Error 200019 problem with USB-6251

[jchsu]

Ok, sorry, I was wrong, but it still does not explain why if I sample the unconnected channel (ai1), I get the expected noise and not the function generator signal (note I still have ai0 connected to the function generator, I am simply not sampling it).

[John_P1]

Hi Jeremy,

The AI channels of the 6251 are actually multiplexed into a single A/D converter.  If you specify to read from multiple channels then they will be included in the scan list of the multiplexer and the hardware will switch between channels.  If only acquiring from one channel, the 6251 won't switch to the unused channels.

Most of the time, ghosting can be traced back to having a high-impedance signal source: an open channel would be an extreme case of this where the impedance could be considered infinite.  The AI channels have a capacitance associated with them, so this "infinite" impedance increases the settling time so greatly that it will look like you are simply sampling the same signal twice.

When you are only measuring the unconnected channel, the signal on your other line is never multiplexed to the A/D converter.  You may still get some cross-talk, but ghosting will not occur in this case.  Typically, measuring an open channel is not desired--if acquiring multiple channels you will very likely experience ghosting, and if only measuring the single channel the measurement will likely drift to the rails of the A/D converter.

The following KnowledgeBase also gives a good decription of ghosting and how it can be prevented:  How Do I Eliminate Ghosting From My Measurements?

-John