Acquiring with PDA, but channels are being swapped, I think.....

[twowheels00]

Equipment:
HP iPAQ h5550 w/ expansion slot
DAQCard AI-16E-4
Powering and collecting voltages from six potentiometers

I have noticed an interesting phenomena with the data I'm getting ready to collect. I have attached an Excel file of a static test I did. By static I mean that the potentiometers are static, no rotation occuring, so output voltages will remain constant, or pretty close. Looking at the file (each column is a channel) it is obvious that the channels are swapping, for example, every 50 or 55 acquisitions the channel in column 1 will change to column 2, and will continue to move down the columns. The first 120 rows are the same, and then it begins. It's very obvious if you watch the locations of the highest and lowest channel voltages (2.08 V and 2.36 V). I hope that explains it well.

So, some more info. I have attached the VI I produced for the task. My task has the following information:

Software: DAQmx Base 1.0.1, LabView 7.1, LabView PDA 7.1
Channels:gain=1, Bipolar, referenced single-ended (they are being compared to the AIGND)
Scan Rate: 200Hz
Number of Scans: 0 (for continuous)
Convert Rate: 50000 Hz

I'm collecting 5 samples/channel at a time

Furthermore, during acquisition, the voltages showing on the VI plot, with a PDA, are not the same as those being stored. My plot shows 4 channels above 2.5V and two channels below 2.5V, for the static test. These voltages were confirmed by using the above DAQ card in a laptop. So, the voltages shown on the PDA plot seem to match those seen on the laptop test panel, but the voltages stored are different (lower, essentially). Just to note, neither the PDA or laptop were plugged into a wall outlet for testing, to eliminate excess noise in the card.

What do you think? I was thinking that somehow the VI or task is changing what channels to collect, or some type of attenuation happening somewhere. I'm really kind of lost here. I discovered this when I was trying to figure out which column in my tab-delimited file corresponded to which potentiometer (if channel 0 was the first column or the 6th column).

I appreciate any help here. This is the final task in my master's thesis, so I'm anxious to figure it out!!

Josh Nordquist

[twowheels00]

To ALl,

Here is the excel sheet with static test data

Thanks!!

Josh

[twowheels00]

Ok,

I ran some tests. I collected each channel seperately. The voltages remain constant this way. But, the voltages collected are not the same as those shown on the plot.

I'm assuming this is a crosstalk issue now. So, how can I wire these to reduce the effect? I have included a wiring schematic, the wiring is really basic, but maybe there are some things I can do to decrease any cross talk effects.

[Otis]

If you just sample 2 channels and read a different number of samples do you see the same behavior? For example, if you read 3 samples would it repeat every 30 samples? If you read just 2 or 3 channels do you see the same behavior?

[twowheels00]

I did change the samples collected from 5 to 6, and noticed the phenomena happend about every 56 acquisitions instead of 50.... I'll try some of these tests

Josh