And last, how close can…
search
There's also an argument to be made that the dynamic power of SETs is better than push-pull. I too cocked an eyebrow when I first heard this one, but Gordon explained: "In push-pull you've got the transformer's core magnetizing, collapsing the field, then magnetizing in the opposite phase, for one thing. And then the core of a single-ended amp has large amounts of DC current in it, and the inductance actually seems to create more power. In a sense, you not only have storage in the power supply, but also instantaneous power stores in the output transformer itself!"
Conclusions
There is no question that the Meridian Digital Theatre system auditioned here performed admirably, and that its individual components were equal to the best obtainable conventional devices. The system was beyond significant reproach in its reproduction of two-channel discs, and the realization of multichannel material was limited only by the source. As an obsessive audiophile, I'd be very happy with the Meridian system, but wonder if another amplifier might have made the speakers sound warmer, or if I could somehow pipe my LPs to the speakers' amps without intervening A/D…
There is no question that the Meridian Digital Theatre system auditioned here performed admirably, and that its individual components were equal to the best obtainable conventional devices. The system was beyond significant reproach in its reproduction of two-channel discs, and the realization of multichannel material was limited only by the source. As an obsessive audiophile, I'd be very happy with the Meridian system, but wonder if another amplifier might have made the speakers sound warmer, or if I could somehow pipe my LPs to the speakers' amps without intervening A/D…
Sidebar 1: Specifications
Reference 800 DVD/CD player: Internal ATAPI/IDE DVD drive. Sampling rates: 32-96kHz, up to 24-bit resolution. DSP-based processing for up/downsampling, error correction, resolution enhancement, gain control (with 72-bit resolution). Configuration as tested: 5 RCA, 2 TosLink, tape monitor digital I/O (ID16); component, composite, and S-video video I/O (VE00); coax and XLR analog outputs with 24/96 D/A conversion (OA20), 2 AES/ EBU & 2 RCA digital outputs (OA20), DVD-Video region decoder (DV10), RS-232 controller (DV00), RS-232 & comm links (C000).
…
Sidebar 2: Associated Equipment
Because, most of the time, I used the Meridian gear as a complete system, only limited comparisons were made to specific reference equipment. However, the resident system at the time of the Meridian auditions was as follows:
Digital sources: California Audio Labs CL20 DVD/CD player, Meridian 508-24 CD player, Mark Levinson No.360 D/A converter.
Preamplifiers: Klyne 6L3.3P, Sonic Frontiers Line 3.
Power amplifiers: McCormack DNA-1 (with Rev.A modifications), Sonic Frontiers Power 2.
Loudspeakers: PSB Gold-i.
Cables: Interconnects: Cardas…
Sidebar 3: Measurements Meridian 800
Looking at the Meridian 800's performance as a CD player from its OA20 digital/analog card, its output impedance was a low 47.5 ohms from the single-ended outputs, and almost twice that figure at 94.8 ohms from the balanced XLR jacks. The player didn't invert absolute polarity with the absolute phase set to "+," and the XLRs were wired with pin 2 positive. Although the volume control operates in 1dB steps up to an indicated "99," the analog output stage clipped with a full-scale digital signal when set to "88." All subsequent measurements were…
But the upper trace shows what happened when I increased the level to "87." The top of the wave in each cycle hard-clips, I assume in the digital filter (a sinewave at the same peak level is reproduced without problem), followed by some full-level RF oscillation. It is extremely unlikely that the Meridian will ever display this tweeter-destroying behavior on 96kHz-sampled music, but perhaps it would be a good idea to keep the volume control below "87."
Channel separation revealed the usual 6dB/octave decrease with frequency, due to capacitive coupling between the channels, but was still…
Sidebar 4: Measurements Meridian DSP5000
Although Kal used the DSP5000s as rear-channel speakers, they are respectable full-range speakers in their own right. I thus performed a complete set of measurements using the DRA Labs MLSSA system, a calibrated B&K microphone, and a dCS 904 A/D converter running at 96kHz sampling to convert the analog test signal to digital.
As a digital-input active speaker, the DSP5000 doesn't have a meaningful voltage-sensitivity rating. However, driving it with a noise signal at -12dBFS and with its volume control set to "70" (out of a possible "99…
In the time domain, the use of DSP-implemented crossover filters results in an excellent, time-coherent step response (fig.15). Note in this graph that, even though the microphone was my standard 50" away, the step doesn't start until the 8.5ms mark---equivalent to an extra time of flight of 4.5ms. This 4.5ms delay is due to the digital filters both in the A/D converter I used and in the DSP5000, but it will be subjectively inconsequential.
Fig.15 Meridian DSP5000, step response on tweeter axis at 50" (5ms time window, 30kHz bandwidth).
Other than the tweeter resonance,…
Sidebar 5: Measurements Meridian DSP6000
For logistic reasons, I measured only the Center version of the DSP6000. However, as this model differs from the side version only in the angle of the head-unit's sidewalls, this set of measured data should be representative of all three speakers.
The DSP6000 appeared to be around 0.5dB more sensitive than the DSP5000 with the same noise signal at -12dBFS and a volume setting of "70." The difference is inconsequential. The balance averaged across a 30 degree horizontal window on the tweeter axis (fig.18) was almost as flat as the DSP5000's…
The '6000's step response (fig.21) is hard to interpret, and I admit I was a little surprised not to see the same time-coherent shape as in the '5000. But after I'd inspected the step responses of the individual drive-units, I realized that what fig.21 shows is first the beginning of the midrange unit's positive-going step just after the 8ms mark, over which is then laid the tweeter's negative-going step. The midrange step overshoots after returning to the time axis, its subsequent positive-going recovery coinciding with the positive-going woofer step. This is not a time-coherent design---if…