% Error is Shifting? - Strain Calibration

[hatchling]

Hi All,

Quick question dealing with an issue I have while calibrating a variety of strain gages.  I'm using LabVIEW 2014 w/ (7) 9236 Modules.

The main problem is that, upon an initial calibration, the Percentage Offset Error gives good-looking values, zeros.  However, once I re-calibrate later, the values jump up to 5, 10, even as high as 15%.  I am not sure what could be causing this.  In both cases, the screen reads "Calibration successful".  Does this mean that I am good to go, regardless of the Percentage offset error value?  I am bothered because anytime I see large % errors, my gut tells me something is wrong...

The images below show the process I am using to calibrate.  I am following the recommended process shown here:  http://www.ni.com/getting-started/set-up-hardware/data-acquisition/strain-gages

When I hook all of the gages up, I then jump into the block diagram, where I open up the DAQ assistant Express VI that I use to select the gages and then calibrate them (see images below)

Once I am in the VI, the following screen is shown.  Here I add the different strain gages via their respective module location:

Once I enter the gage factors/resistance, I continue on to device calibration, which brings up the screen shown below.  I leave both offset nulling and shunt calibration engaged, using the default 100k value of the shunt resistor built into the 9236:

 When I press next, the software takes the first measurement, which displays the screen below.  Things at this point look normal.  I then press calibrate:

And bingo, the first time I do this procedure, I usually get the error % column to read zeros across the board.  The calibration successful indicator gives me good feelings.  HOWEVER...when I repeat this process (a re-calibration) later (perhaps the temperature in the room has changed), I am having trouble.  See below

Here is what happens later on, if I attempt to recalibrate.  I follow the exact same procedure, and seem to be getting large error % values.  Does anyone know what is going on here?  Why would an initial calibration seem fine, and then later give me error after I calibrate again.  It indicates that calibration worked, and the readout values seem correct, but I am irked by this oddity.

Any help would be appreciated, thanks!

[RavensFan]

I think your measured strain values in the last picture look fine.  You went from numbers that were in the hundreds to 1000+ microstrain.  I'm kind of surprised that the next image shows values less than 1 microstrain (all values are e-9 or nanostrain).  The fact that the errors were all perfect 0 is odd, but perhasp roundoff explains it.   In your last picture, your highest number is 2.8 microstrain and all the rest are 1 microstrain or less.  But those percent errors are all very large.

It seems like the routine is comparing your final values (that are small) to a percentage of another number that is still pretty small, buth then giving a large percentage.  Yet, your final value shows Calibration Successful.

I wouldn't worry about your null offset calibration in the last screen.  But something is very odd about its % calculation.

I do notice that the difference between simulated strain and measured strain is very large in the gain adjustment section.  Maybe that is driving the large error in the offset section.  The percentages are very similar between the two.