Stirling Broadcast BBC LS3/6 loudspeaker Measurements
I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the Stirling Broadcast BBC LS3/6's frequency response in the farfield, and an Earthworks QTC-40 for the nearfield responses. (I measured S/N SB88019A.) The testing was primarily performed without the speaker's grille; I did repeat some of the farfield measurements with the grille in place. Stirling specifies the LS3/6's voltage sensitivity as 87dB/W/m; my estimate was significantly lower, at 84dB(B)/2.83V/m.
How the LS3/6's impedance magnitude and electrical phase angle vary with frequency is shown in fig.1. The Stirling is specified as a nominal 8 ohm design; its impedance can be seen to drop below 8 ohms only in the lower midrange, in the presence region, and above 35kHz. It is closer, therefore, to having a nominal impedance of 12 ohms, which will go some way to explaining the lower sensitivitythe speaker will be drawing less than 1W with a drive voltage of 2.83V. With a minimum value of 6.85 ohms at 3.6kHz and a phase angle that remains low except when the impedance magnitude is high, the LS3/6 will also be an easy load for the partnering amplifier.
There is a discontinuity at 160Hz in the impedance traces. Investigating the vibrational behavior of the cabinet walls with a simple plastic-tape accelerometer, I did find a strong resonance at that frequency on the sidewalls (fig.2), as well as a mode at 145Hz on the rear panel. Given the large radiating areas of the affected panels, and the low frequencies and high amplitudes of these modes, I would have expected this behavior to produce some audible congestion. Art Dudley didn't remark on any such coloration, but I do wonder if his feeling that double bass lacked the "snap and temporal sharpness" he is used to hearing from his reference Altec Valencias was related to these modes.
The saddle centered on 40Hz in the impedance-magnitude trace (fig.1) suggests that the large-diameter reflex port on the front baffle is tuned to that frequency. The woofer's output, measured in the nearfield (fig.3, blue trace below 350Hz), does have the expected minimum-motion notch at 39Hz, and the port's output, again measured in the nearfield, peaks in textbook manner between 30 and 60Hz (red trace). The port also rolls off smoothly, disturbed only by a small, low-level peak in the midrange. The woofer's upper-bass response appears to be boosted in this graph. This will be due in part to the nearfield measurement technique, but it does suggest that the LS3/6 has a somewhat underdamped reflex.
Higher in frequency in fig.3, the woofer crosses over to the tweeters (green trace) at about 3.3kHzslightly higher than the specified 3kHzwith high-order filter slopes. There is a very narrow and likely inaudible peak in the tweeters' output around 22kHz, which can also be seen in the Stirling's farfield response, averaged across a 30° horizontal window on the main tweeter axis (fig.4). (I chose this axis for measurement because that is the axis that is the standard 36" from the floor with the LS3/6 sitting on its 17"-high stand.) Other than the upper-bass prominence noted earlier, the LS3/6's overall response is remarkably flat and even. Repeating this measurement with the grille in place reduced the level of the mid-treble region by a couple of dB, but left untouched the levels of the top octave and low-treble regions.
Below 300Hz, fig.4 shows the complex sum of the woofer and port outputs, weighting each in the ratio of its radiating diameter. As is usual with a reflex design, the Stirling rolls off with a steep 24dB/octave slope below the frequency of the port resonance. I suspect that this, together with the slightly higher-than-expected frequency of the port resonance for the speaker's size, correlates with AD's finding that the LS3/6's low frequencies did not sound as weighty as he would have liked.
The LS3/6's horizontal dispersion, normalized to the response on the main tweeter axis, is shown in fig.5. The crossover frequency of 3.3kHz is on the high side for a woofer with a radiating diameter of 6.5", and the dispersion does indeed show a trough to the speaker's sides at the top of the woofer's passband. (This is because the woofer's size is now of the same order as the wavelengths of sound in this region.) As a result, the lower tweeter's wide dispersion at the bottom of its passband appears to produce an excess of off-axis energy in this region. All things being equal, this will render the Stirling's in-room balance somewhat bright, particularly in small, barely furnished rooms. However, things are rarely equal, and given the Stirling's flat on-axis response, this off-axis behavior may just add to the speaker's sense of articulation and detail, especially when the speaker's grille is left in place.
I was surprised to see that, with a small-diameter supertweeter taking over above 13kHz, the LS3/6's top-octave dispersion was not wider than it appears in fig.5. I suspect, however, that the dispersion is being affected by the wide baffle. In the vertical plane (fig.6), a suckout in the crossover region between the woofer and tweeters occurs more than 10° above or below the lower tweeter axis. In addition, a large suckout occurs in the crossover region between the two tweeters 5° above or below the reference axis. This behavior suggests that, to hear the flattest balance, the listener's ears should be level with the Stirling's lower, primary tweeter.
In the time domain, the LS3/6's step response on the lower tweeter axis (fig.7) suggests that the supertweeter is connected in inverted acoustic polarity, while the tweeter and woofer are connected in positive polarity. However, the decay of each drive-unit's step smoothly blends with the start of that of the next lower in frequency, which correlates with the excellent frequency-domain integration of their outputs seen in fig.4. There is a slight ripple in the decay of the woofer's step with a period of 1ms. However, there appears to be no significant delayed energy at 1kHz in the Stirling's cumulative spectral-decay plot (fig.8). Other than a small mode at 4.1kHz, and others just below and above 20kHz, the initial decay in this graph is commendably clean.
The Stirling Broadcast LS3/6's measured performance is what I would have expected from a classic BBC-inspired design updated for the 21st century.John Atkinson