Fluke 1595A - Manual (Page 20)

1594A/1595A Super-Thermometer
Specifications
10
When calibrating an SPRT on the ITS-90, the W
T90
value is measured at required fixed-point temperatures.
This is done by measuring the resistance of the SPRT at a fixed-point temperature followed immediately by
measurement of the RTPW. This is repeated for each temperature point. Once again the ratio accuracy of the
Super-Thermometer, applied to each measured resistance, determines the accuracy of the resulting W
T90
value.
2.2.5.3 Example 1: Measuring an SPRT
This section explains how to calculate the uncertainty of a temperature measurement when measuring a
calibrated 25.5 W SPRT at 157 °C, using the internal 25 W reference resistor in a 1595A. Since the Super-
Thermometer measurement accuracy is directly related to other sources of uncertainty, additional uncertainties
will be included in the calculation for completeness.
Since an SPRT can be measured with different techniques, two different calculations will be presented to rep-
resent the most common and recommended techniques.
2.2.5.3.1 Measuring With Updated RTPW
In this example, the RTPW of the SPRT is measured by a 1595A and entered into the SPRT’s probe definition
in the 1595A. Then the SPRT is measured at 157 °C, in temperature mode, using the coefficients entered in the
SPRT probe definition.
As explained above, this measurement technique primarily uses the ratio accuracy of the Super-Thermometer.
It is equivalent to measuring the ITS-90 W
T90
value and using it to calculate temperature.
The total uncertainty of this measurement is based on six uncorrelated uncertainties. These uncertainties are:
● ●
Resistance ratio accuracy of the 1595A at 157 °C
● ●
Measurement noise at 157 °C
● ●
Resistance ratio accuracy of the 1595A at 0.01 °C (triple-point of water)
● ●
Measurement noise at 0.01 °C
● ●
Reference resistor drift
● ●
Uncertainty of the triple-point of water cell
The following demonstrate how to calculate and combine the listed uncertainties.
2.2.5.3.2 Resistance Ratio Accuracy at 157 °C
The resistance of the SPRT, when measuring at 157 °C, is 41.1 Ω. The ratio of this resistance against the 25
W reference resistor is 1.644. From the resistance ratio accuracy specifications of the 1595A, the standard
uncertainty (k = 1) when measuring a resistance ratio of 1.644 is 0.08 ppm. This is converted to temperature
by dividing 0.08 ppm by 1.0 × 10 6 and then multiplying by 1.644. The result is then divided by W
T90
sensitiv-
ity (dW/dT) at 157 °C which is 0.0038 (found on the SPRT calibration report, see tip below). The final result is
0.000035 °C.
2.2.5.3.3 Measurement Noise at 157 °C
Random error due to measurement noise must be included as an uncertainty. During measurement at 157 °C, it
is observed that the standard error of the mean is 0.000040 °C.
Note: The user must monitor measurement noise and use the actual measured measurement noise in the
uncertainty calculations.
2.2.5.3.4 Resistance Ratio Accuracy at 0.01 °C
Uncertainties related to measuring the RTPW of the SPRT must be included in the analysis. However, RTPW
uncertainties are magnified when applied to uncertainties of temperatures that are above 0 °C. At 157 °C this
magnification is estimated by multiplying RTPW uncertainties by the W
T90
value at 157 °C (1.612 in this
example).
The resistance of the SPRT at the triple-point of water is approximately 25.5 Ω. The resistance ratio against the
25 W reference resistor is then about 1.02. From the resistance ratio accuracy specifications of the 1595A, the
standard uncertainty due to linearity while measuring a resistance ratio of 1.02 is 0.03 ppm. This specification