Labview Signal Clipping?????

[ghillisniper]

I am very new to Labview and am trying to test a project outline for measuring strain on a destructive test sample.  We have tested several samples that should theoretically reach almost 40% strain.  I have gages that, in theory, should achieve up to 20% strain if bonded properly.  Using an NI-9219 DAQ and Labview on a laptop we were able to get up to about 62m on our strain measurement and then the graph would plateau and remain flat.

This is odd because whether the gage failed or the bond failed, the gage should unload the strain some and we should see that graph have some noise in it.  The flat line suggests to me (a very inexperienced user) that the software or the DAQ may be clipping the signal or placing an artificial ceiling on it.   I have searched and searched for any solid information on if this is a possibility and how a person might remove the ceiling or fix the clipping issue if it is in fact what the problem is.

I am also searching for information on exactly what units I am seeing as output.  I have been assured by coworkers that I am looking at millistrain but other resources suggest it may be a voltage measurement.

Thanks for any help that comes my way!

[Matthew_Kelton]

What you are seeing is the module reaching a voltage limit on its input.  How do you have the 9219 configured?  You should either be doing it in LabVIEW or in NI MAX.  If you do it in MAX, LabVIEW will use that setup if you are just doing reads.  You can fully setup the device as well in LabVIEW and that would ignore the setup of the device in MAX.

[ghillisniper]

Most of it was setup for me.  It appears we are working in NI MAX.  We are running the DAQ in a quarter bridge (one lead wire to terminal 3 and one wire to terminal 5).  They set the excitation voltage at 2.5.  We have contemplated expiramenting with that but have not yet.  I'll have to check on other configuration parameters.  Are there any particulars that I should specifically check that would help troubleshoot this?  I don't know exactly what I'm looking at in some cases. 

Thanks.

[Matthew_Kelton]

You want to look at the max and min input limits.  How are you accessing the device in your code?  Did they create a task for you in MAX to use?

[ghillisniper]

I did a little bit of trial and error with it today.  It is in fact running through LabView Signal Express.  The max and min input limits were 1 and -1 respectively.  I don't know what that means; however, because the scaling seems way off on the graph when I test a strain gage.  Initially, we tested a piece and had it max out (plateau) at just over 6m.  I was told that was equivelent to 6% strain which seemed reasonable for the amount of deformation our test piece had.  However, I just tested a strain gage and folded it completely in half and only got up to about 17m.  I also rolled it around a small radius on a handle of a tool and got just over 7m.  The numbers just don't seem to be right or consistent, although there are some issues with bending them back and forth probably.

I was happy to see the signal not plateauing but I'm still a bit lost.  So I guess to try to clarify what I think I know:

I checked the device in MAX then opened LabView SE.

I added a task in LV to measure an analog input and picked strain.  

This brought up the dialoge box where I could put in my gage factor, gage resistance, max and min inputs, etc.  I only adjusted the gage factor and to see what happened, I changed the max input to 2 this morning.  I didn't change anything else.

We have experienced another, possibly unrelated issue, when trying to calibrate the signal.  We select the second tab in that box (not calibrate but I don't recall what it is off the top of my head) that allows shunt calibration and offset nulling.  In there, we did the offset, allowed the softward to measure, and had an error of something like 124%.  We hit the calibration button and wound up with an error well over 1,000%.  The first 2 test runs we did where the signal plateaued, we got reasonable calibration errors.  Since then we have gotten several error messages when trying to calibrate and any time it actual did the procedure we wound up with higher error after calibration than before.

Sorry to confuse the issue even more with more problems, but nobody here seems to really know the answers.  Thanks.

[Matthew_Kelton]

When you go to the strain gauge options/setup, is everything setup correctly?  There are two types of quarter bridges for strain configuration.

You could also try changing it to a pure voltage measurement and confirm that the raw voltages match what everyone expects.

[ghillisniper]

It is quarter bridge type I.  I'm honestly not sure of the difference but was told to use that one.

Good point on the voltage.  I'll see if I can do that and figure out what output is.

[Matthew_Kelton]

You can try the other config and see if it makes any difference.