calibration issue with labview 8.0

[Team21]

Hello.  I am an engineering student at the University of Connecticut.  For my senior design project, my partner and I are using LABVIEW 8.0 to acquire our data from two sensors, a load cell and a LVDT, simultaneously.  We had built the VI already and it ran fine.  Last Monday, my partner calibrated the LVDT using feeler gauges.  This past Friday, I calibrated the load cell using a lever system.  Saturday we set up to do our testing, and somehow the calibration of the LVDT was lost, it was as if it had never been calibrated.  So we calibrated it again with the feeler gauges, and then tried to run the VI.  It said there was an error in the load cell channel, and would not run.  The error said something about a repeated value in the array. We then deleted the calibration we just performed on the LVDT, and tried running it, and the VI ran fine and the load cell worked perfectly.  We then recalibrated the LVDT, and tried running it on its own, without the load cell calibrations, and that worked fine as well.

So basically, we were able to get it running correctly with either one of the two sensors calibrated correctly, but not both at the same time.  Would you be able to give me any advice?  My advisor feels it is probably something simple that we are overlooking. Thanks for your help.

Matt

[johnsold]

Matt,

Your advisor is probably right. However, since most of us are not clairvoyant, it would help if you post your VI so we can see what you are doing.

How are you saving the calibration coefficients?

What interacts between the load cell and LVDT channels?

Do not expect anyone on the Forum to do your project for you. We will help you see what you were doing wrong, but you need to correct the program yourselves to learn anything.

Lynn

[Tom_W_[DE]]

Hi Matt-

I'll assume you're using the Virtual Channel Calibration in MAX.  The error you saw about repeated values in the array likely resulted from accidentally entering the same reference value twice in the table.  Which version of NI-DAQmx are you using?  This is also just a guess, but the loss of calibration data may be caused by not saving the virtual channel after exiting the calibration utility.

I would suggest that you calibrate both channels again, making sure that there are no duplicate entries in the calibration table and that you be sure to save each virtual channel before moving to the next. 

Hopefully this helps-  

[Team21]

OK, i'm using LABVIEW 8.0.    I've never really used LABVIEW that much, i did on a couple labs for other classes, but those the teaching assistants pretty much made the VI's for us.  I've never really had to use it on my own.  And we had a friend, a grad student, help us with this one. The VI is basically just a DAQ assistant inside a loop, outputting to a few graphs. Channel zero is the load cell, and channel one is the LVDT.  To calibrate it, we clicked on the DAQ assistant, clicked which channel, then calibrate on the right side.  Applied various loads (or displacements) to the sensors, and recorded the calibration values. No values were repeated.  After doing that, it asks for a min and a max value or something... what is this that its asking for?

I dont have the VI with me, its saved on the computer in the school lab, and i'm at home.

[Tom_W_[DE]]

Hi Matt-

The min and max values are required to interpolate scaling for the whole range.  If you don't apply simulated load/displacement that corresponds to the measurement limits it will ask you to confirm that you want to just use the ideal conditions.  I was able to replicate the error you seem to be describing by simulating the same measurement and entering it into the table twice.  How much variation do you see between your calibration test points?  Can you apply some large positive test point and large negative test point for reference to see if we can force the calibration to commit?  I was able to successfully calibrate, save, and use an LVDT and strain/load cell task within the same DAQ Assistant so it sounds like some small problem on your end.

Hopefully this helps-

[Team21]

Tom,

I will be going to the lab tomorrow afternoon to hopefully fix my VI.  Do you have a phone number at work that I can contact you on while I have the software open, if I need more assistance?  Thank you for all your help.

-Matt

[Tom_W_[DE]]

Hi Matt-

I'm assuming your University has a site license for LabVIEW, so you can actually contact our applications engineers directly by phone.  There will be someone available to help you out in real-time between the hours of 8:00AM and 5:00PM (CST).  The best way to get in contact is to choose the Phone NI option at http://www.ni.com/support/

Thanks-

[Team21]

I called the support line last night, spoke with Jason.  Come to find out, there's a glitch in the software that if the two sensors are not calibrated with the same number of calbration points, it will return that error.  So sure enough, I recalibrated the LVDT with the same number of points as the load cell, and it ran.

Ok, now i have another question.  The load cell is rated up to 2200 newtons.  For the calibration of it, i was only able to apply 859 newtons to it using the lever system i rigged up. Is it possible for LABVIEW to extrapolate that so i can measure values higher than 859 newtons?  In the DAQ manager, in the load cell channel it says to input the max and min value.  If i try to put 1000 in for the max, it says it cant do that, "You have requested 1000. Valid values end with 859."

I tried standing on the load cell to see how accurate it was, I weigh a little more 859 newtons.  The reading topped out at 859.000 newtons.  It's accurate between 0 and 859, but can i get it to read higher than 859 without doing any more calibrating?

[Justin_M]

Hi Matt,

A couple posts up, Tom mentioned that the max and min you set during calibration determine the entire range (meaning how high LabVIEW will interpolate up to/down to).  What did you set for the max and min during calibration?  You should be able to put 1000 here and have LabVIEW figure out the values for anything past 859 up to 1000.  I'm not entirely clear on this, but it sounds like you are trying to enter 1000 into the max range after setting 859 as the calibration max (two different maximums).  Try recalibrating and increasing that max value there.

Thanks,

Justin M.

National Instruments

[Team21]

Justin.  I wasnt able to physically apply more than the 859 newtons to the load cellduring calibration, because my lever system i rigged up couldnt handle it. 

Seeing as its a linear scale though, i might just try and get a best-fit line from the calibration data relating N to V and put that in under the scaling, that way it can calculate higher values.

Thanks for the help.