Reviewer: Srajan Ebaen
Financial interests: click here
Sources: Retina 5K 27" iMac (4GHz quad-core with Turbo boost, 32GB RAM, 3TB FusionDrive, OSX Yosemite. iTunes 14.4), PureMusic 3.02, Qobuz Hifi, Tidal Hifi, Fore Audio DAISy1, COS Engineering D1, Aqua Hifi Formula, AURALiC Vega
Preamplifier: Nagra Jazz, Wyred4Sound STP-SE Stage II, Vinnie Rossi LIO (AVT module)
Power & integrated amplifiers: Pass Labs XA30.8; FirstWatt SIT1 monos, F5, F6, F7; Crayon Audio CFA-1.2; Goldmund/Job 225; Aura Note Premier; Wyred4Sound mINT; Nord Acoustics NC500 monos; LinnenberG Audio Allegro monos
Loudspeakers: Audio Physic Codex; Albedo Audio Aptica; EnigmAcoustics Mythology 1; Boenicke Audio W5se; Zu Audio Druid V & Submission; German Physiks HRS-120; Eversound Essence; Gryphon Mojo S [on review]
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; Black Cat Cable redlevel Lupo; Ocellia OCC Silver
Power delivery: Vibex Granada/Alhambra on all sourcecomponents, Vibex One 11R on amps/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: 4 x 6m with high gabled beam ceiling opening into 4 x 8m kitchen and 5 x 8m living room, hence no wall behind the listening chairs
Review component retail: $4'900

"The XA25 has an RCA input and no balanced outputs. It is a very simple amp of surprising performance. However, it is the intellectual property of Pass Lab. So it remains under wraps. But I will tell you this so that you can reverse engineer it: three stages; ~700 damping factor; high current; high slew; .00x% distortion; 40µV output noise; big Class A envelope; 25wpc into 8Ω, 100 into 2. And it sounds great." That was Nelson Pass; papa as he's affectionately known on the DIY Audio forum dedicated to all things Pass and present. Today's amp lacks a decimal point. So it belongs not to the point.8 range and it also lacks their power meter. Yet it's got more dissipation surface than the standardized FirstWatt enclosure. And unlike their occupants, it not only power doubles into 2Ω. It's stable down to—pop the cork!—0.5Ω. That's Scintilla turf as long as you don't ask it to dispatch >10 amperes (200 watts peak into 2Ω) at which point protection would clamp down on your excess.

Meanwhile unlike Pass Labs but very much like FirstWatt amps, it gets by with just two output transistors per channel (one per half cycle), not an entire alley. That's the absolute minimum for a push/pull output stage. Here it involves 700w/40A devices of the Ixys HiPerFet type. Those parts clock high power density and efficiency. They don't require a beastly power supply. Cosmetically and conceptually then, the XA25 occupies ground somewhere between FirstWatt and Pass Labs. Of course it remains pure Pass as 55bs of 17x14.4x6" WxDxH hardware that runs ~25C° at the radiators, consumes 240 watts at idle and presents your preamp with a 47kΩ input. Voltage gain is 20dB, bandwidth DC-100kHz, noise sub 50µV, slew rate 100V/µs. "The signal path from input to output was simplified to fewer components. Degeneration, 'the other form of feedback', was eliminated. This increased efficiency and class A operating current for greater class A operation into low-impedance and reactive loads.

"The output stage consists of a single pair of 700-watt new generation transistors operating in push-pull Class A with a new constant-current bias circuit to compensate for temperature drift. They connect directly to the loudspeaker without ballast resistors for lowest possible distortion and highest damping factor. We still use the same new-old stock of complementary cascoded small power Jfets and medium-power Mosfets in the classic current-feedback voltage gain circuit. The amplifier is still DC coupled and has no frequency compensation. The result is faster, lower distortion, lower noise, higher damping and larger Class A operating envelope all with a total of three pairs of push-pull gain transistors. In their ideal state, matched push-pull Fets give perfect square-law cancellation of distortion. That effect is somewhat spoiled by degenerative source resistance used to constrain the 'personality' of the devices. The removal of this form of feedback is an important element in the performance of the XA25. It is accomplished by new approaches to stabilizing gain and bias of push-pull Fets. This gives a lower output impedance, more gain and lower spectral content of distortion as evidenced by measurements. But the more important factor is the difference it makes to the music. We observed this difference in listening tests some years ago. Degeneration removed some of the organic quality to the sound. So it became the subject of further development work. The XA25 measures superbly but its subjective qualities are most special."

For those genuflecting to pope Mammon—who isn't?—the XA25 undercuts the previously most affordable XA30.8 Pass amp by a solid $1'600. The new low for Pass-age on the good ship is $4'900. Hello baby. A shocker then is the real-world oomph into sub 2Ω loads. Baby popped out of the oven with grown-up hairy cojones. But it's no degenerate*. To return to FirstWatt and Pass Labs, Nelson's former kitchen-table venture has since joined the Auburn mothership. It frees our man from personal assembly and production. Pass Labs and FirstWatt are now closer than ever. As AmpMan—watch out, Marvel!—Nelson can focus on dreaming up new circuit permutations, then ironing out wrinkles with iterative auditions, circuit and parts alterations until the book on one design gets closed and another opens. And who knows just how many sketch books Nelsons pencils in at the same time, some never to find mention anywhere?

"I’ve been pondering when and how to release the SIT-3, probably the last commercial SemiSouth SIT amp. It still puts out 10 watts and so on but operates the device in common-drain mode as a follower. The SIT-1 and 2 used the SIT in common-source mode, which resulted in the interesting character of those two amplifiers and highlighted the triode nature of the device. Common-drain lacks voltage gain but offers much lower noise, distortion and output impedance. In the SIT-3 the voltage gain is developed through an input stepup autoformer which gives about 12dB. No feedback of course, same box as ever, lower price. The last piece of business there is a new front end. Other stuff as well. I developed the DEF amp (stands for Depletion/Enchantment mode Follower) which is an interesting self-biasing push-pull output stage without degeneration or a bias circuit. Very cool but requires exotic parts and the result is still a 'regular' sort of amplifier. Also a B1 variant "now with voltage gain!" and some new active crossover circuits, including crossovers for Linkwitz’s LX-Minis as alternatives to the DSP units, DIY modular crossover circuits and Pure Audio Project’s active crossover and the list goes on." When creativity strikes, catching the gifts is the dance.

*Degeneration refers to the use of a resistor in series with the emitter, source or cathode of the gain device. This reduces its transconductance hence voltage which now depend mostly on the ratio of resistance/impedance rather than the device's intrinsic characteristics. Degeneration improves distortion and stability at the expense of gain. Whilst those effects mirror negative feedback, they don't reduce output impedance or increase bandwidth as 'true' (or the other kind of) negative feedback does. This practice predates the invention of negative feedback. It's thus considered separate and usually goes unmentioned in the context of whether a circuit employs any negative feedback or not. So an amp can be advertised as having zero negative feedback yet still use degeneration. The secret sauce of the XA25 is how it controls bias and gain without it.

Because Nelson has always made himself exceptionally available not just to reviewers but to scores of DIYers, it seemed only fitting to feature a photo of him, his wife Jill and their dog Jack, taken at their Sea Ranch home on the Sonoma County coast.