Digital Processor Reviews

With the Moon 891's balanced and single-ended line inputs, the preamplifier preserved absolute polarity at the balanced and unbalanced outputs. With the volume control set to the maximum, the voltage gain at 1kHz was 9.75dB from the balanced output with both balanced and unbalanced inputs, and 9.66dB from both output types with an unbalanced input. (These gains are lower than the specified 13.5dB.) Setting the output level to "Fixed," the gain from both output types was 0dB, ie, unity gain.

The line input impedance is specified as 22k ohms. The single-ended impedance was 11k ohms at 20Hz and 1kHz and 6.2k ohms at 20kHz, and the balanced input impedance was 44k ohms at 20Hz and 1kHz, 23k ohms at 20kHz. The single-ended output impedance was the specified 50 ohms across the audioband; the balanced output impedance was 96 ohms, again at all audio frequencies. The line frequency response was flat up to 30kHz into 100k ohms and 600 ohms (fig.12), gently rolling off above that frequency to reach –3dB at 200kHz, as specified. The very close channel balance and the overall response were preserved at lower settings of the volume control.

Channel separation was superbly high, at >130dB in both directions below 2kHz and still 110dB at the top of the audioband. The wideband, unweighted signal/noise ratio, taken with either the balanced or unbalanced inputs shorted to ground and the volume control set to its maximum, was a superb 99.3dB ref. 2V in both channels. This ratio improved to 113.2dB when the measurement bandwidth was restricted to the audioband, and to 116dB when A-weighted. Fig.13 shows the spectrum of the Moon 891's balanced low-frequency noisefloor at 2V with the volume control set to its maximum. The level of random noise is extremely low, and there are no power supply–related spuriae visible in this graph.

Fig.14 plots how the THD+noise percentage in the Moon 891's balanced output varies with output voltage into 100k ohms. The downward slope of the traces indicates that the distortion lies below the noise up to 4V. Our usual definition of clipping is when the THD+N reaches 1%; however, the Moon's balanced output seemed to be limited to 12.1V, when the THD was 0.045%. The 891 behaved very similarly into 600 ohms, and with the single-ended output into both impedances.

Fig.15 shows how the THD+N percentage changed with frequency at 4V into 100k ohms (blue, red traces) and 600 ohms (green, gray traces). The distortion into both impedances is extraordinarily low, with only a slight rise below 100Hz. As with the digital inputs, the second harmonic was the highest in level, but even into 600 ohms lay at –106dB (0.0005%; fig.16). Intermodulation distortion was also superbly low in level (fig.17).

One of the Moon 891's unbalanced line inputs can be turned into an MC- or MM-compatible phono input. As another reviewer will be auditioning the phono stage, I examined its measured performance in both MM and MC modes. To minimize noise, I connected a wire from one of the Audio Precision's ground terminals to the grounding lug on the Moon's rear panel.

Out of the box, while the MC input impedance and gain were correctly set to 100 ohms and 60dB, the MM input impedance was set to 10 ohms and the gain to 66dB. Using the front-panel buttons and menu, I reset these parameters to the recommended 47k ohms and 40dB. The MM input impedance was 42k ohms at 20Hz, 39k ohms at 1kHz, and 32k ohms at 20kHz. The MC mode's input impedance was 103 ohms from 20Hz to 20kHz. In Variable mode, the MM input's maximum gain at 1kHz was 50.1dB from the balanced and unbalanced outputs. In Fixed mode, the gain was 40.3dB from both outputs. The MC input's maximum gain was 68.1dB in Variable mode and 58.4dB in Fixed mode, again from both output types. Both input types preserved absolute polarity at the balanced and unbalanced outputs.

The phono input's RIAA correction (fig.18) was well matched between the channels and extremely accurate. The ultrasonic response was down by just 1dB at 100kHz. Channel separation was superb, at 100dB in both directions across the audioband. The MM mode's wideband, unweighted S/N ratio, ref. 1kHz at 5mV, assessed at the balanced output with the inputs shorted to ground and the volume control set to the maximum, was an excellent 87.7dB in both channels. Restricting the measurement bandwidth to 22Hz– 22kHz increased the ratio by 2dB, while inserting an A-weighting filter resulted in a ratio of 93.5dB. These ratios were taken with the gain set to 40dB. Resetting the gain to 60dB reduced the ratios by 20dB. The MC mode's S/N ratios, ref. 1kHz at 500µV and with 60dB gain, were less good, at 51dB (wideband), 52dB (audioband), and 56.75dB (A-weighted). To get the lowest noise from the Moon 891's phono input, the gain should not be set too high. Spectral analysis of the MM input's low-frequency noisefloor indicated that no supply-related spuriae were present.

To be sure I wasn't prematurely clipping the preamplifier's output, I examined the phono input's overload margins with the volume control set to "Fixed." Both the MM input's and MC mode's margins were impressively high from 20Hz to 20kHz, at 22.3dB ref. 1kHz at 5mV and 24dB ref. 1kHz at 500µV, respectively. Even with an input around 10dB below the 1kHz overload level, the only distortion harmonic present was the second, at an inconsequential –100dB (0.001%, fig.19). The levels of the intermodulation products with an equal mix of 19 and 20kHz tones were similarly negligible (fig.20).

The measured performance of the Moon 891 is state-of-the-art for both analog and digital inputs!—John Atkinson