Conrad-Johnson Premier Eleven power amplifier Follow-up: the Eleven A
A Follow-up by Wes Phillips appeared in August 1995 (Vol.18 No.8):
In my original review in October 1994 of Conrad-Johnson's Premier Eleven power amp, I waxed enthusiastic over what I felt to be a truly special product. Did I say waxed enthusiastic? I was so besotted that my review read like a love letter to the Muses—specifically Euterpe, Polyhymnia, and Terpsichore.
Tom Norton cast a more dispassionate eye on the unit's measured characteristics—which is the whole point of our specs section—and wondered why the amplifier produced anomalous third-harmonic hash when fed a 1kHz signal at moderate power output (with spikes occurring twice per cycle, yet). TJN speculated that it had something to do with crossover artifacts, and everyone involved decided not to fret over it, since we all agreed that the amp sounded great anyway. Nevertheless, about the time that the monoblock Premier Twelves were introduced, C-J announced that the Premier Eleven had been modified and would henceforth be known as the Eleven A (footnote 1).
Lew Johnson called one day and suggested that I give the Twelves a listen. Boy, was I ever tempted! But, in the course of our chat, he mentioned the mod they'd made to the Eleven. I was dubious. "The chief enemy of good is better," stated Milton Greene—meaning, I suspect, that progress eradicates even things generally conceded to be good. Having experienced too many product "improvements" that destroyed the very attributes I had initially admired, I have always harbored a much more cynical interpretation of his dictum.
"Gee, Lew, I'd rather listen to the Eleven A."
"That could be arranged."
"A pair of them would allow me to bi-amp the Martin-Logan Aeriuses," I purred.
Thus, two Premier Eleven As found their way to my listening room. The first change was apparent as soon as I connected the speaker cables. The Eleven As have hex nuts on the binding posts rather than the cylindrical nuts with two flat sides that had caused me to snap off the whole post as I cinched it tight with an adjustable wrench. Now I could use my Postman. Hallelujah!
"You weren't the only one [to snap off a binding post]," Lew chortled. "The old posts were simply too awkward to use. The new ones come from the same vendor and are composed of the same metal; the only difference is the nut. We didn't want to negatively affect the sound."
"So what other changes have been introduced?"
"You're the critic, you tell me what they do—then I'll tell you what we did!"
Fair dinkum. So as not to introduce too vast a change in the system, I began by using just one Eleven A. The mighty C-J Evolution 20SE preamp fed the amp with signals from my Linn LP12/Ittok/Audio-Technica ML-150 analog rig and from McCormack's SST/DAC 1 digital combo. The Aeriuses were my primary speaker—although ProAc 1Ses and Metaphor 2s also saw duty. Transparent Audio Music Wave Reference speaker cables and Music Wave Ultra interconnects got the signals where they needed to go.
No worries, mate! Everything I valued in the original amp survived intact into the Eleven A. The sense of pace, the illumination of the telling detail, the revelation—as if by artful accident—of the subtlest inner voicing; it was all there. But added to that was now a greater sense of slam. Slam? That's the difference?
Weeellll—it's not exactly that cut and dried. In my original review of the Prem Eleven, I lauded its ability to sort through the dynamic, rhythmic, and timbral complexities of Corigliano's Symphony 1 (Erato 2292-61132-2), so I thought it appropriate to start my listening session there. It sounded better—more emphatic, with slightly less "slop" in the attacks, or leading edges, of transients and tuttis. But I hadn't originally been aware of any "slop" in the leading edges of transients.
Following one of those listening trails that are sometimes inspired by a revelatory audition, I had an intense desire to hear Mahler's Symphony 6, suddenly finding in its despair and desperation (not to mention its "rhythm of catastrophe") a counterpart to the Corigliano. Boulez's recent recording leading the Vienna Philharmonic (DG 445 835-2) has seen a lot of play in my system lately, as I like his rejection of excess emotionalism here. (The Sixth has enough doomed foreboding without the conductor milking it, thank you very much.) Again, I had the sense that the sound was tighter, more controlled—the timpani rang out with authority and clarity, and the Andante's "axe-strokes of disaster" were chillingly massive and unblunted.
Try as I might to find a downside to this gift of articulation, I couldn't. Not that I wanted there to be a price, mind you; I just couldn't believe that the improvement came without affecting what turned me on in the first place. But that's precisely the case: I cannot cite any way in which the original Eleven's virtues have been compromised. My lute recordings still charm with grace; chamber music inhabits my room, gleaming with the revealed sinew of its unobstructed lines; and—now—my big orchestral scores manifest a solidity that I wasn't even aware was (ever so slightly) lacking. And, no—I guess technically speaking it's not so much that the leading edge is sharper or quicker, but rather that it isn't spread out afterwards. The peak isn't blurred—at least, that's how I perceive the effect.
During all of this listening, one other aspect of the amplifier's performance seemed improved: the biasing LEDs didn't flash as much. Now, I might not have noticed this, but my buddy Ruben thinks it his God-given duty to keep me from getting a swelled head. He'll drop by and carefully look the system over, sussing out the changes since he last visited, and—spotting a cable almost forming a coil—will tut-tut quietly. Previously he would always glance significantly at the flickering red lights and pointedly ask, "Lose the adjustment wand?" The Eleven A just didn't seem to quiver its LEDs as coquettishly as the old Eleven. I know it sounds dumb, but it is a difference—one that led me to suspect that the amp was much more stable at maintaining bias.
Feeling that I had fulfilled my part of the challenge, I called Lew Johnson back and demanded, "What did you guys do?"
"All we really did was filter out just a little bit of noise injected into the circuit by the bias circuitry—those LEDs that indicate that the tubes are at their correct bias point. Tom Norton, in the technical-evaluation portion of your original review, attributed the problem to some sort of crossover-notch distortion. It isn't. We put in a very simple little filter to greatly attenuate the noise generated by the LEDs. That's the main change (footnote 2). We also re-routed a couple of ground paths."
"That's it?" Hmmm. "What LED you, sorry, led you to make a change that seems that minor?"
"Well, it was just a technical thing. One of the technicians here spotted an anomalous waveform under certain circumstances and wanted to correct it. Frankly, I didn't consider it to be all that important; I thought we were dealing with what we call a bench phenomenon—something that shows up in testing but that has no consequence in the sound room. So when we came up with a circuit change that eliminated it, my concern, frankly, was to be certain that we hadn't made the sound worse. I guess that gives you some indication of how far off-base I was. I anticipated no change—and was praying that we didn't make it sound worse—so I was stunned at the improvement. This was while we were working on the Premier Twelves, so we had to apply the circuit changes to the Eleven—and it made us go back and redesign the Premier Eights altogether."
"So," I asked, waiting for Lew to congratulate me on my brilliance as an observer, "why do the tubes stay biased more stably now?"
"Offhand, I can't conceive of a good reason for that—except that, because of the nature of the filter that we put in there, it takes a more sustained high-amplitude signal to make the LEDs light. The filter is slowing down the response of the LED circuit; but since it is really only there to indicate a quiescent, or steady-state, condition, it really doesn't matter. Instead of being near instantaneous, we've introduced a second of lag—or less. That tends to smooth it out so that the bright lights don't flash as often."
Dang! Caught confusing correlation with causation again. I hate when that happens. I asked a question from Baba Wawa's book on interviewing: "Have you guys learned any life lessons from this experience?"
"In a sense, yes. One doesn't want to be too quick to make an assumption that some observable phenomenon is too insignificant to affect sound quality. I suppose that oughtn't come as a shock, because so many things that clearly affect the sound of a product don't produce any observable phenomena, as far as I can tell. So when we do have one, it shouldn't surprise us that it does have an effect."
Great answer! If Messrs. Conrad and Johnson found it hard to believe that such a minor mod could have such a profound effect, it was nothing compared to my amazement upon finding out what had produced the differences I was hearing.
I continued to audition the Eleven A under varying circumstances. I biamped the Aeriuses (footnote 3). I auditioned the ultrarevealing WATT/Puppy 5s. I rocked out, blissed out quietly, and ran the gamut musically. I never did perceive the Eleven A as less than the straight Eleven, always preferring the new, revised version. And the cynic in me has been proven wrong once again. Sometimes the chief enemy of good is better.—Wes Phillips
Footnote 1: The Eleven A has been shipping since January 1995. If you're in doubt as to whether you have an Eleven or an Eleven A, look at the serial-number plaque on the rear of the amp. All Eleven As are designated as such there. The cost to upgrade an Eleven to an Eleven A is $275. Speaking of cost, by the time you read this, the amp will list for $3495.—Wes Phillips
Footnote 2: Measurements of the new sample revealed the only distortion component present to be the benign second-harmonic.—John Atkinson Footnote 3 :Highly recommended, by the way.—Wes Phillips