# Accuracy of SCXI 32-Channel Analog or Thermocouple Input Module Reading

Updated Feb 22, 2018

• SCXI-1100
• SCXI-1102
• SCXI-1104
• SCXI-1112
• SCXI-1102B
• SCXI-1102C
• SCXI-1104C

## Issue Details

I am taking a thermocouple reading using SCXI and would like to know what accuracy to expect. What is the possible error that may occur?

## Solution

Absolute accuracy is the specification that will be used to determine the overall maximum possible error of your measurement. This voltage error can then be converted to temperature error by using a standard conversion table for a particular thermocouple type. Absolute accuracy assumes that the SCXI signal conditioning equipment has been calibrated within the last year.  This document will address the expected accuracy of the following SCXI modules:
• NI SCXI-1100
• NI SCXI-1102/B/C
• NI SCXI-1104/C
• NI SCXI-1112

Several factors contribute to the absolute accuracy:
Percent of Reading - Percentage of the actual input voltage often called "gain error".
Offset - Constant value applied to all measurements.
System Noise - Depends on the number of points averaged for each measurement.
Temperature Drift - Based on factors including variations in your ambient temperature.

Based on these components, the formula for calculating absolute accuracy for a given module is:
Absolute Accuracy = ±[(Input Voltage x % of Reading) + Offset + System Noise + Temperature Drift]
Temperature Drift = ±[(Input Voltage x % of Reading/°C)+Offset/°C] x Temperature Difference
All values used in calculating the Absolute Accuracy and the Temperature Drift can be found in a specifications table within the National Instruments catalog or can be found online in the data sheet for the specific hardware. Note that temperature drift effects for these modules are already taken into account unless your ambient temperature is outside of 15-35 °C range

Example Calculation:
The following is an example of calculating Absolute Accuracy for an SCXI-1102 system measuring a type J thermocouple at 100 °C. First we will assume that the ambient temperature falls within the 15-35 °C range, then we will calculate the accuracy accounting for temperature drift.

A type J thermocouple at 100 °C generates 5.268 mV (this value is from a standard type J conversion table or formula). We will assume in our temperature task, we are using a range of ±100 mV or applying a gain of 100. The remaining necessary components to Absolute Accuracy can be found in the SCXI-1102/B/C User Manual.

 Variable Value Value Used in Equation Input Voltage 5.268 mV 0.005268 V Percent of Reading MAX (gain error) 0.02% 0.0002 Offset 15 µV 0.000015 V System Noise (Avg) 5 µV 0.000005 V

Absolute Accuracy = ±[(Input Voltage x % of Reading) + Offset + System Noise + Temperature Drift]
Absolute Accuracy = ±[(0.005268 x 0.0002) + 0.000015 + 0.000005] V = ±21.05 µV
A voltage of ±21.05 µV translates to about ± 0.5°C using a standard formula for J type thermocouples. Thus our range would be 99.5-100.5 °C.

If the ambient temperature may fluctuate outside of the 15-35 °C range, we then need to take into account temperature drift. If we take the same example above in an ambient temperature of 45 °C we would calculate the Absolute Accuracy as follows:

 Variable Value Value Used in Equation Input Voltage 5.268 mV 0.005268 V Temperature Difference 45-35 °C 10 °C Percent of Reading/°C 0.00002 (1/°C) 0.00002 (1/°C) Offset/°C 1.5 µV 0.0000015 (V/°C)

Temperature Drift = ±[(Input Voltage x % of Reading/°C)+Offset/°C] x Temperature Difference
Temperature Drift
= ±[(0.005268 x 0.00002)+ 0.0000015](V/°C) x 10°C = ±16.05 µV
Absolute Accuracy = ±(21.05 + 16.05)µV = ±37.10 µV
A voltage of ±37.10 µV translates to about ±0.8 °C using a standard formula for J type thermocouples. Thus our range would be 99.2-100.8 °C.