Harbeth Super HL5plus loudspeaker Measurements

Sidebar 3: Measurements

I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the Harbeth Super HL5plus's frequency response in the farfield; and an Earthworks QTC-40, with its small, ¼"-diameter capsule, for the nearfield responses.

The Harbeth is specified as having a sensitivity of 86dB/W/m, and my estimate of its voltage sensitivity was the same: 86dB(B)/2.83V/m. Though this is 1dB or so below average, I commend Harbeth for not inflating this specification. Although the Super HL5plus has a specified impedance of 6 ohms, my measurement (fig.1) indicates that the speaker's impedance remains above 8 ohms for almost all of the audioband, and that the electrical phase angle remains relatively small. The HL5plus will therefore be an easy load for the partnering amplifier to drive, and a good match for tubed designs.

615HLS5fig1.jpg

Fig.1 Harbeth Super HL5plus, electrical impedance (solid) and phase (dashed) (2 ohms/vertical div.).

A wrinkle in the impedance-magnitude trace at 150Hz suggests a cabinet resonance of some kind at this frequency. Indeed, investigating the enclosure panels' vibrational behavior with a simple plastic-tape accelerometer (footnote 1) while it was supported on three upturned cones revealed a strong resonance at 152Hz on all surfaces. Fig.2 was taken with the accelerometer at the center of the top panel; this resonance is indicated with the cursor, and there are two more resonant modes an octave higher, at 300Hz. However, while there also some higher-frequency modes on the side and rear panels, these were at different frequencies. This suggests that the idea of using a thin-walled cabinet to maximize the quality of a speaker's midrange reproduction —proposed by, among others, Harbeth founder Dudley Harwood when he worked at the BBC in the early 1970s—does work as promised. However, I do wonder if the mode at 150Hz contributed to Art Dudley's feeling that the HL5plus had a "warm tone."

615HLS5fig2.jpg

Fig.2 Harbeth Super HL5plus, cumulative spectral-decay plot calculated from output of accelerometer fastened to center of top panel (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz).

AD did comment on the Super HL5plus's good low-frequency extension, and the saddle centered on 33Hz in the impedance-magnitude trace suggests that this is the tuning frequency of the port on the front baffle. However, the minimum-motion notch in the woofer's nearfield output (fig.3, blue trace), which is where the cone is held stationary by the back pressure of the port resonance, occurs slightly lower, at 29.3Hz. The port's output, again measured in the nearfield, peaks at the same frequency (fig.3, red trace), and its upper-frequency rolloff is clean and free from resonant modes.

615HLS5fig3.jpg

Fig.3 Harbeth Super HL5plus, acoustic crossover on upper-tweeter axis at 50", with nearfield responses of woofer (blue) and port (red), respectively plotted below 350Hz and 425Hz.

Higher in frequency in fig.3, the crossover of woofer to tweeters occurs at 3.2kHz with what appear to be fourth-order acoustic slopes. The woofer rolls out without any significant breakup issues. While the tweeters roll in smoothly, the response above 12kHz is disturbed by dips and peaks, presumably due to interference effects between the two tweeters on this axis. (The microphone was level with the topmost, smaller-diameter tweeter.) This unevenness can also be seen in fig.4, which shows the Harbeth's response averaged across a 30° horizontal window centered on the upper tweeter axis. However, below the top octave, the HL5plus's response is superbly even, and the slight boost in the upper bass is entirely an artifact of the nearfield measurement technique. Note also the excellent bass extension in this graph.

615HLS5fig4.jpg

Fig.4 Harbeth Super HL5plus, anechoic response on upper-tweeter axis at 50", averaged across 30° horizontal window and corrected for microphone response, with complex sum of nearfield responses plotted below 300Hz.

Fig.5 shows the Harbeth's horizontal dispersion, referenced to its output on the upper tweeter axis, which therefore appears as a straight line. The contour lines in this graph are evenly spaced and smooth, though there is a slight lack of energy at extreme off-axis angles just below the crossover to the tweeters. In the vertical plane (fig.6), a suckout centered at 2.5kHz develops more than 5° above and below the upper tweeter axis. However, there is a little more top-octave energy 5° below this axis, which suggests that the optimal axis on which to listen to the Harbeth is actually level with the lower, larger-diameter tweeter, as AD found in his auditioning. You need to use this speaker with a stand high enough to place its 1" tweeter level with your ears.

615HLS5fig5.jpg

Fig.5 Harbeth Super HL5plus, lateral response family at 50", normalized to response on upper-tweeter axis, from back to front: differences in response 90–5° off axis, reference response, differences in response 5–90° off axis.

615HLS5fig6.jpg

Fig.6 Harbeth Super HL5plus, vertical response family at 50", normalized to response on upper-tweeter axis, from back to front: differences in response 45–5° above axis, reference response, differences in response 5–45° below axis.

Turning to the time domain, the Super HL5plus's step response on the upper tweeter axis (fig.7) suggests that both tweeters are connected in negative acoustic polarity, the woofer in positive polarity. That the decay of the tweeters' step doesn't quite smoothly blend with the start of the woofer's step on this axis also suggests that the optimal listening axis will be slightly below the top tweeter. AD commented that "the Harbeth Super HL5plus sounded conspicuously, even startlingly, clear." It came as no surprise, therefore, to see that the Harbeth's cumulative spectral-decay plot (fig.8) demonstrated a superbly clean decay throughout the midrange and treble. Harbeth's RADIAL2 material does indeed result in a well-behaved woofer cone. The two response peaks between 20 and 30kHz are presumably due to the tweeters' fundamental dome resonances. Note also that this graph was taken on the lower tweeter axis; there is more energy apparent above 12kHz, supporting my conjecture in the discussion of the vertical dispersion that this would be the optimal axis on which to listen to this speaker.

615HLS5fig7.jpg

Fig.7 Harbeth Super HL5plus, step response on upper-tweeter axis at 50" (5ms time window, 30kHz bandwidth).

615HLS5fig8.jpg

Fig.8 Harbeth Super HL5plus, cumulative spectral-decay plot on lower-tweeter axis at 50" (0.15ms risetime).

Other than that lively enclosure, which is a deliberate design decision—note AD's comment about "the consistently truthful, present manner with which they reproduce singing voices"—the Harbeth Super HL5plus's measured performance is beyond reproach.—John Atkinson



Footnote 1: See my article "The Sound of Surprise."—John Atkinson
COMPANY INFO
Harbeth Audio, Ltd.
US distributor: Fidelis AV
460 Amherst Street (Route 101A)
Nashua, NH 03063
(603) 880-4434
ARTICLE CONTENTS

X