Scan lists for splitting sampling intervals on PCIe-6363?

[chris131]

Hi,

I'm trying to set up a data acquisition task on a PCIe-6363 to get data from three analog signals. The first signal needs a sampling rate of 1 Ms/s, whereas 500 ks/s would be sufficient for the other two. Sampling all three at 1 Ms/s exceeds the board's capability. The board is able to sample two channels at 1 Ms/s, but not three.

I came across some discussions of using scan lists, which define the order by which channels are sampled. Would it be possible to set up a scan list on this board that achieves the above, by sampling in the order CH1,CH2,CH1,CH3,CH1,CH2,CH1,CH3, etc...? Here, channel 1 would be sampled at twice the rate of channels 2 and 3. If so, what would be the string to set up such a scan list?

Many thanks for your help!

[Kevin_Price]

This *used* to be possible.  But the capability was taken away sometime after DAQmx 16.0

I don't know why and registered my own complaint / request here.

(I only tested with X-series hardware like yours.  It's conceivable that M-series or other AI devices might still support it.  But I kinda doubt it.)

-Kevin P

[ZYOng]

Maybe it is still possible with Read Multiple Channels Sequentially Using DAQmx?

However, NI 6363 has the sampling rate of 2MS/s for single channel and 1MS/s aggregate for multichannel. Most likely you can only do 500kS/s + 250kS/s + 250kS/s.

[chris131]

Thanks! Alternatively I may go with a second board. I noticed that while the PCIe-6363 allows the measurement at an aggregate of 2 Ms/s even when using multiple channels, it gives a warning that the settling time is violated and the accuracy might be compromised. It's probably better if I don't use it at the rate outside of specs. It would still be useful to be able to configure the channel switching with the scan list, a shame that ability was removed.

[Kevin_Price]

---

@ZYOng wrote:

Maybe it is still possible with Read Multiple Channels Sequentially Using DAQmx?

Unfortunately, I don't think so, at least not for X-series that I did some testing on.  The referenced example was written back in 2009.  Back then DAQmx still supported repetition of a channel in a scan list.  Well, at least for multiplexing boards, which was what I habitually used.   (Also, the focus of the example is on controlling the convert clock rate to force non-default multiplexer switching intervals.)

But if anyone can demonstrate otherwise, that'd be good news!  Post here and also rebut the Idea Exchange entry I linked to.

-Kevin P