APPENDIX A. MEASUREMENT ACCURACY FOR THERMOCOUPLE
61
A.1
Measurement accuracy: Pm
The measurement accuracy Pm depends on the voltage caliber of the appliance (see the previous paragraph)
and on the slope of the thermocouple. You shall take the slope at 0
◦
C while knowing that it will vary as a
function of the temperature, but this variation is generally of the second order for the accuracy calculation.
A.2 Voltage of thermocouple (mV)
Slope of the thermocouples (
µ
V)
Example:
• At 100
◦
C the slope of thermocouple J is 55
µ
V
• At -100
◦
C slope is 30
µ
V .
• So the error is 2 time greater then 100
◦
C to -100
◦
C.
A.2.1
Example Of Accuracy Calculation
You make a welding between -50
◦
C and +50
◦
C with a J thermocouple with compensation of cold welding.
P t
=
P l
+
P s
+
P m
(A.1)
Linearization accuracy
P l
= +0
.
25
◦
C
(A.2)
Compensation of cold welding
P s
= +0
.
5
◦
C
(A.3)
Used caliber
10
mV
(see previous example)
(A.4)
Accuracy for voltage measurement
0
.
1
%
∗
10
mV
+ 10
µV
= 20
µV
(A.5)
Slope of J thermocouple
50
µV
/
◦
C
(A.6)
Accuracy
P m
P m
= 20/50 = 0
.
4
◦
C
(A.7)
Total accuracy
P t
= 0
.
25 + 0
.
5 + 0
.
4 = 1
.
15
◦
C
(A.8)
A.3
Accuracy class - class index
This is one of the most important concepts of the CEI recommendation; it tends to shorten the list of
specifications. To do so, it introduces the concept of PREDICTIVE CLASS that depends on the C CLASS
INDEX.
The normalized values of the class index are: C = 0.1, 0.25, 0.5 and 1.
The intrinsic error (in the reference conditions) does not C% (the manufacturer may also specify this
limit of the intrinsic error as an absolute value (ex. + 5
µ
V) for the first calibers).
The variations (of the measured value) with the variations of one of the influential variables in the nominal
range of use do not exceed:
• C % for the position for the magnetic induction with external source and parasite voltages
• 0.5 C% for the power supply source
• 0.3 C% according to the class index at ambient temperature (0.15% for the 0.25 class).
