Acoustic Energy AE1 loudspeaker 1992 Measurements part 2

The AE1's impulse response (fig.7) is dominated by the ultrasonic—ie, inaudible—tweeter ringing. The step response (fig.8) is also obscured by the ringing from the tweeter, but suggests that both drive-units are connected in inverted acoustic polarity, this confirmed by looking at the step responses of the individual units (fig.9). This graph also shows that the tweeter's output leads that of the woofer by about 250µs.

Fig.7 Acoustic Energy AE1, on-axis impulse response at 50" (5ms time window, 30kHz bandwidth).

Fig.8 Acoustic Energy AE1, on-axis step response at 50" (5ms time window, 30kHz bandwidth).

Fig.9 Acoustic Energy AE1, on-axis step responses at 50" of tweeter (red) and woofer (blue) (5ms time window, 30kHz bandwidth).

The cumulative spectral-decay plot (fig.10) reveals a generally clean initial decay across the band, broken by some mild hash in the mid- and low treble. (Resonant modes reveal their existence in this by appearing as ridges parallel with the time axis.) To see whether some of the treble modes are due to the woofer breakup behavior—the metal-cone woofers used by Monitor Audio, for example, feature a degree of woofer ringing in the mid-treble; see Vol.15 No.6, p.199—I repeated the waterfall measurement of fig.10 on the woofer alone. The result can be seen in fig.11. There is still some odd behavior going on between 1 and 3kHz, but apart from the well-damped residual mode just above 5kHz indicated by the cursor portion and also seen in fig.2, the driver rolls out cleanly.

Fig.10 Acoustic Energy AE1, cumulative spectral-decay plot at 50" (0.15ms risetime).

Fig.11 Acoustic Energy AE1, cumulative spectral-decay plot of woofer at 50" (0.15ms risetime).

Finally, I used a simple accelerometer to examine the cabinet behavior (see Vol.15 No.6, p.205). Listening to the sidewalls with a stethoscope while the speaker was playing pink noise revealed a single resonance in the midrange; processing the accelerometer's output to give a "waterfall" plot (fig.12) indicated it to lie at 469Hz, coincident in frequency with the wrinkle in the AE1's impedance plot (fig.1). This resonance was also detectable on the cabinet's other surfaces, but much lower in level.

Fig.12 Acoustic Energy AE1, cumulative spectral-decay plot calculated from the output of an accelerometer fastened to the cabinet sidewall. (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz.)

What effect would it have on the speaker's sound? It is low in level and, being of high Q, would not be excited all the time. In addition, as its frequency lies well above middle C, it will not obscure the speaker's clarity in the tenor or bass regions. [Fig.13 shows the same graph plotted on a vertical scale consistent with what I later adopted for this measurement, which undercores how low in level these cabinet modes are.]—John Atkinson

Fig.13 Acoustic Energy AE1, cumulative spectral-decay plot calculated from the output of an accelerometer fastened to the cabinet sidewall. (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz.)

COMPANY INFO
Acoustic Energy
8709 Castle Park Drive
Indianapolis, IN 46526
(317) 849-5880
ARTICLE CONTENTS

X