This review page is supported in part by the sponsor whose ad is displayed above
Now was a good time to examine the internals. After removing 10 screws fastening the semi clamshell cover and detaching the ground cable that connects the cover to star ground, the innards showed the expected German Gründlichkeit [fastidiousness - Ed.] - a neat, very symmetrical layout well thought through. At the right rear corner sits the motorized attenuator, left front the input selector, in the middle the transformer. The circuit board is peopled by a wealth of capacitors big and small while the power transistors are located left and right tightly affixed to the heat sinks which are coupled to the side panels. All high-quality parts are easy accessible in case there's a need to. One jumper on the circuit board is visible just behind the large right capacitors. This jumper can be moved in position to make the Model 3 a fixed-gain power amplifier for any signal entering via A/V input number 5. With the jumper in this position, the amplifier remains a fully integrated amp for inputs 1 - 4. Input 5 is thus a selectable home-theater bypass.

Click on photo to open a large 1586 x 1200 version in a new window [187KB]

We installed the Model 3 on a separate Solid Tech amp rack and found that the felt-padded footers were too slick to resist the gravity of our Furutech power cord which tried to topple the amp. So the Model 3 ended up with an extra weight on top and sat on three Boston Audio Tune Blocks (the older square model) with Tungsten carbide bearings. For interconnect we used Crystal Cable in both XLR and RCA variants. To connect to the 107dB Avantgarde Acoustic Duo Omegas, we relied on Crystal Cable leads as well.

After its trip from Germany saw it temporarily deprived of a healthy flux of incoming electrons, we left the amplifier powered up for a few days while our dosage of daily musical vitamins derived from other sources outside Rotterdam. The first thing we noticed upon returning home while hunkering down for a first listen was a slight hiss from the Duo Omega tweeters. Turning up the volume without input signal did not alter the hiss level. We checked all connections but isolated no plausible culprits. It seems that certain Avantgarde electronics suffer a minor embedded hiss. The Model 5 we reviewed earlier did as well. Mind you, we are exceptionally sensitive to high-frequency noise. Computer power supplies, neon bulbs, you name it - we get effected and are very much spoiled by our own dead-quiet gear.

So we contacted Avantgarde with the question why. A prompt reply stated that the observed hiss was "a direct consequence of the new circuitry which changed the spectrum of the inherent residual noise of parts/circuits. Measured under the same conditions as the Model 5, the figures are identical (86dBs of S/N ratio). However, due to the higher bandwidth and resolution, the noise sounds 'clearer', thus more like a hiss than the 'darker' Model 5's noise spectrum." Later Matthias Ruff added: "The Model 5's noise spectrum was more or less a pink noise. The Model 3's spectrum shows a more or less white noise spectrum. And I prefer 'uncolored' noise ..." [while others would argue and hope for 'no' noise especially with an amplifier custom-designed for ultra-efficient speakers - Ed.].

At a later date, we were visited by two Avantgarde men. While they ruled the roost and volume, it became apparent how they are used to far higher SPLs than we are. The hiss then is noticeable only during idle or longer silent music passages. Also remember that with a 107dB loudspeaker system, the 0.4 class A watts produce 103dB at 1 meter, in the hot seat roughly 95dB per channel, adding up to 98dB for both if the loudspeakers are 8 ohms nominal. The Omega version is rated at 16 ohms, thus the Model Three puts out almost a full watt in class A, sufficient for almost 100dB in the seat from two hornspeakers. And then there's the class A/B reserve of another 26 watts.

To get used to the new configuration, we listened for quite a while in casual leisure mode while reading, dining, conversing and working. Ideally, the amount of attenuation -- or in fact intensification -- per degree of turn could be less so that the full 270 degrees making up the range from 0 to 100 might be used more comprehensively. With the remote, the motor on the attenuator helps a lot to attain smaller volume steps. As we learned later, with the Model 3 as preamp only, things get loud very easily.

Once acclimated to the new kid on the block, it was time for serious listening. The usual disc suspects were ready and demagnetized: Hadouk Trio Live a Fip, Antonio Fonseca Zamazu, Mussorgsky/Stokowski Pictures at an Exhibition just to name a few. But we kicked off with a set of test CDs by Eelco Grimm, Checkpoint Audio, unfortunately packaged with only Dutch instructions. The first CD contains music tracks, the second professional measuring tracks which we use to isolate frequency response troughs and peaks - which we didn't. Isolate that is.

High, mid and low frequencies were handled equally well so the first round of more analytical listening could commence. Here we applied ourselves more or less without heart. Was the positioning of instruments appropriate, the image of the instrument as expected and more of such inspections... not pleasure listening. It gave us a good idea of the Model 3's capabilities. From the analytical perspective, things didn't sound as we expected from knowing the Model Three to be a simple push-pull transistor amplifier. They sounded better. There was a touch of benign warmness especially in the higher frequencies. Cymbal crashes with their magnitude of harmonics sounded very brass-like where other amplifier can make them sound like bad brass where too much tin was added to the copper. On the other end of the spectrum, the bass wasn't too tight which hinted at high frequency distortion being well handled.

Thus far, we still listened for pure sounds rather than music, the least fun part of writing reviews. The best was to come when we could allow the analytical brain to step down a little and let the sounds evolve into music. When listening, memory, attention and emotion begin to literally play with the sounds, music emerges. This music is not just a form of art, it is much more. It can be described as a process in which three entities play a role. The performer -- in recorded music, all of the performers, engineers, producers -- the listener and of course, the music as sound. When viewed thus, there is one more or less static part, the performer. With recorded music, it is static because it can be repeated over and over again and is the same each time. The other two components in the process are variables. In case of the listener -- who remains mentally the same during a listening session we hope -- she is part of a feedback loop. When the third component of the process, the music as presented by the particular audio equipment involved, strikes an emotional cord in the listener, perception is altered and any subsequent musical input could be met by an euphonic or euphoric state.