Musical Fidelity CD-Pre24 CD player-preamplifier Measurements
Looking first at the Musical Fidelity CD-Pre24 as a conventional analog preamplifier, the input impedance at 1kHz was a usefully high 126k ohms, this not changing with the volume-control setting, and the maximum voltage gain was 11.9dB. (The unity-gain setting of the volume control was approximately 1:00.) The CD-Pre24 preserved absolute signal polarity, and the output impedance was a low 49 ohms other than at very low frequencies, where it rose slightly to 85 ohms.
The A-weighted signal/noise ratio with the volume control wide open and the input shorted was 89dB ref. 1V output, not quite up to the standard set by the Nagra PL-L (reviewed elsewhere in this issue). Unlike the Nagra, however, the Musical Fidelity's figure worsened only slightly, to 83dB, with an unweighted wideband measurement. Channel separation (not shown) was excellent at better than 90dB below 1kHz, though this decreased with increasing frequency due to capacitive coupling, reaching 60dB at 30kHz, which is still good. The frequency response with the volume control set to its maximum position was down by 0.25dB at 20kHz (fig.1, top pair of traces), this rolloff increasing to -0.4dB with the control set to unity gain (middle traces). Into the low 600 ohm load (bottom traces), the bass response rolled off slightly early, reaching -1dB at 12Hz. Note that a 0.2dB channel imbalance creeps in at the control's lower setting.
Fig.1 Musical Fidelity CD-Pre24, analog frequency response at 1V RMS into 100k ohms with volume control at maximum (top), at unity gain (middle), and into 600 ohms (bottom). (Right channel dashed, 0.5dB/vertical div.)
Fig.2 shows that the CD-Pre24 could swing almost 10V into 600 ohms, however. The downward slope of the traces in this graph shows that, below 8V or so, the measured level of distortion in the preamp's output is dominated by noise. Musical Fidelity components have always featured extremely low distortion levels; fig.3 shows how the THD+noise level varies with frequency at 5V output, above the level required to drive a typical power amplifier to clipping but high enough for noise not to dominate the reading. There is a slight but inconsequential rise in THD apparent at the top of the audioband into both 100k ohms and 600 ohms, but a greater rise below 80Hz into 600 ohms. This suggests that the CD-Pre24 has to work hard at combinations of high levels, low frequencies, and low loads, with the third harmonic dominating (fig.4)—if "dominating" is the appropriate word to describe something happening at -94dB (0.002%).
Fig.2 Musical Fidelity CD-Pre24, THD+N (%) vs output voltage into (from left to right): 600 ohms, 1k ohm, 10k ohms, 100k ohms.
Fig.3 Musical Fidelity CD-Pre24, analog stage, THD+N (%) vs frequency (Hz) at 5V into 100k ohms (bottom at 10Hz) and 600 ohms (top).
Fig.4 Musical Fidelity CD-Pre24, spectrum of 50Hz sinewave, DC-1kHz, at 2V into 600 ohms (linear frequency scale).
With 1kHz driven at a more reasonable level into a more reasonable load (fig.5), only the subjectively innocuous second harmonic is apparent above the noise floor, but this lies at a totally trivial -109dB! A similar spectrum (fig.6) reveals low levels of intermodulation products resulting from an equal mix of 19 and 20kHz tones.
Fig.5 Musical Fidelity CD-Pre24, analog stage, spectrum of 1kHz sinewave, DC-10kHz, at 1V into 8k ohms (linear frequency scale).
Fig.6 Musical Fidelity CD-Pre24, analog stage, HF intermodulation spectrum, DC-24kHz, 19+20kHz at 1V into 8k ohms (linear frequency scale).
Turning to the CD-Pre24's digital circuitry, the channel separation (not shown) was about 20dB worse than for analog sources, though still good at better than 90dB below 1kHz. The frequency response for CD replay (not shown) was flat from 10Hz to 10kHz, but drooped slightly at the top of the audioband, reaching -0.25dB at 20kHz. Like other recent digital components from Musical Fidelity, the CD-Pre24 didn't apply the appropriate equalization for pre-emphasized discs, which will be boosted by 9dB above 10kHz as a result.
The CD-Pre24 successfully locked on to external 96kHz- and 88.2kHz-sampled datastreams. However, its frequency response for these sources was curtailed to the CD's 22kHz bandwidth, suggesting that the unit's sample-rate converter chip downsamples to 44.1kHz for these sources and that they are not fed directly to the 96k-capable DACs. Confirming this, only a 44.1kHz datastream is available from the digital outputs even when the CD-Pre24 is driven by a 96kHz source, as Mikey found.
Fig.7 shows spectral analyses of the Musical Fidelity's analog output while it decodes external 16- and 24-bit data representing a dithered 1kHz tone at -90dBFS. The left channel (solid traces) behaves as expected, with the increase in word length lowering the noise floor by up to 15dB; this suggests an ultimate resolution of 18.5 bits, which is excellent. However, if you look at the right-channel performance in fig.7 (dashed traces), not only does the noise floor not drop with increased bit depth, but odd-order distortion harmonics make an appearance. This suggests that the right-channel data are being truncated to 16 bits!
Fig.7 PSB Stratus Gold i, spatially averaged, 1/3-octave response in JA's listening room.
Fortunately, data from the internal CD drive were handled correctly, but I was worried enough by this to obtain a second sample of the CD-Pre (footnote 1). As shown by fig.8, this unit (serial number 10048) did handle 24-bit data correctly in both channels. However, like Mikey's sample (serial number 10000) it still downsampled external high-sample-rate data.
Fig.8 Musical Fidelity CD-Pre24, sample 2, 1/3-octave spectrum of dithered 1kHz tone at -90dBFS, with noise and spuriae (from top to bottom): 16- and 24-bit external PCM data (right channel dashed).
Footnote 1: My thanks to neighbor Wes Phillips for the loan of one of the samples he had received for his on-line review at OnHiFi.com.—JA