The Smaller Advent loudspeaker Measurements
I performed a complete set of measurements on the two samples of the Smaller Advent, serial nos. 47877 and 47106, the latter manufactured on October 30, 1972, according to the tweeter chassis. They were very similar, though 47106 was slightly less sensitive than 47877 as measured on the tweeter axis: 84.5dB(B)/2.83V/m vs 85.5dB(B)/2.83V/m. Both figures are 2–3dB below the average for modern speaker designs. The Advent's impedance, however, was not too demanding of current, as can be seen from the impedance graphs for 47877 (fig.1) and 47106 (fig.2). Despite the specification, the impedance remains above 4 ohms at all frequencies.
Fig.1 The Smaller Advent (SN 47877), electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)
Fig.2 The Smaller Advent (SN 47106), electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)
The small wrinkles in the impedance traces suggest the presence of cabinet resonances of some kind. The Smaller Advent's enclosure was indeed quite lively, both samples showing a series of strong resonant modes in the midrange (fig.3).
Fig.3 The Smaller Advent (SN 47106), cumulative spectral-decay plot calculated from the output of an accelerometer fastened to the center of the cabinet's side panel (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz).
The impedance graphs suggest that the woofers of both samples were tuned to the same frequency, 46Hz, which implies modest bass extension. However, as shown by the frequency-response traces (fig.4), the second-order rolloff of the sealed-box alignment results in more mid- and low bass than would be expected from the woofer tuning frequency: there is useful output down to 30Hz or so. The apparent rise in upper bass will be due in part to the nearfield measurement technique, but the woofer alignment does appear a little underdamped, corresponding to Bob Reina finding the speaker to sound "warm" (though this might also stem from the lively cabinet walls).
Fig.4 The Smaller Advent, anechoic response on tweeter axis at 50", averaged across 30° horizontal window and corrected for microphone response, with the nearfield response of the woofer, SN 47877 (red), SN 47106 (blue).
Other than the 1dB difference in sensitivity, noticeable in the midrange in fig.4, the two samples of the Smaller Advent are very similar in overall response. The balance is relatively even through the midrange to the low treble, but there are many small peaks and dips apparent. The mid-treble is plateaued down by 3dB, which will tend to make the Advent's treble sound lifeless, as BJR noted in his auditioning comments. This might be offset by the vicious peak apparent at 13kHz, though the speaker's output drops like a rock above that frequency. On pink noise, this top-octave peak was audible as a distinct whistle. I suspect that it correlates with Bob's dissatisfaction with the treble's "grayish, grainy quality."
The response graphs in fig.4 were taken with each Advent's grille in place, which actually provides the baffling for the tweeter. Fig.5 shows that the grille boosts the speaker's output by an average of 3dB in the upper midrange and mid-treble. Unlike modern designs, the Smaller Advent is best auditioned with its grille in place.
Fig.5 The Smaller Advent, effect of the grille on the tweeter-axis response (5dB/vertical div.).
With its asymmetric and recessed drive-unit placement and its use of a 9.5" woofer and a wide baffle, the Smaller Advent's lateral radiation pattern (fig.6) is perverse, with a lack of off-axis energy in the upper midrange and many ridges apparent in the treble. Above 10kHz, the 2" tweeter's output drops rapidly to the speaker's sides, though as the on-axis response also disappears in the top octave, this might be thought academic. The limited dispersion might ameliorate the effect of that 13kHz peak in-room, though I would also expect the Advent's high treble to lack air. In the vertical plane (fig.7), the Smaller Advent is relatively uncritical of listening axis at high frequencies, though a suckout develops at 1.8kHz (the crossover frequency?) at even small angles above and below the tweeter axis.
Fig.6 The Smaller Advent, lateral response family at 50", normalized to response on tweeter axis, from back to front: differences in response 90–5° off-axis on woofer side of baffle, reference response, differences in response 5–90° off-axis on tweeter side.
Fig.7 The Smaller Advent, vertical response family at 50", normalized to response on tweeter axis, from back to front: differences in response 45–5° above axis, reference response, differences in response 5–45° below axis.
The Advent's step response (fig.8) indicates that the two drive-units are connected in the same, positive, acoustic polarity, the tweeter's output leading that of the woofer. The smooth integration of the two drive-units' steps in the time domain correlates with the smooth blend (seen earlier) of the drive-units' responses on the tweeter axis. The picture presented by the Smaller Advent's cumulative spectral-decay plot (fig.9) is less positive. As expected, the response peak at 13kHz is associated with a ridge of delayed energy, but smaller resonant ridges are evident throughout the treble.
Fig.8 The Smaller Advent (SN 47106), step response on tweeter axis at 50". (5ms time window, 30kHz bandwidth.)
Fig.9 The Smaller Advent (SN 47877), cumulative spectral-decay plot at 50" (0.15ms risetime).
Some aspects of the Smaller Advent's measured performance were positive. For example, it reminded me of the benefits for bass performance of a well-tuned sealed box, compared with the reflex designs that are now ubiquitous. But despite its relatively even balance through the midrange and low treble, the Advent's tweeter is a total underperformer compared with modern designs, and its cabinet is way too resonant.—John Atkinson