Hi MAK,
Sorry if I missed your problem initially. We initially tried to build a board with the ADuC using thru-hole parts around the chip, on a SIMM socket card. We had lots of noise, which we resolved by going to a surface mount board. What I ended up doing was duplicating the Analog Devices evaluation board layout as closely as possible. This cut down the noise immensely. However, it sounds like you are seeing this noise even on the eval board. The other question is what is your analog noise? In our case, we are mapping +/- 5 volts to 0 to 2.5 volts, so 1 LSB = 2.44mV. We see 1 to 3 LSBs of digital noise, so as long as our analog noise is below about 5mV, it is below our digital noise, and we don't see it. You may want to investigate that also.

Best Regards,

Josh