I have two small batches (2 pieces each batch) of AD9851 BRS DDS chips. The two batches were obtained from different sources at different times. All driven with 30 Mhz clock sources, using the internal 6X multiplier, same circuitry, same firmware but different results. Batch 1 performs to datasheet but batch 2 result in incorrect output frequencies, requiring a different frequency control word to get the same frequency. The number needed is as if the master clock is at 120 Mhz not 180 Mhz.(30 *6). Batch 2 also shows what looks like distortion at the center point on the sine output rise and fall and this translates to noise on the comparator outputs at the 50% amplitude point of the rise and fall. Distortion and noise are most prevalent between 50Khz and 2 Mhz output frequencies. Batch 1 are labled AD9851 BRS and Batch 2 are AD9851 BRSZ. I amstumped!