Green Mountain Audio Diamante loudspeaker Measurements

Sidebar 3: Measurements

The Diamante's impedance magnitude and phase are shown in fig.1. The single rise in the bass, to 15.5 ohms, suggests that the ATL loading does act as a true transmission line rather than as some kind of perverted reflex, though its frequency, 51Hz, implies limited low-frequency extension overall, as SS found during his auditioning. When the speaker is used with an amplifier having a high output impedance, this 51Hz impedance peak will result in an effective boost at this frequency. I suspect that this is why SS found his Atma-Sphere tube amplifier not to be a good match for the Diamantes.

Fig.1 Green Mountain Diamante, electrical impedance (solid) and phase (dashed) (2 ohms/vertical div.).

Varying between 5 and 6 ohms between 110Hz and 14kHz, with a very low phase angle, the impedance curves suggest that the speaker will be a very easy load for the partnering amplifier. The voltage sensitivity, however, estimated at around 82dB/W/m using B-weighted noise, is very low. Moderately powered amplifiers will not drive the Diamante to comfortably loud levels without strain.

Turning to the frequency domain, fig.2 shows the nearfield response of the woofer spliced to the far-field response on the midrange axis, averaged across a ±15° horizontal window. (I understood from Roy Johnson during his visit to Santa Fe that listening with your ears level with the midrange unit, 34" from the ground, was optimal.) Again, the Diamante's bass is revealed as being quite limited in extension, the graph revealing an anechoic –6dB point of 50Hz, the box-tuning frequency. The rollout slope, however, is the gentle 12dB/octave typical of a sealed enclosure, and in a room of normal size will be boosted by boundary effects.

Fig.2 Green Mountain Diamante, anechoic response on midrange axis at 45", averaged across 30° horizontal window and corrected for microphone response, with nearfield woofer response below 200Hz.

Of greater concern is the shelved-down mid- and high-treble region seen in fig.2. As well as a slight suckout in the crossover region—due to the tweeter and midrange outputs not quite adding in phase on this axis at the 45" microphone position—the tweeter bandpass is about 3dB too low in level. I repeated the measurement at a 90" measuring distance, to confirm that I was not looking at a position-dependent effect—Roy had been careful to specify a listening distance of 8' or more—but there was no appreciable change in the measured response. The effect of this tweeter shelving will be to make the Diamante sound too sweet, too mellow in absolute terms; though this certainly won't be unpleasant, it is a departure from strict accuracy.

All the measurements so far were made with the standard H-200 crossover module. Substituting the H-100 module and subtracting the acoustic response of the H-200 Diamante gives the curve shown in fig.3. As can be seen, replacing the H-200 with the H-100 module shelves the tweeter down by another 1.5dB, which will make the speaker sound even more mellow—probably a good thing with cheap CD players and the like.

Fig.3 Green Mountain Diamante, effect of H-100 crossover module, normalized to response on midrange axis with H-200 module.

If the drive-units aren't blending perfectly on the midrange axis, where do they do so? Fig.4 shows how the response changes from a microphone position ranging from level with the top of the ATL base (the front curve) to 15° above the enclosure top (the rearmost curve). There's more crossover-region energy on or below the woofer axis, which is too low for comfortable listening. More importantly, however, there's significantly more treble energy directed up above the loudspeaker. This, presumably, is why SS felt the Diamante to not work terribly well in his large room with its live ceiling; the "bounce" of this treble energy from the ceiling to the listening position would make the sound too bright. Steven did his listening with his ears level with the Diamante's tweeter, 38" from the floor. Fig.4 implies that, though this deepens the 3kHz suckout—perhaps correlating with Steven's feeling that the speaker lacked information—it does add a bit of mid-treble energy, which perhaps is why he didn't remark on an over-mellow high treble.

Fig.4 Green Mountain Diamante, vertical-response family at 45", normalized to response on midrange axis, from back to front: differences 15°, 10°, and 5° above cabinet top; difference level with cabinet top; difference on HF axis; reference response; difference level with ledge on baffle; difference on woofer axis; difference level with top of base unit.

Laterally (fig.5), the Diamante maintains its response within a ±15° window, but the suckout in the crossover region deepens further to the speaker's side. This dispersion is not as good as that implied by the specification. With the speakers set up in my room, I took a spatially averaged 1/3-octave curve (fig.6), calculated by averaging 120 individual measurements taken for left and right speakers separately over a 72"-wide by 36"-high window centered on the listening position. The midrange is evenly balanced, but again, the treble can be seen to be smoothly sloped-down compared, for example, with the B&W Silver Signature that I review elsewhere in this issue. And this is what I heard. While the Diamantes threw a fairly deep soundstage, with good image focus, I felt their tonal balance to be too mellow. With fizzed-up recordings, like the Linda Ronstadt/Emmylou Harris/Dolly Parton Trio CD (Warner Bros. 25491-2), this came as a welcome relief; but with naturally miked, naturally balanced recordings, such as the Wilson Audio Beethoven violin sonata LP (W-8315), it was too much of a good thing too much of the time.

Fig.5 Green Mountain Diamante, horizontal-response family at 45", normalized to response on midrange axis, from back to front: differences 90° through 5° off-axis; reference response; differences 5° through 90° off-axis.

Fig.6 Green Mountain Diamante, spatially averaged 1/3-octave response in JA's listening room.

Turning to the time domain, fig.7 shows the Diamante's step response on its tweeter axis, where SS auditioned it. The ideal step response, the speaker's output when presented with a DC voltage step, is represented by a vertical line away from the time axis, followed by a slow, slightly concave decay back to it. The small up/down spike in fig.7 is the tweeter output, followed by the midrange and woofer outputs, with their longer rise-times. Obviously the Diamante doesn't offer a time-coherent performance on this axis.

Fig.7 Green Mountain Diamante, step response on HF axis at 45" (5ms time window, 30kHz bandwidth).

Moving down to the midrange axis (fig.8), the tweeter output is moved back slightly in time so that it almost coincides with the midrange output. It's still a bit early, however. Finally, moving the microphone down so that it was level with the slight ledge on the front baffle, between the midrange and bass units, gives the best step response (fig.9), the slow rise of the midrange output being overlaid with the tweeter output to give a good step shape.

Fig.8 Green Mountain Diamante, step response on midrange axis at 45" (5ms time window, 30kHz bandwidth).

Fig.9 Green Mountain Diamante, step response on "ledge" axis at 45" (5ms time window, 30kHz bandwidth).

To examine whether this performance was truly time-coherent on this low axis, I looked at the speaker's excess phase performance. Excess phase is the difference between a speaker's actual phase response on any particular axis and the theoretical phase response it would have calculated from its amplitude response by a mathematical operation called the Hilbert Transform. In a truly time-coherent design, the excess phase would be zero across the audio band—of the speakers I've measured, only the Thiel CS2 2 and Dunlavy Audio Labs SC-IV have gotten close to this paradigm. The Green Mountain's excess phase response meets a pretty good ±15° tolerance between 150Hz and 10kHz, but the shape is not flat, implying that the speaker is still not completely time-coherent, at least at 45".

The cumulative spectral-decay, or waterfall, plot gives an indication of how clean a loudspeaker's decay is. Any resonant modes where energy is smeared out over time as ringing will reveal themselves as ridges parallel to the time axis. Fig.10 shows the Diamante's plot: The initial decay is clean, but the midrange unit has a resonant mode noticeable at 2kHz, the cursor position, while the tweeter appears to have a strong mode present around 5kHz. I would expect the former to add a slight nasality, the latter a slight brightness, to the speaker's sonic signature.

Fig.10 Green Mountain Diamante, cumulative spectral-decay plot on midrange axis at 50".

Finally, I used a simple accelerometer to examine the resonant behavior of the Green Mountain Diamante's cast enclosure. (This is nothing more than a quantitative version of the classic knuckle-rap test.) While the top cabinet shows a low-level mode at 160Hz, present on side and back walls and perhaps related to the internal geometry of the speaker, the strongest mode could be found on the rear wall opposite the midrange unit. This is high in frequency, however, at 480Hz, and should only occasionally be excited by music (as opposed to test tones). The base unit was much more live than the top enclosure when judged by the knuckle-rap test, but because it's both isolated by the top enclosure's feet and is only excited by the energy in the transmission lines, its vibrational behavior is minimal.

As with some other low-order–crossover designs—the Unity Audio Signature 1 reviewed last year (Vol.16 No.5, p.119) by Robert Deutsch comes to mind—the listening axes on which the speaker is time-coherent and offers the most neutral response don't coincide. All in all, though its measurements were somewhat mixed, I was quite impressed with the Diamante. Although I occasionally thought I caught a hint of lower-midrange congestion, the Diamante doesn't appear to suffer from the normal transmission-line ill of a colored, lumpy upper bass. But with its mellow tonal balance and well-defined soundstaging, the Diamante is a good, if pricey, first design from a new company. Its enclosure is beautifully finished—a work of art in its own right. I look forward to hearing what Green Mountain comes up with in the future.—John Atkinson

Green Mountain Audio
310 South 25th Street
Colorado Springs, CO 80904-3500
(719) 636-2500