Question re asynch AI, NI DAQCard-AI-16XE-50

Hi list,

Being new I need some steering in the right 
direction!

I am attempting asynchronous AI using a NI 
DAQCard-AI-16XE-50.
After sorting out some REALLY silly problems I 
had, I now have one (probably equally silly) 
where I can't figure out what it is that I am 
missing.

I read four AI channels on subdevice 0, I used 
the demoprogram cmd.c and set my own values for 
freq, n_scan, n_chan, subdevice, channel, range 
and aref where suggested.

I read 1000 scans from 4 ea 16-bit channels in a 
successive read of 1024 bytes each 7 times, and 
the last read is 832 bytes (a total of 8000 
bytes). Range is set to 0 (=[-10,10]V).

What puzzles me now is that the last read FIFO 
buffer is always different from all the previous 
ones. It *looks* like I get two's complement in 
all but the last FIFO buffer, and straight
binary in the last read (but a value of approx 0 
volts on the oscilloscope gives ~32000, so the 
range still appears to be [-10,10]V). All four 
channels behave the same, and if I choose to read 
512 byte "chunks" the last 512+320 bytes are 
"odd".

Is there anyone out there who could hint me as to 
what I am missing, or where it would be a good 
idea to start looking?

I'm really really out of my depth here... 
As you can probably tell from the mail anyway :-)

Yours thankfully and apologetically,
Carin Cronander

Received on 2003-05-14Z10:05:48