Sumo Andromeda power amplifier Sumo Andromeda II in June 1991

Thomas J. Norton reviewed the Sumo Andromeda II in June 1991 (Vol.14 No.6):

The two-channel stereo amplifier is still the most typical configuration found in most systems, and the Sumo Andromeda II sticks with that convention. While the monoblock is becoming more popular, there's no free brunch; monoblock design results in higher costs to the manufacturer and thus to the buyer, all else being equal.

The Andromeda II is a pure complementary design using full-wave, balanced bridge circuitry. What this means for the user is that there are no common grounds at the output of the amplifier—meaning that caution must be exercised not to connect the output grounds of the two channels either together or to the system ground. The only current loudspeakers I know of in which this might cause a problem are the Polk SDA designs. Polk does, however, market an isolating device which makes their loudspeakers compatible with the Andromeda. There are also a number of loudspeaker switch boxes which are incompatible with the Sumo for the same reason. Also, since the Andromeda II is already internally bridged, it cannot be bridged externally to turn it into a higher-powered monoblock.

The Sumo is also a balanced differential design throughout, from input to output. But while positive and negative inputs are provided to make use of this balanced operation, an XLR input is not. In order to make use of the balanced inputs, separate interconnect cables must be run to the positive and negative inputs (using a preamp which provides separate positive and negative output jacks). The amplifier may also be driven single-ended with one pair of interconnects by inserting shorting plugs (provided) into the negative inputs. All of my listening to the Sumo was done in this mode.

Internally, power is fed to the circuitry by a single large transformer having separate secondary taps feeding separate rectifiers and filter capacitors for both the front end and output stages. Four pairs of individually matched, 150W MOSFETs provide the output for each amplifier channel. The rise-times of these devices is said to be 50 nanoseconds (ns). The output stages operate in class-A for the first 12W, reverting to class-AB above that point. These stages are linearized by Sumo's proprietary TL (transconductance linearization) error-correction circuit, which, by means of a pair of matched bipolar transistors, compares the gate and source of the MOSFETs and provides an error-correction signal to the driver stage if required. This TL circuit also provides the bias voltages for the output stage.

A pair of fully complementary differential amplifiers performs as the voltage gain section. All low-level stages are operated in class-A. Overall, Sumo claims that even without feedback (a modest 10dB of overall feedback is actually used), the inherent, open-loop distortion of the Andromeda II is less than 0.2% (power output not specified) with an open-loop bandwidth (3dB down point) of 50kHz. No protection circuitry or current limiting is used in the Andromeda II, with the exception of the power-supply rail fuses on the rear panel (the owner's manual refers to circuit breakers, which is incorrect).

Fit and finish
Both amplifiers are built to good commercial standards. Neither will exactly make you want to replace the top chassis panel with a plexiglass cover—that type of showcase can be had at much higher prices—though the Andromeda II has a slightly neater layout. The latter uses a number of the same push-on internal connectors that Corey Greenberg complained about in his recent review of Sumo's Polaris (Vol.14 No.4). While I understand (and to a degree agree with) CG's preference for hard-wired connections in place of these not-too-reassuring-looking terminals, various types of push-on and screw-on fasteners are used in even considerably more expensive amplifiers. They do have the undeniable advantages of simplifying assembly and maintenance.

The Andromeda II's input jacks did not inspire confidence, flexing too much under stress (the flexure was not at the male-female junction, but at the socket-to-chassis point). I never encountered any loss of electrical contact, but was concerned that the weight of typical audiophile cables might cause eventual failure. Sumo informed us shortly before press time that later-production Andromeda IIs now use sturdier input jacks.

My only minor complaint about the cosmetics of the Sumo's finish, which seemed to collect more fingerprints than the FBI. The marks do come off easily with a soft, damp cloth, however. In compensation, the Andromeda II is somewhat more elegant, looking less "industrial" than the Adcom 565 that I also review this month.

But none of our readers would (I hope) buy an amplifier for its appearance. What counts is...

The sound
My evaluation of the Andromeda II got off to a bit of a slow start. If you read my review last month of the Dynavector XX-1L cartridge, you'll have already received something of a preview. Driving the Apogee Stages, with the Dynavector on the front end, resulted in a sound which was rather lean and slightly too crisp. CDs were better balanced, and the Benz Micro MC-3 was even better suited to this particular combination of loudspeaker and amplifier. The latter was, in truth, a lovely combination. But I also know that the Benz measures down somewhat at the top end—not just where your dog might notice, but in the 10–12kHz area. Using a combination of vinyl and CD (though emphasizing the latter using both of the CD players listed earlier in this report), I gradually began to home-in over the next several weeks on what I felt to be the sonic signature of the Sumo.

It was not to be an easy quest. Switching to the B&W Matrix 801s, and back to the Dynavector cartridge (which had been returned to my reference system to allow me to finish its evaluation), the sound was now fuller in the midbass and not at all tipped-up on top. A trace of excess warmth was noted in the upper bass which spilled over into the lower end of male vocals, giving them a bit too much body. This was more of a problem with CD than with LP via the Dynavector cartridge. Things improved noticeably in the midbass when the (considerably more expensive) Wadia CD player/processor was substituted for the one-box Sony CD player. The low end tightened up noticeably, the top end became more open, and the overall focus sharpened up. There was some loss of immediacy in the lower midrange—a particular strength of the Sony player and perhaps (for me) its most appealing quality, but otherwise things were starting to come together.

To see if I could cure what I judged to be still some excess in the mid and upper bass, I switched next to the Snell C/IVs. In my current listening room the latter lacked a bit of the majesty of the 801s and were a bit less forward and "present" in the midband, and perhaps a shade less extended at the very bottom. But the Snells never displayed any obvious shortcomings in low-end extension, and the substitution of the C/IVs for the B&Ws (the latter used, as is my normal practice, without any outboard bass equalization) resulted in no "withdrawal" symptoms. They were, in fact, more open in the midbass than the 801s had been. Here the Andromeda II began to come into its own. The low end sounded powerful and gutsy; the Dorian Pictures was deep and rich. Low-frequency pipes had that shuddering, lusty growl which adds immeasurably to the feeling of realism. The top end was open and airy, with a definite but not overdone sparkle. Though the word "crisp" appears often in my listening notes, it was always used positively. There was nothing hazy or misty about the Sumo's top-end sound. MOSFETs have taken a bit of criticism for allegedly being somewhat veiled at the top. That was not my finding here.

There were, to be certain, times when I felt the top end of the Andromeda II to be just a shade too analytic. This was true more often when I used it with the Apogee Stages, almost never true with the B&W 801s, and an occasional problem—but not a particularly serious one—with the Snells. What did cause me a degree of concern, with all of the loudspeakers, was the Sumo's tendency to sound somewhat lean in the lower midrange. Voices in particular were stripped of some of their natural body. This might sound in conflict with my statement that the Andromeda II's upper bass was slightly full—an observation first made (as related above) with the B&Ws and later confirmed through the Snells. But the thinning was above the upper-bass region. The result was a noticeable reduction in (here comes that phrase again) "palpable presence." It not only affected voices, but appeared to somewhat reduce the overall "oomph" of the sound. The Sumo was clearly a very powerful amplifier, as its bass response indicated. But the "power region" of, say, a symphony orchestra lies not in the bass, but in the lower midrange.

I had compared the Andromeda II with the Muse 100 that Corey Greenberg had liked so much in April during my listening sessions with the B&W 801s. The Muse lacked the Sumo's upper-range "air," was a bit flatter in front-to-rear perspective, and could not play as loud without strain. But the Muse bettered the Sumo in handling this "power region." It simply sounded more immediate. And while the Sumo had more punch in the low bass than the Muse, the latter, again through the 801s, was a shade tighter in the mid- and upper bass.

Still, there was no escaping the Andromeda II's clear, open, transparent sound. At its best, I found listening to the Sumo a compelling, involving experience.

Monoblocks or single-chassis: Sumo vs Adcom
In theory, separating the two channels of a stereo amplifier should yield dividends. Interchannel crosstalk will certainly be reduced. Heat buildup will be less due to a (usually) more open, less crowded architecture. Monoblocks can be (but aren't always) of a more practical size. And the amplifiers may be located closer to the loudspeakers. In practice, I have some doubts as to whether these advantages—except for the last—will always be significant. The only way to be certain, of course, is to compare two identically designed amplifiers—identical, that is, except for one being split into two monoblocks, the other being integrated as a one-chassis stereo amplifier. Even then, I suspect the results will differ depending on whether or not we compare modest, cost-efficient designs or all-out assaults on the state of the art. The monoblock configuration should, ironically, benefit lower-cost designs the most—the type of design in which it is least likely to be used.

Neither the Adcom GFA-565s nor the Andromeda II are exactly budget amplifiers. But neither are they all-out, price-no-object efforts. Both involve some degree of carefully balanced compromise—though I suspect that their makers will not agree that this compromises the results—to keep costs below the audio stratosphere.

They do not sound alike, as what I have written to this point should make clear. I found the Andromeda II to be, overall, the more involving performer. Though not without some reservations: the Adcoms displayed some pretty fancy footwork of their own. On the King's Singers' The Beatles Connection (EMI CDC 7 49556 2), the GFA-565s were hard to fault. The sound was very clean and sweet—certainly not analytical or in any way bright or artificial. The voices were up-front and very much "there." The Andromeda II was, in contrast, more laid-back through the midrange, less immediate. But the latter sounded more three-dimensional, with a more "see into" soundstage. Individual voices were easier to hear, massed voices had a more realistic texture and were less homogenized.

On Leo Kottke's That's What CD (Private Music 2068-2-P), the Adcoms came out ahead in more convincingly portraying the timbre of the lower midrange, in presenting properly weighted voices and instruments. Imaging was sharp, and on the two vocal cuts Kottke's, ah, rather strange voice (footnote 1) was realistically full-bodied and tactile. From the upper midrange and into the highs, however, the Sumo sounded, overall, "faster" and more detailed than the Adcoms, with a more layered, open, and transparent soundstage. The latter was the less "tube-like" of the two amps, though neither is likely to win over die-hard tube lovers. As to low-frequency response, I was rather surprised to find myself leaning, in the end, toward the less powerful, single-chassis Sumo.

On the Dorian Pictures, both amps will do a suitable magic-fingers treatment of your backside in the lowest frequencies. But the Sumo appeared to have just a bit more weight and gutsiness in its overall low-end performance. In part this was a function of its more open top end (definitely audible in the presentation of this recording's ambient space), its more evident overtone structure providing the harmonics which provide focus to even the lowest frequencies. This quality was also apparent on percussive bass. There's a sharp bass-drum impact near the end of "Pops Hoedown" on Telarc's Round-Up (CD-80141). With the Adcoms, the word "impressive" came quickly to mind as I heard it (over the Snell C/IVs in this case). With the Sumo, I bellowed out an involuntary and quite audible "Yeah!" following the same passage. The actual sonic difference was not that dramatic. My reaction was.

But the Sumo is, based on my experience, a bit touchier in the matching department than the Adcom. Though it was at its best through the Snell C/IVs and (in a briefer audition) the Nestorovics, I was never entirely happy with it through the Apogee Stages—especially when used (as already noted) with an analytical cartridge. And I found its low end to be a trace too full through the B&W 801s in my listening room, a fullness at least partially the fault of the loudspeaker/room match (I've obtained better results from the B&Ws in prior listening spaces). The Sumo's revealing, unforgiving, airy but somewhat crisp top end, combined with its slight leanness in the lower midrange, make it, to this listener, the more open, transparent window on the sound. But if you provide it with the wrong combination of associated equipment and program material, it can sound less than sweet; even, on occasion, overly analytic.

At its best, the Andromeda II soared for me in a way that the Adcoms never quite managed.

I was favorably impressed by the Sumo Andromeda II. It offers an attractive alternative to spending megabucks on system go-power, but it isn't really a Krell (or whatever) in sheep's clothing. But the Sumo Andromeda II did surprise me. It was a bit more idiosyncratic than the Adcom in its matching requirements, but its open, transparent, yet powerful sound finally won me over. For this listener, the Sumo pushed the right buttons.—Thomas J. Norton

Footnote 1: Described by Kottke himself as sounding "like goose farts on a foggy day."—Richard Lehnert