Wadia 27ix & Wadia 270 transport Measurements part 4

Intermodulation distortion was low (fig.12), with the 1kHz difference component below -90dB from peak level. (Note that, with this test signal, the peak level is 6dB higher than the level of either of the individual tones.) However, you can see images of the tones reflected back down from the sampling frequency between 12kHz and 16kHz. The price paid for the Wadia filter's excellent time-domain behavior is poorer rejection of ultrasonic spuriae. Interestingly, the intermodulation products rise a little into a 600 ohm load (fig.13), but so do the aliased images.

Fig.12 Wadia 27ix, balanced outputs, HF intermodulation spectrum, DC-24kHz, 19+20kHz at 0dBFS, 100k ohm load. (Linear frequency scale, 20dB/vertical div.)

Fig.13 Wadia 27ix, balanced outputs, HF intermodulation spectrum, DC-24kHz, 19+20kHz at 0dBFS, 600 ohm load. (Linear frequency scale, 20dB/vertical div.)

Using the Miller Audio Research Jitter Analyzer, I tested the Wadia components' jitter performance both as connected with the ST-optical links and with the D/A processor driven by a PS Lambda transport. (The Miller rig is a combination of proprietary software written for a National Instruments PC card-based measurement system and a low-jitter CD-R carrying a special test signal.)

I first investigated the effect of the clock link. The 270's clock error was just 3 parts per million, which indicates superb accuracy. Without the clock link, and with the Wadia 270 and 27ix connected by an ST-optical datalink, the measured jitter level was 248 picoseconds peak-peak—somewhat higher than I have measured with other great-sounding CD playback systems. The dCS Elgar, for example, driven by a Meridian 500 transport, achieved an astoundingly low 142ps. When I added the clock-link connection, which meant that the 270 transport now derived its reference clock from the crystal controlling the 27ix's D/A converters, the measured jitter dropped only slightly, to 234ps.

X