|
|
楼主 |
发表于 2011-7-4 11:57:35
|
显示全部楼层
Reviewer: Srajan Ebaen
Financial Interests: click here
Source: APL HiFi NWO 3.0-GO; Ancient Audio Lektor Prime; Raysonic Audio CD-168
Preamp/Integrated: Supratek Cabernet Dual; ModWright LS-36.5 with PS 36.5; Melody I2A3; Woo Audio Model 5; Trafomatic Experience One [on loan]
Amp: 2 x Audiosector Patek SE; 2 x First Watt F4; Yamamoto A-08s; Fi 2A3 monos; Yamamoto A-09s
Speakers: Zu Audio Definition Pro; DeVore Fidelity Nines; WLM Grand Viola Monitor with Duo 12; Rethm Saadhana; Zu Presence [on loan]; Mark & Daniel Maximus & Ruby Monitors w. OmniHarmonizer
Cables: Ocellia Silver Signature loom; Crystal Cable Ultra loom; Zanden Audio proprietary I²S cable; Crystal Cable Reference power cords; double cryo'd Acrolink with Furutech UK plug between wall and transformer; Stealth Audio Indra and Meta Carbon
Stands: 2 x Grand Prix Audio Monaco Modular 4-tier
Powerline conditioning: 2 x Walker Audio Velocitor S fed from custom AudioSector 1.5KV Plitron step-down transformer with balanced power output option; Furutech RTP-6 on 240V line feed
Sundry accessories: GPA Formula Carbon/Kevlar shelf for transport; GPA Apex footers underneath stand, DAC and amp; Walker Audio Vivid CD cleaner; Walker Audio Reference HDLs; Furutech RD-2 CD demagnetizer; Nanotech Nespa Pro; Acoustic System Resonators and front wall sugar cube matrix
Room size: 16' w x 21' d x 9' h in short-wall setup with openly adjoining 15' x 35' living room; concrete floor and ceiling, concrete/brick walls
Review Component Retail: $3,000
elson's infamous Kleinhorns used to evaluate a large range of wideband drivers. These horns have since been gifted to an artists' collection. One of Nelson's current speaker experiments centers around a Feastrex D9nf in an open baffle.
Nelson Pass. FirstWatt. Kitchen table venture. Surely all this has been active long enough to no longer catch our readers in the dark. For a light switch reminder, reviews of the F1 thru F4 plus Aleph J have all the relevant information covered solid in our archives. Here's what bears repeating though. By design, all FirstWatt products are tailored to certain very specific uses. As the name gives away, those are focused on the higher-sensitivity speakers that come on song within the very first watt to consume little power.
"Jack of all trades but master of none." "Being everything to all people." Not! So exclude, eliminate, focus, optimize, perfect. Arrive. That would be limited editions of 100, with each amp model -- and five line-level products are in the chute next -- custom-tailored for very narrow applications. Single-ended and push/pull transconductance amps. A low-power amp with JFETs for output devices. A pure power buffer with zero voltage gain. And, as the odd one out by being a perfectly normal amp in this group of eccentricities, a new installment of the classic Aleph circuit with JFETs at the input, hence Aleph J. That's been Nelson's output under FirstWatt thus far.
Common to it all have been the chassis; class A operation; twin 18V secondaries; hand assembly by the maestro himself; and a very generous transfer of the circuit schematics into the public domain once the limited editions had sold out. To recap, with FirstWatt, Nelson Pass is beholden to no one but his own creative genius. His livelihood and economic viability are secured by Pass Labs. FirstWatt is thus the quintessential mad scientist's lab at the edge of the mainstream abutting another universe. You just never know what to expect next. Except that five follows four. In this universe.
In his own, would the man pull another stunt as he did with the Aleph J? Not that way, this time. The latest in the F series (which might stand for fun or fabulous or far out) is the F5 which follows numerically on the heels of the F4 buffer/follower amp. The F6 thru 8 have already been conceptualized as higher-power mono versions of the 3, 4 and 5 respectively. FirstWatt is about narrow paths less traveled, not milking repetition in overgrazed fat meadows. "I think it will be a little while before FW starts repeating itself" quipped Nelson laconically. So watts up with the 5? What particular niche and needs does it fill? (For a great read on Nelson's earlier career by the way, jump over to Thomas J. Norton's 1991 interview with the man in the Stereophile archives.)
It introduces a few firsts for FirstWatt: feedback*; and a full 50 stereo watts. The simplified circuit shows a DC-coupled 2-stage current source amp with 15.15dB into 8-ohm voltage gain, specifically a complementary MOSFET common source output stage (as previously seen in Nelson's Zen Variation 5 DIY circuit, here biased at 1.3A and class A up to 2.6A) driven from a complementary JFET common source with +/-24V rails. On the subject of bandwidth, check out the square wave performance into 200kHz - scarily close to perfection. Might the lack of signal path capacitors have anything to do with that?
When I first saw that graph, I had to know - was this ultra-low distortion and speed due to newfangled parts or circuitry? The latter. Nelson describes himself as a circuit mechanic. Parts are important but to him, circuit architecture is senior. Where modifiers will throw money at designer parts, a true master of the craft goes to the source. He manipulates how the electrons flow, resets the levies and dikes and flow valves in the three-dimensional construct of an electronic circuit. This enforces very specific electron behavior to establish ideal relationships between voltage, current, distortion, bandwidth, amplification factor, ground plane, noise, impedances. It writes the 'molecular code' for how a circuit behaves. That would seem rather more creative than dressing up parts.
* Nelson Pass: "As I have previously pointed out elsewhere and in the F5 write-up, the term current feedback probably came from the marketing departments at the chip companies when they began selling high-speed op amps using this approach."
The distortion performance at 1 watt is below 0.002% and rises to a negligible 0.05% at 20 watts which would turn low-power SETs green with envy. Poster Klaus on the Pass DIYaudio.com forum suggested "would the master be offended if I called the F5 a modern FET-based Hiraga? I see some similarities, current output (open loop), current feedback, straight two-stage design (if one wants to count a diamond buffer and a CFP as one stage, resp)." to which Nelson replied: "Not at all. It's a topological classic - the devil is entirely in the details." Other posters called the circuit "an amp so simple, even a cave man could build it". "Of the Zen and FirstWatt series and for damping factor, the F5 would be the best since it has a DF of 80 and will drive 2 ohms without burping. I don't believe, however that I would point it at a speaker with both a low impedance and a low sensitivity. Reactance is not an issue. The performance is slightly more robust into a fully reactive load vs a resistive load of the same impedance. The F5 has the lowest distortion yet achieved by either a Zen or First Watt amplifier, descending to 0.001% below 1W. Suffice it to say that the F5 has more power, more damping, more current, more bandwidth, lower distortion and higher input impedance than the other Fs. Many will think it sounds better."
Those looking for ground-breaking circuit innovation will come up short. The F5 takes wing at the zenith of a master's career where, regardless of craft, less becomes more. A simple sketch conveys more meaning than a technically advanced photo-realist painting. A short haiku nails more essence than a florid novel. It's about trimming the fat. It's about distilling a circuit down to its barest essence while seeing and manipulating the interrelatedness inside that simplicity with new clarity.
Nelson has penned a how-to F5 article in AudioXpress. "One of the aims of these articles is to get people to build amplifiers". With 50 stereo watts, low output impedance and feedback, the F5 appears to move away from the FirstWatt center focus and toward the periphery.
But the distortion figures suggest clearly that the core of the credo -- the ability to track the most minuscule signal fluctuations inside that very first watt -- isn't merely retained. It's arguably been heightened with a high-speed wide-bandwidth circuit of uncommon linearity. The culprit or boogie man in this instance is feedback. It's also a contentious and misunderstood topic on which Nelson is penning an upcoming paper for our readers.
"This circuit employs feedback in a dual pair of low impedance voltage dividers -- low impedance feedback has been (incorrectly) referred to as 'current feedback' -- and one of the charms of this arrangement is that unlike the classic two transistor differential pair, the drive current available exceeds the bias of the input stage. Each JFET has its own feedback so there are two separate and independent feedback loops to this amplifier. The input impedance is 101 Kohm and the output impedance 0.1 ohms for a damping factor of 80. The noise is about 30uV. Because the output stage employs only two output devices per side which are capable of very high current, deliberate limiting is set by 3-watt power resistors to the source pins of the MOSFETS to increase thermal stability and serve as convenient current-sensing elements. The amplifier with 24V supply rails is good for a 50-watt peak into 8 ohms or about 2.5 amps. For a 4-ohm load we would want 5 amps and for a 2-ohm load, 10 amps. Since we only have two output devices, we probably are best off stopping there to ensure against damage into dead shorts. Current limiting gets a bad rap in general, but I think it's a matter of where and how the limits are set. In toto, a nice little amplifier without too many parts that sounds great."
In the end, that's the part which matters most to you and me. Nelson's haiku instructions to thinking DIYers are for the hands-on crowd. Amazingly here, detailed instructions on how to roll your own are out prior to first commercial units. For the F5, Nelson's very sporting generosity has put DIYers on equal footing with FirstWatt. To hell with commercial considerations. Off to fun 'n' games. That's been the prevailing attitude of this kitchen table effort all along: "The B1, a buffered "passive" using my new Jfet buffer is in production, that is to say I have stuffed one up and it works. I have the boards for another 10, so it will be waltzing shortly. The specs are fabulous and it will be really interesting to see if the buffer solves the age-old passive problem that people have. I think it will, insofar as the concept worked very well for our INT150 integrated, which is really just an X150.5 with a selector and buffered volume control." If it takes two to tango, Pass Labs and First Watt are partnering up very nicely.
Act 2. Having now reviewed every FirstWatt amplifier made save the Aleph J, something's plain. They all dance in a very small circle of freedom around a tether pole staked through neutrality: variations on a sonic theme; how to reach Rome from yet another direction. I'm reminded of the Indian saint Ramakrishna. He attained ecstasy following one line of esoteric teachings. Then he pursued an entirely different discipline only to encounter the same ecstasy. Again. All roads lead here became his personal verification. The ecstatic could thus assure devotees of various religions that their different beliefs didn't matter. In the end, they'd all get consumed in and erased by the Luminous.
But the material form remains a filter even with a highly sublimated personality. Flavors continue. Nirvana in the body isn't stark white. It's not neutral boring nothing. So the F3 is sweeter. A mini triode dose. As exceptionally transparent as it is to the preceding preamp literally doing the driving, the F4 is more limpid. It doesn't 'grip' but floats and flows. The F5 articulates. That very first impression held. I was spinning Instrumental 1, a custom compilation by the good doctor in Berkeley who had co-financed Alon Wolf's Magico Ultimate hornspeaker project as one of three original commissioners. The doc has the best musical tastes of anyone I've ever met - simply by perfectly overlapping with my own. Instrumental 1 is minimalist gossamer in-the-zone stuff expertly sequenced. Artists include Eleni Karaindou, Markus Stockhausen, Herbert Joss, Anouar Brahem, Arild Anderson, Paolo Fresu, Terje Rypdal, Ketil Bjørnstad and others.
Having come directly off intense inspections of Yamamoto's reference amp with various 300Bs when Nelson's second silver-faced model landed, I was shocked by how ridiculously low its playback level could be without causing any serious participatory losses. I got deeply involved at whisper levels without crank-reflex intrusions. Regardless of which 300B I rolled, the A-09S -- whose 27dB gain is far higher than the F5's -- needed higher playback levels to come on song. This demonstrated how the... um, purity of nano-level distortion in the transistor amp (which a valve lover would call a wholesale stripping away of benign and desirable THD) actually served the music better. It was more intelligible.
Allow me a brief detour to stress the point. Certain erotic movies celebrate their protagonists as pursuing ever more outlandish means to sexual gratification. It's as though their continued use of toys, fetishes, strangulation and pain severely desensitized them to require ever more bizarre measures of stimulation. The less one feels, the more effort must be made to elicit the desired response. The same occurs with junk foods' high saturation of salt, sugar and fats. Once a palate has been dulled by such assaults, eggs or veggies not tarted up with salt and sauces become tasteless and boring - for simply tasting like themselves. The F5 invites such ruminations and applies them to the habits of us valve addicts. Compared to a Halcro's nonexistent distortion, tube THD seems completely justified to our kind; in fact, outright necessary. Why then is it that faced with the F5's impossibly low distortion (impossible certainly for any tube amp), thermionic liberties suddenly seem far less justified or necessary? Are valvers addicted to unnatural stimulation?
I don't have the answer. It's an open question to point at something real but elusive. The technical purity of the F5 isn't flat. Its feedback causes no dryness; its proud specmanship has real relevance to the experience. But it certainly doesn't sound like my glowing amps. As you prime their pump, they come into their own with great tone density. Viewed from the F5's seat, there's less resolution within that density. There's less separation; less residual grunge around a faint muted trumpet, less swishiness on swirling cymbal brushes, less ephemeral glitter on bell trees. Something congeals. Greater loudness distracts from it but truly subdued midnight levels with the SETs obscure. One can no longer enter into the music. A door closes. One stands outside and removed. The glowing bottles clearly suffer when my wife is already dreaming and I'm still chasing the tunes and the most immaterial of fades.
Viewing things from the strengths of the glowing bottles when their amplitude is sufficient to skip over that get-real hurdle, the F5 is leaner. Less padding of tonal girth. But it doesn't sound lean per se. It's completely beyond threadbare (and granted, I ran the 12dB-gain tubed ModWright LS/PS 36.preamp at 3:00 o'clock + on the above Zu Presence). There's more treble tintinabulation to re-use Nelson's crafty term. That action doesn't saturate timbres like 2nd-order octave doubling does with tubes. It instead increases finesse of insight and subtle energy. After all, upper harmonics are high in frequency and subdued in output. To track them requires extension and ultra-low noise floor. Then the live factor survives when you dim the volume lights. Loudness is often a substitute to create the illusion of seeing more. Then we forget what we really don't see. It's compensation. It proves out insufficient illumination at quieter levels. In plain speak, that's inferior and ultimately insufficient resolution. The need for SPLs nearly always signifies it.
This segues back to the all-roads-to-Rome statement. To your tubular scribe, all FirstWatt amps are examples for how superior resolution can truly serve -- and not distract from -- the music. Tubes can serve in their own way. That makes for a different experience. It shifts the focus to tone colors, textures and a certain kind of inner voluptuousness. The F Series of Nelson Pass amps focuses instead on the 'outer' voluptuousness of contributing complexity riches. How many tiny spinning wheels and belts and gears make up the musical construct, make it all happen? Calling one inner and one outer is bound to mislead of course. It's all part of the same music.
Since I brought up Ramakrishna earlier, perhaps it's appropriate to speak in terms of perceptional doors from which the listener enters the experience. If the (good) tube experience is ultimately a melting in the heart for expanded feeling (anahat, the heart chakra), the superior transistor experience is an expansion of the mind where seeing occurs (ajna, the third eye). They're different experiences but not superior or inferior one to the other. This could be a meditator's attempt of describing the same phenomena which sound doctor Nelson Pass talks about as 2nd and 3rd-order type amplifiers. Needless to say, this is all quite subtle, dancing in a small circle of freedom around the center point of neutrality. Suchness as the Buddhists would say. Without filters. The remaining F amp filters are very transparent.
In the world of FirstWatt, the F1 had the strongest 3rd-order flavor. Super articulation. Separation. Crispness. And drive. The F1 sounded driven. Propulsive, a clear personality. The F4 and F5 are rather more relaxed entries in that domain, with the F5 slightly more gathered up and sorted than the F4. The buffer/follower amp is the most kif of the bunch as the stamboulis would look for in an opium den - supremely relaxed, in a go-with-the-flow state. The F2 and F3 are sweeter and 'rounder', with the F3 deepest into that particular terrain, the F2 just slightly. In general, numbers 3 to 5 strike me as closer to the center than the first two.
Take anything massed -- a chorus, symphonic strings à la Samuel Barber -- and as predicted by Mr. Pass, the degree of sight the listener enjoys into that layered action from the 5 is brilliant. Take anything speedy and sharp -- rapidly plucked strings, blatty brass, drum kit workouts -- and the F5 is clear and incisive but never brutal and hard. Except for the F1, none of the FirstWatt amps struck me as bass monsters. As a transconductance amp, the F1 could only be properly used on crossoverless widebander speakers. Those are never really endowed in that sense, making what the F1 got from them just shy of freakish but still quite relative in the bigger scheme of what's ultimately possible from 15" woofers.
Take soundstaging -- depth, width, localization, stability -- and the F5 is a champ. No drift, no wander, no blur. No etch or grippiness neither. The F4 appears even 'floatier' but those are very subtle matters appreciable only in a direct comparison. Which of course is what the man already said in his own words. Nelson is definitely equipped to not just design amplifiers but review them objectively. Take microdynamics -- those heart palpitations caused by a singer's or instrumentalist's unexpectedly emphatic inflections -- and I'd have to call the F5 superior to my most cherished valve amps. If that's directly related to circuit speed, the F5 certainly lives up to that obnoxiously ideal square wave graph we admired earlier. This returns us, again, to that ability for remaining involving at very subdued levels. There's enough wiggly, jumpy, kicking live factor left to be interesting and gratifying. It's all there, not lost in any THD fog.
At elevated levels, over the Zus and on vocal peaks, I did detect the occasional glimmer of spiciness, of likely response peaks in their widebanders which my usual valve amps don't bring out. On my Lowther DX-55-based Rethm Saadhanas, this happened as well but not objectionably so, just a presence-region tweak for some enhanced if ultimately not entirely honest fun but clearly intrinsic to the speakers, not amp. Nelson has done extensive measurements of such drivers and could tell you precisely what operational aspects of the F5 contribute. I'll merely confirm that yes, if your widebanders are spicier than you wish, the F3 would remind you less if at all.
The monster SuperPAC tweeters on my 98dB WLM Grand Viola monitors -- paralleled horn-loaded paper units high-passed at 800Hz -- told their own tales of F5 treble elucidation. This amp suffers neither high-frequency phase shifts nor premature coagulation. If infrasonics help recreate the full scale of audible recorded space per se, high frequencies light up the space around individual performers. The deliberately wide-dispersion WLM 'tweeters' (their low crossover point makes them rather more than traditional treble units) do this holographic lock to a much higher degree than the Zus. The power response of the upper end is higher too as is the perceived energy of the ambient field. In plain speak again, they soundstage like demons. The F5 builds out this asset with panache. Soundstage freaks with the right speakers to heighten this stereophonic illusion will be over the six moons and on the seventh. Way wide and deep far-out stuff.
It's no surprise that the F5 is no overdamped control freak. Unlike the S.A.C. il Piccolos from Germany with their 80dB of feedback and claimed full-bandwidth damping factor of 20,000 which fairaudio.de described as "dry as dust and super slamming in the bass", the F5's so-called damping factor, though higher than all other F amps, is still modest in the scheme of things. It's innocent then of cyborg bass, my term for entirely unnatural low frequencies that sound hard as nails or concrete. Truthfully (I've waited to tell you this so you'd keep with this review), you could take my F4 report; add 15dB of voltage gain and 2-ohm happiness; season with just a touch more articulating action, separation and concomitant 'visibility'; add a smattering of treble energy and presto - brothers from the same mother.
Ditto for operational noise. All F amps are quiet like the proverbial graves. That's perfect for noise-critical applications with high-sensitivity speakers. Mum's the word as well on the 240V mains transformer inside the F5. With my wall power often about 10% above its rated value, some iron protesteth and goes into hum. Not this one. Even the piercingly blue eyes of earlier F models have gotten turned down by popular request so the power LEDs no longer sear your retinae. Super-clean bill of health all around for the F5.
Act Three. This is where the main actor bursts into song while being stabbed in the back. Opera and all. We call it the conclusion. So, Nelson also is right about the 1-hour improvement. I didn't clock it but after between one to two spun CDs, the amp arrives and settles down. Could I tell the F5 runs on inferior feedback-fossil fuel compared to the other Fs? No. I have tested tube amps with selectable feedback where even small amounts of 2dB were clearly audible. Usually but not always, those did sound best with zero feedback on the speakers I used. Feedback then did dry out things, curtailed elasticity and swing if you will. However, I hear none of those drawbacks here.
Just as ultra-low measured harmonic distortion can sound sterile but doesn't here, feedback can but needn't. Hell though if I have a clue as to why, when and how come not then. It's interesting coming from tubes when first encountering the F5 in a well-dialed system that won't get embarrassed by greater honesty. While I've always known that tubes introduce 'pleasing distortions' (the main challenge finding machines that do this subtly enough to flavor, not become headline intruders), Nelson's FirstWatters from the F3 on have caused some head scratching and second-guessing.
Particularly for midnight listening, ears keen as Spock's, lover dearest sleeping next door, the F5 stomps my SETs. At regular levels, triode tricks catch up to make their own case. For me at least, then it's a question of what flavor experience I crave. If it's a highly visual one, the transistors; if a melted heart, the tubes. I admit this commentary isn't the stuff of scales and tape measures. The level of game Nelson Pass engages these days when commercial considerations are out the window means that this brazenly subjective approach as an experiencer becomes mandatory. After all, the graphs already show us ruler-flat response with distortion so low, we'd assume it below our ear/brain threshold.
Addressed solely at us nutters in the Tube Corps, ultra-low distortion isn't intrinsically boring and uninvolving, clean as a whistle but bereft of juiciness. It's about more though than chasing textbook perfection no matter the means. This is no Halcro. Yet transistor goodness needn't mimic valve amps to somehow clone their behavior before a transistor amp can make its very own kind of magic. Depending on how your personal journey through audiolandia has proceeded, this could all be self-explanatory. In fact, you might wonder how a reviewer could take six busy years before coming to that conclusion. Um - we're not all the Silver Surfer. But Oliver Stone's "Greed is good" speech for Gordon Gecko comes to mind. How it stood normal values on its head. So the F5 could stand the aural beliefs of valve fanciers on its head; not for sounding the same but because it achieves equal validity -- arguably superior in aspects as explained, merely different in others -- via such contrarious means. 20dB of feedback. Pah. Or perhaps the very reason for the F5's sterling showing? If feedback is bad, tell that to Mr. Pass and my zero-feedback ears. Glorious confusion. Or as Gurdjieff is rumored to have muttered under his last breath: "You're all royally screwed now".
Curtain time. The F5 brings the Nelson Pass sonics explored in the present FirstWatt range to 25wpc needers with enough gumption for 2 ohms. Be aware though that 15dB of amplifier gain isn't very much. To serve up 90dB speakers in style, you'll want 20 - 25dB in your preamp . Not that those are FirstWatt's focus. But the F5 could be a back-door opportunity for those not subscribed to the hi-eff religion. You'll simply no longer be running the amp inside that first watt where its distortion drops off the map. Nelson predicted that some listeners would like the F5 best of the bunch. I certainly see why. As much as an audiophile floozie can at any given day, I might just share that opinion in fact. It's unmistakably the most universal FirstWatt yet, with the high-power 'muscle' versions of F3 thru F5 already announced to welcome those with even harder-of-hearing speakers into the fold. Consider this SET freak deep inside the fold already. The old Zen master is cooking and with his latest amp, even cavemen and Geico customers are reportedly capable of rolling their own. What a gift to the DIY community - and those of us reluctant to inhale solder fumes. Thank you, Mister Pass! ... who, it is only proper, shall have the last word on this matter:
"How do these effects come about? I really don't know. My experience so far is that good ideas, technique and philosophy all help but you have to build things and play with them at length, and if you can't hear or you don't put in the time, then you're out of luck. Edison was right - invention is 10% inspiration, 90% perspiration."
[ 本帖最后由 davidxtb 于 2011-7-4 14:26 编辑 ] |
|
发表于 2011-7-4 10:44:47
提升卡
变色卡
