NI 9219 quater bridge accuracy

[Ivo_D]

Hello all,

I have a real problem in understandig, how the accuracy of the quarter-bridge-measurement with the NI 9219 ist calculated.

I'm using a 350 ohm strain gauge für strain measurement.

The absolute accuracy is usually given by:

Absolute Accuracy = (Gain Error * Reading) + (Offset Error * Range) + Noise + Drift

The problem is, that in the NI 9219 Datasheet no range is given. The selected range in the NI MAX is only used to improve the resolution of the ADC, or am I wrong? Further more, the quater bridge is sourced by a not specified current. In this thread is explained, that the NI 9219 uses a constant voltage excitation of 150 mV (Datasheet page 13) to source this current.

Can someone explain me, how I calculate with these informations the accuracy of my strain measurement?

Best ragards,

Ivo

[AndreasGareis]

Hello Ivo,

what do you mean that the input range is not given? It's listed in the datasheet in table 1 on page 8 depending on measurement type.

Best regards

[Ivo_D]

Hello Andreas,

thank you for your reply. Yes, for the most measurement types is a range given. But for the quarter bridge measurement are only the nominal resistances of the strain gauge listed (120 ohm, 350 ohm). For the strain measurement, only the relative change of the strain gauge resistance is of interest. And for this change, no range is given, which I could use for my accuracy calculation.

First I calculated this way: accuracy=measured strain * gain error + signal input range (MI MAX) * offset error . But I doubt this is the right way.

All in all: I dont know how to calculate the absolute accuracy for the quarter bridge measurement. Do you have a hint, where I can finde informations about this?

Best regards

Ivo

[Ivo_D]

Hello all,

I have spend some more time in thinking about my problem and I have conceived a new approach for calculating the absolute accuracy for the quarter bridge measurement.

My problem was, that in the equotation:

Absolute Accuracy = (Gain Error * Reading) + (Offset Error * Range) + Noise + Drift

the offset error of 2400 ppm for 350 ohm results in a offeset error of 0.85 ohm. For a strain gauge with GF=2.05 and 0.004 strain this would be a error off approximately 30%.

But now I realised, that LabVIEW/NI MAX does a Null-offset calibration, when I adjuste my "bridge" for the strain measurement. So my thought is, that I can simplify my equation to:

Absolute Accuracy = (Gain Error * Reading)  + Noise + Drift

because of the compensated offset with the initial voltage.

Am I right with my thought?

Best regards

Ivo