The input to the 8405A inputs comes from a two channel DDS set to output two 10MHz signals 0.01 or 0.001 Hz apart to create a rotating phase with 100 or 1000 seconds period. The output of the DDS has been assessed and its phase linearity over time is better then 1e-12 The  +/- 0.5 V phase output is sampled at 0.1 s interval with a 6.5 digit multimeter. The +/- 0.5 V is translated to +/- 50 ns and input to Timelab. The linear residue of a full rotation is calculated and appears to have a 3e-9 magnitude non linearity Is this to be expected? The ADEV @ 1 s of the 1000 s rotation is about 1e-11. Is this expected? Can these values be used in a comparison with other phase measurement devices? Erik.