Zu Druid 6

TL; DR: VIVID! Sean finally built one for himself.

In amateur astronomy, we have orthoscopic eyepieces. Orthoscopic optics are designed for absence or near absence of dimensional distortions in the optical view, and they use relatively few glass elements for high light transmission, scant aberrations and low-to-no light scatter. Orthoscopic eyepieces are the standard for critical visual astronomy of planets, due to the undistorted detail, minimal glass and resulting contrast rendering fine details observable at high magnifications. Newer optical designs emphasize other characteristics, particularly wide fields of view for a more immersive experience in observing the night sky, but designing for wide fields trades away some orthoscopic properties to gain something more spectacular and useful in a different way, to the point where intentional spatial distortions are introduced or accepted in order to tame worse ones when field of view goes very wide.

Not wanting to go all the way down another hobby’s wormhole here, I’ll leave the matter of orthoscopics at that as a way of shifting your mindset to orthosonics for a similar agenda to design for authentic sound in hifi audio. It’s more difficult than it seems.

Last year marked 50 years elapsed since I spent my first dollar of my own money on gear to replay music. 2018 also marked 50 years elapsed since I spent the first dollar of my own money on guitars. I’ve been buying recordings longer than both. I mention this because one thing has remained absolutely consistent in all this time, and counting: at any given time, and for any given category of products, there are only a small handful that are worth buying. There’s always lots of choice, but choice isn’t the same as worth. I am aware of the full panoply of audio gear but am undistracted by most of it, because….well…..most of it at any price isn’t compelling.

Especially loudspeakers, which mostly remain very far away from representing music and the instruments (including the human one) that produce the sounds of music. If mid-to-high-end hifi buyers were primarily driven by music concerns rather than gear, or what’s cool, or whatever makes the best success statement, or whatever else you’ve got that isn’t strictly endorsing of objective music replay, the entire market could be served by just two speaker makers: Audience with their ClairAudient line, and Zu Audio. I wish I could say otherwise, but these are the only makers of speakers fully delivering authentic music fidelity today, i.e. replaying music orthosonically. There are a few additional contenders who nearly get it right, at various price strata: JohnBlue, Voxativ, 47 Labs’ Lens, Quad with their electrostatics, and Konus are some who cover between 270° and 330° of the full 360° circle of fidelity. Which is great, and that much completion of the circle can deliver beautiful sound. But only Zu and Audience have mastered making speakers that can effortlessly deliver altogether holistic music fidelity, authentically.

Send your Magicos, YGs and KEF Blades to the smelter. Torch everything Wilson, Focal, Devore, Maarten, Revel and Dynaudio, let alone those horrible JBL Century 100 reissues we see rising from the dead now. Throw every crossover-intensive speaker into the woodchipper and send the ground-up mulch to your local toxic waste facility. Fake news is mostly a trumped-up distraction, but fake fidelity – that’s real. The orthoscopic authenticity cornered by Zu and Audience is rooted in both companies having developed, within any practical considerations, uncompromised full-range drivers coupled to insightfully-engineered cabinets. In loudspeakers, adroitly-chosen materials, along with astute mechanical, resonance, acoustical, structural and construction decisions, preserve the essential simplicity of a full-range driver design, emitting music without exaggeration.

And that’s the whole distinction, because let’s face it -- most of the supply side of the hifi realm is in the exaggeration business. Moreover, much of the consuming public likes it that way! For every “hungry-ear” buyer who seeks the artificiality of being in the band or inside the piano, or gulped down by the singer to be a swallowed resident inside the body’s resonant cavity, there is a designer and manufacturer willing and eager to tilt the ratio of transient-to-tone, or the inverse, away from real and aggressively toward spectacle. It’s endemic, and buyers are as much at fault as the designers & sellers in the industry. For people who want more transient detail than is actually present in a performance or from an instrument, orthosonic speakers aren’t for them. If you want a Fender guitar amp running 6V6 tubes to sound like a Marshall stack, or if you think Johnny Cash sounding 12 feet tall is authentic, you don’t want a true-sound, orthosonic, hifi speaker.

If you need or want a surprisingly convincing speaker that can be held in one hand, buy a pair of ClairAudients. But if you want something with room-filling shove, you want Zu. So, I will leave Audience behind here, since this commentary is about Zu, particularly the Druid 6.

This year marks my 15th year of continual Zu in my hifi systems. I started with a pair of used Druids that had been updated by Zu to 2004 configuration. They turned out to be one of the first ten pair of Druids made back in 2000. Those Druids were the first convincing find in my then 30+ years quest for a convincing crossoverless speaker. The 2004-spec Druid was certainly not perfect, but it was a revelation. Imperfect because it was a bit of a narrowcaster, and was soft on the top end. A revelation for its completely coherent presentation, octave-to-octave balance, consistent transient behavior, striking intimacy, pinpoint imaging, absence of crossover-point pinching, plus it was a tone monster. Druid was effectively a dynamic driver Quad ESL-57 but with dynamic range, energy and willingness to be abused. It was the first speaker to ever pass my test for reproducing specific electric guitar/amp combinations with true fidelity. I immediately added an early pair of Definitions to another hifi system but kept those Druids on my secondary system until Druid 5 was introduced, along the way getting periodic upgrades installed, culminating in Druid 4-08.

In parallel, I made the Definition > Definition 2 > Definition 4 migration, yet the Druid never lost relevance. Definitions favored scale; Druids favored intimacy and they both overlapped on tone, coherence, speed, holistic transient and tonal behaviors and both excelled in delivering shove – that quality of musical dynamism that gives music projection and reach. In Zu’s case you get shove without gobs of power.

Druid 5 updated and expanded on the core Druid proposition but extended the top end for more harmonic completeness (courtesy of the Radian compression supertweeter replacing the older Zu tweet), deepened bass response, added greater agility and speed to the main driver, upped transparency, broadened the spatial presentation, boosted shove and put the Druid form factor in league with Definition by further killing cabinet talk. Druid 5 was the first Druid that had sufficient scale to double as a movie speaker, if the room wasn’t too big. Frankly, when Druid 5 debuted, it struck me as nearly perfect in practical terms, and about as much as one could expect from that form factor. Beyond Druid 5 lay incrementalism.

Or so I thought, then.

Enter Druid 6. Why has it taken me seven months to write this commentary? Because Druid 6 may as well be ground zero for a new vector of Zu. It completely unshackles the single-FRD / no-sub form factor of the Griewe – Druid configuration from the physical limits of its original form. Definition 1 had scale and a lot of cabinet talk. Definition 2 sacrificed some spatial scale to slay the cabinet talk. Definition 4 restored the spatial scale and tamed the cabinet talk to deliver a notably more objective speaker. Druid 6 has the spatial scale of Definition 1, which original Druids couldn’t come close to. It has the tone density trademark of all Druids and beyond what any Definition delivers. Snap and dynamic shove set a new Zu standard. Nuance and transparency Windex the entire presentation. Druid 6 is unbelievably quick, with the dynamic agility of Tiny Archibald confusing basketball opponents in his prime. And there is a fairly dramatic improvement in bass depth, character, impact and texture. I never had a real impulse to add subwoofers to any of my Druids, but then I had another system with Definitions and their powered sub-bass modules to satisfy deep bass cravings. Still, it just seemed counter to the simplicity and elegance of Druid to clutter the room with more boxes or towers to get bass fundamentals few recordings actually include. It’s probably only a half-octave further bass extension over Druid 5 but it sounds so convincing and fundamental that a sub is really only called for by a bass fetishist. All this adds up to a speaker that is a larger improvement over Druid 5 than Druid 5 was over v4-08. Which is saying quite a lot.

If you’ve traipsed through the Druids sequence as I have, that might not be your first conclusion, because on everything most obvious – high frequency response, neutrality, greater shove and improved bass, Druid 4-08 was left behind and there wasn’t a single aspect in which Druid 4-08 was better than Druid 5, other than you could buy a pair for much less cash. Druid 6 improves all these particular aspects over Druid 5 somewhat less so than the last generational shift, but that all sums, along with a specific and new Druid quality, to a new speaker that makes the greater leap over its predecessor. The new quality is “vividity." “ Yeah, that’s not a word, but I’m using it here. My TL;DR for this assessment is “Vivid.” And that is what slams your mind when you wire up Druid 6, especially after the 100 – 200 hours burn-in period needed when new.

Basil Hayden bourbon is 80 proof. Technically, that’s a little short for the category. But if I’m having whiskey before the sun is down, I often start there. It’s easygoing and doesn’t front load you for the evening. Druid 5 was like Basil Hayden. Forgiving, convincing, mellow. Works with almost any mood, music, room, aesthetic, amplification. Druid 6 is more like a 137 proof George T. Stagg bourbon. You need to be ready for it, and it’s going to seem a little loud on your palate. If you leave everything the same and just wire ‘em up, a pair of Druid 6 will hit you like a hot whiskey. Burning, flavorful….vivid. And unfailingly revealing. Druid 6 is a deity’s-honest truth of music presentation. So much so, some of you Druid aficionados out there might (will) prefer Druid 5. Depending on what you have today, Druid 6 will force upgrades, or at least changes in your system, upstream of the speakers. The most likely change Druid 6 will force is in amplification. Depending on how your DAC is voiced (don’t kid yourself, they’re all voiced) you’ll be contemplating a change there too. Phono cartridge? Could be, but less likely than DAC. Preamp, least likely to require a change unless a change in power amp argues for it. In my case, Druid 6 brought to a halt 15 years of Audion SET amplification with Druids, and really a total of 20 years of continuous SET listening in my secondary system. Why? Because the extra half octave or so of bass response is just a bridge too far for really clean, bloat-free bass from any zero-feedback SET amp I know of. SET works well with the deeper-plunging the Definition series because in the Def, a solid-state plate amp has a grip on the sub-woofer driver(s). The tonal qualities of the sub bass are derived from the power amp outputs (and characteristics) but the actual sub driver control is a product of the plate amp’s damping and grip. In Druid 5, the ~34 Hz bottom limit doesn’t reveal the limitations of low-damping factor, SET bass. My Audion Black Shadow 845 amps have excellent bass on Definitions and Druid 5. On Druid 6, bass extension is just enough further to bring SET deep bass control deficiencies to the forefront. Same with SET & PSET 300B.

So, I have to replace my beloved Audion Golden Dream monoblocks (300B PSET) to enjoy Druid 6 without distracting and illusion-undermining, zero-negative-feedback bass bloat. More on that near the end of this commentary.

But, if you’re forced into an electronics revision once in 15 years because of a significant loudspeaker advance, one can’t really complain. For the first time since first exposure to Zu Druid, I recommend alternatives to single ended triode amplification. As I will outline in a topical addendum at the end of this, I recommend single-ended pentode/tetrode, push-pull triode and some high-coherence push-pull tetrode and pentode amplifiers. Or if you’re tube-phobic, some solid-state amps one might want to consider include, Pass, First Watt, 47 Labs Gaincard, M2Tech Crosby, etc.

Druid 6 is the first Druid to serve as a convincing transducer for symphonic music. With the right amplification it will even play viciously intense metal without choking like you’re trying to squeeze toothpaste from its tube too quickly, forcing the tube itself to burst. No prior Druid could quite do that. On the other hand, Druid 6 loses nothing to prior Druids in the lone-performer-with-guitar genre, in fact drawing you into closer intimacy with the tone and textures of expression while maintaining the right spatial distance. It doesn’t exaggerate like the bulk of this industry’s loudspeaker offerings, inflating definition like you’re inside the instrument to feign intimacy.

Druid 6 is also easier to set up than any prior Druid. It might be more demanding of amplifier matching but the formerly-fussy floor-to-plinth gap height adjustment is much less hyper-critical now, for attaining correct bass. Because of its dispersive scale, getting a convincing soundstage is a less obsessive proposition in placement and toe-in. There will be rewards for taking the obsessive route, but you don’t have to. And one of the best small touches is that Druid 6 default footers are flat discs, layered with a thin polymer, and attached via a ball-mount that threads into the plinth. You can spec the plinth to be drilled through so the height of the feet can be adjusted from above with a hex wrench! Fantabulous. But the big advantage sonically is that the ball mount acts as a bearing for resonance dissipation. You can get spikes to anchor the speakers to a firm flooring under thick carpet, but for hard floors the default ball-mounted, top-adjustment, polymer-interfacing flat-disc footer is the bee’s knees, sonically and ergonomically. And it’s kind to your floors.

The design and execution emphasis on Druid 6 was the cabinet. The full range driver is improved and I won’t say that’s not important, but the complete revision to the cabinet materials, construction and the mounting of the drivers (including the FRD’s torque-tensioned anchoring to the rear of the cabinet, complementing the baffle attachment and its beefier surround), and the massive plinth are collectively a major advance in Druid resonance control, energy channeling and, by extension a murderous spree annihilating cabinet talk.

Figuratively Druid 6 wastes nothing from the signal in its transducing obligations. Of course it does waste something, but compared to what’s come before, it sounds like it doesn’t. Druid 6 is direct, declarative, clear, orthosonic. It snaps and crackles like real life sounds. Instruments and people have body and breath in correct proportion. Vocal fry sounds exactly like its coming from the Millennial voice behind it. Real acoustic guitar (if amplified, microphoned, not piezo-electric) sounds authentic in tone and in attack : resonance proportion. Voices, instruments and people are sized realistically. Everything is in high resolution in a natural way, without faked detail. Presentation is bursty and serene, strong and laid back, as is authentic to the performance.

It sounds perfect, but there’s a backside to the coin. Druid 6 is not forgiving of bad recordings, cheesy mixes, idiotic mastering. It lays bares faults in performance, recording decisions by the engineer, mastering, pressings (if vinyl). It is intolerant of runaway digititis. The manic sawing of digital compression loses the loincloth producers try to cover it with. The nasty bits of degraded fidelity traded away for expense reduction are naked for all to see.

You can ameliorate the relentless truthfulness of Druid 6 by choices upstream. A more forgiving phono cartridge or more sonically elastic phone preamp or DAC. Get the right tubes for your new configuration. Back off sheer resolution a trace with Zu Mission cables instead of Event 2. Or not. Go for the unmitigated, all-nude, vivid Druid 6 experience. Hear the beauty and the beast in modern recordings. You’ll thrill to music recorded in the pre-multi-track, pre-all-solid-state studio era. Recordings from the days of vacuum tube consoles and mics, with performers in the same room, have a holistic, tuneful sound mostly lost today, except for recordings from a few performers. Those recordings have a vividness perfectly transduced by Druid 6 and projected into your domestic space. With Druid 6, a modern, congested, over-processed, excessively-compressed assault is revealed for the noisefest it is. If you can’t handle that, you need Druid 5 or you need an obfuscating amp for poor recordings, and another amp for great ones.

Which means Druid 6 isn’t for everybody, and that’s a good thing. Because Zu can’t make as many of them as it can build Druid 5 in the same period of time. And Druid 6 costs about twice as much as its esteemed predecessor. Hence both are in the line concurrently. And this makes sense. In my TL;DR I noted “Sean built one for himself.” Sean has visited me often enough that I’ve had many hours of listening with him present. I’ve seen Sean spin vinyl of music he loves irrespective of recording or pressing quality, able to note the sonic offenses and still set them aside to let himself be infused by the tunes. And when the worst disk is done playing, he thinks about how to get more out of it.

This makes Druid 6 its own contradiction. It is the least forgiving Zu speaker ever, and yet the most fun. It’s seductive and off-putting in equal measure, depending on the quality of the source material. And yet its uncanny PRaT pulls you in to ignore what’s wrong with the source and revel in what’s right.

Zu’s finish quality is higher than ever, running with the best. The jewelry adorning the cabinetry is all functionally mandated, designed to look fab and is perfectly machined. While it sounds like this is as good as Druid can get, there’s no doubt Sean will take Druid further in coming years, but Druid 6 is a high plateau on which it can be parked, bought, enjoyed and admired while another speaker in the Zu line gets the new foundation of materials and build techniques laid down by Druid 6. In the meantime, in Druid 6, Sean Casey is making the one, true, full-range, orthosonic speaker. It doesn’t exaggerate, nor does it shade the truth. It neither spotlights nor romanticizes. It doesn’t make rough, wooly music silken, and it won’t make velvet sounds abrasive. Druid 6 presents Tom Waits and Maria Callas with equal authenticity. Yo Yo Ma and Joe Satriani are equally convincing. Frank Sinatra and Hound Dog Taylor are nothing but themselves.

A (lengthy) note on amplifiers for Zu.

If you’ve read anything prior I’ve posted here about Zu, you know I regard the amp-speaker interface and combination the fulcrum of fidelity for any Zu-based system. As I referenced earlier in having to abandon SET amps with Druid 6, buying this speaker requires careful consideration about the mated amplification. In the past 18 months another development required a complete revision to my power amplification in my Definition 4 system as well, so I have put extended effort into surveying alternate amps for both Zu systems over the past year or so.

In 2017 I had solar panels installed on my roof, and then later that year added Tesla Powerwall batteries. The panels were installed on the area of the roof directly over my living room where the Zu Definition 4 system lives. It’s a one-story rancher, wood construction house, so not much other than air, wood and sheetrock umbrellas my Definition 4 system from the shower of solar RFI. Apart from any RF emissions from the panels, the system has two wireless internet connections: a Wi-Fi connection to my mesh network by the Tesla Powerwall controller, and a cellular connection by the solar system inverter. At the same time, to accommodate the exterior Wi-Fi needs as well as prepare for wireless Roon endpoints to two separate systems, I changed my house Wi-Fi from a Google wireless router to an Eero mesh network.

The result was that my SET amps in the Definition 4 system proved perfect antennas for the shower of RFI bathing my living room, and no tactic for quieting that worked. The first step was to get an active preamp out of the system, so I sold off my crazy-good Melody p2688 tube preamp, which removed about half the problem. It was replaced by the splendid Luxman AT-3000 TVC from the early 1990s – a real work of magnetic art. Once that was added, I had quiet with push-pull tube and solid-state amps. At the time, I still had Druid 5 on the other system, and its location sharply contained the new RFI problem, so my Audion SET amps could still be used there. Then Druid 6 arrived to undermine my commitment to Audion SET. So, now what, for amplifiers?

About a year ago, after years of great difficulty finding a phono preamp that can make an Allnic Puritas phono cartridge sound correct (Allnic’s own phono preamps do not), I ventured a radical experiment to try the M2Tech Joplin Mk2 phono ADC, which I mated to the M2Tech Young Mk3 DAC, both powered by the Van der Graaf linear power supply. For this signal, The ADC converts analog to 24/192 digital and the DAC converts the processed phono analog digital signal to analog at 24/192 decoding. With RIAA done in the digital realm, and gain adjustable by 1 db increments, and a good range of options for cartridge loading, I finally got the Puritas to sound musically convincing. This led me to take a flyer on the M2Tech Crosby power amplifiers as a temporary fix to my Definition 4 power problem so I’d have something good to listen to while I took my time trying alternate tube amps. I bought two Crosby to run as bridged monoblocks. These are Class D amps using ICE modules with an M2Tech proprietary input section for better sound than most ICE-based amps. Into 8 ohms, one Crosby outputs 60w/channel. Run as a bridged monoblock, Crosby outputs 180w. Into the Def4’s 6-ohm load, Crosby should be good for ~270w each.

With a pair of Crosby amps and the Luxman TVC in the Def4 system, I had a quiet system again. In fact, dead-quiet. Quieter than ever! Without the RFI-induced gurgling, sputtering, spitting and whirring coming through my RF-antenna SET amps, I could listen in peace and embark on a tube amp odyssey as I had time.

Chalk up the M2Tech Crosby Class D amp as wonderfully-Zu compatible. This is the best Class D sound I’ve heard, period. Maybe only rivalled by the 47 Labs Gaincard, which has much less power. I prefer these to most bi-polar solid-state amps, only a few of which have somewhat better musicality, usually at less output. Such amps are generally associated with Nelson Pass. So if you are tube phobic or have any other reason to need or prefer a smooth, bursty, dynamic, musically-convincing, high-definition, solid state amplifier for Zu, consider the relatively affordable M2Tech Crosby.

Back to tube amps. At the same time I was considering tube amp alternatives to my long-time Audion SET amplifiers, I also have to note that I regularly am contacted for advice on affordable tube amps for Zu speakers as well. Now, I have a pair of Quad II Jubilee monoblocks, which were the last edition of Quad II amps produced in the UK. They sound quite good on any Zu speaker, so when I got Druid 6 and found them not ideal for SET in the bass region, it was easy to slip them in the Druid 6 system, or to move them over to Definition 4 when I wanted some tube-amplified music. The Quad II pair are always a great backup amp for me, and the closest-to-SET seamlessness I’ve heard in a push-pull amplifier, simple circuit and all that.

But Quad II is built for KT66, with 6L6 an alternative. I had mine pretty tweaked through tube selection: Tubestore Preferred Series 274B rectifier, Mullard mesh plate EF86, Sylvania NOS 6L6 long ago sourced from Mesa Boogie in the form of their legendary STR-415 power pentode. As an alternative I also used cryo’d Tube Doctor KT66. The Quad II sound seriously good but still only 12-15w and the pair could be more elastic. I would love them to have more of the Plasticman kaPow! reach & slam that my Audion Black Shadow 845 SET amps muster.

My system racks are setup for monoblocks, so all stereo amps are ruled out. Meanwhile, as I started my tube amps odyssey, more requests for affordable amp recommendations trickled into my email. Which is why I want to let you know about Ling Xiao Nan and his hand-built, affordable, amplifiers. Xiao Nan designs and builds his amps in the Guangzhou region of China. He is self-taught on the subject of vacuum tubes and amplification, and been building amps for over a decade. He’s also a guitar player, but he’s proud to point out that he is a “full-time amp builder.” You can find Xiao Nan offering amplifiers on eBay under his Tube Fantasy brand, but that’s only a fraction of what’s possible with him.

Online, Xiao Nan generally sells clones of vintage circuits, particularly pre-war and post-war German cinema amplifiers, the Quad II and the Williamson designs. The circuits are faithful with only a few component value deviations for modern speakers, particularly in cathode capacitors. When he is cloning an amp circuit, he builds his transformers to the original electrical spec but uses his own preferred winding techniques. Xiao Nan machine-winds his transformers on his least expensive amps, and hand-winds them in everything else not-very-much-more expensive. He keeps cost down first by building in China, but also by keeping to a Quad II-sized chassis for most of what he offers, in black or natural aluminum, or a similarly-proportioned upsize for designs with larger transformers. Ling Xiao Nan builds monoblocks.

I first learned about Tube Fantasy when I was out on my patio on a cold-for-Los-Angeles night in December with a glass of Corbin Cash Rye, Tidal on my iPhone through my i.am+ Buttons, idly flipping through tube amps on eBay. A listing for a Quad II clone pair at a ridiculously low price stopped me cold, especially since I had gotten a request to recommend a cheap tube amp for Zu earlier that day. They were so inexpensive I decided to do the Zu community a public service to evaluate them. If someone was selling a pair of Quad II clones for a few hundred bucks, I had to hear them, for better or worse.

I’m going to comment on those Quad II clones shortly, but in email correspondence about the sale and in asking some questions I had about the amps, I quickly got to know Xiao Nan beyond the usual eBay transaction, and in doing so I learned about other amps he makes. Perfect. I now had a custom amp builder to let me affordably tour circuits and tube types to bracket myself and zero in on where to finally land in replacing my Audion SET amplifiers. This led to me acquiring the following monoblock pairs in the span of 3-1/2 months: Quad II clone, Williamson push-pull, 2a3 triode push-pull, Klangfilm KLV-204 clone (F2a beam power tetrode-based) single-ended tetrode, LS50/GU50 single-ended pentode, 300B-or-2a3 push-pull. And I am considering getting a pair of Telefunken V69 clones.

At some point, some of these amp pairs will be sold, but the luxury of auditioning a range of circuits and tube types on a relatively modest cash outlay is considerable. And you know what? Xiao Nan might make you a pair of stepping-stone amps, or he might make you the last pair you need. He certainly can. Xiao Nan tends toward simpler circuits, high quality parts, emphasis on transformers and pentode/tetrode designs, though he builds triode push-pull anytime you want. He is quick to build, his soldering is clean and careful. Nothing is needlessly bulked up for faux masculine appeal.

Ling Xiao Nan and Tube Fantasy are in many respects the Sean Casey and Zu Audio of tube amps. Both represent strong points of view, are music-driven and constantly tinkering for better sound, and delivered value is high.

Here’s a sampling of Ling Xiao Nan / Tube Fantasy prices, monoblock pairs, shipped to USA (DHL), amps only, no tubes:

QUAD II Clone (18w), $499

Williamson Clone (25w wired triode / 45w wired pentode), $688

2A3 push pull triode (13w), $688

Klangfilm KLV204 Clone, single-ended-tetrode (10w), $688

LS50/GU50 Direct Coupled, Single-ended Pentode (13w), $1000

In cases where Xiao Nan supplies any of the tubes, he specifies so. Otherwise, you source tubes and stuff the amps when you get them. Every amp Xiao Nan builds gets burned in and listened to for five hours before shipping. All my amps have arrived trouble-free electrically. One had stripped transformer cover threads due to customs overtightening during disassembly-reassembly under inspection, and they also damaged a faceplate. Xiao Nan immediately ordered replacements for me.

Xiao Nan does not pot his transformers both for maintainability and because potting compromises high frequency performance. You may hear a trace more mechanical hum up close than from, say, a true Quad II. But music-on, this is irrelevant. The amps are quiet electrically. Common to all of Xiao Nan’s amplifiers is exceptional transparency, very high definition, fast transient speed, exceptional soundstaging and depth. For whatever reason, with any of his amplifiers, I am getting more spatial depth in the soundstage than with any amplifiers I’ve had connected to my Zu systems, regardless of price, and that includes the many amps visitors have brought to audition on Zu speakers. That fast, transparent, dynamic, tuneful sound I valued particularly in Audion SET is here. Further, in the push-pull amps, crossover notch grunge is vanishingly low. Some notes on specific amps:

Quad II Clone: Fast, open, dynamic sound. Because Xiao Nan winds the output transformer differently than Quad did, this also changes the feedback behavior somewhat. The result is greater audible difference between KT66, 6L6 and EL34 in the clone compared to genuine Quad II. With EL34, the clone is relentlessly revealing, hi-def and blindingly fast, which is great for excellent recordings/masters/pressings, but not forgiving of flawed wax or digitalis-digital. With KT66 the clone is closest to the original Quad II sound, which was not strictly romanticized but does possess some vintage warmth. The perfect balance is found with the NOS 6L6. With this power tube, the clone is smooth, defined, dynamic; delivering Druid-like tone density and lots of dynamic punch for its power. The amp takes some time to burn in and bloom. About 120 hours will do it. It will sound thin in the bass region until then, but progressively less so.

Williamson (Triode-Mode): Dynamically stronger than the Quad II or its clone. Refined, smooth, high definition, solid deep bass on Definition 4 or Druid 6. Same basic sonic difference between EL34, 6L6, KT66, as in the Quad Clone. Again, I settled on NOS 6L6. In the Williamson, the Quad II and QII Clone, my preferred rectifier is the Tubestore Preferred Series 274B. It gets the most tone, space, definition and dynamic punch from all three amplifiers. This is an exceptional tube amp, that lives up to the Williamson amp legend from shortly after WWII. In fact, it’s the best Williamson implementation I’ve heard and easily walks over a Marantz 8B.

2a3 Push-Pull: Inexplicably bursty and elastic for its 13 watts. And my pair have only 8 ohms windings. On 16-ohm Druid 6 they sound dynamically huge. Excellent bass discipline compared to any SET amp, while retaining all the triode magic, though still quite objective. No slow, lazy, old-school triode amp sluggishness. This amp, Xiao Nan’s custom circuit at my request, gives up nothing in definition, speed and punch to the tetrode and pentodes in his other push-pull amps. The rectifier you want is NOS 5v4.

Klangfilm KLV-204: This is an early post-war German cinema and studio amp. It was used as a monitoring amp in recording studios (so high definition required) and as a reserve amp for cinemas using the larger KLV-402 & 502 amplifiers. It lives up to its high definition requirements. I have known about the F2a tetrode tube for decades and never managed to get my hands on any of the Klangfilm or Telefunken amps that used it. I was excited to get a chance to hear the Shindo Cortese, which promptly disappointed me like most other Shindo amps I’ve heard. But this KLV-204 clone is exciting to listen to! Only 10 watts (I had my pair wound for 16 ohms to use with Druids), it sounds dynamic beyond its means. No detail gets past it. Midrange is tone-dense like a great triode but burstier. Among the quickest, most agile amps I’ve heard at any price. Like Druid 6, this KlangClone is vivid, tonally, dynamically, and spatially. With Druid 6 it is musically amazing on an excellent recording, but the x-ray truth of the combination can render a poor recording too distracting to listen to. Badly recorded or mastered bass, particularly, shunts the illusion of musicality. This is one of my favorite amps ever heard, but it is a specialist and Xiao Nan builds it affordably enough to use it selectively. The F2a tube, btw, is a German Post power tetrode built for 10,000 hours of life. They often go longer. As NOS and vintage tubes go, they aren’t cheap, but they aren’t ridiculously expensive, either, given how long they last.

LS50/GU50 Single-ended Pentode: More like the sound of the Klangfilm KLV-204 than any of the others. Crisp, clear, fast and transparent. Also dynamic beyond what’s expected from its mere 13 watts output. Clean top-to-bottom. A little bass shy compared to the push-pull amps but there’s a remedy. In most tube amps, particularly those with cathode bypass caps, there is a trade-off between ultimate bass performance, and ultimate top end. When faced with this, Xiao Nan prioritizes top end linearity over bass linearity. But making adjustments to the chosen values for cathode bypass caps can land you in your zone. You can do this after getting the amps or if you communicate to Xiao Nan your general preferences, he can reliably adjust for the right value. You just have to communicate effectively. Overall, Xiao Nan’s LS50/GU50 single-ended pentode monoblocks are sonically in the same realm as the F2a-based Klangfilm KLV-204 clones, but not quite so MRI in unsparing revelation. The LS50 has more latitude in acceptable recording quality than the KlangClone. Given the Definition 4 sub-bass extension, Xiao Nan suggest snipping the cathode bypass cap on the E180F driver tube to bring bass performance into better balance with the speaker. But I have to say before doing that, the intrinsic bass from this amp on Def4 maybe a little light, but it is very well presented in terms of character, definition and attack.

300B/2a3 Push-Pull: I commissioned Xiao Nan to build for me a pair of monoblocks that can use either 300B or 2a3 output tubes, in push-pull configuration. I have a quad of KR 300B Balloon tubes held back from my recent sale of my Audion Golden Dream 300B monoblock amps, which I wanted to use in push-pull configuration with either my Druid 6 or Definition 4 system to make continuing use of those magnificent tubes. Xiao Nan indulged me with a custom design. Those amps are arriving in the next day or two. Comments pending if anyone is interested.

Telefunken V69 Clone: Probably last in my series of bracketing amps from Xiao Nan will be a pair of clones of the legendary Telefunken V69 Cinema amp. This is a full Class A, push-pull, tetrode amplifier using a pair of the F2a tetrode tubes in each monoblock. Like the Klangfilm KLV-204, input and driver tubes are the equally-revered EF-12 small-signal pentode. Class A output is 25 watts per monoblock. I will decide in the next day or two whether I will order these, put them in the mix and then vet the whole shebang for what stays and what goes.

You can find Ling Xiao Nan on eBay as seller “fatkit83-8.” If you private message me, I can offer his direct contact information for amp inquiries.

Of course the genuine Quad II remains a highly-viable tube amp for any Zu speaker. It is coherent, musical and also delivers more dynamic shove than its diminutive specifications might lead you to expect. Ling Xiao Nan gives you alternatives, and plenty of them, for less cash. And similar to like-minded Asian originators like mhdt (DACs), Jasmine Audio (phono preamps and amplifiers), and Melody (preamps and amps), all of Ling Xiao Nan’s implementations are informed by music listening and in-depth inquisitiveness seeking more fidelity than he previously delivered at any given time. Just like Sean Casey.

Ask questions if you have them.


Hi Phil, Marc here. Fascinating review, a lot to take in, esp yr conclusion that SETs reached their limitation on controlling the early 30s bass extension of the VIs.
Hi Marc,

Well, at least you have it now!

SET well done is fine down to ~34Hz. It works in Definition deeper because the driver control is handled by the solid state amp. Druid 6 challenges the zero NFB SET bottom end discipline of SET directly controlling a driver. I didn't want to reach this conclusion, but the truth outs. The upside is that after all these years, it forces a new amplifier adventure in discovery!

Phil, this is of particular interest to me, because as of currently, I'm listening to my Defs IV sans subs.

A combination of impossible to eradicate plate amp hum issues, and my inability to truly ever integrate the Defs subs in two rooms now, but my new room allowing the FRDs to extend lower, means that FTTB at least, I'm using my Defs a little like Druids w two FRDs and no subs support.

So if Druids VI truly break the 35 Hz barrier in room, they may be a better choice for me than Definitions VI, Experiences I or Dominances II.

Too bad this could force a move from SETs. I am eyeing the new Pass Labs amp putting out 30W that has been glowingly reviewed, maybe combine that w LTA MicroZOTL based pre.
Check out the XTAC by Audiomachina and you will see most if not all of your concerns addressed properly and without compromise. 
I have checked and "fatkit83-8" is NOT a seller on EBay. And "Tube Fantasy" produces nothing of audio worth on EBay. 
This whole original thread reminds me a great deal of the Tekton spankfest threads of late. 
>>I have checked and "fatkit83-8" is NOT a seller on EBay. And "Tube Fantasy" produces nothing of audio worth on EBay.<<

Funny, he's right there on my eBay app. He doesn't always list his brand name.

>>Check out the XTAC by Audiomachina and you will see most if not all of your concerns addressed properly and without compromise.<<

[Laughing hysterically...] No, that's exactly wrong what I don't want.

>>So if Druids VI truly break the 35 Hz barrier in room, they may be a better choice for me than Definitions VI, Experiences I or Dominances II.<<


I don't think a Definitions owner will want to give them up for Druid 6. Especially after reading everything you've done to reach your current music equilibrium. But are you saying you essentially have the Def4 sub turned off? I can't explain your sub amp noise. They are absolutely silent in my Def4s. 

Spirit, note that Nelson says the SIT-3 amp prefers a lower impedance speaker, so you may want to get load resistors from Zu if you go that route (see Herb's SIT-3 review).  Sean told me he uses the resistors when demoing with Pass amps at shows.

FYI, I run the a similar setup to you in one of my systems as I have Zu DefHeads.  Sounds killer with my cheapie 3w SET and I don't miss the low bass at all!
Sean, my conversations w Sean haven't really got to grips w my issue. But that's seperate from subs integration issues.
So you read the actual description of these speakers and the white paper? Or did you looked at the photo and made an informed decision? I treated your post with respect although my initial impulse was to do the opposite. Read your original post again and then ask yourself who should be laughing hysterically. I am calling you out and I want you to tell me precisely what about the XTAC that caused you to laugh hysterically. BTW I have heard the Zu speakers many times both at shows and in owners rooms.  
Because the XTAC takes much of what is good about the Zu design and takes these to a much higher and logical level. This of course assumes that Phil even knows anything about the design of these speakers and what is good and bad about the Zu designs. I have a very hard time taking the OP seriously but I am trying to be kind and would genuinely like to discuss things with him at greater length.
>>So you read the actual description of these speakers and the white paper?<<

Yes, I read the white paper. I agree with most of the diagnostic observations about what's wrong with speaker types to-date. Further, the author's time-domain focus, and his dislike of crossovers parallel views I've articulated for decades.

OK, I was a little flip citing my amusement. This is a new thread, hence it might attract some folks who haven't read anything else I've posted here in the past. If you had, you'd already know that I don't agree that ganged drivers and amps in a line-array implementation are best advancements against the problems outlined in the Doctor's diagnostic summary.

This approach just introduces a new irritant or distraction to replace older ones it solidly attempts to eliminate. It's also for most people impractical. But it's also an honest effort and I don't question the designer's sincerity. If XTAC makes sense to you -- and your aural mind when listening -- enjoy the system.

A few further random notes:

1/ Bob Hovland offers an isolation transformer that improves the sound of amps I've tried with it. The design principle is that if an amp is fed power from a balanced power isolation transformer that is designed to operate close to its saturation point, the connected amp will sound, in Bob's words, "more relaxed." I auditioned a pair of Bob's proprietary isolators, found this to be true, and bought them. Each one is a simple black, crinkle-finish box, about 6.75"d x 5"w x 4.25"h. Using Bob's isolation xformers improves dynamic elasticity and removes an anxiety factor that you can put your finger on when you first remove Bob's iron after an initial extended listen. The effect improves further with burn-in. It's not dramatic, but it's musically beneficial. This is relevant to Zu owners, since most use relatively modest power amplifiers. The isolation xformers are rated for 2.5A at 120VAC, safely handling peak demands up to 4A. They are favorable to every tube amp I've tried them with, and with my M2Tech Class D monoblocks as well.

2/ Ling Xiao Nan can also make an upgraded Quad II Clone, using a higher-spec, larger output transformer. This also requires his larger chassis, so cost is higher than the standard true clone. I haven't heard it but he built a pair on a hunch, and reports that the uprated QII Clone now sounds better than the Williamson, and compared to the standard clone, delivers deeper, better bass with a more relaxed top end.

3/ I recently learned something new about the Quad II, explaining its particular synergy with Zu Druid. The Quad II has a multi-tap OPT but the chassis only has room for a pair of binding posts. Most Quad II have their outputs wired to the OPT 8 ohms tap. So why does the amp work so well driving the 16 ohms Druid? The curiosity of a gentleman in Australia answers the question. From the 8 ohm tap, measured at 1 kHz and driving 8 ohms, Quad II peaks at ~20.5w, THD < 1.5%. Peak power is actually produced at 10.5 ohms, reaching just shy of 22w. The 8 ohms tap driving 8 ohms makes up to 7 watts in Class A. However, if the 8 ohms tap is driving 16 ohms, peak power declines only slightly to ~18.5w, but ~14.5w are available Class A. If you use Quad II on Zu Druids, any version, don't have your output terminals wired to the OPT 16 ohms tap. If you did that so the 16 ohms tap drives 16 ohms, power would peak at ~21w but only ~7.5w would be available Class A.

You can get similar benefits for 6 ohms Definitions. There is an undocumented, "illegal" 4 ohms tap on the Quad II OPT. If you are driving Definitions with Quad II, make the change. The 4 ohms tap driving 4 ohms yields peak power just over 20w, but only 8w are available Class A. However, the 4 ohms tap driving 6 ohms causes only a modest peak power decline to 18w, but up to 12w are available Class A. If the 8 ohms tap drives 6 ohms, peak power is the same ~18w, but only ~5.5w are available Class A.

The author of the study does not recommend using the Quad II 16 ohm taps for any load less than 32 ohms, and there aren't many such speakers. But I have to point out that the QII 16 ohms tap would be sensational for double LS3/5a (15 ohms version) wired in series for 30 ohms, ala the mid-70s Absolute Sound Double Advent System. I had a Double LS3/5a system, powered by Julius Futterman OTL amps that peaked at 32 ohms, 40 years ago.

Here's the link on Quad II details, with lots more than I described above:


Phil, your comment on the Hovland isolation transformers. I use balanced power in my system, the Westwick Audio 8K tranny, providing 8kVA on tap. 

Are you suggesting I use Bob's black boxes to my subs, esp w my perennial hum issues?

Btw, yr thoughts on opting to not use the onboard Definitions Class D subs, and going to seperate outboard subs.

The advantages being total control of placement for smoothest in-room response and not being restricted to subs only where the main towers are, getting away from the inherent issue I have of Zu down-firing subs onto a flexy 35x30 Victorian floor built c.150 years ago, and potential benefits dispensing w Class D and going to Class A/AB.

I agree your vast suspended, older, flexy floor presents a special challenge for bass energy, especially for a down-firing sub arrangement. I don't think sub-bass performance is assured to be better with Class A/B than well-implemented Class D, but it's possible.

I've found it is impossible to know with certainty whether *any* power supply precursor will solve a noise problem, hum or otherwise. You just have to take an educated guess and try something, fully prepared to be disappointed. But Hovland's isolation transformer works and they aren't monstrously expensive.

Phil, I can't find any online reference for the Hovland transformer. Can you point me?

Bob builds to order and he doesn't have a web site to order from. Email Bob at roberthovland@earthlink.net.

The best way to "find" Ling Xiao Nan is to google fatkit83-8. For some reason, nothing appears when fatkit83-8 is put into the eBay website search engine. 

I have heard the Druid VI at Capitol Audiofest, and it is clearly the best Druid yet. However, not sure that I would trade my Definition 4s (driven by an 300b SET), especially in my huge listening space - although the fit and finish and cabinet design of the Druid VI is a step above the Definition 4. 
GSM, my Def 4s are in a big space too, 18x48, 800 sq ft/5000 cub ft.
Concur on Druids maybe struggling.
roberthovland@earthlink.net is definitely Bob's email address. Try it again and not it's not a .com, but a .net. Ancient email provider here.

If you have further trouble, I'll have Bob email you.
Phil, any thoughts of the Cube Audio Nenuphar, which like the Druid 6 uses a single FRD, but no supertweeter, and promises bass extension similarly down to mid-30's Hz.
I've had no opportunity to hear the Cube Nenuphar. I'm aware of it. By all accounts, compared to comparable Zu it lacks bass foundation, and it is only93db/w/m efficient, 8db less than Zu. So for a given amp the burst factor will be reduced. On the other hand the driver is reputed to be quick and agile. So, no basis for a fact-based response. But once fully acclimated to Zu, I'd be very reluctant to give up 8db of efficiency in two speakers vying for crossoverless supremacy.

Oh, and 50% pricier than Druid 6.
Phil, has Sean ever talked about the possibility of doing away w the Radian 850? Ie finding a FRD that covers beyond 11-12kHz all the way to/past 20k?
Well these are not typical array speakers and the special amplification is critical. By using drivers and different size you open up a whole series of problems which are avoided by using multiple identical drivers. What I find so interesting are the speakers you tout as having solved the most problems. Now I havent heard some of the newer Zu speakers but the old  ones were rife with problems some of which appear to have been addressed in the new lines. I guess I will have to spend some time in this room at Axpona. To date not always the best sound but among the best guys with probably the best music. 
As a Vietnam vet and infantryman  the only thing  M-60 means is the
U.S . Army's deadly machine-gun .If this was done for a laugh, it ain't funny .
I'm not sure why a $75k+ Audiomachina speaker is being brought up on a $10k Zu thread.

Post removed 
The M-60s are the 60 watt per channel OTL amps that are made by Atma-sphere. Although Ralph the owner is a really good guy and far more tolerant than I, I doubt you will get an apology from him for choosing this name for his amp. Thanks for your service but really you are way out of line!
>> Well these are not typical array speakers and the special amplification is critical. By using drivers and different size you open up a whole series of problems which are avoided by using multiple identical drivers. What I find so interesting are the speakers you tout as having solved the most problems. Now I havent heard some of the newer Zu speakers but the old ones were rife with problems some of which appear to have been addressed in the new lines. <<

I fully understand what they are. I also grok that the "special amplification" is critical to the design, if one believes it's critical to own that design in the first place. It's still an array speaker. I don't agree that ganged drivers and ganged amps are desirable. And in any case, the corners implementation is a serious limitation. And as asked elsewhere, why is this speaker relevant to a Druid 6 thread?

Yes, drivers of different sizes introduce or leave unsolved some problems. Druid 6 has one driver covering most of the musically-relevant frequency range, with a supertweeter for harmonic completeness. Simple, really. Is it perfect? Since it's a loudspeaker, no. It's just what I say it is.

Were 2004 Druids "rife with problems?" Perhaps. All speakers have been "rife with problems" compared to an ideal. Zu was just less-rife with problems than most of what else the market offered then. There are also differences between theory and practice. If you think another speaker is better, you should buy it. Any comments I made about Zu in 2004 and 2005 were made in that temporal context. Any comments made in 2010, 11, 12, 13, 14, same. Now it's 2019. A 2019 Zu Druid 6 has come a long way since a 2004 Druid Mk "2.5". It is theoretically possible to design a better speaker. In practical terms where are they? Very, very scarce, are these theoretically better speakers.

What I can say about the Audiomachina speaker is that a ganged array introduces its own problems, which I don't want in the mix. And in any case the corners implementation scheme is unusable to me even if I wanted to accept the results of a line array. The speaker might make someone else perfectly happy! Perhaps you, jsautter. As I wrote, Druid 6 isn't for everyone.

Post removed 
I didnt write a treatise on a particular speaker claiming it to be the second coming discounting other designs in the process. I didnt start a discussion on the XTAC. See if you see some similarities between your thread and the recent Tekton threads.
Post removed 

Perhaps but I am not, to date, a fan. If Phil likes SETs on the Zus he really should hear your M-60s powering these speakers.
Post removed 
>> I didn't write a treatise on a particular speaker claiming it to be the second coming discounting other designs in the process. I didnt start a discussion on the XTAC. See if you see some similarities between your thread and the recent Tekton threads. <<

OK, I see what's eating you. I discounted other designs. Well, I surely did. I've been discounting, for example, crossover-based speakers for eons. I have no idea what's going on in Tekton threads and if there are similarities between this thread and those, it's entirely coincidental.

You are certainly welcome to write a treatise on Audiomachina. Somehow you seem bothered I wrote something long about Zu. It's just a review on an audio forum. What could be less consequential to the world at large? Put in the same time and effort I did with Druid 6 and lay out for us what you hear from Audiomachina. That's what the forum is for.

I don't make any claim to being neutral about speaker designs. You won't find me putting any real time and effort into reviewing crossover-based speakers because they are all seriously compromised and it's intrinsic, not just a matter of execution. I'm also on record long-term that the hifi industry broadly went off track a few decades ago, leaving very little of real value worth buying in the marketplace. So no one should be surprised by this Druid 6 commentary. If Audiomachina is a convincing speaker to you, make your case if you want, but otherwise certainly enjoy them!


>> If Phil likes SETs on the Zus he really should hear your M-60s powering these speakers. <<

I have. Quite a good combination.
Post removed 
When we compared SETs to the M-60s it wasnt even close. Not a single aspect of the presentation of the SETs or Push Pulls for that matter were even close. The P.P.s from the outset were vague and the SETs homogenized everything at even moderate volumes. The speakers used were 96db eff. and I believe a nominal 12 ohm load never going below 8. I havent spent enough time with the XTAC to really write anything. I will say that if the XTACs are correct then everything else is wrong. Pretty startling the differences between these and all other speakers I have heard or owned. The XTACs do not sound like traditional arrays.  
I don't know who "we" are. but unless you were comparing M-60s to Audion Golden Dream, Black Shadow or Golden Night specifically representing SET, I don't really have an opinion. There are a limited number of good sounding SET alternatives. Audion really stand alone in SET, which I will go so far as to say that the Audion SET amps are much more like Atmasphere OTL in speed, neutrality, transparency and soundstaging, than they are to other SET amps. Similarly, there are great push-pull tube amps but many more are mediocre to just OK. So without knowing exactly what amps represented push-pull, it's hard to know what you heard. I like the M-60 on a variety of listenable speakers including any Zu Druid. For a long time I had Futterman monoblocks made by Julius himself -- I picked them up from the living, breathing maestro in 1975. They were my main amplifiers for close to 15 years. But I digress...

It is not for reasons of sound quality that M-60s are not in my hifi systems. But overall, if you heard push-pull tube amps as vague, and SET as homogenizing, then question first the p-p and SET amps you chose to audition. I certainly could select amps that commit those sins, but I wouldn't own them.

I have in the past made the observation that Zu is right and that makes most other speakers wrong prima facie, so I accept the postulate wrt Audiomachina. I don't know what you think "traditional arrays" sound like, as every array design I've heard over several decades sounds significantly different from one another. But an array of ganged drivers and amps is still an array, and still has the challenge of acting in true unity. Nevertheless, all of this is moot to me as the corner design is unimplementable in any dwelling I've ever lived in or expect to. And I think a lot of people are in the same boat.


We used 300B amps by Audio Note and Mastersound. A 32B integrated amp by Mastersound (branded Vaic) which really came alive when we bypassed the internal preamp with a CAT. The PP amp was a VAC phi 200. All comparisons were done with the CAT ultimate preamp until we put in an Essence Jasper preamp made and designed by Dale Pitcher over 25 years ago. The effect the Jasper had on this system is the subject for a later date. 
>> We used 300B amps by Audio Note and Mastersound. <<

With those amps representing SET, I concur with your assessment.

>> The PP amp was a VAC phi 200. <<

Oy. That's trouble.