Not getting accurate readings off a calibrated load cell

[Eric Martell]

I'm using a Cooper Instruments 200 lb load cell connected to a SCB-68 connector box hooked up to a 6036e DAQ card, and I'm having trouble calibrating it.  (I've done this before, but it's been a while, and something just isn't working right.)  So, I'm using the calibration wizard (the channels are set as Custom Voltage with Excitation), hanging a bunch of known weights from the cell, entering the values, and all that is going fine.  However, when I go to use the load cells (in a very simple VI that just reads the value from the cell), it's reading consistently low by 2-3 pounds.  I'm using a 10 V excitation signal, and I set the low end of the Signal Input Range to 0 (it kept wanting to set -5=-5 in the calibration when the low end was -5V, and that was even worse).  I've tried modifying a custom scale, and that isn't fixing the problem either.    Has anyone encountered this before or have any ideas on what else we can try?

Thanks,

Eric Martell

[mallorim]

Hi Eric,

A couple of questions that will help to troubleshoot this issue:

1) When you just run the task in Measurement and Automation Explorer (MAX) do you see the right values? Or is the calibration off there too?

2) When you are in the Calibration Wizard and committing values, are the numbers jumping around or do they stabilize? What settings did you select for averaging?

3) Are you using any type of signal conditioning?

Thanks for the information. Regards, Mallori M.

[Eric Martell]

Thanks for responding.

To answer your questions:

1) We didn't calibrate the cell in MAX.  When we just look at the data coming from the cell, it's pretty noisy, and MAX reflects that.  However, when we use the calibration in LabVIEW, it's a lot less jumpy, just inaccurate.

2) In the Calibration Wizard, the numbers seem pretty stable (out to 4 decimal places).  I'm not quite sure what you mean by averaging within the Wizard.

3) We're not using any signal conditioning.  We actually own a signal conditioner, but it's being used for research right now, and we're trying to use our backup hardware for a presentation at a conference.

Thanks,

Eric Martell

[mallorim]

Hi Eric,

This is a curious problem. Is it every point that has a constant offset or does the offset vary from point to point?

Also what is the model number and sensitivity of your load cell? Have you gotten it to read correctly before? You mentioned that the values were stable out to the fourth decimal place, but haw significant is this decimal place? For instance, when I open the calibration wizard, the values can jump around depending on what I am measuring. If you load cell has a sensitivity of 1 mV/lb, then this could be large variations depending on what the decimal point was when you saved the value.

With regards to my mentioning averaging, when you first click the calibrate button in the DAQ Assistant, the first screen has a Number of Samples to Average control and a Rate control. If you leave the # of samples to average at 1000, but change the rate to 10,000, do the values begin to vary?

Regards, Mallori M.