Sharp SM-SX100 digital integrated amplifier Measurements part 4

Increasing the measurement bandwidth to 200kHz and changing the input data to 16-bit "digital black" gave the traces shown in fig.11. Again, the high noise floor in the audioband and the 500kHz spectral component can be seen, but so can the full extent of the noiseshaping. The noise peaks at -35dB around 120kHz, even after its processing by the amplifier's internal passive filter!

Fig.11 Sharp SM-SX100, 1/3-octave spectrum of digital black, with noise and spuriae, 16-bit data (right channel dashed).

Finally, I assessed the Sharp's rejection of wordclock jitter using the Miller Audio Research Jitter Analyzer, another "virtual" instrument. The device under test is fed data representing a high-level 11.025kHz tone with the LSB toggled at 229Hz. Its analog output, set to a level of around 1.75V RMS, is connected to the DSP card, and Paul Miller's software package averages 64 32k-point FFTs and searches the resultant spectrum for symmetrical sidebands.

The results are shown in fig.12. The measured jitter level was an enormous 25 nanoseconds, about 200 times higher than the best PCM converters. None of this jitter was due to the data themselves, but instead consisted primarily of low-frequency sidebands—at ±10.7Hz (purple "1" numeric markers), ±15.6Hz (purple "2"), and ±27.3Hz (purple "3")—and their harmonics. There is also a relatively strong pair of sidebands present at the AC supply frequency of ±60Hz (brown "8").

Fig.12 Sharp SM-SX100, high-resolution jitter spectrum of analog output signal (11.025kHz at -6dBFS with LSB toggled at 229Hz). Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz. Grayed-out trace is Sony SCD-1.

And as with the earlier spectral analyses, fig.12 shows the Sharp's noise floor to be around 24dB higher than the best conventional players, which implies a noise-restricted dynamic range for the SM-SX100's CD playback of 12 bits. (For comparison, the grayed-out trace in this graph is a spectral analysis performed on the Sony SCD-1's analog output.)

On the face of it, an amplifier accepting digital input data and operating entirely within the digital domain is a very attractive idea. But as Sharp's SM-SX100 reveals, it takes heroic engineering to make it work, and there are still some compromises involved, particularly in achieving sufficient dynamic range. But it is important not to read too much into the Sharp's measured performance. MF thought the amplifier's sound generally superb, which strongly implies that much of its idiosyncratic behavior is subjectively benign. But I do wonder if his dislike for the sound of the amplifier's PCM digital inputs was connected with the high jitter seen in fig.12.—John Atkinson