Cary Audio Design CAD-572SE monoblock power amplifier Page 2
As my ears became accustomed to this affordable power amplifier, it was hard not to be aware of the performance gap between it and the cost-no-object CAD805C reference. As is frequently the case with SE triode amplifiers, it was rather too easy to find differences—a reviewer can be swayed too easily. Of course, some of the differences arose from the load interface—the interaction of the different and significant output impedances of these amplifiers with different speakers. Fortunately, these aspects are largely quantifiable by measurement, and can be accounted for when weighing up the subjective characteristics.
Then there are the other subjective differences—for example, those associated with component type and quality, topology, overall electronic design, the degree of negative feedback, and/or the "class" of the output stage. In addition, with low- and zero-feedback amplifiers there may be present some simple low-order distortion rather like that found in loudspeakers. This may well be audible, if not generally destructive to the overall sound quality. However, when the distortion is audible, it may be associated with subtle changes in timbre heard as minor shifts in apparent frequency response rather than as a loss of clarity or an imposed roughness. Variations in the spectral "shape" or harmonic distribution of an amplifier's distortion are also responsible for differences in sound quality.
Given that most loudspeakers have a generalized impedance curve that peaks in the upper midrange, often dips in the 2-4kHz presence range, and then rises somewhat at high frequencies, an SE amplifier designer might unconsciously or even consciously choose particular tubes, then configure and bias them so that the result sounds better-balanced with typical speakers. He or she might then use the mild tonal-balance shift due to a particular nonlinearity to balance the frequency-response error caused by matching the tube to the speaker impedance (footnote 1).
Subjectively, the available power sounds rather more than the 572's paper specification suggests. Still, this is not a big amplifier; with suitable speakers, the 572SE plays quite loudly enough for row-30 concert-hall realism, but is undeniably inadequate for stadium-rock effects unless you plan to use your speakers as headphones. If you're a rocking headbanger and don't have 8 ohm horn speakers of +96dB voltage sensitivity or the equivalent, look elsewhere.
I tried the 572SE with the compact, average-sensitivity (86-87dB/W/m) Sonus Faber Electa Amator IIs, but met with failure. Listening to this speaker can be a joyful experience, but it needs a hefty kick—in my opinion, +75Wpc—and the 572SEs barely roused the Amators from their slumber. Conversely, the Wilson Audio WITT 2, only a few dB more sensitive by my measurement, took quite kindly to the small Cary. While the combination could not generate the kind of wallop of which the WITTs are otherwise capable, the result was musically satisfying—thanks in part, no doubt, to the WITT's kind and respectably uniform impedance curve.
Again, without requiring or demanding high sound levels, the 572SE worked very well with Quad's ESL-63 electrostatic, the very natural resulting sound suggesting that this is how the '63s should always have sounded. My reference midclass compact loudspeakers, the Spendor SP2-2s, though getting on a bit now (my pair is probably at least 10 years old), offered quite good load matching and a sensitivity similar to the Wilsons'. I liked this combination with the CAD-572SE; the Spendors had never sounded so even-tempered or as liquid.
In contrast to many other contenders, the 572SE offers genuine transparency, delivering fine low-level detail and deep, see-through soundstages. Detail was nicely illuminated not only in the central region, but even at the back of the virtual stage. This quality, a hallmark of membership in the high-end club, qualifies the 572SE for admission.
As regards dimensionality, the 572SEs threw a stereo image as wide as it was spacious, and imbued with substantial reverberant air and atmosphere. Focus was strengthened by the neutral perspectives constructed by the even, well-balanced tonal quality. The 572SE never threw the sound forward, and offered high resolution, very good ambience, and a strong commitment to a recording's musical and performance qualities. Here, absolute accuracy is less important. I feel, than the communication of the atmosphere of the original recorded event.
The bass fell short of the 805C's standard, and well short of that of the better solid-state designs. However, this criticism requires qualification: Though there was some noticeable shortfall in slam and extension when viewed objectively, the bass wasn't bad at all down to 50Hz. My opinion of the 572SE's generally good quality was reinforced by its tight, truthful character on smaller bass percussion, especially jazz double bass. Though Stanley Clarke had to be reproduced on a small scale, Mingus proved no trouble at all for the 572SE.
The 572SE didn't quite convey the absolute midrange homogeneity of an exemplary 300B design—but then, neither does the 805C. The 572SE's midrange nonetheless sounded very, very good on a small selection of medieval voices and well-recorded vocal solos. An early Joni Mitchell album proved well worth revisiting to hear just how well this amplifier allowed performers to sing naturally and with great expression.
On the other hand, the Cary's mids sounded slightly lightweight—but only by comparison with the finest references. To all intents and purposes, the 572SE was neutral and easy on the ear; there was no obvious audible disadvantage in the small degree of feedback employed in the amplifier to stabilize its operation. In any case, the use of negative feedback is by no means the deciding factor in the determination of sound quality, even though if it is used excessively or inappropriately, can be harmful to fidelity.
The Cary 572SE was free of any glare, grain, or harshness. Its treble was nicely proportioned, evenly balanced, high in resolution, low in grain, and complemented the midrange very well.
Dynamically speaking, the CAD-572SE didn't sound small. Dynamics were well presented, while the contrast between soft and loud, and the expression of fine playing, were both properly communicated. It wasn't as masterful as the CAD805C, but it still set a high standard, this evinced by a lively, upbeat character with a very good sense of rhythm and fine timing. For example, all sections of jazz percussion were set nicely in time to drive a clean, solid beat.
As with other singled-ended amplifiers, the 572SE could be driven well past its clipping point (as monitored with a peak-hold power meter), and while there was certainly some sonic deterioration past the overload point, this was surprisingly mild. In moderation, clipping overload either passed unnoticed or had little effect on overall performance. In practice, the 572SE played about twice as loud as its 20W specification would suggest. The clipping sound was soft and of low harmonic order, in contrast to the "knocking on the glass" sound of high-feedback clipping. The way the music goes on playing even in clip indicated that the 572SE has good overload recovery behavior, a matter often neglected by designers. This is an area in which Cary can claim some significant skill.
Certainly the finite output impedance of an amplifier such as this will undergo some interaction with your chosen loudspeaker—unless it's one of those choked with electrical compensation in order to provide a uniform impedance for the amplifier. This interaction has to be taken into account, but overall I feel that this small amplifier attains significant greatness. (For those who follow my personal scoring system for overall merit, the Cary 572SE achieves a handsome 35 points.)
I found the CAD-572SE to be a musically lively performer with pleasing dynamics; it preserved the life of performances and readily held my attention. The more I listened to it, the more I came to value this design as a smaller-scale version of the 805C itself.
This smaller design proved to be a true half brother to the mighty CAD805C, specifically where the "half" concerns the rated watts output. Given our approximately logarithmic hearing characteristic, that 3dB or half-power difference is a relatively small step in audible volume. The 572SE isn't an 805C, although its inherent musicality—the combination of resolution, transparency, dynamic expression, and rhythm—compares well to that of the 805C. It isn't as authoritative, load-tolerant, or vibrant as its big brother, but what a great amplifier it must still be to withstand such a comparison.
The 572SE did well in the lab for this class of amplifier, has very good build and finish, and creamed the sound-quality tests—in short, I found it to be quite a gem, an amplifier to cherish. It strongly hints to me that I should abandon the audio rat race and listen to more music and less gear.
In its own right, the 572SE is a topnotch performer, a class-winning SE triode amplifier that deserves to become a classic. It's got that ability to draw attention away from itself and encourage one to focus one's ears on the music. I'll happily cast my vote for a Class A rating in Stereophile's "Recommended Components" listing.
Footnote 1: Through this first layer of variation, such amplifiers exhibit a "core performance" in which traditional aspects of sound quality still hold true. This was amply demonstrated when I recently tried a relatively well-known brand of European zero-feedback SE triode amplifier. While the genre's superficial attractions—creamy tonality and low listener fatigue—were evident, a representative example from this range also sounded comparatively slow, undynamic, and lacking in transparency.—Martin Colloms