BBC LS3/5a loudspeaker Harbeth Measurements
The LS3/5a's impedance magnitude and phase are shown in fig.1. Dropping to 6 ohms only in the top treble octave, it shouldn't give the partnering amplifier any hernias. The sealed box is tuned to 76Hz, the impedance reaching 31.2 ohms at that frequency. There is a very slight wrinkle in the impedance trace at 350Hz, probably not visible at the printed scale of the graph, which coincides with the frequency of the main cabinet resonant mode (see later). The speaker's sensitivity is low at around 82.5dB/W/m; it won't accept high power inputs without the woofer bottoming.
Fig.1 Harbeth LS3/5a, electrical impedance (solid) and phase (dashed) (2 ohms/vertical div.).
Fig.2 shows the quasi-anechoic response averaged across a 30 degrees horizontal window, spliced to the nearfield woofer response, while fig.3 shows the significant effect of the grille on that response. It can be seen that using the speaker with the grille on, as recommended, gives the flattest overall response. The woofer is 6dB down at a reasonable 60Hz, with a 2.5dB hump between 120Hz and 180Hz, this adding warmth and a semblance of power to orchestral sound.
Fig.2 Harbeth LS3/5a, anechoic response on tweeter axis at 45", averaged across 30 degrees horizontal window and corrected for microphone response, with nearfield woofer response plotted below 300Hz.
Fig.3 Harbeth LS3/5a, effect of grille.
Laterally, presumably because of the recessed baffle, the speaker develops a rather peaky mid-treble off-axis (fig.4), which might be expected to make the room sound a little bright. Vertically (fig.5), the outputs of the two drivers remain integrated as long as the listener is between the woofer axis and the top of the cabinet. Above and below that range a crossover suckout develops, but the severity of the suckout above the speaker implies that tall stands are best. In my listening room, the spatially averaged traces for the two different LS3/5as sitting on 24" stands are shown in fig.6. Both are a little shy in the bass, below the somewhat exaggerated upper-bass region. (This is partly a room effect.) The older pair can be seen to have more energy in the top two audio octaves, tying in with my listening impressions—though the Harbeth pair, which I preferred overall, had a little more brightness-region energy.
Fig.4 Harbeth LS3/5a, horizontal response family at 45", 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.
Fig.5 Harbeth LS3/5a, vertical response family at 45", normalized to response on tweeter axis, from back to front: differences in response 15 degrees-5 degrees above reference axis; reference response; differences in response 5 degrees-10 degrees below reference axis.
Fig.6 Harbeth LS3/5a, spatially averaged, 1/3-octave response in JA's room.