From a strictly technical viewpoint, the Monica 2 is a poor design. First, all asynchronous reclocking adds jitter equal to the period of the reclocking frequency. In this case, the 80MHz reclock adds approximately 8ns rms jitter. Furthermore, the reclock jitter has a linear distribution where all possible values have equal probabilities. In other words, the likelihood that any particular clock edge occurs 8ns later than it should is the same as any particular clock edge occurring 1ns late or any other deviation from the correct time. The jitter from the CS8414 receiver is spec’d at less than 200ps rms. The natural jitter of the PLL in the receiver has a Gaussian distribution where the closer a value is to the correct time, the higher its probability. In other words, most of the clock edges will be at or very close to the mean frequency of the PLL.
Again, discussing non-oversampling DACs from a strictly technical viewpoint, the period of the DAC sample clock should be identical to the period defined by the sample rate of the digital audio data. Any deviation in the period during the D-to-A conversion results in distortion in the analog signal. A 1KHz sine wave at 0dBFS reclocked at 80MHz has a calculated S/N about 5.5dB lower than the same signal without reclocking.
If you thought reclocking somehow reduced or eliminated jitter, think again. If you prefer the sound of a reclocked DAC, enjoy.
