load cell measurement

[SeanJ]

My colleague has selected a Tedea Huntleigh Model 615 load cell for a project.  It is a 300kg load cell.

We wish to measure the force exerted on the load cell.  I am considering purchasing a USB 6009.  It has 14 bit analog inputs, so 300kg/16384 =  18g measurement resolution.

We will be powering the load cell with a 12V supply.  I intend to use the setting as shown below on the DAQ Task on a 6008.

 

The rated output is 2mV/V.  With a 12V supply, the maximum output will be 24mV, meaning that each step will be 24mV/ 16384 = 1.46uV.  Is the USB 6009 capable of measuring voltage in this step size?

Apart from the superior resolution, are there any other benefits of using the NI 9237?

Thanks,

Seán

[Henrik_Volkers]

The 600X are not intended to be used for bridge sensors without additional preconditioning... so better use a 9237.

Don't mix up resolution and accuracy!   And for resolution calculations the input range has to be used ...  so the 6009 has +-1V minimum range  divided by 14 bit resolution to +-24mV output for +-300kg... even worse, the spec tell me a typical 1.53mV accuracy for the +-1V range !!    And the input impedance...    ah... useless!!

And read the spec about the exitation power ... external supply ...

[SeanJ]

Hi Henrik,

Many thanks for your reply.

Looking through the specification, output impedance of Model 615 is 350 ohms.  Analog input impedance of 6009 is 144k ohms.  There is an obvious mismatch so we will not get maximum power transfer of the signal and experience reflections.  

However, it doesn't state whether this matters to the 6009, given most cable is 50 or 75 ohm impedance is this an issue for the 6009?  Perhaps it has an isolator on board that dampens reflections?  

Our resoltution is now (2000/24) * (300,000g/(10^14)) = 1524g, (please correct me if wrong).  For our intents and purpose, this may be ok, we need to know approximately the right value, ±1kg is ok.  

I guess the question is whether the accuracy of 1.53mV is consistent, i.e. can we calibrate it out?

To be honest, the more I am writing, the more I am beginning to feel the 9237 is the best option!  

Ciao,

Seán

[SeanJ]

Please dismiss my comments about impedance matching!

If only th edit button worked, the higher the impednace the less it impacts the measurement system.  Still in the case of a load cell where it will have to be calibrated does it matter.  It seems to be a tickier area than I had anticipated, possibly should have dodged the bullet!