Spendor S3/5se loudspeaker Measurements
As expected, Spendor's little S35/se is not very sensitive, 2.83V raising just 81.5dB(B) at 1m—a dB less than the BBC's original LS3/5a. However, its plot of impedance magnitude and electrical phase against frequency (fig.1) indicates that it is a kind load for the partnering amplifier to drive. The impedance drops below 8 ohms only below 40Hz and in the lower midrange, and stays above 10 ohms for much of the audioband. The peak of 28 ohms at 71Hz indicates the tuning of the sealed box, which in turn implies limited LF extension.
Fig.1 Spendor S3/5se, electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)
A wrinkle just below 250Hz in the impedance traces suggests the presence of some kind of cabinet resonance. (Both this and the next measurement were taken with the speaker sitting on upward-pointing cones, which allows cabinet resonances to develop to their fullest.) Fig.2, calculated from the output of an accelerometer fastened to the back of the speaker just above the electrical terminals, shows that this panel does have a high-level resonant mode at 242Hz, as well as another just below 400Hz. Fortunately, this panel faces away from the listener. Modes at 200Hz, 300Hz, and 360Hz can be found on the side and top panels, but these are much lower in level. However, I do wonder if they are associated with the touch of chestiness AD noted on voices, despite the fact that he used Blu-Tack pads rather than spikes as the interface between the speaker and its stand. (An article in Stereophile's September 1992 issue showed that Blu-Tack was very effective at damping cabinet resonances.)
Fig.2 Spendor S3/5se, cumulative spectral-decay plot calculated from the output of an accelerometer fastened to the cabinet's back panel above the terminals. (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz.)
Fig.3 shows the individual farfield responses of the tweeter and woofer, spliced to the nearfield woofer response below 355Hz. The crossover frequency can be seen to lie closer to 4kHz than to the specified 3.5kHz, but there is broad overlap between the two drivers. The high- and low-pass slopes look close to 18dB/octave, and the drivers are reasonably flat within their passbands. The slight rise in the upper bass is almost entirely due to the nearfield measurement technique; the speaker is probably tuned to be maximally flat, with a -6dB point of 55Hz with respect to the level at 1kHz.
Fig.3 Spendor S3/5se, acoustic crossover on tweeter axis at 50", corrected for microphone response, with the nearfield woofer response plotted below 355Hz.
Fig.4 shows how these individual drive-units add up at a nominal farfield point on the tweeter axis, averaged across a 30 degrees horizontal window. Again, assuming the apparent rise in the upper bass is a measurement artifact, the response rises slightly from the lower midrange through the mid-treble. AD didn't note any coloration, so I assume this balance aids articulation and the presentation of recorded detail rather than being heard as an energy excess.
Fig.4 Spendor S3/5se, anechoic response on tweeter axis at 50", averaged across 30 degrees horizontal window and corrected for microphone response, with the nearfield response of the woofer plotted below 300Hz.
What did surprise me was the slow rollout in the top octave. In anything but a small room—and from the time I once spent enjoying music in AD's room, I can testify that it is not "large"—the speaker will sound rather mellow. AD's room is also more live-sounding than, say, my own, which is probably why he was not concerned by a top-octave reticence. However, he did note that cymbals shimmered "more politely" compared with the basic S3/5, which might be a function of this behavior.
For comparison, the black trace in fig.5 is the S3/5's response averaged across a 30 degrees horizontal window on its tweeter axis, spliced to its nearfield woofer response. Despite the cabinets of the two speakers being identically sized and the woofers looking the same, the S3/5 is tuned to a higher frequency—84Hz vs 71Hz—with a slightly less damped alignment. Some of the rise in the upper bass will therefore be real, which I assume was done to better balance the S3/5's less reticent top octave (see later). AD didn't remark on any differences in the LF region, however. The individual drive-unit responses are plotted in red in this graph, and can be seen to cross over at 4.5kHz. But note that the '3/5's tweeter is balanced a couple of dB "hotter" than that in the 'se. All else being equal, that will go some of the way toward explaining AD's description of the differences in the speakers' trebles.
Fig.5 Spendor S3/5, anechoic response on tweeter axis at 50", averaged across 30 degrees horizontal window and corrected for microphone response, with the individual responses of the tweeter and woofer (red trace above 300Hz), and the nearfield response of the woofer (red trace below 300Hz).
The big factor that stops things being equal, of course, is the speaker's dispersion. Fig.6 shows how the S3/5se's balance varies to the sides of the tweeter axis. The radiation pattern is remarkably even, something that ties in with stable, accurate stereo imaging. Despite its top-octave rollout, that 0.75" tweeter does maintain its HF output to 20 degrees off-axis, which will ameliorate the audibility of the rollout in rooms of small to medium size. However, the tweeter does increasingly roll off above 12kHz at more extreme off-axis angles.
Fig.6 Spendor S3/5se, lateral response family at 50", normalized to response on tweeter axis, from back to front: differences in response 90 degrees-5 degrees off-axis, reference response, differences in response 5 degrees-90 degrees off-axis.
The S3/5's lateral dispersion (not shown) is basically similar, but with its higher crossover frequency, the beginnings of an off-axis flare between 5 and 10kHz can be seen, due to the woofer's starting to beam in the octave below that region. This will accentuate the audibility of the '3/5 tweeter's slightly elevated level compared with the 'se's.
In the vertical plane (fig.7), a suckout in the crossover region develops at off-axis angles greater than +5 degrees and -10 degrees, suggesting that the user's stands should place the tweeters at ear level for the best results. However, I note that AD was not bothered too much by differences in listening height.
Fig.7 Spendor S3/5se, vertical response family at 50", normalized to response on tweeter axis, from back to front: differences in response 45 degrees-5 degrees above axis, reference response, differences in response 5 degrees-45 degrees below axis.
In the time domain, the '3/5se's step response (fig.8) indicates that the tweeter and woofer are connected with the same positive polarity, the former slightly but inconsequentially leading the latter in arrival time at the microphone. The cumulative spectral-decay plot on the tweeter axis (fig.9) is very clean, particularly in the tweeter region, though a low-level ridge of delayed energy can be seen at 4.5kHz.
Fig.8 Spendor S3/5se, step response on tweeter axis at 50" (5ms time window, 30kHz bandwidth).
Fig.9 Spendor S3/5se, cumulative spectral-decay plot at 50" (0.15ms risetime).
There were no smoking guns to be found in the S3/5se's measurements, though it is fair to point out that its maximum loudness will be limited by increasingly nonlinear behavior compared with larger, more sensitive speakers. But when used within its dynamic limits in a small, not too absorptively furnished room, the little Spendor S3/5se will give, as AD found, excellent performance.—John Atkinson