KEF Reference Series 103/4 loudspeaker

By now most readers will be familiar with the relatively new tuned-cavity method of low-frequency loading. Such designs have popped up all over the place of late, especially in those little satellite/woofer systems, but KEF can rightly lay claim to generating the design's theoretical basis, as JA described in his review of the KEF R107/2 loudspeaker in Vol.14 No.5 (May 1991). Essentially, the technique consists of loading the rear of a woofer in a conventional fashion—usually a sealed box—but also loading the front of the driver into another enclosure, ducted to the outside. Basically, the design acts as a bandpass filter with its response centered on the port-tuning frequency. The rolloff is smooth and rapid on either side of this frequency, providing a natural low-pass characteristic but thereby virtually mandating a three-way system. If properly designed, this configuration offers a number of theoretical advantages. The radiating element is actually the air in the port, which is low in mass. Low distortion is possible, as is relatively high sensitivity.

In the 103/4 there are three internal cavities, and two 6.5", fiberglass-reinforced, paper-cone bass drivers. Each of the latter is loaded into its own sealed chamber on one side; on the other side each driver fires into a central coupled cavity porting to the outside. The two bass drivers are linked by a solid metal rod visible through the large, front-mounted port. This rod is said to cancel the drive-units' identical but opposing mechanical vibrations. The same is said to be true of the woofers themselves.

I don't entirely agree with this explanation. The woofers, though mounted in opposition to each other—that is, the front of one and the rear of the other fire into the main ported cavity—must each fire in-phase into this cavity. Meanwhile, the cavity behind each loudspeaker is simultaneously being driven in opposite phase by the other side of the woofer cone. In terms of extraneous cabinet vibration, this means that the pressures in the central, common cavity are positive at the same time that the pressures in the top and bottom cavities (the "back" enclosures of the individual drivers, if you will) are negative. The external radiation generated by these flexing walls will, to an extent, cancel out, but this cancellation—given the complexity of vibrating systems—will certainly not be perfect. Vibrating membranes seldom act as true pistons, instead displaying numerous breakup modes which vastly complicate their behaviors. The internal connecting rod would seem to function largely as an additional brace, minimizing the motions of the two woofers relative to each other. The design's requirements, however—three chambers and a connecting rod—appear to contribute inherently to a rigid structure.

692kef.inside250.jpgThe mid and high end of the 103/4 are covered by a single Uni-Q drive-unit. Similar in concept to the slightly larger driver in the KEF Q60 reviewed last month, the Uni-Q incorporates a 1" dome tweeter into the apex of a 6.5" midrange driver at the position normally occupied by the dustcap. This coincident drive-unit has been made possible by new magnetic materials developed over the past few years—in this case, a neodymium/iron/boron alloy—which are extremely powerful and can thus be made extremely small and still provide a strong enough field to give a drive-unit a usable sensitivity.

Other features of the Model 103/4 include a Conjugate-Load-Matched crossover network—which minimizes overall system impedance variations—and drive-units for each pair of loudspeakers matched to within 0.5dB of each other for left/right symmetry. These features are common to all models in the KEF Reference Series. There are also two pairs of good-quality five-way binding posts (for bi-wiring if desired) on the cabinet rear, spaced to accept double banana plugs (rare in British designs).

KEF also makes available an optional Kube 200 active equalizer for the 103/4. The 200 has front-panel contour controls for bass and treble, and is said to extend the –6dB point of the 103/4 to 20Hz. A number of possible hookups for this device are provided, including extra direct and equalized outputs, the latter for possible bi-amp use, the former to drive other amplifiers where equalization is not desired. A simpler Kube 100 has recently been made available which dispenses with the treble contour control and sells for $200, half the price of the Kube 200.

I took one look at the 103/4s and decided that they needed to be raised off the floor. The height of their tweeters is considerably below my listening-ear height. I set them up the same way I'd mounted the B&W 804s that I reviewed recently: on 9"-tall Hales Audio stands, which are, effectively, solid blocks filled with sand. Tiptoes and Tonecones were used between loudspeaker and stand and between stand and floor, making for a quite rigid setup.

The KEFs impressed me immediately with their big, open soundstage and tight definition. There was a step-into transparency through much of the frequency range which produced a lively, clear quality. Beginning with the bottom end, the overall extension was not particularly pronounced; the specified –6dB at 38Hz seems reasonable. But the low end is tight, particularly with the optional Kube (which I'll get to shortly). There was never any feeling of a lack of bass; though the bottom octave was clearly not there, I seldom missed it. I did appreciate the clarity higher up in the bass register. The 103/4s exhibited a trace of warmth in the mid- to upper-bass region, particularly when the program material exhibited any complexity in this range, but it never distracted.

The optional Kube 200 should, in theory, dramatically enhance the 103/4's low-bass performance. I was very surprised, therefore, to find that it was not very effective in extending the subjective low-end cutoff. At first I found this hard to accept; the measured boost provided by the Kube below 50Hz is quite pronounced. But try as I might, I could not obtain much beyond a subtle enhancement in the very bottom at the low-frequency contour's center position—just a hint of a bit more going on there (and I do mean just a hint). It was easy to confirm this with the Kube inserted into the Rowland Consummate's tape-monitor loop; it could be switched in and out by remote control from the listening position. I even set the Wadia transport to repeat particularly potent bass passages, switching the Kube in and out between repetitions. It made little difference in the bottom-most octave's effective response.

Increasing the Kube's low-frequency contour control to maximum did result in an audible but undramatic enhancement on program material with extended low-end content—mostly organ—but it also subtly softened the bottom end and made me fear, however unnecessarily, for the loudspeakers' safety. I do know that I would not recommend using the Kube with LPs; most LPs and LP-based systems contain a fair amount of extreme low-frequency garbage. Boosting them by the amount provided by the Kube is a recipe for trouble.

What did clearly change with the Kube, and for the better, was the tightness and detailing in the mid- and upper bass. The slight warmth evident without the Kube nearly vanished when the latter was engaged, noticeably opening up the overall sound. This was a clear result of a small frequency-response dip introduced by the Kube just below 100Hz (in the setting at which I most used it). The Kube was also remarkably free of other audible artifacts. Perhaps, just perhaps, it extracted a very small loss of upper-range transparency. But I definitely preferred the sound with the Kube in the circuit because it did enhance the clarity of the midbass, and therefore of the sound in general.

The reason for the Kube's relative lack of effectiveness in the lowest bass—despite the strong boost it provides—would seem to lie in the relatively sharp rolloff typical of a bandpass woofer design. There are also undeniable physical limits to 6.5" woofers, no matter how clever the design. While it was possible to clearly hear and measure the increased LF output with the Kube by listening to pure tones, its effectiveness with music was considerably less. The 103/4 will not, nor is it designed to, compete in low-end grunt with KEF's flagship R107/2.

Further up the frequency range, I heard a clear, open midrange not entirely free of coloration: I noted an occasional cupped-hands quality, particularly on some (but not all) male vocals. The overall midrange clarity did, however, contribute to that tightly focused, well-layered soundstage.

Higher up in the frequency range, however, the 103/4 exhibited problems. While different in character from those noted in the Fried R/4 (also reviewed in this issue), they nevertheless produced, in the KEFs, a rather crisp, slightly top-heavy quality which lent an impressive clarity to the sound, but which tended to be overly analytical. Percussion had excessive zip, and isolated HF notes pushed unnaturally forward. My listening notes are a bit conflicted on this point; the problem seemed to be higher in frequency than that in the Frieds. It did not have a consistently negative impact on the upper range of the human voice, yet did intrude on reliable recordings with strong high-frequency contents. My notes describe On the Banks of the Helicon (Dorian DOR-90139) as "nothing like a Record to Die For" (footnote 1).

Nor did the Kube's HF contour control solve the problem, use of it resulting in a softening of the entire high-frequency range and a dilution of the 103/4's normally excellent focus. This control is intended to tame overly live acoustic environments (mine does not qualify for that description), not equalize the loudspeaker.

To get a better handle on the problem—which seemed slightly less intrusive than the similar one in the Frieds but was, after all, in a more expensive loudspeaker—I compared the 103/4/Kube combination directly with the B&W 804, a loudspeaker of comparable size and price that I reviewed last November.

The B&Ws excel in soundstaging, sweet and extended (though not obviously "airy") highs, and a surprisingly extended, if a bit soft, low end. They're a bit forward in sound, have a hint of midrange nasality, but otherwise excel in vocal reproduction, floating a realistic, palpable image in front of the listener. Compared with them, the KEFs do not appear to go quite as deep or as powerfully into the bass, even with the Kube, but compensate with a tighter, more sharply defined (especially with the Kube at its bass contour mid-setting) mid- and upper bass. The B&Ws produce less tightly focused images, and are decidedly less crisp and more musically sweet at the top end. The KEFs are less forward than the B&Ws, with a somewhat leaner, less immediate midrange. They are also slightly more colored through the midrange, and less three-dimensionally tactile. The KEFs' subjectively leaner balance appears to make what coloration there is more obvious. Between the two loudspeakers, I find the B&Ws more musically involving, the KEFs more lean. I'd opt for the first; your priorities may be different. Ideally, one would like to combine the best qualities of both loudspeakers. I liked the KEFs' clarity, but found them a little relentlessly analytical over the long haul.

In a final attempt to tame the top ends of the KEFs, I listened to the loudspeakers with the Audio Research Classic 120 monoblocks. While it did help—providing a degree of previously missing sweetness, though at the expense of some image softening—the trebles remained a bit problematical. It was decidedly less musical than that in either the Paradigm Studio Monitors or the B&W 804s. I also tried a less expensive but promising amplifier which only recently arrived, the PS Audio 100 Delta. While noticeably crisper than the Audio Research, the PS Audio also seemed a bit sweeter on top than the Mark Levinson No.23.5—though the jury is still very much out on this. The PS Audio could not, however, control the low end quite as well as the Levinson.

I have to say that I was very favorably impressed by the KEFs' inner clarity and soundstage. The KEF probably comes closest to a recommendation as it stands, but I would certainly also recommend care in auditioning them.

The bottom line, however, is this: While I found these loudspeakers to be fun dates, I would hesitate to bring them home to Mother.

Footnote 1: It was one of my two selections for our feature in the January 1992 issue.
US distributor: GP Acoustics (US) Inc.
10 Timber Lane
Marlboro, NJ 07746
(732) 683-2356

kensargent's picture

I think what you mean is something other than that the woofers in the KEF 103/4 operate IN phase. For there to be any pressure generated in the central chamber, and thereby generation of any velocity at the port, the drivers must be OUT of phase, both electrically and mechanically. That is to say that as one of them moves away from its' frame, the other moves toward its' frame. As a result, the pressure in the central chamber is increased in one half of a cycle, and decreased in the other half of a cycle. This is the actual arrangement that would create the situation you describe: when the pressure is at maximum in the central cavity, it will be at minimum in the opposing cavities at the ends of the enclosure.

The rod connects the woofers' frames, surely, as this arrangement would, in fact, reduce vibration as the manufacturer states, but only if the cones were moving opposite each other. If the woofers were moving in phase, the rod would serve only to equalize, or average, the vibration, but not reduce it.

There are some arrangements in professional loudspeakers that put the woofers out of phase both mechanically and electrically, on an ordinary baffle, with one facing into the cabinet (front-loaded) and one facing out (rear-loaded.) In this arrangement, the advantage is said to be a cancellation of nonlinearities caused by the woofers' suspension, with the result reportedly being measurably lower distortion.