This review page is supported in part by the sponsor whose ad is displayed above

Reviewer: Srajan Ebaen
Source: Zanden Audio Model 2000P/5000S; Opera Audio Reference 2.2 Linear
Preamp/Integrated: ModWright SWL 9.0SE; Music First Audio Passive Magnetic; Bel Canto Design PRe3
EQ: Rane PEQ55 active merely below 40Hz
Amp: 2 x Audiosector Patek SE; Yamamoto A-08S; Canary Audio CA-308s; FirstWatt F3 & F1; Bel Canto e.One S300; Eastern Electric M-520
Headphones: AKG K-1000 w. hardwired Stefan AudioArt harness; audio-technica W-1000
Speakers: Zu Cable Definition Pro in custom lacquer; Anthony Gallo Acoustics Ref 3.1
Cables: Zanden Audio proprietary I²S cable, Zu Cable Varial, Gede, Libtech and Ibis; Stealth Audio Cable Indra, MetaCarbon & NanoFiber [on loan]; SilverFi interconnects; Crystal Cable Reference power cords; double cryo'd Acrolink with Furutech UK plug between wall and transformer; Crystal Cable Ultra loom [on loan]
Stands: 1 x Grand Prix Audio Monaco Modular five-tier
Powerline conditioning: 2 x Walker Audio Velocitor S fed from custom AudioSector 1.5KV Plitron step-down transformer with balanced power output option
Sundry accessories: GPA Formula Carbon/Kevlar shelf for transport; GPA Apex footers underneath stand, DAC and amp; Walker Audio Extreme SST on all connections; Walker Audio Vivid CD cleaner; Walker Audio Reference HDLs; Furutech RD-2 CD demagnetizer
Room size: 16' w x 21' d x 9' h in short-wall setup, with openly adjoining 15' x 35' living room
Review Component Retail: $13,995/pr

To kick off in full color -- 300B lore nearly demands it -- moon man Edward Barker hasn't been cuckoled by his wife. Rather, he's been canaried by Californians. Nope, that does not imply anything untoward. No being shanghaied to San Pedro or shystered out of his purse. All it means is that he's turned into a highly satisfied owner of Canary Audio's CA-903SE four-box preamp and a pair of Canary Audio CA-339 monos. Edward calls this particular valve kit second only to the truly esoteric, Kondo in his case. But unlike Kondo, it's still within reach of mere mortals. Going for broke since and immortality, British Kondo importer Kevin Scott of Definitive Audio has been hovering by the side lines, awaiting Edward's wallet to give birth -- rather than out -- so that his standing order for a Kondo preamp and matching amps may get dispatched. Because the lure of the ultimate has already become a factor in Barker's current Canary condition, today's review of the very same amps that still grace his rig fell on me. All 140 lbs of it. I volunteered for two specific and interrelated reasons. First, Edward tells me that in the circles he travels, Canary Audio has garnered a rather enviable rep. For their push/pull 300B amps. Upon learning that I own a pair of Canary CA-308s, our man in the UK didn't volunteer any informed opinion. Nothing whatsoever.
Considering the nearly fanatical fervor SETs inspire over their push/pull brethren -- in the circles I travel in -- this upside-down scenario stopped me short. If the UK was hipper to Canary Audio's push/pull 300Bs than single-endeds, did they know something I didn't? Did this explain Canary's catalogue which majors in push/pull amps and very much minors on single-endeds? Which segues into the
second point. If so, what better testing ground than matching my 300B SET monos against a P/P equivalent from the very same maker? This had near ideal makings to learn something that wouldn't really be accessible any other way.


Sure, Manley Labs actually builds such a switchable-on-the-fly beast with 300Bs. In fact, I've reviewed it. Yet detractors and personal doubts argue against it. A circuit topology and output transformers which are supporting both SE and P/P operation couldn't possibly be truly optimized for either. The juxtaposition I had in mind -- of CA-308s vs. CA-339 monos -- undermines that argument. You could still debate that for an even tighter match, the push/pull entry should be "single-ended", i.e. not use paralleled devices per phase just like our proposed SET of choice avoids. But, Canary wanted their quartet 300B amp reviewed, not the twin.



My Zu Definition Pro speakers boogie down the house on considerably less juice than an 8-watt single 300B. The amplifier comparison thus wouldn't be unfairly weighted against the smaller amp. Because these speakers are rated at 300 watts power handling -- meaning they can and will take it unlike other high-sensitivity designs -- the 50-watt push/pull amps wouldn't find themselves disadvantaged either. And coming from the same design house, both mono pairs would share transformer and other parts vendors to equalize inherent differences enough. There'd be sufficient merit for such a comparison in the first place.


Single-ended KT88 or EL34 amps are far rarer than push/pull variants. Exactly the opposite holds true for 300Bs. Perhaps the most famous of all 300B push/pull amps -- in the US at least -- has been VAC's 30/30 in its various incarnations. An evolved concept thereof presently sees itself revived in Art Audio's forthcoming Vivo. That's authored by Kevin Carter who had a hand in the later versions of the VAC amp. He was commissioned by Art Audio's Joe Fratus to add a high-power 300B amp to his stable that wasn't a paralleled SET like Wyetech's 18-watt Sapphire for example.



Canary Audio meanwhile has reserved the concept of paralleling 300Bs for phase splitting. They've taken that to an arguably unreasonable limit with their $27,995/pr Reference One. To wit, a pair of 80-watt monos. How many 300Bs does it take to make eighty conservatively rated 300B watts? Four pairs - per push/pull channel. You and I both feel the pain. Instantly. That's a ream of paper on output tubes alone. Especially since they're not included in the purchase price. And today's CA-339s still use half that stout amount. Tube wise then, a single Reference One becomes a pair of CA-339s. And a single 339 mono becomes a pair of CA-330 [$8,250]. Those are Canary's entry-level 300B monos. The CA-306 is the $5,195 stereo version thereof.



The lineup of 300B Canary amps thus progresses from one valve per side in the CA-308 SETs to two per side in the stereo and mono push/pulls, to then four and eight. Not ohms but tubes. Some 300B vendor somewhere is very happy with this state of affairs. And while Western Electric might hope for your business, the $4,800 it takes to outfit the CA-339s with genuine issue current-production WEs could prove prohibitive to most. So let's be clear from the get: Anyone contemplating ownership of the 339s better be prepared to foot the bottle bill without cringing. Don't purchase a stylish go-faster car only to complain that it burns up your gas allowance in a hurry. One comes with the other.
Rated at 50 triode watts and weighing in at 70 lbs/ea., the 339s follow Canary's cosmetic leit motif of champagne fascias and top plates. Those are set off against substantial black sheet metal with square perforations. Unlike its SET stable mate but exactly like all of Canary's remaining 300B amps, the 339 tucks its valves behind the upfront transformer casing to avoid eye contact with the glowing matter. I can see glass lovers flinch. Whatever visible emissions they may have been expecting will primarily come off the backlit azure display that doubles as status indicator and shows company name and model-specific text. Around back we find Canary's oversized custom binding posts that are easily tightened by hand, plus the requisite low-level RCA socket and IEC power inlet. The power mains is on the front. Individual bias trim pots adjacent to the 300Bs should be set such that they read out 0.55V for each tube via a digital multimeter set to 20V DC scale.
The eight-deep tube forest sits military-style in straight formation - two
outer rows of 300Bs parade behind a 5U4GB rectifier each to give cover to a small inner row of two 6SN7GTBs input/driver valves. A mere 0.75 volts will drive the amp to full output for a claimed frequency response of 10Hz to 80kHz +/-1dB. Dimensions on deck are 12" W x 24.5" D x 9" H, seriously expanded over my CA-308s and turned around. Build quality and parts choices like Hovland capacitors are very high as is common for this firm. Cosmetics are typical for exposed valves and very cleanly executed. We can once again assume that the circuit itself doesn't break new ground but is simply done right. Don't expect Berning or Butler-style innovation. Canary Audio is about solid, not extravagant - solid value, solid implementation of proven principles. Let others be the groundbreakers and experimenters.

To set up the match, let us briefly touch on conceptual and technical matters that surround push/pull and single-ended. Advocates of SE wave around circuit simplicity like holy water and then intone the lack of recombinant behavior whereby push/pull, by definition, must reassemble two separately amplified phase halves of the wave form. "Let no man tear apart that which the music gods created singly and then try to patch it back together without seams." That might as well be the core credo of that religion.

Proponents of push/pull take up their banners behind lower output impedance, improved linearity, distortion cancellation and higher power. From their fortified position, they then rag on SET's unpredictable frequency-vs.-impedance behavior as unfit to signal fidelity; and on a general lack of damping factor as piss poor insufficient for the majority of speakers on the market. Single-ended guys sing the praises of zero feedback while push/pullers tolerate judiciously applied feedback without balking (and often get quieter amps for it).

Devotees of single-endeds must carefully pick their copasetic speakers and often buy into a modicum of noise. That plainly limits eligible speaker options from the power perspective while the sensitive types that may apply will tend to magnify residual noise issues even more. To justify such limitations on their menu, fearless embracers fancy themselves capable -- blindfolded no less -- to hear the deleterious effects of phase splitting. Their mantra becomes "the simpler the better". Detractors meanwhile call them blindly in love with harmonic distortion. And on it goes forever, just as with any other dispute erected on limited vision.

None of us are free of assumptions and preconceptions. We all try to understand our own experiences rather than letting them simply be. Since we nearly always lack the full picture, our explanations tend to be full of holes. What if one simply threw theory and beliefs out the window? How would essential differences between these two basic amplifier topologies telegraph (always admitting that this could never be a clear-cut apples-to-apples comparison but at best only a reasonable approximation)? The following page provides answers within this one specific context.