Why do I see phantom "ZERO" values using NI 9211?

[dav2010]

Hi,

I am trying to troubleshoot some crashing issues with a VI that depends on measuring voltage of 2 strain gauge sensors.  The VI crashes after 30-40 minutes of operation and the voltage output of the strain gauge becomes non-sensical at the time the VI crashes.   In particular, the voltage that the VI displays is huge (about 10x normal) and doesn't make any physical sense.  Once the VI is stopped and restarted, the output voltage returns to a realistic value immediately.  For example, during operation, the strain gauge is outputting about 6-9mV, and suddenly it jumps to 79mV and locks up until the VI is stopped.   So, I decided to test the power supply that the strain gauges use to see if there is any problem, with the power supply.

So, I used one of the strain gauges, and logged the output voltage from the strain gauge for a few hours, with no motion on the strain gauge, and no change to the supply voltage.  The strain gauge set up was left in the lab with all the lights out, no people, no animals, no wind or weather to disturb them.  Also, the strain gauge was unloaded.  There is some drift from 2.8mV to 2.4mV over time, but, I am more concerned about the ZERO values that show up here.  What is causing them?  Actually, they are not ZERO, they are 0.00001mV readings.

Is there a way to filter out these very low values?  Anything less than 0.5mV is probably not a real value.  The value at 2.5mV is the unloaded strain gauge.

I am using cDAQ 9174 and NI 9211 to measure the voltage of the strain gauge.

Thanks,

Dave

[StevenHowell]

Dave,

Can you post your code? Is it a LabVIEW application? 

Steven Howell

Certified LabVIEW Developer

LabVIEW Consultant

Arcus Technology Services, LLC.

steven.howell@arcustechservices.com

[dav2010]

Sorry, I meant to attach the code.  Thanks for reminding me!  By the way, I ran the test again, and got this data, running the exact same test with the same test code.  The issue appears to be intermittent, which is how my VI failures are; sometimes they fail within 10 minutes, and sometimes it takes 30-40 minutes.  This is a very tough problem for me.

Dave

[johnsold]

Dave,

You appear to have the sample rate set to 3 samples per second and then in a loop with a 5 ms Wait you read one sample and write it to the file. That is attempting to read 200 samples per second and write each of them separately to the file. The Write to Spreadsheet File VI opens and closes the file each time it is called.  Your disk drive must hate you.

The Excel file you posted contains 22591 rows of data collected over 3 hours and 10 minutes according to the note in the file. (3*60+10)*60 = 11400 seconds. So you are recording less than two samples per second.  The way you have the error handling set up you will only see an error if the last iteration generates an error. Set your loop to stop on Error OR the stop button. See if you get an error from the DAQ Read when the zeros occur. Also connect a chart the the data line coming from the Read VI (connect to the same wire as the Voltage indicator. That way you can see whether the zeros are coming from the DAQ Read or due to something in the file write process.

You shoudl look at the Producer/Consumer Design Pattern. This allows you to separate the DAQ timing from the file timing by using two loops.  Consider reading more samples at a time but less often and writing to the file in larger chunks.

Lynn