I get some impossible readings during a data acquisition

[Maxime]

First of all, I am using a NI PCI 6110 device and Labview 6.1 (under Windows XP).

Ok, here's what happens:
I am using basic VI's (AI Config, AI start, AI Read and AI clear) to do data acquisition on 3 analog channels. The AI Read is inside a "for" loop, and is executed only twice. Parameters for the data acquisition may be changed by the user before the acquisition. The data obtained is saved in a .xls file.

The problem I have right now is that the first few (about 1 to 5, it changes randomly) data acquisition done using the VI are generally perfect (note: for testing purpose, the parameters are the same for all the data acquisition and the sampling frequency is 5 or 10 kHz). Then, one of th
e data acquisition will have one impossible reading. By impossible, I mean a voltage of either 1250 or 1312,5 V on a 120 V RMS line. Once I get that in a data acquisition, if I still don't change any parameters, that strange reading will always be at the exact same place in other data acquisition. If I change the parameters, the odd reading will simply be at another place.

The only way I've found to get rid of that reading is to restart the computer. Once it has restarted, I can get 1 to 5 good data acquisition, then it starts again.

I've attached images of two different Excel files containing bad data. I was using the same parameters for both acquisitions. The last column of data indicates the state of a relay (used as a switch; 170 means connected and 0 means disconnected). It is analog data transformed, once the acquisition is over, into a kind of digital data. It is sometimes (not always) affected by the same problem, except that the strange value is 0. The current
values do not seem to be affected.

One last note: the scan backlog of the AI read always indicates 0 and I can confirm that the bad data is coming directly from AI read; it does not appear during data processing.

Any ideas would be appreciated; thanks!

[abu]

try filter the signal first before it goes to your pc... and see what happen....

[Maxime]

I really can't explain why, but I just solved this problem...

The only thing I changed is that the number of scans acquired was calculated from the measured frequency of the signal (which is always close to 60 Hz) and from a user input (an integer from 0 to 190). Instead, I replaced that frequency by a constant (60 Hz). Now, for some reason, I don't have that problem anymore.

What made me think of this solution is because my VI worked in the past, and I remembered one of the last modifications I made, which was to use the frequency to calculate the number of scans.

Like I said, I don't know how the number of scans affected the data acquisition to the point of creating impossible values.