Before we get to the meat of how the S5 sounds, a brief slap on the back to Mark Ward, for trusting his transistor baby to this unabashed valve maniac for its coming-out review, with JA at Stereophile apparently hot on my heels. It pleases me to no end that despite my well-publicized allegiances to the Great Vacuum inside them dainty glass bottles, makers like Coda feel comfortable enough with the moons' cut-one-way fairness to present me with such contrary opportunities to broaden my own horizons. There's nothing as dangerous in HighEnd audio as painting yourself into a corner of cushy notions or rigid beliefs. In the end, it doesn't at all matter how you arrive - as long as you get close enough to that giant sugar cube in the sky to actually grab a lick or two before listening to a live performance and be reminded again of the blasted differences that remain regardless of money or cleverness.

That said, let me come right out and state unequivocally that the Coda S5, like the Bel Canto eVo 4 GenII, is one of those rare sand amps that I could happily live with for the long haul. No, it doesn't sound like tubes. Yes, there are things premium implementations of tubes do that it doesn't do. Yes, there are things it does that elude most tubes, others that elude valves altogether. At the end of the day though -- when arguments fade, tempers chill and the music takes over to remind one and all what it's all about -- this amp's take on the music is plainly convincing and viable, with specific core strengths that will appeal to a definitive cadre of music lovers. For one, this amp's hellaciously fast and tracks subtle dynamic swings extremely well. For another, the minor treble reticence noted in the Unison vis-a-vis the Tripath amp has been lifted for cloudless skies into the stratosphere.

For yet another (point, chit, marker?), the S5 doesn't do grit, etch, layer compaction or even subtle chalkiness which, I hate to say, seems to be a distinct and remaining element of most solid-state treble performance (or, conversely, invites darker voicings to remedy the chalkiness but in turn diminishes air and spaciousness). Further on the marking scale, the amp's bass performance isn't unnaturally ripped to telegraph 'monster amp' -- something apparently beloved by some but abhorred by your scribe -- and its general mien is one of finesse and agility. Needless to say, Coda provides for would-be arc-welders with their 12.5 and 30.5 as well as higher-power S Series amps. However, I personally do not favor the overly dry slam-dunk symptom down low that sometimes results from excessive power - too unrealistic and gewaltig to these ears. Give me finesse and articulation but don't kill the rotund warmth in the bass by 'chiseling' its tombstone.

Which is to say, the S5 -- happily for me -- doesn't fall into the camp of Luke Manley's religion wherein the VTL evangelist talks about the advantages and sound of massive gobs of power. To a certain extent though not as refined, the Mesa Baron had that kind of sound, too - massive, dense, with endless macro-dynamics and that ability to swell to eternity. In that regard, the Coda amp is more nimble and light-footed, more focused on transparency than ultimate density. Yet despite this deep transparency and low-level detail revelation, despite its natural lack of valve bloom (harmonic extortion if you will), the S5's speed, attack and transient cutting power do not combine into thin or lightweight. Put into color temperatures, the Coda in fact resides ever so slightly on the warm side of the fence which, combined with its dynamic elan and jump factor, makes for an invigorating and athletic presentation. Think slightly lean in a healthy rather than undernourished way - blazing reflexes, excellent stamina, a high metabolism; not so much rock climber as a dancer or fencer. Power in motion. Definitely not a quarterback or Rugby ace no matter how fast on their feet.

Still having the Eastern Electric MiniMax preamp in-house for an eventual review match-up with its amplifier sibling allowed me to experiment with a valve/transistor pre/power setup and -- using leaner rather than more voluptuous tubes -- expand the final harmonic envelope just enough to gingerly nudge the scales toward more density without undermining the control and speed of the sand amp. Hey, this harmonic thang is simply something transistors, by comparison, don't do as well as the bottles. The op-amp-for-an-output device scheme can mimic it at the expense of ultimate openess, at least in the designs I've heard. This simply by way of illustration in what tonal area the S5 diverged from my base line reference. As indicated earlier, this is no mar but simply character, a matter of not white and black but much finer gradations in-between.

Despite its massively paralleled architecture, the Coda amp proved exceptionally quiet even on my spherical sonar precision microscopes from Germany. Likely precisely because of its massively paralleled architecture, it could maintain its composure under any conditions which, surprisingly, involved especially the micro scale where assumptions, in terms of output devices, might dictate the less the better to keep things simple. Super computers, from what I've read, achieve their operational speeds in exactly the same super-paralleled fashion. Raising speaker sensitivity can simply be achieved by paralleling drivers. The more drivers, the less each has to work. Less excursions equal higher speeds for any given signal amplitude. Less excursions also equate to far more linear pistonic behavior, better stop/go behavior without overshoots. In fact, the most dynamic speakers seem to barely move their transducers even when things get really loud.

Something similar seems to be the case with Lauchli's amplifier architecture. The S5's power reserves are so drastic in their over-build as to allow conversion to 5 times as powerful monoblock status in the afore-mentioned S12 guise, simply by changing bias current via new emitter resistor values.Talk about low-impedance head room! Popular conceptions look at head room as grace under duress. This seems to imply something less graceful in the absence of duress, as though muscle and moxy were all nice and good for a brawl but a hindrance during sophisticated quick-witted flirting. I'm sure that examples of this exist. Alas, the S5 is emphatically not one of them. Rather, its specific grace is active in the smallest of gestures, and perhaps these smallest of gestures become so relevant because of the over-kill output stage? Because these optimized devices are free to track the miniscule rather than having to concern themselves with bench presses? Whatever may be the case, these thoughts arise because of the way the S5 sounds. They are thus factual descriptors even if, ultimately, they should prove incorrect on the engineer's bench as the true cause for these audible effects.

I know what you're thinking. There's always a problem with high-power push-pull valve amps. The daisy-chained tubes, even should they start out perfectly matched, soon drift and diverge. They no longer act as one but like a string section wherein everyone plays just slightly different enough to blur fast runs. I'm admittedly sensitive to this argument as well. Since the Coda amp runs surprisingly cool for a bona fide Class A design, thermal nonlinearities with the transistors are a non-issue. And... truth be told, I'm not an engineer and don't ever want to delve into those dry waters. Let's just say that the S5 does fly in the face of my single-ended religion but works; and whether despite or because of it I don't need to understand to appreciate that it works. Which serves as a welcome break to hand the mike to Eric Lauchli for a brief sketch on the 'how' on his Precision Bias Class A topology that's the real bedrock upon which the S5 is founded:

"The Precision Bias topology was developed to provide a much more efficient and elegant implementation of Class A amplification. Precision Bias requires a specific set of parameters for the output devices and the surrounding circuitry. Essentially, we are fine-tuning and precisely calibrating circuit parameter control to yield an exceptionally smooth cooperation between the positive and negative output banks. Actually, one could copy the basic schematic of these amplifiers and yet end up with a less than desirable result unless every last detail of the implementation was also duplicated. This is a rare case where an inherent non-linearity of the devices at the extremes is used to good advantage.

At lower power levels where most music and listening occurs, the load is shared equally by all devices. As we reach the limits of bias current at higher power, the Precision Bias implementation allows one polarity of outputs to gradually increase its share of the load while the other backs off until one bank has full control of the load and the other is off. Because of the tight control of circuit parameters coupled with high speed, linearity and near perfect matching of the output devices, the transition occurs with virtually no effect on the transfer function of the output stage as a whole. The abrupt and often complex changes that occur in conventional designs when one polarity of devices switches in and out is effectively eliminated. This is borne out by the fact that no feedback is needed or used to maintain the low levels and low harmonic orders of THD, even at high power into low impedances.

The practical effect of Precision Bias is that a transfer function is achieved that is virtually indistinguishable from conventional very high-current Class A bias levels. Results are equal or superior to a standard "Class A into 2 ohm" design, but without their extreme size, heat, complexity or power consumption. The Class A rating of all Coda Precision Bias amplifiers is the power level at 8 ohms where all output devices remain on throughout the signal swing.

The application of Precision Bias results in an ultra smooth transition from one polarity of devices to the other. This occurs at very high power levels into low impedance loads, seamlessly and without sonic degradation. The end result is that for a given bias current, we can reach much higher power levels before one bank of outputs switches off, giving us excellent efficiency compared to more conventional implementations."

For those technically inclined, there you have it. For those who aren't, the important practical consideration to take away is that the S5 is as far removed from those flambe dishes which rivet fancy diners as is a tepid soup that's been left to chill to near room temperature. While the former is explosively hot -- our exaggerated visual cue for common Class A amps with their massive arsenals of heat sinks attempting to syphon off excessive throwaway thermal energy -- the latter is barely "live". You'd return the lukewarm soup to the chef with a bitter complaint. Alas, on your equipment rack and particularly if you live in hot climates, cool-running amps are the monkey's banana. And for all its hot'n'heavy Class A bias, the S5 is one cool cat, even after 24 hours of non-stop signal.