Pioneer Elite TZ-9 loudspeaker Measurements
Sidebar 3: Measurements
Looking at the way the TZ-9's impedance changes with frequency (fig.1), the LF enclosure-resonance peak lies at 47Hz, with the tuning of the ports revealed by the impedance minimum half an octave lower at 32Hz. The speaker would best be characterized as a 4-ohm design. The sensitivity was high, at around 91dB for a 2.83V input (measured using a 1/3-octave pink-noise band centered on 1kHz). The TZ-9 will go very loud with moderately powerful amplifiers (as long as they have no difficulty driving 4-ohm loads).
Fig.1 Pioneer TZ-9, electrical impedance (2 ohms/vertical div.).
Fig.2 shows the TZ-9's response to a unidirectional rectangular impulse 55µ;s long at a 1m measuring distance. (The horizontal scale is 5ms; 32 samples were averaged because of the poor signal/noise ratio of the test signal (footnote 1); a DC offset has been introduced to show as much of the pulse as possible; and the ripples before the onset of the pulse, which are due to the tail of a previous pulse, should be ignored.) This response was taken in front of the tweeter, which corresponds to just below the listening axis in my room. Interpretation of a raw impulse response is fraught with difficulty, but I conjecture that the initial negative-going drop is probably due to the tweeter, with then a well-defined rise in the opposite direction due to the midrange dome, which is connected with opposite polarity. The overshoot of the midrange output is overlaid with what is possibly further output from the tweeter, followed by a damped but complete cycle from the woofer with an equivalent frequency of around 550Hz. With the exception of this delayed arrival, the time behavior is pretty good, suggesting good HF transient performance. The tail of the impulse is broken up by a mild degree of HF ringing, the main component of which has a period of approximately 150µs; ie, a frequency of just under 7kHz.
Fig.2 Pioneer TZ-9, impulse response on tweeter axis (5ms time window).
To confirm my description of the impulse arrival behavior, I repeated the measurement, this time with my hand pressed against the baffle over the tweeter. The midrange unit impulse (not shown) starts in a positive-going fashion, with an equal overshoot on the opposite side of the time axis, the end of this pretty much coinciding with the much lazier rise of the woofer outputs, which are connected with the same polarity as the midrange. There was still quite a lot of HF ringing overlaying the tail of the midrnage unit's impulse response; I assume, that this is a characteristic of the midrange unit.
The FFT-derived frequency response drived from the fig.2 impulse response indicated a slight falling trend with frequency, with an excess of energy both in the exact region around 500Hz where the front woofer crosses over to the midrange dome, and at 700Hz. There is also a slight lack of energy an octave above the upper crossover region. A very narrow notch at the base of this suckout at 7700Hz was considerably deeper when the measurement was repeated with the mike midway between the midrange dome and the tweeter, though the depth of the overall suckout between 6 and 10kHz remained constant. This suggests that this notch is due to cancellation, most probably between the direct sound from the tweeter and a reflection of that sound from a nearby obstacle. I doubt that the overall notch will be audible, however. The 500700Hz lift also became more apparent with the mike moving toward the midrange axis.
Fig.3 shows the TX-9's spatially averaged room response. This curve has proved, on a comparative basis, to predict reasonably well what is heard, performing a measure of integration of the direct sound of the loudspeaker with the reverberant field in my listening room. The steady-state frequency responses of the speakers in the listening area are measured using pink noise and an Audio Control Industrial SA-3050A 1/3-octave spectrum analyzer with its calibrated microphone. Nine sets of six averaged measurements are taken independently for left and right loudspeakers, at a distance of just over 2m in a window 72" wide and varying from 27" to 45" high. The response shown in fig.3 is the average of these measurements, weighted slightly toward the sound heard at the listening position. This spatial averaging is intended to minimize the effect of low-frequency room standing-wave problems (below 500Hz or so) on the measurement; the individual measured spectra also give an idea of the off-axis behavior of the speaker under test.
Fig.3 Pioneer TZ-9, spatially averaged, 1/3-octave response in JA's listening room.
The midrange in-room is very flat, but the low bass has rather too much energy in the 3080Hz region, with the fast roll-out below 30Hz typical of a reflex design. This ties in with the subjective feeling of "too much bass" on double bass and male spoken voice. The low-frequency extension is excellent, however, as is to be expected from such a large enclosure. Measured with spot sinewaves in the nearfield, midway between the front woofer and the ports gave an approximate 6dB-down point of 25Hz. No wonder organ recordings came over well! The lower midrange seems quite smooth, with only a faint hint of the usual depression in the 200400Hz region due to destructive interference between the direct sound and the primary reflection from the floor between the speaker and the measuring microphone. The in-room sound is smooth throughout the treble region.
As a further aid to diagnosis, I carried out individual nearfield measurements on the front woofer, one of the ports, and the dome midrange unit. The result, with approximate scaling applied to each set of level measurements to reveal the crossover points, can be seen in fig.4. The dome midrange can be seen to roll-out acoustically with a 24dB/octave character below 480Hz. As expected, the port output is centered on 32Hz, where the speaker's impedance is at a minimum, with the woofer taking over above 50Hz or so. The front woofer extends almost an octave higher in frequency than the rear woofer, with its upper 6dB point lying around 500Hz. Below 200Hz or so, there will be an equal contribution to the speaker's response from the rear woofer, so the woofer response in-room will be at least 3dB higher than that plotted in fig.6, correlating with the in-room measurement that indicated a rather bass-heavy sound.
Fig.4 Pioneer TZ-9, nearfield responses of midrange unit (green), front woofer (blue) and port (red).
Listening to pink noise revealed that what cabinet output there was was centered in the 240280Hz range. I was surprised to find that the sidewalls and rear panel were quite lively in this region, presumably because the early roll-off of the rear woofer means that the reaction-canceling action of the tie-rod is less operative in the lower midrange and above. Certainly, on pink noise, a slight "hoot" in this region was quite audible, particularly when the listener was off to the side of the main axis. This would suggest that the decision made by Pioneer's engineers not to provide any bracing of the side panels was not wise.John Atkinson
Footnote 1: Because noise is random, each doubling of the number of samples increases the desired signal by 6dB but the noise by only 3dB. Thus, averaging 32 samples will increase the S/N ratio by 15dB.