Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

accuracy of C-module 9203

Solved!
Go to solution

The NI page for the 9203 C-module states an accuracy of 0.049 mA on a range of 0-20 mA.  (or app. 800 in its -20 +20 mA mode). Is this a typo or is this module really not very accurate?

The operating manual does not mention this accuracy at all, but instead mentions factors such as offset and peak to peak input noise and "INL" on a scale of LSB's.

------------------------------------------------------------------------------------
Seriously concerned about the Labview subscription model
0 Kudos
Message 1 of 3
(3,272 Views)
Solution
Accepted by topic author aartjan
Hi aartjan,

The operating manual you linked gives the accuracy specifications in the tables on pages 12-13 as a Percent of Reading (gain error) value and a Percent of Range (offset error) value. Using the values given in the first table for unipolar range, you will see that the calibrated max values for the device will be ±0.18% of reading and ±0.06% of range, with the range being 21.5 mA. This corresponds to an accuracy of 0.0018*20mA+0.0006*21.5mA=0.049mA in the worst case. It lists typical calibrated values as ±0.04% of reading and ±0.02% of range in unipolar mode. Looking at the calibrated max values from the table for bipolar mode would yield an accuracy of 0.002*20mA+0.0009*43mA=0.079mA. Hope this helps,
Daniel S.
National Instruments
Message 2 of 3
(3,264 Views)

A very clear explanation. Thanks!

It would mean the following: I would attach absolute pressure sensors that yield 4-20 mA on a range of 800-1200 mbar, or 25 mbar per mA. The worst possible reading error would be 0.049*25= 1.225 mbar. And, if I understand Gain error correctly, this would be only at full range. Furthermore, I may expect the real gain and offset error to be more according the "typical" error percentages, since I expect no extreme ambient temperatures during the measurement.

------------------------------------------------------------------------------------
Seriously concerned about the Labview subscription model
0 Kudos
Message 3 of 3
(3,259 Views)