JMlab Micron & Micron Carat loudspeaker 1991 Measurements
Figs.1 & 2, made with the Audio Precision System One, show the impedance amplitude and phase of the first and second samples of the JM Micron, respectively. Both drop to just below 4 ohms in the upper bass, the overall impedance being nearer to a 6 ohm specification. The tuning of the ports is revealed by the minimum at 60Hz in both graphs, suggesting a relatively restricted bass response. As might be expected from the different tweeters, the manner in which the impedances vary with frequency are very different in the treble. The sensitivity, assessed with an octave-wide band of pink noise centered on 1kHz, was around 88dB/W/m, which is high for a minimonitor.
Fig.1 JMlab Micron Carat, sample 1, electrical impedance (solid) and phase (dashed) (2 ohms/vertical div.).
Fig.2 JMlab Micron Carat, sample 2, electrical impedance (solid) and phase (dashed) (2 ohms/vertical div.).
Turning to the time domain, figs.3 & 4 show the impulse response of the two speakers on the tweeter axis. Note the significant amount of complicated high-frequency, presumably audible ringing imposed on the tail of the Kevlar-tweetered speaker's impulse (fig.3). The titanium dome (fig.4) is much better behaved in this respect and what ringing there is has been pushed above the audio band.
Fig.3 JMlab Micron Carat, sample 1, impulse response on tweeter axis at 1m (5ms time window, 30kHz bandwidth).
Fig.4 JMlab Micron Carat, sample 2, impulse response on tweeter axis at 1m (5ms time window, 30kHz bandwidth).
To the right of fig.5 is shown the response of the original Micron on the tweeter axis, averaged across a 30 degrees lateral window. The ugly behavior of the Kevlar dome tweeter above 14kHz is well-documented. Of more importance, however, is the region an octave to either side of the 4kHz crossover point. The uneven low treble is evident, as is the poor integration of the drivers on this axis. This is presumably the root cause of the timbral errors noted by DO. To the left of fig.5, the nearfield low-frequency responses of the Micron's woofer and port are approximately matched to the averaged on-axis response. The woofer can be seen to roll off below 120Hz, with the port output centered on 60Hz, as suggested by the impedance measurement (fig.1)—DO's in-room measurements yielded a bass response that was flat to 65Hz at the listening seat.
Fig.5 JMlab Micron Carat, sample 1, anechoic response on tweeter axis at 1m, averaged across 30 degrees horizontal window and corrected for microphone response, with nearfield woofer and port responses plotted below 300Hz.
Fig.6 shows the identical curves for the second sample of the Micron. The new titanium-dome tweeter lacks both the top-octave rise and the ragged low treble of its Kevlar-domed sibling, which explains why DO was so much more enamored of its sound. The level matching between the quasi-anechoic midrange and treble curve and the nearfield woofer and port curves can only be approximate, yet the indication was that the woofer of the second sample, despite its impedance plot and port tuning being almost identical to the first sample's, actually gave a little more upper-bass energy (this is why I've raised its level in fig.6). Certainly DO felt the second sample to sound warmer than the first, and considerably better at reproducing the body tone of the cello and the power region of the orchestra.
Fig.6 JMlab Micron Carat, sample 2, anechoic response on tweeter axis at 1m, averaged across 30 degrees horizontal window and corrected for microphone response, with nearfield woofer and port responses plotted below 300Hz.
Looking at how the balance changes with listening axis, there were only minor changes with the measuring microphone moved from an axis level with the center of the woofer to one level with the top of the cabinet. Above that, however, a significant suckout appeared in the crossover region, suggesting that the JMlab should definitely be used with tall stands, 24" probably being the minimum unless you have a very low listening chair. The horizontal behavior is shown in fig.7, with good, even dispersion up to 10kHz, and a decline in output off-axis above that frequency. This kind of radiation pattern is typical of a good minimonitor and undoubtedly contributes to the excellence of its soundstaging. (It isn't all that contributes, however: the ragged low treble of the first sample will have smeared the imaging in this region.)
Fig.7 JMlab Micron Carat, horizontal response family at 1m, normalized to response on tweeter axis, from back to front: differences in response 30 degrees-7.5 degrees off-axis; reference response; differences in response 7.5 degrees-30 degrees off-axis.
The MLSSA "waterfall" plot of the original Micron (fig.8) reveals that the sonic signature of the Kevlar tweeter is quite obvious as a series of strong resonances above 14kHz. A major dip in the amplitude response is evident around 4kHz, with other resonant problems noticeable at 1.4kHz and in the crossover region. The waterfall plot for the second sample (fig.9) is still rather hashy in the low treble, but confirms the overall more even balance of the titanium tweeter in this region.—John Atkinson
Fig.8 JMlab Micron Carat, sample 1, cumulative spectral-decay plot at 1m (0.15ms risetime).
Fig.9 JMlab Micron Carat, sample 2, cumulative spectral-decay plot at 1m (0.15ms risetime).