How do I set strain gauge resistance on 9237

[CharlieM]

Paulo,

The resistor in the 9944 is a dummy 120 Ohm resistor that will allow you to easily complete a quarter bridge connection. You will still need to connect your strain gauge to the terminals of the 9944. I found this link helpful.

[Jimmy P]

I have been looking at some of these posts about how to calibrate the 9237 for the offset and shunt. It looks like you say that it is best to :

1.      Place a precision 100 KOhm resistor in parallel with my gauge by coding “Shunt Cal Enable = True” on the FPGA

2.      Take an analog reading

3.      Remove the precision 100 KOhm resistor in parallel by coding “Shunt Cal Enable = False” on the FPGA

4.      Take an analog reading

5.      Compare the two readings and ensure that the strain values returned indicate a resistance change of having your strain gauge of some resistance shunted by 100 KOhms.

Looking at my attachment is this the right idea that I have? I am unsure of how to take an analog reading of the voltage. Can you explain this and is this the correct set up to do this in the FPGA and then go about calibrating the module? Thanks for the help.

Jimmy

[CharlieM]

Jimmy,

What you have in that image will not result in a proper calibration. After the initial LSB, Offset and Data Rate steps, as seen in the Getting started example, you need three separate analog readings, each with their own start and stop. You will need one normal reading. You will need to turn on Shunt Cal, take a reading and then turn off Shunt Cal. You will need to turn on Offset Cal, take a reading and then turn off Offset Cal. The Offset and Shunt Cals make physical changes in the bridge. It is up to you to take reading and make the appropriate calculations.

[Jimmy P]

Thanks for the reply, Charlie. I think I may understand the idea behind the calibration, but I am unsure of the set-up. First, when setting up this calibration procedure what I first need is a test material with a strain gage and an applied strain. Is this correct? I would think that this is my normal reading. Second, when you say analog reading do you mean just measure the voltage and then i would need to convert this to strain from the strain formula? I will use a 9944 quarter bridge completion. How would I take the "normal" reading without hooking up the 9944? It looks like the shunt cal. and offset cal. only deal with placing the resistor in parallel with the gage. So I guess I am unceratin about the specific steps to take. Would you or someone else who is reading this post be able to explain in fine detail what needs to be done?

Thank you.

[CharlieM]

1.      Shunt Cal and Offset Cal are calibrations for your final set-up and wiring. Thus, you need your strain gauge glued to your test material and wired to the 9237/9944. .

2.      All your readings, measurements, analog readings… (whatever you want to call them) are to be taken with your test material in a relaxed, unstressed state. These readings from the 9237 are a ratio of channel voltage to excitation voltage. The transition from a binary reading on the FPGA VI to a useful value is shown in the 9237 getting started example.

3.      You should have the 9944 hooked up at all times.

4.      The details of what is physically done by enabling Shunt Cal and Offset Cal is described in the manual. You can use these features to develop your own custom calibration procedures.

[Jimmy P]

Thanks for the help, Charile.

I have been able to zero the offset and then I am using a 100K ohm resistor to do the shunt calibration. I am doing this calibration for shunt and null by hand as I look at averages from an excel graph of the data. The way I am doing this is just adding values to the voltage value after my convert from binary. I have the null working fine, but how would I adjust the shunt voltage so that it is exactly what I should be getting? The equation I am using is from http://digital.ni.com/public.nsf/allkb/9D5648C81141022B862561AA004B7383?OpenDocument . You mention doing it in the FPGA and I would assume that would take care of everything, but is there an easy way to do this manually? If I adjust the value for the shunt, it changes my zero null calibration. What would you recommend?

Jimmy P.

[rex1030]

I have been struggling with the offset calibration as well. I have been trying to find out from the app engineers how it works but I keep getting generic, useless, answers. There is no mention of how it works in the manual, which I am beginning to expect with this module. I have been looking into how some of our DAQmx modules handle offset calibration, but assuming thats how this module does it hasn't helped much.

How, exactly, does the offset calibration work? Step by step, what does it do to the circuit and how does it change subsequent strain values? Please find out for sure before you reply. Thanks Charlie!

By the way, i really appreciate your continuing help with our calibration problems on the 9237 because I have had app engineer after app engineer mearly repeat what the manual says then give up (with my support threads and the threads of my coworkers). So, thanks charlie.

[CharlieM]

This example program shows how the Shunt Cal Enable and Offset Cal Enable I/O Properties of the NI 9237 can be used to calibrate a system. The system that this example calibrates for is a quarter bridge configured strain gauge system at rest.

The calibration performed in this example takes three unique readings. The first is a normal analog reading of the system. The second reading is done with the Shunt Cal Enable I/O Property set to true. The last reading is done with the Offset Call Enable I/O Property set to true.

The difference in the first two readings simulates a strain change by shunting one of the arms in the bridge with a precision 100 KOhm resistor. That change is read as a strain and compared to a theoretical calculation based on knowledge of the resistance of the strain gauges used. This test will indicate the effect of the resistance of the wires used to complete the bridge. The longer the external wires to the strain gauge, the greater the resistance. Such an increase in resistance can cause inaccuracies in strain readings. A normal % error should be less than 1%.

The difference in the first and last readings indicates how unbalanced the bridge is. The last reading, with the Offset Cal Enable I/O Property node set to true, returns a reading of the bridge perfectly balanced. It does this by internally disconnecting the external leads and shorting out the analog input channel. If the first reading of the unstrained system has a reading near the limit of the module, the system may move outside of the module’s range when it is loaded. The NI 9237 has a range of -25 mV/V to 25 mV/V.

This example requires LabVIEW 8.5, LabVIEW RT 8.5, LabVIEW FPGA  8.5 as well as the NI RIO 2.3 or 2.3.1 Driver.

[Jimmy P]

Charlie,

Thanks a lot for the reply and your explanation. Would it be possible to send an example of what you attached with LabVIEW 8.2? This is the latest version of LabVIEW that I have.

Jimmy P.

[CharlieM]

Jimmy,

This has not been tested and will probably have some broken arrows that you will need to clean up. But, it should give you the basic idea.