Naim CD5x CD player Measurements
As the Naim CD5x is an upgraded version of the CD5, which Michael Fremer reviewed in April 2001, I was interested in seeing what had changed in the player's measured performance. All the measurements were taken from the RCA jacks with the DIN socket turned off with the remote. The maximum output level at 1kHz was about the same as the CD5's, at 2.107V, and the output's absolute polarity was correct. The player's output impedance was still extremely low across most of the audioband, at 2 ohms, but still rose at extreme low frequencies, presumably due to the size of coupling capacitor used. The value of 610 ohms at 20Hz will mean that preamplifiers presenting loads of less than about 5k ohms should be avoided if the low bass is not to sound rolled off.
The CD5x's frequency response is nicely flat (fig.1, top pair of traces), though a subjectively inconsequential rolloff of -0.5dB is evident at 20kHz. And while the original CD5 featured a relatively large balance error when playing back pre-emphasized discs, the CD5x was fine in this respect (fig.1, bottom traces). Channel separation (not shown) was superb, at better than 110dB below 4kHz, and still 90dB at 20kHz. Error correction was not quite up to the superlative standard set by the CD5, but was still excellent, the '5x producing no audible glitches in its output until the gap in the data spiral reached 1.25mm in length.
Fig.1 Naim CD5x, frequency response at -12dBFS into 100k ohms, with de-emphasis (bottom) and without (top). (Right channel dashed, 0.5dB/vertical div.)
Fig.2 shows a spectral analysis of the Naim's output while it played back CD data representing a dithered 1kHz tone at -90dBFS. The trace is free from harmonic products, but some very-low-level power-supply-related spuriae can be seen. These are way too low to be audible, but the Naim did seem very sensitive to the grounding arrangement between it and my Audio Precision System One, at least from its RCA jacks. The picture was very similar when I looked at the spectrum of the Naim's output while it decoded "digital black" (fig.3), though note the absence of the usual rise in ultrasonic noise due to the noiseshaping used to wrest resolution from sigma-delta DAC chips.
Fig.2 Naim CD5x, 1/3-octave spectrum of dithered 1kHz tone at -90dBFS, with noise and spuriae, 16-bit CD data (right channel dashed).
Fig.3 Naim CD5x, 1/3-octave spectrum of "digital black," with noise and spuriae, 16-bit CD data (right channel dashed).
The linearity error of the Burr-Brown PCM1704 DAC chip used by the CD5x (fig.4) is negligible down to -100dB, and below that level is dominated by the dither noise of the 16-bit test signal. The Naim's reproduction of an undithered 1kHz sinewave (fig.5) is excellent, therefore.
Fig.4 Naim CD5x, left-channel departure from linearity, 16-bit CD data (2dB/vertical div.).
Fig.5 Naim CD5x, waveform of undithered 1kHz sinewave at -90.31dBFS, 16-bit CD data.
Though the measured distortion level was very low, at just 0.005% with a full-scale 1kHz tone, there were more higher-order harmonics present than I would like to have seen (fig.6). The even-order harmonics were overall predominant, with, most unusually, the eighth the highest in level, at -89dB (0.0033%). This graph was taken into a lowish impedance, 4k ohms, but the picture didn't change significantly into 100k ohms (not shown). The distortion increased into very low impedances; like the CD5, the CD5x is best used with preamps that don't load their sources with much less than 10k ohms. Though the 1kHz difference product resulting from an equal mix of 19kHz and 20kHz tones was very low, almost -100dB (0.001%), there were higher-order intermodulation distortion products visible in the output spectrum (fig.7). Although these are all too low in level to have audible consequences, I was surprised to see them.
Fig.6 Naim CD5x, spectrum of 1kHz sinewave, DC-1kHz, at 0dBFS into 4k ohms (linear frequency scale).
Fig.7 Naim CD5x, HF intermodulation spectrum, DC-25kHz, 19+20kHz at 0dBFS into 4k ohms, CD data (linear frequency scale).
Finally, the CD5x's word-clock jitter, assessed with the Miller Audio Research Analyzer, was a low 241 picoseconds, almost all of this stemming from sidebands at ±749Hz (fig.8, purple "3" markers) and ±1380Hz (red "5"). The latter coincide with the sixth harmonic of the signal content, a 229Hz squarewave at the LSB level. The spurious tone at around 9.8kHz is associated with the host PC I use for the measurement hardware, but another, higher-level tone can be seen at 7790Hz (blue "102"). This is probably a characteristic of the Naim player, but I have no idea what it is due to. It is still 100dB down from peak level, so is most likely inconsequential.
Fig.8 Naim CD5x, high-resolution jitter spectrum of analog output signal (11.025kHz at -6dBFS sampled at 44.1kHz with LSB toggled at 229Hz). Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz.
"Good rather than great" is how I would sum up the CD5x's measurements, though this didn't prevent AD from enthusing over its sound. But from my experience of how the CD5's measured performance improved when it was powered from Naim's external FlatPack supply, I can't help but wonder if that would improve the CD5x as well.—John Atkinson