This review page is supported in part by the sponsors whose ad banners are displayed below
MSA sonics: A common signature for Mosfet outputs is what Stereophile's Sam Tellig cunningly coined Mosfet mist. It differentiates a generic class of one—transistor amplifiers—into many classes of Mosfets, bipolars and power JFets. Each is accompanied by certain core signature traits disregarding obvious generalizations and exceptions. Mosfets are routinely billed as 'tube like'. This points at their gentler transients, softer edges, fluffier textures and concomitant warmth. Bipolars tend to be cooler, more lit up, dry and incisive. Power JFets in amplifiers are still rare but FirstWatt's F3 and J2 suggest that they combine bipolar articulation and illumination with Mosfet textures to position in the middle.


Solid-state amp designers thus have their own if you will 'pentode' and 'triode' choices to make and numerous suppliers of transistors to consider. Nagra's focus on measurably good engineering eliminates single-ended operation in their book—that's why even their 300B amp had to be push/pull—but appreciates that so-called single-ended push/pull with just one device per phase eliminates paralleled matching errors for greater precision. Hence the MSA is limited to four transistors total. It's as minimalist/purist as push-pull circuits can get.

To gather feedback on speaker system performance—handicapped as it fundamentally is by room interactions and the innate time/amplitude/THD distortions of loudspeakers—a superior headphone reference system is useful. I rely on a fully maxed-out Woo Audio Model 5 300B SET with top glass from Synergy Hifi and TJ/Full Music connected directly to my Weiss DAC2. Transducers are ALO Audio-recabled beyerdynamic T1s and Sennheiser HD800s. 600-ohm loads seeing millivolt signal are quite idealized conditions for direct-heated triodes to actually behave as the ultra-low distortion super linear devices they theoretically can be but often aren't. The crossover-less single drivers in headphones which are quasi direct-coupled to our ears meanwhile max out resolution and clarity beyond what most speaker systems can pull off on their finest days.

For well-recorded 'adult instrumental' music, the Norwegian Øystein Sevåg can always be relied upon. His Caravan album displayed on my iMac screen below is no exception. "Wind" is a minimalist Garbarek-type number with Bendik Hofseth on saxophone and built up over washes and drum groove. While most the textures are mellow, Paolo Vinaccias' rim shot and kick drum accents crack real hard over my headphones. Their suddenness or steepness is akin to touching a hot burner and withdrawing your hand. Lightningy. The peakiness or loudness rise is akin to the involuntary shouted curse that usually accompanies getting burnt. It's startling, raw and entirely uncensored.

The MSA tracked this dynamic envelope better than two Mosfet amps from FirstWatt and ModWright (M2 and KWA-100 respectively) but did it neither as fiercely as my headphone system nor as violently as far more powerful direct-coupled class A amps like a big Karan would. On transient sharpness or edginess—the lightningy aspect—the MSA exhibited Sam Tellig's Mosfet mist. It was less honed and blistery than FirstWatt's F5 for example. Texturally, such sharp attacks didn't peel out from soggier surrounding thicket to the same distinctive extent. For dynamic jump or startle factor however, Nagra's 60-watter was actually quite surprising. It delivered the swing voltages. I'm inclined to regard its switch-mode power supply as enabler of this subjective speed. For transient pungency or whipping action meanwhile, the MSA was mellower. It clothed its attacks in some thin cotton. It didn't maximally highlight textural differences between a snarly popped e-bass wire and a legato bowed violin string. That was recognizable Mosfet behavior which incurs some mild homogenization. This usually falls under prettification.

In most reviews attempting to define the Nagra house sound, variations on sumptuous dominate. This suggests that the Swiss are very concerned with what's vaguely grouped under tone or timbre. Of the three components supplied, the MSA exhibited the least of this yet clearly carried distilled remnants versus something cooler and more lit up like my F5. I'd characterize the MSA as microdynamically pushing the envelope of the more affordable but quality Mosfet amps I had on hand. Macrodynamically it will be power limited at least in stereo mode yet relative to neighbourly peace and what I can realistically listen to, this never became an issue and was relevant only vis-à-vis my headphone reference.

If the most useful reviews condense a component's performance into one or two core features the writers are certain will translate across various systems to not bother with details that wouldn't survive outside their own setups, I'd single out for the MSA its handling of massed strings. Be it the elegiac Adagio for Strings by Samuel Barber, the Adagio of Prokovief's 5th Symphony, any Bruckner Andante or the saucy mayhem of a large Budapest gypsy ensemble doing a csardas or doïna, the Nagra indulged that typically (controlled!) tube thang of layering and burnished tone colors.

Blistery Flamenco guitar arpeggios and rasgueados, close-mic'd drum set workouts or brutal club fare with hard-as-nails bass beats weren't a special forté. Piano, violin and cello as the three classical instruments most cruel to transistor amps were. That's because the MSA focused on the sustain of tones while its excellent S/N ratio tracked decays very keenly. The Nagra then differed a bit from precedents in my circle of acquaintances. It didn't compromise good dynamic reflexes with its deliberate orientation towards minor warmth and mellower edges. Often warmer amps become cuddly and portly. They slow down. The MSA didn't. It did however refuse to get ugly. Music that only sounds authentic with maximum grit and bright shards underwent a Swiss civility makeover. That's part of this package.

This squarely applied also to other speakers like Aurelia's Cerica from Finland which were in at the same time. If you think legato not staccato, you've pegged the MSA's essence. If you think tube like, you will have to reference not the Yamamoto A-09S but something like my white Serbian EL84 push/pull monos. Their two local feedback loops of 0.3 and 4dB through the interstage and output transformers respectively 'mow grass' very deliberately (cancel THD). My current assessment of the Kaivalyas pegs them as about 30% tube, 70% transistor. I'd place the MSA on similar turf. It's likely why in past Nagra presentations, I've always preferred this amplifier to the company's own valve amps. Their Mosfet Stereo Amp distills just enough tube flavor to really suit this reformed valve appreciator.