Meridian 203 D/A processor Measurements
The 203's output level when decoding a full-scale 1kHz sinewave was 2.358V (left channel) and 2.346V (right), slightly higher (1.4dB) than the 2V standard and Meridian's specification. Frequency response (fig.1) was flat, but with some very small amplitude irregularities (passband ripple) above 1kHz. However, these are sufficiently low in amplitude to be inaudible. De-emphasis error (fig.2) revealed a 0.3dB negative error at 4kHz, correcting at 8kHz, then another 0.3dB negative error at 16kHz.
Fig.1 Meridian 203, frequency response at –12dBFS into 100k ohms. (Right channel dashed, 0.5dB/vertical div.)
Fig.2 Meridian 203, frequency response at –12dBFS into 100k ohms with de-emphasis. (Right channel dashed, 1dB/vertical div.)
Looking at the spectral content of the 203's output when decoding a –90.31dB, 1kHz sinewave produces the plot in fig.3. The levels of both distortion harmonics and noise are generally quite low, but with some power-supply–related noise apparent at 60Hz and 120Hz. In a foreshadowing of the 203's excellent linearity measurements, the –90.31dB signal can be seen to reach exactly the –90dB horizontal division. Indeed, the 203's extraordinary linearity is shown by the Departure from Perfect Linearity graph in fig.4. This is remarkable performance and the best I've measured. However, it is not surprising in light of the Bitstream conversion method, which boasts good low-level linearity as one of its primary advantages. The left channel only is shown in this graph, the right channel being identical. Playing the "Bonger Test" from the Chesky Test CD confirmed the good low-level performance, with a minimum of converter artifacts audible.
Fig.3 Meridian 203, 1/3-octave spectrum of dithered 1kHz tone at –90dBFS, with noise and spuriae with 16-bit CD data. (Right channel dashed.)
Fig.4 Meridian 203, left-channel departure from linearity, 16-bit CD data (2dB/vertical div.).
Channel separation was similarly impressive. I measured better than 120dB across most of the band, rising only slightly to 100dB at 10kHz. Again, this is excellent performance, especially in a relatively inexpensive converter. The output impedance was to specification at 12 ohms across the band.
The 203's reproduction of a dithered, 1kHz sinewave at –90.31dB is shown in fig.5, and is overlaid with a little HF noise. The 1kHz, 0dB squarewave (fig.6) is typical of Philips linear-phase filtering, showing the unclipped "ringing" on the waveform tops. The 203's impulse response (not shown) reveals that the 203 is non-inverting.
Fig.5 Meridian 203, waveform of undithered 1kHz sinewave at –90.31dBFS, 16-bit CD data.
Fig.6 Meridian 203, waveform of 1kHz squarewave at 0dBFS.
Overall, the 203 performed very well on the bench, especially in linearity and channel separation. I would like to have seen less de-emphasis error, however.—Robert Harley