This review page is supported in part by the sponsors whose ad banners are displayed below


Before we start, some Pass basics. In Nelson's universe the primary distinction between properly designed amplifiers is the relative weighting of remaining harmonic distortion. It's either predominantly 2nd, predominantly 3rd or a clever mix of the two. Anything higher—4th and beyond—is to be shunned. That's because higher-order artifacts become progressively more dissonant. That makes slightly larger percentages of 2nd/3rd THD preferable to far smaller percentages of the higher harmonics. Nelson's approach to avoiding the latter is keeping it simple. It means a minimum number of gain stages; no or low feedback; and class A bias.


Having commercialized amps which champion either the 2nd or the 3rd harmonic, he knows that listener bias is quite evenly divided. About equal numbers of people favor one over the other as tracked by sales. The second harmonic is typical for triodes and single-ended circuits, the third goes with pentodes and push/pull circuits. One flavor plays it lusher, sweeter, softer, warmer and fuzzier. The other does it leaner, tighter, cooler, more incisive and separated out. The ideal isn't loads of either. The ideal is to keep distortion as low as possible. What due to given parts can't be avoided should remain tightly confined, not bleed into the higher harmonics where it becomes complex and thus musically damaging.

Here is visualization from a perhaps unexpected source, namely class D. Xuanqian Wang of Hong Kong firm AURALiC furnished before/after graphs to show how proprietary modifications to stock Hypex modules reduce higher-order artifacts from the 4th to the 10th harmonic whilst slightly increasing the 2nd and 3rd. Trading lower higher THD for higher lower THD relates directly to why Xuanqian believes his implementation sounds better (he also adds input transformers, linear power supplies and more circuit tuning). Such 0.0000 figures might have you wonder. Is human hearing really sensitive enough to be bothered? Yes is what many listeners insist is the case. Even very small doses of higher-order THD taste bad.


Nelson is far from alone here. Where he perhaps goes farther than most is dedicating an entire brand to the pursuit of extreme circuit simplicity by relinquishing high power and ultra-low output impedance. Low power gets away with two gain stages or less, higher output impedance gets away with no feedback or little. These factors reduce or eliminate a higher-order footprint. In that sense all FirstWatt amps are variations on a theme. They circle the same idea of what constitutes good sound. They do so at close distance from the center but from slightly different directions or perspectives. A final distinguishing feature are their output devices. Due to small runs and an unapologetic enthusiast's focus, some FirstWatt amps use rare or even unobtainium transistors of which Nelson either collected a personal batch as NOS tube fanciers would; or commissioned production of his very own. Which segues neatly into...

... SIT2 vs. F6. Let these amps warm up for an hour before you sharpen any pencil (half an hour might suffice but I played it safe). Fixed bias is internally calibrated to be dead on for thermally stabilized operation. Temps have to come up first to meet their bias sweet spot. Comparative context were my easy-drive 12-inch 3-way sealed Aries Cerat Gladius with my customary front-end of quad-core iMac, PureMusic 1.89g in NOS-style 176.4kHz upsampling, SOtM's battery-powered super-clock'd USB bridge, Metrum Hex in AES/EBU mode and Nagra Jazz.
As my preamble suggested, certain FirstWatt amps are like deciding between two gorgeous identical twins. They look very similar and could seem impossible to tell apart. At first. Once you get to know them better, personalities emerge. Same genome, different expressions. With these two I struggled with the hidden mole that would give the game away. One eventual tell was the F6's superior woofer control. But that was a cheap trick. It meant having to wait for a minor incident of additional bass bloom/resonance. A golden ear (cough!) shouldn't have to (dammit!) wait for some bloody bass transient to mosey along and betray lower than ideal damping. What if a cut caused no such blem to begin with? Now one's lovely twin didn't bend down to expose her secret mole in the small of the back. Aluminum ear? Damping was the tell though. It was a general personality expression like a minor unease with speaking completely freely. The other twin was more unselfconscious. More gush, less self editing. As a result there was more magnetic presence. And that felt bigger and more here than there.

Regular hifi terminology—tonal balance, treble/midrange/bass, soundstage—is about primitive markers of physicality. It's no more advanced than distinguishing people by sex, race, height and body type. Once we dig deeper for character and presence, things get less easily quantifiable. Even so they remain valid and are arguably more relevant since anything high-end should have the primitive stuff licked. Things ought to look like they are and sound like themselves. The trick with the less quantifiable stuff is cause. What creates our subjective but repeatable impression so we can properly communicate it to others?

The F6 was more damped all over. Again, this had the SIT2 seem bigger. But forget your tape measure. Base physical markers of stage width and depth didn't account for it. I can hint at it with projection. It's how a crack orator tweaks your attention across space. To him you pay rapt attention whilst his learned colleague equally versed in the topic and perhaps technically even more brilliant has you daydream. The former isn't more handsome. He doesn't speak louder. He has no fancier words. He doesn't really say anything new or different than his colleague. What then is it about his delivery that reaches out and creates some magnetism or a bridge between discourse/audience whilst the other is an island each listener must reach by their own volition and effort?


This explanation of projection itself doesn't track. The F6 didn't put me to sleep. It didn't require extra effort to enjoy. Yet something about the phenomenon relates just because the SIT2 had more of it. Like the other twin its delivery was more magnetic. It had more gush, more elasticity, more freedom. The F6 sounded more damped. Controlled. Contained within itself. Yet the two looked so much alike. I was in fact quite shocked by how much given that the SIT2 exploits Nelson's exotic transistor. It demonstrates how the maestro's colossal experience has learnt how to hit the bull's eye any number of ways; and how what he considers the bull's eye no longer wavers as it might with a younger designer who still explores the overall map looking for a nail to hang his hat on.

Put another way, the F6's directness was slightly less acute though I really couldn't claim that this came from sounding softer and the SIT2 spicier. I didn't hear any significant difference in how either handled transients and decays (crisper attacks often coincide with a sense of higher clarity or cutting power). Whilst those textural aspects are subtler than the primitive stuff, this was a further layer down and maddingly elusive to extricate, isolate and define. It was easier to say what it wasn't. The M2 had always struck me as the most retrograde amp of the lot, the one which harks back to earlier days when Nelson's sonic vision wasn't as distilled and crystallized yet. I'd blamed its softness and minor vagueness at its use of signal-path transformers since it paralleled what I hear from tube-coupled valve amps. Voltage-gain transformers notwithstanding, the F6 did not exhibit this quality. It very much was a latter-day FW amp and as such educated by the most recent super-sophisticated precedents of the SITs. Quite SIT-ish yet without the super transistor. And the demise of SemiSouth had even managed to put an end also to the SJEP 120R100 power Jfet which Nelson stuck into this F6.