Reviewers: Marja Vanderloo & Henk Boot

Sources: CEC TL5100 CD transport with ultra clock; Linn Unidisk 1.1 universal player & Linn Sondek CD12 [in for review]; Audio Note DAC-2 Signature, modified, no digital filter
Preamp/integrated: TacT RCS 2.0 room control system; Audio Note Meishu, modified, with AVVT 300B output tubes; Avantgarde Acoustic Model 5 [in for review]
Speakers: Avantgarde Acoustic Duo, internally wired with silver;
Audio Note AN/Jsp, silver wired
Audio Note AN/Vx interconnects; Siltech Paris interconnects; Gizmo silver interconnect; Qunex 75 reference interconnect [in for review]; Crystal Cable CrystalConnect Reference interconnect, CrystalDigit S/PDIF RCA/RCA and RCA/BNC, CrystalSpeak Reference, CrystalPower Reference AC-Eur/IEC [in for review]; Gizmo silver LS cable
Power line conditioning:
Omtec PowerControllers
Sundry accessories:
IAR carbon CD damper; Denson CD demagnetizer; TacT RCS calibrated microphone and software; Exact Audio Copy software; Compaq server w/Windows Server 2003 and XP; wood, brass and aluminum cones and pyramids; Gizmo's Harley Davidson cap; silver Buddha head
Review component pricing: $3,970

In audio land, there's a religion that proclaims hornspeakers best suited to low-powered tube amplifiers sporting direct- heated, single-ended triode circuits which, most of the time, results in beautifully built amps designed around 2A3 and 300B output valves. Last year's VSAC show in Silverdale was a fine example of what this belief can cause.

We too have followed this path. Our big and bulky, highly modified Audio Note Meishu integrated amp happily feeds our music to a pair of super-sensitive Avantgarde Duos. The 300B output tubes are capable of spitting out a very decent 8 watts, enough to cause serious ear damage when fed unattenuated to 105dB loudspeakers. As reviewers, we have thus far concentrated on tubes amplifiers. Many of their kind have done their thing in our environment. In our opinion, even the earliest CDs can be enjoyed without shattering the enamel of your teeth if you use tubes. Their typical quality of rounding out the highs coupled to friendly forms of distortion create a certain warmth. Some tube amps even go a step further to cause a sort of laziness that makes everything sound very mellow.

One day a Riccardo Kron design landed on our doorstep for review - a pair of KR VT850 monoblocks to be specific. This 850 is a hybrid design with a Mosfet stage driving high-dynamic KR842 output tubes. Even though the late Riccardo was a pur sang tube man, he combined solid and hollow state technologies to reach his goal of the best music reproduction possible. He used solid state for all preceding gain stages and only at the very end, beautiful tubes were entrusted to handle the signal. We loved the results, especially the complete effortlessness of the reproduction and its capability to really deliver slam when needed. A whack on a drum became a real whack. So solid state certainly has special capabilities when used with care.

Back to another show, CES 2004. Here Avantgarde-USA combined BAT tube amplification with the experimental Super Duos, experimental in the sense that the Duos now sported six 10-inch woofers by adding two sub modules per channel. The sound resulting from a combination of setup, room treatment and equipment was warm, rounded and laid back. The system made the listener contemplative and serious - classical perhaps?

Completely different was the Duo system next door where the German manufacturer of Avantgarde exhibited. It was clearly more dynamic and involving, making us an active part of the music rather than passive listeners. Our copy of Rchestra's Weeshuis grabbed us by the cohones (or equivalent) once processed by this system - less room treatment, a different setting of the crossover/attenuator parameters of the subwoofers (with 'only' 2 drivers per channel) and a very different amplifier. Avantgarde Germany played their new Model5 integrated or vollverstärker. We listened for quite some time and liked the overall sound. We made a deal with Armin Krauß, Avantgarde's customer support engineer, for a review model. Some weeks later and delayed by the fact that another reviewer bought his Model 5 and thus a new amplifier had to be burned in at the factory, the German parcel arrived in Rotterdam.

Compared to the big black box of the Meishu, the Model 5 measures only 13.4" x 14.9" x 2.8" or, for us metrics, 340 x 379 x 72mm. The whole 10mm thick chassis, 2mm cover and 10mm fascia are powder-coated grey for a friendly, non-intrusive appearance. The top of the cover is adorned with an engraved company logo. The front sports two round knobs flanking an elongated oval display to look a bit like a smiley face a child could have drawn. The left knob is the input selector, the right the volume control. The display is special. Against its white background, two indicator needles display the selected input and volume. In standby, the left needle points upward, the right down. When in standby, a bright LED illuminates the dial a glorious red. The asso-ciation of the bright red and 'vacant' sign we associate with the Red Light District cannot be a coincidence. Once the Model 5 is 'occupied', the red LED switches off and a bright white LED takes over. However, do not worry. The bottom of the Model 5 houses a dimmer for the LEDs. The LED also switches off when the remote selects mute. Yes, the Model 5 arrives with a remote to control volume up/down and mute. That's it, no further thrills here. The design of the metal remote matches the amplifier - stylish and robust.

The back of the Model 5 offers 5 pairs of RCA inputs, one pair of record outputs, one pair of XLR outputs, WBT speaker terminals, an IEC power inlet and the mains switch. The impedance for all five inputs is 10kOhm. Input five can be con-figured as an A/V thru-put that now assigns volume control to an HT processor. The XLR outputs are preamp-level and designed to drive powered loudspeakers like Avantgarde's own active Solos. A further thoughtful feature is the display's auto turn-off when in A/V mode.

The Model 5 is rated at 400 milliwatt or 0.4 wpc of class A into 8 ohms. We love this spec. Where else do you see such pride in low figures? When 0.4 watts are not sufficient, the amp switches to class A/B and delivers up to 27 watts into 8 ohms. The lucky ones who use high-sensitivity, high impedance loudspeakers like Avantgarde's 19-ohm Trio Omega do even better. The Model 5 then delivers 1.1 watts in class A and a whopping 38wpc max at 20 ohms.

It is obvious that the Model 5 has been built for the kind of high-performance loudspeaker like the horns mentioned. The 0.4 watts of Class A power already creates outputs of 100dB/1m with our Duos, plenty uncomfortable for our ears. Before we let the music speak, we must delve a little deeper into the design of the Model 5 to report on certain novel circuit implementations.

The technical genius behind Avantgarde Acoustic is Matthias Ruff. Inspired by the very minimalist 47Lab Gaincard, he started to think about a simple amplifier that would be perfectly copasetic with high-sensitivity loudspeakers. Many users of such speakers have been unhappy with the available choices for matching amplifiers. Hugely powerful amps sound great with low-efficiency speakers but not highly sensitive ones. As we have indicated above, the first half watt is the most important then. That is not an operational range most amplifiers are happy in. Instead, they start to blossom much further up the power ladder. In the micro-power regions, they sound sluggish, thin and asleep.

Accordingly, Matthias dreamt of an integrated amplifier that would match the resolution, detail and dynamics of his horns and come in at a very affordable price. Class A operation -- where the higher bias eliminates the zero-crossing distortions of push-pull designs -- was a necessity. After the first 0.4 watt, class A/B with its slightly higher distortion becomes far less critical because this power is then merely used as headroom and for signal peaks - the slam and transients of climaxes. Due to its voice coil motion, every loudspeaker 'talks back' to the amplifier feeding it. The official term for this talk-back is electromotive force or back EMF. To counter this effect, many amplifier designs add a form of suppression called negative feedback that is used to lower the output impedance and thus increase the damping factor. Most push-pull transistor amplifiers use the principle of high-gain/high-feedback to control their output. The input signal is amplified to a level much higher than necessary for listening. The resultant signal is compared to the input signal and all aberrations are subtracted before the remaining signal is fed to the loudspeaker terminals. This process is somewhat akin to shoot first, apologize later.

In order to amplify the signal to the excess level necessary for feedback, modern solid-state amplifiers use several consecutive gain stages, each consisting of one or more transistors. We know that every component and transistors in particular add some distortion to the signal. Now place several gain stages in series. Each consecutive stage amplifies the distortion of the previous one while adding its own. For most designers, that's no problem - at the end of the ride, the global feedback loop puts unwanted distortion into the garbage.

However fast such signal processing occurs by the standards of the human nervous system, each additional stage within the signal path adds a time delay. Matthias speaks of slow forward amplification. This causes the feedback supposedly cleaning up the mess to just miss the intended distortion and instead of cancelling, add distortion to the signal while theory supposes that distortion is minimized. The signal is again amplified in the next gain stage, the negative feedback loop tries to clean it up again ... etc. In computer terms, this looping is called thrashing - two steps forward, one backward. [They have a similar version for Capitol Hill called the Potomac 2-step - Ed.]. Add to this rather constant looping the modulation that unpredictable musical signals add to the mix via back EMF. This becomes constantly variable distortion and is very audible to our hearing system, not only as slowness but larger textural granularity and blurring. These time delays and phase shifts create inferior detail resolution, curtailed dynamics and blunted transients. Some call this sum of distortion the color of the amplifier. Others would call it a great backwards step from the sought-after straight wire with gain ideal.

To counter these disadvantages of ordinary push-pull solid-state amplifiers, Matthias developed a new circuit dubbed DSSF. DSSF stands for Dual Stage Single Feedback and, to quote Matthias, "forwards fast, backwards fast and little". In the DSSF circuit of the Model 5, there are only two transistors in the signal path of each channel. For a push-pull amplifier, things couldn't be simpler. One transistor forms the first driver stage and operates completely in class A. The amplified signal follows its short way through the circuit to a second transistor, a power MOSFET, that will switch between class A and A/B modes. Both transistors have there own very short linearization circuits or local feedback loops. This enables any aberration to be corrected right at the source rather than the end of the whole signal path as with global feedback. Two further transistors regulate the high bias voltage but are outside the signal path.

The deliberate choice for only two amplification stages reduces the time a signal take to pass through the circuit and eliminates the unavoidable distortion of extra gain stages. Another result of this speed is an ample bandwidth of 220KHz. The display on the front therefore proudly states Ultra Bandwidth Amplifier.

Matthias believes that his DSSF circuit requires a certain form of overall feedback, albeit extremely little. The circuit is fast enough to have a high damping factor even with complex loads. Matthias inserts his feedback loop between the drain of the power stage transistor and the emitter of the driver stage transistor. No matter where the distortion arises -- from the circuitry itself or the loudspeaker via back EMF -- with this concept, the signal only has to back through one single stage, the driver stage. From there it's two steps back to the output - two stages forward, one step back. The conclusion is that this circuit is fast. At least the theory makes perfect sense.