waves from WLAN etc.) but it can ruin the sound by coupling to the front end of your system. As a medium-powered amp, Allegro uses less than 30 watts in operation and less than 0.5 watts in standby, giving any green-party supporter a good feeling."