a strange difference between S and M series

[Graziano]

Hi Everybody,
I'm facing a strange problem....

My purpose is to acquire a sinusoidal signal, and to detect the phase of the signal, with a resolution inferior to 1 milliradian. I did a program in LabWindows for an S-series board (NI 6143), and it works really well. I made the program run with a new board, installed on the same PC, which is an M-series, that has a higher Sampling Rate (NI 6251). Sampling at the same frequency of the old case, 250 kS/s, some strange behaviours appear on the detected phase, it seems to show saw-teeth behaviour, with an amplitude of some milliradians, not suitable for my purposes...

The sampling clock is given from an external DDS, in both cases.

At different sampling rates the phenomenon remains. I'd like to know if it's tied to the different internal mode of operation of the two boards, or where does this phenomenon come from....

Thanks in advance,

graziano

[Dr._Imad]

M series / E series boards do not do simultaneous sampling. Their operation is muliplexed, where the first channel is first scanned , then the next, then next etc... So you have interchannel delay, which is not significan. However, for the accuracy you may be looking at, the delay is siginificant and shows as aphase delay (fake) between two signals.

The S series boards are not multiplexed. They perform simultaneous sampling, with dedicated ADC.