Celestion 100 loudspeaker Measurements

Sidebar 2: Measurements

The Celestion 100's plot of impedance amplitude and phase (fig.1) reveals a relatively easy speaker for an amplifier to drive, with a minimum value of 6.3 ohms around 150Hz. The sealed box is tuned to 60Hz, the frequency of the amplitude peak in the bass, while a slight wrinkle in the trace at 400Hz implies a cabinet or drive-unit resonance at this frequency.

Fig.1 Celestion 100, electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)

Looking at the individual quasi-anechoic responses of the drive-units (fig.2) reveals the acoustic crossover to lie around 2.2kHz, as specified, with the woofer rolling out smoothly. To the left of figs.2 and 3 is the response of the woofer measured in the nearfield. A classic sealed-box response, it rolls off gently below 70Hz. To the right of fig.3 is shown the response on the tweeter axis averaged across a 30° horizontal window. Basically smooth, it features a slight rising trend broken by a degree of peakiness at the top of the woofer's passband. Note also the large pair of peaks above audibility due to the tweeter's "oil-can" resonance.

Fig.2 Celestion 100, acoustic crossover on tweeter axis at 48", corrected for microphone response, with the nearfield woofer response plotted below 300Hz.

Fig.3 Celestion 100, anechoic response on tweeter axis at 44", corrected for microphone response, with the nearfield woofer response plotted below 300Hz.

Fig.4 reveals how the 100's sound on the tweeter axis changes as the listener moves from in front of the speaker to the side. (Only the differences are shown, which is why the on-axis response at the rear is depicted by a straight line.) The mid–high treble can be seen to become increasingly depressed to the side, but notice also the response gully developing in the high midrange. This is due to the woofer becoming increasingly directional as the driving signal approaches the top of its passband, while the tweeter radiates more or less hemispherically at the bottom of its passband. Whether this behavior will lead to an in-room sound that is rather bright or not will depend on the room furnishings (or absence of same) and the proximity of the speakers to the sidewalls. In the vertical plane, the 100 is best listened to on or just below the tweeter axis, as can be seen from fig.5.

Fig.4 Celestion 100, lateral response family at 44", normalized to response on tweeter axis, from back to front: differences in response 90–15° off axis, reference response, differences in response 15–90° off axis.

Fig.5 Celestion 100, vertical response family at 44", normalized to response on tweeter axis, from back to front: differences in response 10–5° above axis, reference response, differences in response 5–10° below axis.

Taking 10 individual 1/3-octave responses for each loudspeaker in a 72" by 24" grid centered on the listening seat in my room and averaging them gives the curve in fig.6. The peaks and dips below 300Hz are residual room modes that haven't been removed by the spatial averaging, while the overall response can be seen to slope slightly upward through the midrange and low treble, this correlating with the slightly lean LF balance noted in the auditioning. Despite the small enclosure, there is useful bass output to below the 40Hz band in-room, the woofer's relatively slow, 12dB/octave rolloff being compensated to some extent by room reinforcement (as explained by Martin Colloms last December in "Basso Profundo").

Fig.6 Celestion 100, spatially averaged, 1/3-octave response in JA's Santa Fe listening room.

In the time domain, the Celestion 100 has a classic impulse response (fig.7) for its type, a slightly lazy decay from the highish-order crossover being overlaid with ultrasonic ringing from the tweeter. Despite the presence of grooves on the baffle plates and the rounded baffle edges, there are still some early reflections noticeable on this plot about 1ms and 1.5ms after the initial transient. The impulse responses of the individual drive-units (not shown) and the step response (fig.8) reveal them to be connected with the same polarity.

Fig.7 Celestion 100, impulse response on tweeter axis at 44" (5ms time window, 30kHz bandwidth).

Fig.8 Celestion 100, step response on tweeter axis at 44" (5ms time window, 30kHz bandwidth).

The cumulative spectral-decay or "waterfall" plot (fig.9) shows a relatively clean decay in the treble, though this is to some extent due to the program's autoscaling raising the graph's noise floor to accommodate the tweeter's ultrasonic resonance. There is a hint of some untoward behavior at the cursor position just below 5kHz as well as at the top of the woofer region, between 1 and 2kHz, which might tie in with the slight coarseness heard in the low treble at high playback levels.

Fig.9 Celestion 100, cumulative spectral-decay plot at 44" (0.15ms risetime).

Finally, despite the well-braced construction, an accelerometer reveals a strong cabinet mode to be present at 410Hz (fig.10), this coincident with the wrinkle in the impedance plot (fig.1). However, this mode is much less strongly excited on the sidewalls than it is on the rear panel; it therefore might only have a minor effect on sound quality (though I assume it correlates with the "clouding" occasionally heard on baritone voice and on solo cello in its lower registers).—John Atkinson

Fig.10 Celestion 100, cumulative spectral-decay plot calculated from the output of an accelerometer fastened to the center of the cabinet's sidewall (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz).

10 Timber Lane
Marlboro, NJ.
(732) 683-2356