Texas Instruments | A Basic to Thermocouple Measurements | Application notes | Texas Instruments A Basic to Thermocouple Measurements Application notes

Texas Instruments A Basic to Thermocouple Measurements Application notes
Application Report
SBAA274 – September 2018
A Basic Guide to Thermocouple Measurements
Joseph Wu
ABSTRACT
Thermocouples are common temperature sensors used in a wide variety of commercial and industrial
applications. While slightly less accurate than resistance temperature detectors (RTDs), thermocouples
cover a wide temperature range, are self-powered, and have a fast response time. Their simple
construction make them inexpensive and durable. Because of the small sensor voltage and low noise
requirements, delta-sigma analog-to-digital converters (ADCs) are ideal data converters for measuring
thermocouples. This application report gives an overview of thermocouples, discussing theory of
operation, functionality, and methods in temperature measurement. Many circuits are presented showing
thermocouple connections to precision ADCs. Different topologies focus on biasing thermocouples for the
ADC input and for burn-out measurements.
1
2
3
Contents
Thermocouple Overview .................................................................................................... 3
Thermocouple Measurement Circuits ................................................................................... 14
Summary .................................................................................................................... 36
1
Thermocouple Voltage ...................................................................................................... 3
2
Thermocouple Responses .................................................................................................. 4
3
Thermocouple Construction Types ........................................................................................ 6
4
Type-K IEC-EN 60584-2 Tolerance Class Errors ....................................................................... 8
5
Resistor Biasing of a Thermocouple ..................................................................................... 10
6
Voltage Biasing of a Thermocouple
List of Figures
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
.....................................................................................
Thermocouple and Cold-Junction Measurement Conversion to Temperature .....................................
Comparison of Interpolation Errors Using Various Lookup Tables ..................................................
Burn-out Detection Using Resistor Biasing .............................................................................
Burn-out Detection Using BOCS .........................................................................................
Thermocouple Measurement Circuit With Pullup and Pulldown Resistors .........................................
Thermocouple Measurement Circuit With Biasing Resistors Attached to the Negative Lead ...................
Thermocouple Measurement Circuit Using VBIAS For Sensor Biasing and Pullup Resistor ....................
Thermocouple Measurement Circuit With VBIAS for Sensor Biasing and BOCS .................................
Thermocouple Measurement Circuit With REFOUT Biasing and Pullup Resistor .................................
Thermocouple Measurement Circuit With REFOUT Biasing and BOCS ...........................................
Thermocouple Measurement Circuit With Bipolar Supplies and Ground Biasing .................................
Thermocouple Measurement Circuit With Two-Wire RTD Cold-Junction Compensation ........................
Thermocouple Measurement Circuit With Thermistor Cold-Junction Compensation .............................
Thermistor and Linearization Responses Over Temperature ........................................................
Linearization of Thermistor With Parallel Resistor and Voltage Divider ............................................
Linearized Output of Thermistor Circuit .................................................................................
Thermocouple Measurement Circuit With Temperature Sensor Cold-Junction Compensation .................
10
11
12
13
13
15
17
19
21
23
25
27
29
31
32
32
33
34
List of Tables
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
1
www.ti.com
1
Common Thermocouple Types ............................................................................................ 4
2
Characteristics of ITS-90 Thermocouple Direct Polynomials to Determine Voltage from Temperature
3
ITS-90 Temperature Coefficients for a K-Type Thermocouple ........................................................ 5
4
Thermocouple Tolerance Class Information ............................................................................. 7
5
Conversion From Voltage to Temperature for the LMT70 ............................................................ 35
.........
5
Trademarks
All trademarks are the property of their respective owners.
2
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Overview
www.ti.com
1
Thermocouple Overview
Thermocouples are temperature measurement sensors that generate a voltage that changes over
temperature. Thermocouples are constructed from two wire leads made from different metals. The wire
leads are welded together to create a junction. As the temperature changes from the junction to the ends
of the wire leads, a voltage develops across the junction.
Combinations of different metals create a variety of voltage responses. This leads to different types of
thermocouples used for different temperature ranges and accuracies. Choosing a thermocouple often is a
function of the measurement temperature range required in the application. Other considerations include
the temperature accuracy, durability, conditions of use, and the expected service life.
1.1
Seebeck Voltage
In 1820, Thomas Johann Seebeck discovered that when a metal bar is heated on one end, a voltage
(known as the Seebeck voltage) develops across the length of the bar. This voltage varies with
temperature and is different depending on the type of metal used in the bar. By joining dissimilar metals
that have different Seebeck voltages at a temperature sensing junction, a thermocouple voltage (VTC) is
generated.
The dissimilar metals are joined at a temperature sensing junction (TTC) to create a thermocouple. The
voltage is measured at a reference temperature (TCJ) through the two metals. The leads of the
thermocouple are required to be at the same temperature and are often connected to the ADC through an
isothermal block. Figure 1 shows a thermocouple constructed from two dissimilar metals with the
thermocouple leads connected to an isothermal block.
Metal A
+
VTC
Thermocouple
TTC
í
Metal B
TCJ
Isothemal
Cold-Junction
Block
Figure 1. Thermocouple Voltage
The connection of the thermocouple to an isothermal block is important for the temperature measurement.
For an accurate thermocouple measurement, the return leads of different metals must be at the same
known temperature.
Any connection between two different metals creates a thermocouple junction. Connections from the
thermocouple to the ADC should be simple and symmetric to avoid unintentional thermocouple junctions.
These additional junctions cause measurement errors.
As the thermocouple signal connects to the ADC integrated circuit, each step along the path can
encounter several additional thermocouples. This becomes a measurement problem if there is a
temperature gradient across the circuit. Each connection from wire terminal, to solder, to copper trace, to
IC pin, to bond wire, to chip contact creates a new junction. However, if the signal is differential, and each
of the thermocouple pairs are at the same temperature, then the thermocouple voltages cancel and have
no net effect on the measurement. For high-precision applications, the user must ensure that these
assumptions are correct. Measurement with differential inputs include unintentional thermocouple voltages
that do not cancel if the thermocouples are not located close together, or if there is a thermal gradient on
the board or device.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
3
Thermocouple Overview
1.2
www.ti.com
Thermocouple Types
1.2.1
Common Thermocouple Metals
All dissimilar metals used to construct a thermocouple display a change in voltage from the Seebeck
effect, but several specific combinations are used to make thermocouples. The thermocouples can be
classified into two different construction types: base metal thermocouples and noble metal thermocouples.
Base metal thermocouples are the most common thermocouples. Noble metal thermocouples are
composed of precious metals such as platinum and rhodium. Noble metal thermocouples are more
expensive, and are used in higher temperature applications.
Regardless of metal lead, each thermocouple type is designated a single letter to indicate the two metals
used. For example, a J-type thermocouple is constructed from iron and constantan. With each type, the
thermoelectric properties are standardized so that temperature measurements are repeatable.
Thermocouple leads and connectors are standardized with color plugs and jacks, indicating the type of
thermocouple. Different colors for insulation and lead wires also indicate the thermocouple grade and
extension grade. Table 1 lists several common thermocouple types and their characteristics.
Table 1. Common Thermocouple Types
Thermocouple
Type
Lead Metal
A (+)
Lead Metal
B (–)
Temperature
Range (°C)
EMF over
Temperature
Range (mV)
Seebeck
Coefficient
(µV/°C at 0°C)
J
Iron
Constantan
–210 to 1200
–8.095 to 69.553
50.37
K
Chromel
Alumel
–270 to 1370
–6.458 to 54.886
39.48
T
Copper
Constantan
–200 to 400
–6.258 to 20.872
38.74
E
Chromel
Constantan
–270 to 1000
–9.385 to 76.373
58.70
S
Platinum and
10% Rhodium
Platinum
–50 to 1768
–0.236 to 18.693
10.19
1.2.2
Thermocouple Measurement Sensitivity
The National Institute of Standards and Technology (NIST) has analyzed the output voltage versus
temperature for the various types of thermocouples. Figure 2 illustrates the typical responses for these
same thermocouple types.
Thermocouple Voltage (mV)
80
Type J
Type K
Type T
Type E
Type S
60
40
20
0
-20
-400
-200
0
200
400
600
800
Temperature (qC)
1000
1200
1400
1600
1800
Figure 2. Thermocouple Responses
Several polynomial equations are defined by the International Temperature Scale of 1990 (ITS-90)
standard that correlate the temperature and voltage output. This data is found on the NIST website at
http://srdata.nist.gov/its90/main/. These equations are used to calculate the thermoelectric voltage from
temperature or to calculate temperature from the thermoelectric voltage
4
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Overview
www.ti.com
1.2.2.1
Calculating Thermoelectric Voltage from Temperature
Direct polynomials construct the equations to calculate the thermoelectric voltage from a known
temperature. These equations have a form shown in Equation 1.
n
E=
6 c (t
i
i
90)
i=0
where
•
E is in microVolts and t90 is in degrees Celsius
(1)
Table 2 summarizes the polynomial orders and the respective temperature ranges for the types of
thermocouples.
Table 2. Characteristics of ITS-90 Thermocouple Direct Polynomials to Determine Voltage from
Temperature
(1)
1.2.2.2
Thermocouple Type
Temperature Range (°C) for Polynomials
Polynomial Order (1)
J
–210 to 760, 760 to 1200
8th, 5th
K
–270 to 0, 0 to 1370
10th, 9th, + a eb(t – c)^2
T
–200 to 0, 0 to 400
7th, 6th
E
–270 to 0, 0 to 1000
13th, 10th
S
–50 to 1064.18, 1064.18 to 1664.5, 1664.5 to 1768.1
8th, 4th, 4th
For type K thermocouples above 0 °C, there is an additional term to account for a magnetic ordering effect
Calculating Temperature From Thermoelectric Voltage
Making the reverse conversion, Inverse polynomial functions calculate the temperature based on the
thermocouple voltage. The equations for inverse polynomial functions are of the form shown in
Equation 2.
t90 = d0 + d1E + d2E2 + … + diEi
where
•
E is in microVolts and t90 is in degrees Celsius
(2)
As an example, the inverse function for a K-type thermocouple is shown in Table 3. Polynomials are
constructed over three smaller ranges of the full temperature range. For each range, the temperature is
described with a high order polynomial.
Table 3. ITS-90 Temperature Coefficients for a K-Type Thermocouple
Temperature Range:
−200°C to 0°C
0°C to 500°C
500°C to 1372°C
Voltage Range
−5891 μV to 0 μV
0 μV to 20644 μV
20644 μV to 54886 μV
d0
d1
d2
d3
d4
d5
d6
d7
d8
d9
0.000 000 0
2.517 346 2 x 10–2
–1.166 287 8 x 10–6
–1.083 363 8 x 10–9
–8.977 354 0 x 10–13
–3.734 237 7 x 10–16
–8.663 264 3 x 10–20
–1.045 059 8 x 10–23
–5.192 057 7 x 10–29
0.000 000 0
508 355 x 10–2
7.860 106 x 10–8
–2.503 131 x 10–10
8.315 270 x 10–14
–1.228 034 x 10–17
9.804 036 x 10–22
–4.413 030 x 10–26
1.057 734 x 10–30
–1.052 755 x 10–35
–1.318 058 x 102
4.830 222 x 10–2
–1.646 031 x 10–6
5.464 731 x 10–11
–9.650 715 x 10–16
8.802 193 x 10–21
–3.110 810 x 10–26
Error Range
0.04°C to –0.02°C
0.04°C to –0.05°C
0.06°C to –0.05°C
Table 2 and Table 3 show the complexity of direct and inverse polynomial equations. The mathematical
operations used to calculate these high order equations without loss of precision can take a significant
amount of computational processing with high resolution, floating-point numbers. This type of computation
is generally not suited for embedded processing or microcontrollers. In many cases, it is far more efficient
to determine the temperature through interpolation using a lookup table.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
5
Thermocouple Overview
1.2.3
www.ti.com
Thermocouple Construction
Thermocouples come in several different construction types as shown in Figure 3. Thermocouple leads
are protected by a layer of insulation and often have a protective sheath at the thermocouple junction tip
to protect the sensor element.
Exposed Thermocouple
Grounded Thermocouple
Insulation
Sheath
Thermocouple
Junction
Ungrounded Thermocouple
Figure 3. Thermocouple Construction Types
A thermocouple without a protective sheath is known as an exposed thermocouple. This allows for a small
sensor, with direct heat transfer from the measured object. This type of thermocouple gives a fast sensor
response.
In a grounded thermocouple, the sensor is welded to the sheath. Often this sheath is composed of metal,
which also allows for heat transfer, but adds an extra protection for harsh and difficult environments.
However, because the thermocouple is welded to the metal sheath, there is electrical contact. This makes
the measurement susceptible to noise from ground loops.
An ungrounded thermocouple is isolated from the sheath, adding a layer of insulation between the
thermocouple the measured object. This type of thermocouple has the slowest of the temperature
responses because there is an isolation layer.
As mentioned, both grounded and exposed thermocouples have faster temperature responses because of
the excellent heat transfer of metal contact. However, with direct metal contact there is electrical contact
between the measurement circuit and anything the thermocouple contacts. This may cause ground loop
problems with the measurement.
If the ground of the circuit is at a different electrical potential than the contact from the thermocouple, then
the measurement circuit may be disrupted. As an example, a grounded or exposed thermocouple may
contact earth ground, which may not be the same as the ADC ground. This can cause a variety of
problems, including bad measurement data or even damage to the circuit. Even if the earth ground and
ADC ground are identical, the thermocouple may not be in the range of the PGA. When using an exposed
or grounded thermocouple, ensure that the thermocouple contact does not disrupt signal or measurement
integrity.
6
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Overview
www.ti.com
1.2.4
Tolerance Standards
Temperature measurement accuracy and range depend on the type of the thermocouple used and the
standard followed by the manufacturer. The International Electrotechnical Commission standard outlined
in IEC-EN 60584 contains the manufacturing tolerances for base metal and noble metal thermocouples. A
parallel standard used in the United States from the American Society for Testing and Materials is
described by ASTM E230. Table 4 shows the tolerance of different thermocouples based on different
standards and tolerance classes.
Table 4. Thermocouple Tolerance Class Information
Thermocouple
Type
Tolerance Class
IEC-EN
60584-2
J
ASTM E230
ANSI MC96.1
IEC-EN
60584-2
K
ASTM E230
ANSI MC96.1
IEC-EN
60584-2
Temperature
Range (°C)
Thermocouple Error (°C)
(Larger between two columns)
Class 1
–40 < T < 750
±1.5°C
±(0.004 · |T|)
Class 2
–40 < T < 750
±2.5°C
±(0.0075 · |T|)
Class 3
–
–
–
Special
0 < T < 750
±1.1°C
±(0.004 · |T|)
Standard
0 < T < 750
±2.2°C
±(0.0075 · |T|)
Class 1
–40 < T < 1000
±1.5°C
±(0.004 · |T|)
Class 2
–40 < T < 1200
±2.5°C
±(0.0075 · |T|)
Class 3
–200 < T < 40
±2.5°C
±(0.015 · |T|)
Special
0 < T < 1250
±1.1°C
±(0.004 · |T|)
Standard
–200 < T < 0
0 < T < 1250
±2.2°C
±2.2°C
±(0.02 · |T|)
±(0.0075 · |T|)
Class 1
–40 < T < 350
±0.5°C
±(0.004 · |T|)
Class 2
–40 < T < 350
±1.0°C
±(0.0075 · |T|)
Class 3
–200 < T < 40
±1.0°C
±(0.015 · |T|)
Special
–200 < T < 0
0 < T < 350
±0.5°C
±0.5°C
±(0.008 · |T|)
±(0.004 · |T|)
Standard
–200 < T < 0
0 < T < 350
±1.0°C
±1.0°C
±(0.015 · |T|)
±(0.0075 · |T|)
Class 1
–40 < T < 800
±1.5°C
±(0.004 · |T|)
Class 2
–40 < T < 900
±2.5°C
±(0.0075 · |T|)
Class 3
–200 < T < 40
±2.5°C
±(0.015 · |T|)
Special
–200 < T < 0
0 < T < 900
±1.0°C
±1.0°C
±(0.005 · |T|)
±(0.004 · |T|)
Standard
–200 < T < 0
0 < T < 900
±1.7°C
±1.7°C
±(0.01 · |T|)
±(0.005 · |T|)
Class 1
0 < T < 1600
±1.0°C
±[1 + 0.003 · (|T| –
1100)]
Class 2
–40 < T < 1600
±1.5°C
±(0.0025 · |T|)
Class 3
–
–
–
Special
0 < T < 1450
±0.6°C
±(0.001 · |T|)
Standard
0 < T < 1450
±1.5°C
±(0.0025 · |T|)
T
ASTM E230
ANSI MC96.1
IEC-EN
60584-2
E
ASTM E230
ANSI MC96.1
IEC-EN
60584-2
S
ASTM E230
ANSI MC96.1
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
7
Thermocouple Overview
www.ti.com
As an example, Figure 4 graphically shows the error of a type-K thermocouple with the IEC-EN 60584-2
tolerance classes. At higher temperatures, the thermocouple error becomes significantly greater.
Tolerance Error (qC)
10
5
Class 2
Class 3
Class 1
0
-5
-10
-400
-200
0
200
400
600
Temperature (qC)
800
1000
1200
1400
Figure 4. Type-K IEC-EN 60584-2 Tolerance Class Errors
Thermocouples show a wide range of error dependent on the tolerance class. However, few of these
thermocouples have error tolerances better than ±1°C. For this reason, RTDs are preferred for
applications requiring higher precision and accuracy. It common to use 16-bit ADCs for thermocouple
measurements and 24-bit ADCs for RTD measurements.
1.3
Thermocouple Measurement and Cold-Junction Compensation (CJC)
As discussed earlier, the thermocouple generates a voltage related to the temperature difference between
the thermocouple junction and the leads to attached to the cold junction at the isothermal block (see
Figure 1). However, the voltage created from the thermocouple is non-linear depending on the
temperature of the cold junction. Cold-junction compensation is required to accurately determine the
thermocouple junction temperature based on the cold junction temperature.
With cold-junction compensation, the leads of the thermocouple must be at the same known temperature.
In thermocouple measurement systems, there is a cold-junction block which connects the thermocouple
lead to the ADC measurement. This block holds both thermocouple leads at the same temperature and is
often a connector made from a large metal mass, with thermal capacitance. In some applications, it may
be sufficient to maximize the copper fill around the junctions of the PCB, layering the connection between
metal fill between top and bottom layers. Because air currents may affect the temperature, an enclosure
around the block may be necessary.
An accurate measurement of the cold junction block acts as the reference temperature of the coldjunction. This reference measurement is often made through a diode, thermistor, or RTD. If the reference
temperature at TCJ is known, then the thermocouple temperature at TTC is computed based on the
thermocouple voltage. The process of accounting for TCJ is called cold junction compensation because it is
generally assumed that TCJ is the cold temperature.
In the classical method of setting the cold-junction temperature the leads of the thermocouple are placed
in an ice bath, ensuring that the reference temperature is 0°C. However, in most systems the cold-junction
temperature is measured separately with a device such as an RTD or thermistor.
Once the reference temperature is measured, the thermocouple voltage for that temperature (relative to
0°C) can be determined and added to the measured voltage on the thermocouple leads. This
compensation gives the voltage that would have been developed if TCJ had been at 0°C. Note that this
voltage is required when referencing the NIST charts, since the chart values are specified relative to 0°C.
8
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Overview
www.ti.com
Thermocouple voltages are non-linear with temperature. Therefore you cannot simply add the temperature
of the cold-junction to the temperature computed from the thermocouple voltage. To accurately determine
the thermocouple temperature, the proper method is to:
1. Convert the cold-junction temperature (TCJ) to a voltage (VCJ)
2. Add the cold-junction voltage to the measured thermocouple voltage (VCJ + VTC)
3. Convert the summed cold-junction voltage and thermocouple voltage to the thermocouple temperature
(TTC)
Conversion tables and polynomial equations used to determine thermocouple temperature from the
thermoelectric voltage is found at the NIST website at http://srdata.nist.gov/its90/menu/menu.html.
1.4
Design Notes
The following sections describe different considerations for designing a thermocouple measurement with a
precision ADC. The discussion starts with the operational range of the ADC, setting up the circuit, making
measurement conversions, and performing cold-junction compensation. Each section describes different
considerations that help make a more accurate measurement.
1.4.1
Identify the Range of Thermocouple Operation
The thermocouple voltage is very small and requires a low-noise precision ADC for measurement.
Referring back to Figure 2, different thermocouples have different output voltage ranges. Using a K-type
thermocouple operating from –270°C to 1370°C as an example, the thermocouple voltage would range
from about –6.5 mV to 55 mV.
Because many precision ADCs have onboard programmable gain amplifiers (PGAs), this measurement
signal can be amplified for a more precise measurement. Using this thermocouple output voltage range
and the reference voltage, calculate the maximum gain allowed without over-ranging the PGA. Many
precision ADCs have an onboard PGA with gain settings in factors of 2. Many precision ADCs also have a
precision voltage reference. Voltage measurements for thermocouples require precision references with
low noise. Reference error directly impacts the measurement accuracy. The reference voltage, combined
with the PGA determine the input range of the measurement.
As an example, with a maximum input of 55 mV, the PGA gain can be set to 32. This results in an
equivalent input signal of 1.76 V. Using a 2.048-V internal reference voltage, this maximizes the ADC
input range without over-ranging the PGA.
1.4.2
Biasing the Thermocouple
After calculating the PGA gain, consider the PGA common-mode input range. Many PGAs are
implemented similar to the front end of an instrumentation amplifier. This requires that the common-mode
voltage of the input must be within the PGA range of operation. With increasing PGA gain, the commonmode input voltage is limited so that the amplifier output does not go into either the positive or negative
supply rails. Consult the device data sheet for specific absolute or common-mode input voltage ranges. In
most cases, setting the input to the mid-point of the analog supply voltages ensures the thermocouple is
within the range of the PGA.
Thermocouples require biasing to set the sensor voltage DC operating point. There are a number of ways
to bias the thermocouple. The most common method for thermocouple biasing is using large resistors tied
to either end of the thermocouple as shown in Figure 5. The opposite end of the resistors are then tied to
either supply. This method sets the thermocouple operating voltage at mid-supply assuming that the
resistors are equal, and that the thermocouple voltage is relatively small.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
9
Thermocouple Overview
www.ti.com
AVDD
AVDD
R
Thermocouple
ADC
R
Figure 5. Resistor Biasing of a Thermocouple
Resistor values are generally chosen to be from 500 kΩ to 10 MΩ depending on the input current.
Different ADCs have different magnitudes of input current. If the resistance is too high, the biasing current
becomes too small compared to the ADC input current of the resistors. Consider the ADC input current
when selecting resistor values as this may offset the bias point.
If the thermocouple leads are long, then resistor biasing may create additional error. Long resistive leads
will react with the bias current to develop a voltage error in the measurement. In another biasing method,
the negative thermocouple lead is connected to a known voltage source, as shown in Figure 6. Using a
voltage source removes the bias current passing through the thermocouple. Only the ADC input current
remains, which is usually orders of magnitude lower. In many cases, the ADC reference or an external
reference may be used for biasing. Similarly, many ADCs have a VBIAS line that can be used to attach a
specific analog input to a voltage generator through the input multiplexer of the ADC.
AVDD
Thermocouple
ADC
+
±
Figure 6. Voltage Biasing of a Thermocouple
Similarly, if the ADC uses a bipolar supply, the negative thermocouple input can be tied to ground. Using
the ground establishes the input at the mid-point of the supply and sets the bias point within the PGA input
range.
Regardless, all of these methods establish a DC operating point for the thermocouple measurement. Many
of the later sections of this application note discuss different circuit topologies for biasing the
thermocouple.
1.4.3
Thermocouple Voltage Measurement
After setting the gain and putting the thermocouple in the input range of the PGA, measure the
thermocouple voltage with the ADC. If you have a 16-bit bipolar ADC and a set PGA gain, calculate the
thermocouple measurement voltage with Equation 3. Typically, the reference voltage is the equivalent to
the positive full scale.
VTC = (Output Code • VREF) / (Gain • 215)
10
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
(3)
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Overview
www.ti.com
Start with this example as a K-type thermocouple. Also assume the system settings are VREF = 2.048 V,
PGA gain = 32, and a 16-bit bipolar ADC. If the ADC reports back a data reading of 31CFh (12751d), the
thermocouple voltage can be calculated from Equation 4.
VTC = (12751 codes • 2.048 V) / (32 • 32768 codes) = 24.904 mV
(4)
Using a conversion table, this voltage would be determined to be 600°C. However, this is the correct value
only if the cold-junction temperature is known to be 0°C. To get the actual thermocouple temperature you
need to determine the cold-junction temperature and make the conversion to the voltage.
1.4.4
Cold-Junction Compensation
There are many ways to determine the cold junction temperature. RTD measurements are often used to
get a more accurate temperature reading for the cold-junction measurement. There are also thermistors
and other semiconductor temperature sensors that can be used to get a cold junction measurement.
Regardless of how it is done, the cold-junction measurement must be accurate. Any error in the coldjunction measurement directly adds to the error in the thermocouple measurement.
Returning to the original example, assume the cold junction is measured to be 25°C. Using the K-type
thermocouple table, this is the equivalent to 1.000 mV of thermoelectric voltage. To get an accurate
temperature measurement of the thermocouple voltage, you would add the thermocouple voltage to the
equivalent cold-junction voltage.
V = VTC + VCJ = 24.904 mV + 1.000 mV = 25.904 mV
(5)
Now that the equivalent thermoelectric voltages have been added together, return to the table and find the
equivalent temperature. With some interpolation, the resulting temperature of the thermocouple is about
623.5°C. The thermocouple voltage is non-linear and depends on the cold-junction voltage.
For accuracy, cold-junction compensation requires that the junction temperature is converted to the
thermoelectric voltage for the measurement. The tables and equations start with an assumption of a 0°C
cold junction. Calculation requires a specific conversion when the cold-junction is not at that temperature.
As mentioned in the previous section, the proper method to calculate the thermocouple temperature
follows.
A simple addition between the equivalent thermocouple temperature and the cold-junction temperature
would have resulted in 625°C. This would have produced a 1.5°C error because of the thermocouple nonlinearity over temperature. The only way to compensate for the non-linearity of the thermocouple curve is
to convert the cold-junction temperature to the equivalent voltage, sum the thermocouple measurement
voltage and cold-junction equivalent voltage, and convert the result back to temperature.
As noted earlier, if the cold-junction is held at 0°C (as if held at the temperature by an ice bath), then the
equivalent cold-junction thermoelectric voltage is 0 mV. This allows for a direct conversion of the
thermocouple voltage to temperature.
1.4.5
Conversion to Temperature
Conversion of data from the ADC requires both measurements of the thermocouple voltage and the coldjunction temperature. Often measurements of each are interleaved to ensure an accurate measurement of
both. The flow diagram in Figure 7 shows the conversion method to determine the actual temperature of
the thermocouple based on the ADC measurements.
MCU
ADC
Thermocouple
Voltage
VTC
Cold-Junction
Temperature
TCJ
Convert Temperature
to Voltage
VCJ
6
V
Convert Voltage
to Temperature
T
Figure 7. Thermocouple and Cold-Junction Measurement Conversion to Temperature
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
11
Thermocouple Overview
www.ti.com
As previously mentioned the cold-junction temperature (TCJ) is first converted to a thermoelectric voltage
(VCJ). This voltage is then added to the ADC measurement of the thermocouple voltage where
V = VCJ + VTC. This voltage sum is converted to temperature to determine the temperature of the
thermocouple sensor.
Conversions from temperature to voltage or from voltage to temperature may be calculated though the
polynomial equations explained in Section 1.2.2. Coefficients may be stored in the microcontroller to make
these calculations for each ADC conversion.
An alternative to processor-intensive calculations is to use lookup tables for a simple linear interpolation of
these polynomials. Temperature and voltage ranges may be evenly broken up for making conversions.
Precision Thermocouple Measurement with the ADS1118 describes using lookup tables with different
numbers of table entries to calculate the thermocouple temperature. Figure 8 shows the conversion error
that can be expected from linear interpolation using a lookup table for a K-type thermocouple from 0°C to
+500°C. As the number of lookup table entries exceeds 16, the improvement in accuracy become smaller
and smaller.
Linear Interpolation Temperature Error (°C)
1.5
4 Lookup Table Entries
8 Lookup Table Entries
16 Lookup Table Entries
32 Lookup Table Entries
1
0.5
0
−0.5
−1
−1.5
0
50
100
150
200
250
300
Thermocouple Temperature (°C)
350
400
450
500
G000
Figure 8. Comparison of Interpolation Errors Using Various Lookup Tables
1.4.6
Burn-out Detection
If a thermocouple has failed or burned out, system designers may want to some indication that this has
occurred. A full-scale ADC reading for an opened sensor can be used to help make this determination.
This measurement may be part of a normal measurement or the system may require an interim
measurement as a periodic check.
Methods of thermocouple biasing can allow for automatic burn-out detection. As an example, when the
thermocouple is biased through resistors attached from each thermocouple lead to either supply shown in
Figure 9. The small resistor currents hold the thermocouple at a DC bias point if the thermocouple is intact
and produces a small voltage. If the thermocouple has burned out and become high impedance, the
resistors pull the voltage of each lead to either rail. This would cause the ADC input to be greater than the
full-scale range and cause the ADC to give a full-scale reading (7FFFh for a 16-bit ADC). Again, the
resistors should be small enough to overcome the input bias current of the ADC.
12
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Overview
www.ti.com
AVDD
R
AVDD
Burned out
thermocouple
ADC
R
Figure 9. Burn-out Detection Using Resistor Biasing
In different methods of thermocouple biasing, there may be no pullup resistor to enable a burn-out
measurement. Many precision ADCs include burn-out current sources (BOCS), where a current source is
used to pull up on the positive analog input and a current sink pulls down on the negative analog input as
shown in Figure 10. These current sources are used to pull apart the positive and negative leads of a
thermocouple during a burn-out condition. This is useful even if the thermocouple is biased from only one
end by a DC source, such as a reference voltage or VBIAS line. The positive burn-out current source pulls
the positive lead high enough so that the ADC reads a positive full-scale reading.
AVDD
Burned out
thermocouple
ADC
Figure 10. Burn-out Detection Using BOCS
In general, the burn-out measurement using BOCS should be a separate measurement than the normal
temperature measurement. Using the BOCS may induce error when the thermocouple is not burned out.
Extra current may lead to self-heating of the sensor. Additionally, there are often RC filters at the front end
of the ADC. Because the BOCS are sourced from the device, additional current flowing from BOCS
creates an error voltage as they pass through the series filter resistors.
When enabling the BOCS for a burn-out measurement, ensure that there is additional time for voltage
setting between measurements. Filter capacitance may require time for voltage settling in both the burnout and temperature measurement.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
13
Thermocouple Measurement Circuits
2
www.ti.com
Thermocouple Measurement Circuits
NOTE: Information in the following applications sections is not part of the TI component
specification, and TI does not warrant its accuracy or completeness. TI’s customers are
responsible for determining suitability of components for their purposes. Customers should
validate and test their design implementation to confirm system functionality.
The following sections describe thermocouple circuit topologies with delta-sigma ADCs. Because
thermocouple measurements are primarily simple voltage measurements, these circuit examples focus
mainly on different circuit topologies for biasing the thermocouple and burn-out detection. The Design
Notes section may be used to guide the design with the following system topologies. For each topology,
determine the PGA setting based on the thermocouple operating range, consider the necessary biasing
and PGA input range, and determine the cold-junction compensation. Burn-out detection is also described
with the following system topologies. Cold-junction measurements are discussed at the end of the
application note.
Conversion results are shown with a generic 16-bit bipolar ADC, using the positive full-scale range of the
device. Conversions with 24-bit ADCs are similar in calculation. Results are shown as functions of the
reference voltage and gain of the PGA. Conversion to temperature depends on the linearity and error of
the individual thermocouple sensor, and the cold-junction compensation.
As mentioned in previous sections, conversion tables to determine thermocouple temperature from the
thermoelectric voltage is found at the NIST web site at http://srdata.nist.gov/its90/menu/menu.html.
14
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
2.1
Thermocouple Measurement With Pullup and Pulldown Bias Resistors
In this topology, the thermocouple DC voltage is biased using matched pullup and pulldown resistors. This
is a common method for biasing and allows for burn-out detection.
2.1.1
Schematic
AVDD
AVDD
Mux
Thermocouple
AIN0
ADC
PGA
AIN1
Internal
Reference
AVSS
REFOUT
Reference
Mux
REFCOM
REFP0
REFN1
Figure 11. Thermocouple Measurement Circuit With Pullup and Pulldown Resistors
2.1.2
Pros and Cons
Pros:
• Simple biasing
• Biasing resistors allow for burn-out detection without separate measurement
Cons:
• Requires two external resistors for biasing
• Biasing current flows through the thermocouple and resistive leads, creating additional error
2.1.3
Design Notes
The measurement circuit requires:
• Pullup and pulldown resistors
• AINP and AINN inputs
• Internal reference or an external voltage reference
• Isothermal cold-junction connection and measurement
Figure 11 shows the most common method of thermocouple biasing. Matched resistors are attached to
either lead of the thermocouple to set the DC biasing for the input signal. A first resistor pulls the positive
lead of the thermocouple to AVDD, and a second resistor pulls the negative lead of the thermocouple to
AVSS. Because the measured thermocouple voltage is small, the bias current can be approximated as the
supply voltage divided by the two biasing resistors. If the resistors are matched, the thermocouple voltage
is biased to the mid-point of the analog supply. Setting the biasing near the mid-point of the supply
ensures that the input voltage is within the input range of the PGA. Consult the ADC data sheet for
specific PGA common-mode and absolute input ranges.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
15
Thermocouple Measurement Circuits
www.ti.com
Resistor values are large to reduce the amount of current passing through the thermocouple and the
thermocouple leads. Bias current reacting with long resistive leads create an additional voltage which is
measured by the ADC as an error voltage. However, the bias current must be large enough so that the
resistor current is significantly larger than the input current of the ADC. If the bias current is small or close
to the level of the ADC input current, the DC bias of the thermocouple is offset from the mid-point of the
supply. Biasing resistor values are typically from 500 kΩ to 10 MΩ.
An additional error in the measurement comes from the input current of the ADC. An extra voltage error is
seen as the ADC input current reacts with the series input filter resistors and any series resistance
associated with the input multiplexer. Because this current cannot be removed, it is important to select an
ADC with a low input current and calculate the contribution of this error to the measurement.
The biasing resistors are also used for burn-out measurement. In the case of a burned out thermocouple,
the positive input is pulled to AVDD while the negative input is pulled to AVSS. This creates a large
voltage across the analog inputs, over-ranging the ADC. If the ADC is over-ranged, the ADC value would
read 7FFFh (assuming a 16-bit bipolar ADC), showing a full-scale reading to indicate a burned out
thermocouple. To ensure that the ADC reports a full-scale reading, verify that the biasing resistors are low
enough so that they can pull against the input biasing current of ADC and yield a voltage larger than the
input full-scale voltage. Because the biasing resistors are always in place, a separate burn-out
measurement is not needed.
Unless the cold junction is at 0°C, there should be a separate cold-junction measurement. This
measurement can be done through several different methods, using either an RTD, calibrated thermistor,
or a variety of integrated circuit temperature sensors.
2.1.4
Measurement Conversion
Using the ADC internal voltage reference or external voltage reference, the output code is converted to
the measured thermocouple voltage. An output code of 7FFFh may indicate a open sensor.
VTC = (VREF • Code) / (215 • Gain)
(6)
Measure the cold-junction temperature and convert the temperature to the equivalent cold-junction
thermoelectric voltage. Add the thermocouple voltage to the equivalent cold-junction voltage.
V = VTC + VCJ
(7)
Convert the resulting voltage (V) to temperature determine the exact thermocouple temperature.
2.1.5
Generic Register Settings
•
•
•
•
•
16
Enable the internal reference or use an external reference, set ADC reference
Select multiplexer settings for AINP and AINN to measure the leads of the thermocouple
Enable the PGA, set gain to desired value
Select data rate and digital filter settings
Settings for cold-junction compensation measurement
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
2.2
Thermocouple Measurement With Biasing Resistors Attached to the Negative Lead
Another topology for biasing the thermocouple would be to attach the biasing resistors to a single end of
the thermocouple. This removes any biasing current through long resistive thermocouple leads, reducing
any added voltage error, but requires a separate burn-out sensor measurement.
2.2.1
Schematic
AVDD
AVDD
Mux
Thermocouple
AIN0
ADC
PGA
AIN1
Internal
Reference
AVSS
REFOUT
Reference
Mux
REFCOM
REFP0
REFN1
Figure 12. Thermocouple Measurement Circuit With Biasing Resistors Attached to the Negative Lead
2.2.2
Pros and Cons
Pros:
• Removes the voltage error from biasing current with long resistive thermocouple leads
Cons:
• Requires enabling of burn-out current sources and separate measurement for burn-out detection
2.2.3
Design Notes
The measurement circuit requires:
• Biasing resistors attached to the negative lead of the thermocouple
• AINP and AINN inputs
• Internal reference or an external voltage reference
• Burn-out current sources for a separate sensor burn-out measurement
• Isothermal cold-junction connection and measurement
Similar to the first design, this topology uses matched resistors to set the biasing for the thermocouple. In
this example, matched resistors are used to bias only the negative end of the thermocouple. As long as
the current through the resistors is significantly larger than the ADC input current, the resistor biasing
places the negative input near mid-supply. Setting the biasing near mid-supply ensures that the input
voltage is in the range of the PGA.
Attaching the bias resistors to only the negative thermocouple lead eliminates the bias current flowing
through the thermocouple. This removes the error that comes from the bias current reacting with the
resistive leads of the thermocouple.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
17
Thermocouple Measurement Circuits
www.ti.com
Removing the bias current flowing through resistive thermocouple leads does not eliminate all errors. As in
the previous topology, there is still error as the ADC input current reacts with the series input filter
resistors and any series resistance associated with the input multiplexer of the ADC. Because this current
cannot be removed, it is important to select an ADC with a low input current and calculate the contribution
of this error to the measurement.
In the previous design, biasing resistors pulled apart the inputs in the case of a burned out thermocouple.
In this design, one lead is still set to mid-supply, while the second lead is left unconnected. Because there
us no current to pull up on the positive thermocouple lead, burn-out detection requires a second
measurement with a change in setup for the ADC. To detect a burned out or open thermocouple, the
burn-out current sources in the ADC are enabled for a separate burn-out current measurement. The burnout current sources should not remain on for the normal measurement. These current sources, reacting
with the series input filtering resistors and series resistance in the multiplexer add a large additional error.
Burn-out current sources may be set to various levels, depending on the ADC being used. Verify that the
burn-out current level is high enough so that an open input creates a full-scale reading (7FFFh, assuming
a 16-bit bipolar ADC) for burn-out detection.
Unless the cold junction is at 0°C, there should be a separate cold-junction measurement. This
measurement can be done through several different methods, using either an RTD, calibrated thermistor,
or a variety of integrated circuit temperature sensors.
2.2.4
Measurement Conversion
Using the ADC internal voltage reference or external voltage reference, the output code is converted to
the measured thermocouple voltage.
VTC = (VREF • Code) / (215 • Gain)
(8)
Measure the cold-junction temperature and convert the temperature to the equivalent cold-junction
thermoelectric voltage. Add the thermocouple voltage to the equivalent cold-junction voltage.
V = VTC + VCJ
(9)
Convert the resulting voltage (V) to temperature determine the exact thermocouple temperature.
Burn-out detection requires BOCS to be enabled and a separate measurement. An output code of 7FFFh
may indicate a open sensor.
2.2.5
Generic Register Settings
•
•
•
•
•
•
18
Enable the internal reference or use an external reference, set ADC reference
Select multiplexer settings for AINP and AINN to measure the leads of the thermocouple
Enable the PGA, set gain to desired value
Select data rate and digital filter settings
Enable burn-out current sources for a separate burn-out measurement (optional)
Settings for cold-junction compensation measurement
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
2.3
Thermocouple Measurement With VBIAS for Sensor Biasing and Pullup Resistor
Another topology for biasing the thermocouple requires enabling the VBIAS generator in the multiplexer of
the ADC. The VBIAS is attached to the negative lead of the thermocouple, setting the thermocouple to a
mid-supply voltage. A pullup resistor is attached from the positive lead of the thermocouple to AVDD. This
pulls the positive input away from VBIAS during a burn-out condition, yielding a positive full-scale ADC
reading.
2.3.1
Schematic
AVDD
AVDD
Mux
Thermocouple
AIN0
ADC
PGA
AIN1
VBIAS
Internal
Reference
AIN2
AVSS
REFOUT
Reference
Mux
REFCOM
REFP0
REFN1
Figure 13. Thermocouple Measurement Circuit Using VBIAS For Sensor Biasing and Pullup Resistor
2.3.2
Pros and Cons
Pros:
• Uses VBIAS to set up the sensor DC voltage
• Pullup resistor to AVDD allows for burn-out detection without separate measurement
Cons:
• Requires an extra resistor as a pullup for burn-out detection
• Biasing current flows through the thermocouple and resistive leads, creating additional error
• Requires an extra input multiplexer line for connection to VBIAS
2.3.3
Design Notes
The measurement circuit requires:
• A single pullup resistor attached to the positive lead of the thermocouple
• Enabled VBIAS voltage attached to the negative lead of the thermocouple
• AINP and AINN inputs, and an AINx connection for the VBIAS connection
• Internal reference or an external voltage reference
• Isothermal cold-junction connection and measurement
In many precision ADCs, a bias voltage generator provides a DC input voltage for unbiased sensors such
as thermocouples. This VBIAS voltage may be connected to the sensor through the multiplexer to the
ADC input pins. For most devices, this VBIAS may be set to a voltage of (AVDD – AVSS) / 2. This
provides a mid-supply voltage used to set the sensor bias in the middle of the input range of the PGA.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
19
Thermocouple Measurement Circuits
www.ti.com
A single pullup resistor may be attached to the positive thermocouple lead for burn-out detection. In the
case of a burned-out thermocouple, negative lead is still set to mid-supply, while positive lead is pulled up
to AVDD. As in the previous designs, the pullup resistor is generally large to keep the bias current low.
Any bias current reacting with the lead resistance of the thermocouple becomes an error in the
measurement. However, the biasing current must be large enough to overcome the ADC input current. If a
burn-out condition exists, the pullup resistor must be able pull the positive analog input high enough above
VBIAS to give an ADC full-scale reading (7FFFh, assuming a 16-bit bipolar ADC).
As in the previous topologies, the biasing resistor must be high to keep the bias current low. Bias current
reacting with the resistive leads of the thermocouple is measured as an error voltage. Also, the ADC input
current reacts with the series input filter resistance and multiplexer resistance to add another
measurement error.
While it is possible to connect VBIAS directly to the measurement negative input (AIN1 through the ADC
multiplexer), that particular configuration may not yield precise results. The biasing current flows from the
pullup resistor, through the thermocouple, into the input, and finally is sunk into the VBIAS connection.
The bias current reacting with the series filter resistor (and any series resistance in the input multiplexer)
causes a significant error in the measurement. In the configuration shown in Figure 13, the VBIAS drives
the thermocouple lead from an external pin, allowing the bias current to bypass the input filter resistance.
Unless the cold junction is at 0°C, there should be a separate cold-junction measurement. This
measurement can be done through several different methods, using either an RTD, calibrated thermistor,
or a variety of integrated circuit temperature sensors.
2.3.4
Measurement Conversion
Using the ADC internal voltage reference or external voltage reference, the output code is converted to
the measured thermocouple voltage. An output code of 7FFFh may indicate a open sensor.
VTC = (VREF • Code) / (215 • Gain)
(10)
Measure the cold-junction temperature and convert the temperature to the equivalent cold-junction
thermoelectric voltage. Add the thermocouple voltage to the equivalent cold-junction voltage.
V = VTC + VCJ
(11)
Convert the resulting voltage (V) to temperature determine the exact thermocouple temperature.
2.3.5
Generic Register Settings
•
•
•
•
•
•
20
Enable the internal reference or use an external reference, set ADC reference
Select multiplexer settings for AINP and AINN to measure the leads of the thermocouple
Enable VBIAS on a separate analog input pin attach to the negative lead of the thermocouple
Enable the PGA, set gain to desired value
Select data rate and digital filter settings
Settings for cold-junction compensation measurement
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
2.4
Thermocouple Measurement With VBIAS For Sensor Biasing and BOCS
Similar to the circuit in Section 2.3, this design uses the VBIAS for the sensor biasing. However, external
resistors are not used for the burn-out measurement. A separate burn-out measurement is made after
enabling the burn-out current sources in the ADC. Without the external biasing resistor, there is no
additional voltage error from the biasing current passing through the thermocouple, filter resistors, and any
resistance in the ADC multiplexer.
2.4.1
Schematic
AVDD
Mux
Thermocouple
AIN0
VBIAS
ADC
PGA
AIN1
Internal
Reference
AVSS
REFOUT
Reference
Mux
REFCOM
REFP0
REFN1
Figure 14. Thermocouple Measurement Circuit With VBIAS for Sensor Biasing and BOCS
2.4.2
Pros and Cons
Pros:
• Uses VBIAS to set up the sensor DC voltage
• Does not require an external VBIAS connection through the ADC multiplexer
• Does not require external components for biasing or burn-out measurements
Cons:
• Requires enabling of burn-out current sources and separate measurement for burn-out detection
2.4.3
Design Notes
The measurement circuit requires:
• Enabled VBIAS voltage attached to the negative lead of the thermocouple
• AINP and AINN inputs
• Internal reference or an external voltage reference
• Burn-out current sources for a separate sensor burn-out measurement
• Isothermal cold-junction connection and measurement
As in the previous design VBIAS provides a DC input voltage for unbiased sensors. This VBIAS voltage
may be connected to the thermocouple negative input through the multiplexer and is typically set to a
voltage of (AVDD - AVSS) / 2. As mentioned previously, there is no additional voltage error from the
biasing current passing through the thermocouple, and any series input resistance. However, there may
be some small error from the ADC input current reacting with the same elements. Consult the device data
sheet for information about ADC input current.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
21
Thermocouple Measurement Circuits
www.ti.com
Because there is no current to pull up on the positive thermocouple lead, burn-out detection requires a
second measurement with a change in setup for the ADC. To detect a burned out or open thermocouple,
the burn-out current sources in the ADC are enabled for a separate burn-out current measurement. The
burnout current sources should not remain on for the normal measurement. These current sources,
reacting with the series input filtering resistors and series resistance in the multiplexer add a large
additional error.
Burn-out current sources may be set to various levels, depending on the ADC being used. Verify that the
burn-out current level is high enough so that an open input creates a full-scale reading (7FFFh, assuming
a 16-bit bipolar ADC) for burn-out detection.
Unless the cold junction is at 0°C, there should be a separate cold-junction measurement. This
measurement can be done through several different methods, using either an RTD, calibrated thermistor,
or a variety of integrated circuit temperature sensors.
2.4.4
Measurement Conversion
Using the ADC internal voltage reference or external voltage reference, the output code is converted to
the measured thermocouple voltage.
VTC = (VREF • Code) / (215 • Gain)
(12)
Measure the cold-junction temperature and convert the temperature to the equivalent cold-junction
thermoelectric voltage. Add the thermocouple voltage to the equivalent cold-junction voltage.
V = VTC + VCJ
(13)
Convert the resulting voltage (V) to temperature determine the exact thermocouple temperature.
Burn-out detection requires BOCS to be enabled and a separate measurement. An output code of 7FFFh
may indicate a open sensor.
2.4.5
Generic Register Settings
•
•
•
•
•
•
•
22
Enable the internal reference or use an external reference, set ADC reference
Select multiplexer settings for AINP and AINN to measure the leads of the thermocouple
Enable VBIAS and attach to the negative lead of the thermocouple
Enable the PGA, set gain to desired value
Select data rate and digital filter settings
Enable burn-out current sources for a separate burn-out measurement (optional)
Settings for cold-junction compensation measurement
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
2.5
Thermocouple Measurement With REFOUT Biasing and Pullup Resistor
Similar to the Section 2.3 circuit, this design uses the internal reference to bias the thermocouple instead
of the VBIAS connection. While this voltage may not be exactly at mid-supply, it should be close enough
to set the sensor common-mode voltage to within the input range of the PGA.
2.5.1
Schematic
AVDD
AVDD
Mux
Thermocouple
AIN0
ADC
PGA
AIN1
Internal
Reference
AVSS
REFOUT
Reference
Mux
REFCOM
REFP0
REFN1
Figure 15. Thermocouple Measurement Circuit With REFOUT Biasing and Pullup Resistor
2.5.2
Pros and Cons
Pros:
• Uses the internal reference to set up the sensor DC voltage
• Pullup resistor to AVDD allows for burn-out detection without separate measurement
Cons:
• Requires an extra resistor as a pullup for measuring a burn-out
• Biasing current flows through the thermocouple and resistive leads, creating additional error
2.5.3
Design Notes
The measurement circuit requires:
• A single pullup resistor attached to the positive lead of the thermocouple
• Enabled internal reference voltage attached from the reference output pin (REFOUT) to the negative
lead of the thermocouple
• AINP and AINN inputs
• Isothermal cold-junction connection and measurement
Another feature of many precision ADCs is an internal reference. The internal reference is often used as
only the ADC reference. However, if the reference is buffered and brought out of the device to a pin, it can
be used to bias a thermocouple. While this reference voltage may not be at exactly the mid-supply
voltage, it is likely in the input range of the PGA. Consult the ADC data sheet for specific PGA commonmode and absolute input ranges.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
23
Thermocouple Measurement Circuits
www.ti.com
First, enable the internal reference voltage. The REFOUT line is then attached to the thermocouple
negative input, while a resistor is used to pull up the thermocouple positive input to AVDD. As in the
similar design using VBIAS, the large pullup resistor is used for burn-out detection. If the thermocouple
has burned out and become high impedance, the ADC over-ranges and gives a full-scale reading.
A single pullup resistor may be attached to the positive thermocouple lead for burn-out detection. In the
case of a burned-out thermocouple, the negative lead is still set to mid-supply, while the positive lead is
pulled up to AVDD. As in the previous designs, the pullup resistor is generally large to keep the bias
current low. Any bias current reacting with the lead resistance of the thermocouple becomes an error in
the measurement. However, the biasing current must be large enough to overcome the ADC input current.
If a burn-out condition exists, the pullup resistor must be able pull the positive analog input high enough
above VBIAS to give an ADC full-scale reading (7FFFh, assuming a 16-bit bipolar ADC).
As in the previous topologies, the biasing resistor must be high to keep the bias current low. Bias current
reacting with the resistive leads of the thermocouple is measured as an error voltage. Also, the ADC input
current reacts with the series input filter resistance and multiplexer resistance to add another
measurement error.
Unless the cold junction is at 0°C, there should be a separate cold-junction measurement. This
measurement can be done through several different methods, using either an RTD, calibrated thermistor,
or a variety of integrated circuit temperature sensors.
2.5.4
Measurement Conversion
Using the ADC internal voltage reference or external voltage reference, the output code is converted to
the measured thermocouple voltage. An output code of 7FFFh may indicate a open sensor.
VTC = (VREF • Code) / (215 • Gain)
(14)
Measure the cold-junction temperature and convert the temperature to the equivalent cold-junction
thermoelectric voltage. Add the thermocouple voltage to the equivalent cold-junction voltage.
V = VTC + VCJ
(15)
Convert the resulting voltage (V) to temperature determine the exact thermocouple temperature.
2.5.5
Generic Register Settings
•
•
•
•
•
24
Enable the internal reference, set as ADC reference
Select multiplexer settings for AINP and AINN to measure the leads of the thermocouple
Enable the PGA, set gain to desired value
Select data rate and digital filter settings
Settings for cold-junction compensation measurement
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
2.6
Thermocouple Measurement With REFOUT Biasing and BOCS
Similar to the previous design, the thermocouple is biased with an external connection to the internal
reference from the ADC. However, the burn-out detection is made through a second measurement with
the burn-out current sources enabled.
2.6.1
Schematic
AVDD
Mux
Thermocouple
AIN0
ADC
PGA
AIN1
Internal
Reference
AVSS
REFOUT
Reference
Mux
REFCOM
REFP0
REFN1
Figure 16. Thermocouple Measurement Circuit With REFOUT Biasing and BOCS
2.6.2
Pros and Cons
Pros:
• Uses the internal reference to set up the sensor DC voltage
• Does not require extra components for burn-out measurement
Cons:
• Requires enabling of burn-out current sources and separate measurement for burn-out detection
2.6.3
Design Notes
The measurement circuit requires:
• Enabled internal reference voltage attached from the reference output pin (REFOUT) to the negative
lead of the thermocouple
• AINP and AINN inputs
• Burn-out current sources for a separate sensor burn-out measurement
• Isothermal cold-junction connection and measurement
As in the previous topology, the REFOUT line is attached to the thermocouple negative input for sensor
biasing. However, instead of using a resistor connected from the positive lead to AVDD, the burn-out
current sources are enabled only during a burn-out sensor measurement.
There is still error as the ADC input current reacts with the series input filter resistors and any series
resistance associated with the input multiplexer of the ADC. Because this current cannot be removed, it is
important to select an ADC with a low input current and calculate the contribution of this error to the
measurement.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
25
Thermocouple Measurement Circuits
www.ti.com
To detect a burned out or open thermocouple, the burn-out current sources in the ADC are enabled for a
separate burn-out current measurement. The burnout current sources should not remain on for the normal
measurement. These current sources, reacting with the series input filtering resistors and series
resistance in the multiplexer add a large additional error.
Burn-out current sources may be set to various levels, depending on the ADC being used. Verify that the
burn-out current level is high enough so that an open input creates a full-scale reading (7FFFh, assuming
a 16-bit bipolar ADC) for burn-out detection.
Unless the cold junction is at 0°C, there should be a separate cold-junction measurement. This
measurement can be done through several different methods, using either an RTD, calibrated thermistor,
or a variety of integrated circuit temperature sensors.
2.6.4
Measurement Conversion
Using the ADC internal voltage reference or external voltage reference, the output code is converted to
the measured thermocouple voltage.
VTC = (VREF • Code) / (215 • Gain)
(16)
Measure the cold-junction temperature and convert the temperature to the equivalent cold-junction
thermoelectric voltage. Add the thermocouple voltage to the equivalent cold-junction voltage.
V = VTC + VCJ
(17)
Convert the resulting voltage (V) to temperature determine the exact thermocouple temperature.
Burn-out detection requires BOCS to be enabled and a separate measurement. An output code of 7FFFh
may indicate a open sensor.
2.6.5
Generic Register Settings
•
•
•
•
•
•
26
Enable the internal reference, set ADC reference
Select multiplexer settings for AINP and AINN to measure the leads of the thermocouple
Enable the PGA, set gain to desired value
Select data rate and digital filter settings
Enable burn-out current sources for separate burn-out measurements (optional)
Settings for cold-junction compensation measurement
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
2.7
Thermocouple Measurement With Bipolar Supplies And Ground Biasing
Similar to biasing the thermocouple with an external voltage source, biasing can be done by connecting
the negative lead of the thermocouple to the ground, while using bipolar supplies for the ADC.
2.7.1
Schematic
+2.5 V
AVDD
Mux
Thermocouple
AIN0
ADC
PGA
AIN1
Internal
Reference
AVSS
REFOUT
Reference
Mux
REFCOM
REFP0
REFN1
-2.5 V
Figure 17. Thermocouple Measurement Circuit With Bipolar Supplies and Ground Biasing
2.7.2
Pros and Cons
Pros:
• Uses the ground line to set up the sensor DC voltage
• Does not require extra components for burn-out measurement
Cons:
• Requires bipolar analog supplies
• Requires enabling of burn-out current sources and separate measurement for burn-out detection
2.7.3
Design Notes
The measurement circuit requires:
• Bipolar supplies, with a ground node connected to the negative lead of the thermocouple
• AINP and AINN inputs
• Internal reference or an external voltage reference
• Burn-out current sources for a separate sensor burn-out measurement
• Isothermal cold-junction connection and measurement
In this topology, the negative lead of the thermocouple is connected to ground. However, the supplies are
bipolar supplies, which puts ground inherently at mid-supply. This is very similar to using VBIAS to bias
the thermocouple, because the ground is inherently mid-supply. Because the input is set to mid-supply,
the input range is within the range of the PGA.
Without the pullup resistors, there is no error created from the bias current flowing through the resistive
leads of the thermocouple. However, there is still error as the ADC input current reacts with the series
input filter resistors and any series resistance associated with the input multiplexer of the ADC. Because
this current cannot be removed, it is important to select an ADC with a low input current and calculate the
contribution of this error to the measurement.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
27
Thermocouple Measurement Circuits
www.ti.com
In this design, one lead is still set to mid-supply, while the second lead is left unconnected. Because there
no current to pull up on the positive thermocouple lead, burn-out detection requires a second
measurement with a change in setup for the ADC. To detect a burned out or open thermocouple, the
burn-out current sources in the ADC are enabled for a separate burn-out current measurement. The
burnout current sources should not remain on for the normal measurement. These current sources,
reacting with the series input filtering resistors and series resistance in the multiplexer add a large
additional error.
Burn-out current sources may be set to various levels, depending on the ADC being used. Verify that the
burn-out current level is high enough so that an open input creates a full-scale reading (7FFFh, assuming
a 16-bit bipolar ADC) for burn-out detection.
Unless the cold junction is at 0°C, there should be a separate cold-junction measurement. This
measurement can be done through several different methods, using either an RTD, calibrated thermistor,
or a variety of integrated circuit temperature sensors.
2.7.4
Measurement Conversion
Using the ADC internal voltage reference or external voltage reference, the output code is converted to
the measured thermocouple voltage.
VTC = (VREF • Code) / (215 • Gain)
(18)
Measure the cold-junction temperature and convert the temperature to the equivalent cold-junction
thermoelectric voltage. Add the thermocouple voltage to the equivalent cold-junction voltage.
V = VTC + VCJ
(19)
Convert the resulting voltage (V) to temperature determine the exact thermocouple temperature.
Burn-out detection requires BOCS to be enabled and a separate measurement. An output code of 7FFFh
may indicate a open sensor.
2.7.5
Generic Register Settings
•
•
•
•
•
•
28
Enable the internal reference or use an external reference, set ADC reference
Select multiplexer settings for AINP and AINN to measure the leads of the thermocouple
Enable the PGA, set gain to desired value
Select data rate and digital filter settings
Enable burn-out current sources for burn-out measurements (optional)
Settings for cold-junction compensation measurement
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
2.8
Cold-Junction Compensation Circuits
In the previous thermocouple circuits, cold-junction compensation was not discussed. The following
sections show several examples of cold-junction temperature measurements using other input channels
for the ADC. Several different temperature sensors are shown in different circuit topologies.
Regardless of the temperature sensor being used, ensure that the cold-junction measurement accurately
measures the temperature of the isothermal block connecting the leads of the thermocouple.
2.8.1
RTD Cold-Junction Compensation
As presented in other application notes, the RTD temperature measurement can be used for cold-junction
compensation. There are several different configurations for the RTD, but the one presented in Figure 18
is for a two-wire RTD. In most cases, RTD measurements potentially have the best accuracy.
2.8.1.1
Schematic
AVDD
AVDD
Thermocouple
AIN0
IDAC1
AIN1
ADC
PGA
AIN2
AIN3
IDAC2
2-wire
RTD
AIN4
Internal
Reference
Reference
Mux
Mux
AVSS
REFOUT
REFCOM
REFP0
REFN1
RREF
Figure 18. Thermocouple Measurement Circuit With Two-Wire RTD Cold-Junction Compensation
2.8.1.1.1
Design Notes
The RTD is a temperature sensor that changes resistance over temperature. There are several different
types of RTD construction, but the resistance for any given temperature is well characterized. RTDs are
often used to make precision temperature measurements. Figure 18 shows a 2-wire RTD circuit topology
for making a temperature measurement used for cold-junction compensation.
The measurement circuit requires:
• Single dedicated IDAC output pin
• AINP and AINN inputs
• External reference input
• Precision reference resistor
An IDAC current source drives both the RTD and the reference resistor, RREF. Because the same current
drives both elements, the ADC measurement is a ratiometric measurement. Calculation for the RTD
resistance does not require a conversion to a voltage, but does require a precision reference resistor with
high accuracy and low drift.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
29
Thermocouple Measurement Circuits
www.ti.com
With IDAC1, the ADC measures the voltage across the RTD using the voltage across RREF as the
reference. This provides an output code that is proportional to the ratio of the RTD voltage and the
reference voltage as shown in Equation 20. Ratiometric measurements only produce positive output data,
assuming zero offset error. For a fully-differential measurement, this is only the positive half of the fullscale range of the ADC, reducing the measurement resolution by one bit. The following equations assume
a 16-bit bipolar ADC, with ±VREF as the full-scale range of the ADC.
Output code = 215 • VRTD / VREF = 215 • IIDAC1 • RRTD / (IIDAC1 • RREF)
(20)
The currents cancel so that the equation reduces to Equation 21.
Output code = 215 • RRTD / RREF
(21)
In the end, the RTD resistance can be represented from the code as a function of the reference
resistance.
RRTD = Output code • RREF / 215
(22)
The measurement depends on the resistive value of the RTD and the reference resistor RREF, but not on
the IDAC1 current value. Therefore, the absolute accuracy and temperature drift of the excitation current
does not matter. In a ratiometric measurement, as long as there is no current leakage from IDAC1 outside
of this circuit, the measurement depends only on RRTD and RREF. ADC conversions do not need to be
translated to voltage. Assuming the ADC has a low gain error, RREF is often the largest source of error.
The reference resistor must be a high accuracy precision resistor with low drift. Any error in the reference
resistance becomes a gain error in the measurement.
There are many different types of RTDs and several different construction forms. For more detailed
information about RTD measurement, see A Basic Guide to RTD Measurements.
2.8.1.1.2
Measurement Conversion
Output Code = 215 • Gain • VRTD / VREF = 215 • Gain • (IIDAC1 • RRTD) / (IIDAC1 • RREF) = 215 • Gain • RRTD / RREF
RRTD = RREF • Output Code / (215 • Gain)
2.8.1.1.3
•
•
•
•
•
•
30
(23)
(24)
Generic Register Settings
Select multiplexer settings for AINP and AINN to measure the RTD
Enable the PGA, set gain to desired value
Select data rate and digital filter settings
Select reference input to measure RREF for ratiometric measurement
Enable the internal reference (the IDAC requires an enabled internal reference)
Set IDAC magnitude and select IDAC1 output pin to drive the RTD
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
2.8.2
Thermistor Cold-Junction Compensation
The thermistor is another temperature measurement element often for the cold-junction compensation. In
general, thermistors have a more limited range of temperature measurement and have a response that is
non-linear. Linearization techniques are often used to get more accurate readings over an even more
limited temperature range.
2.8.2.1
Schematic
AVDD
Thermocouple
AVDD
AIN0
AIN1
ADC
PGA
AIN2
R1
AIN3
Internal
Reference
Reference
Mux
Mux
Thermistor
R2
AVSS
REFOUT
REFCOM
REFP0
REFN1
Figure 19. Thermocouple Measurement Circuit With Thermistor Cold-Junction Compensation
2.8.2.2
Design Notes
Similar to the RTD, thermistors are sensors that have a resistance that vary with temperature. The
thermistor may be a PTC type (positive temperature coefficient) or an NTC type (negative temperature
coefficient). The resistance varies significantly with temperature and are far more non-linear than the RTD,
but are used for a more limited temperature range. Figure 20 shows an NTC thermistor measurement
used for cold-junction compensation. This example thermistor has a resistance of 5 kΩ at 25°C. Two
resistors are added to the circuit for linearizing the measurement at a cold-junction temperature near room
temperature.
The measurement circuit requires:
• AINP and AINN inputs
• Enabled internal reference for the ADC and driving the thermistor circuit
• Precision resistors for thermistor linearization circuit
For the topology shown in Figure 19, the thermistor circuit is driven by the ADC internal reference. R2 is
added in parallel with the thermistor resistance to give a more linear response near room temperature.
Figure 20 shows a plot of the resistance versus temperature for the linearization.
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
31
Thermocouple Measurement Circuits
www.ti.com
20000
RT
R2
RT || R2
Resistance (:)
15000
10000
5000
0
-20
-10
0
10
20
30
Temperature (qC)
40
50
60
70
80
Figure 20. Thermistor and Linearization Responses Over Temperature
The NTC thermistor has a resistance (RT) that is non-linear over temperature. At low temperatures, there
is a large change in the resistance for a small change in temperature. At higher temperatures, there is a
small change in resistance for a large change in temperature. As mentioned earlier, the thermistor has a
resistance of 5 kΩ at a temperature of 25°C.
R2 has a resistance of 10 kΩ that is constant over temperature. By adding R2 in parallel, the resulting
resistance is linear for a smaller range of operation. For this measurement it is acceptable because the
cold-junction temperature is at a moderate value compared to the thermocouple measurement
temperature.
Adding R1 as a voltage divider to the parallel combination of R2 and the thermistor resistance, measuring
R1 gives a positive temperature coefficient to the thermistor measurement (measuring across the
thermistor and R2 results in a negative temperature coefficient). Thermistor linearization is shown in
Figure 21.
REFOUT
+
Positive Temperature
Coefficient Measurement
R1
+
Thermistor
R2
Negative Temperature
Coefficient Measurement
Figure 21. Linearization of Thermistor With Parallel Resistor and Voltage Divider
It is likely that this measurement does not require PGA amplification. If the PGA is enabled, ensure that
this measurement over temperature is within the absolute and common-mode input ranges of the PGA.
Note that many PGAs are not be able to measure the ground node. For this example, the ADC measured
the positive temperature coefficient voltage across R1. If the negative temperature coefficient
measurement is required, then R1 is placed at the bottom connected to the ground node and REFOUT
drives the parallel combination of the thermistor and R2.
Choosing R1 to be a resistance of 5 kΩ, the output voltage (measured across R1) be 1.024 V at a
temperature of 25°C. This assumes that the reference voltage is 2.048 V. The output voltage measured
from the thermistor circuit is shown in Figure 22.
32
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
1.6
Output Voltage (V)
1.4
1.2
1
0.8
0.6
-20
-10
0
10
20
30
Temperature (qC)
40
50
60
70
80
Figure 22. Linearized Output of Thermistor Circuit
The temperature response may be adjusted so that the result is more linear at different temperatures.
Adjusting R1 can set the best linearity for a higher or lower temperature for easy calculation with the best
sensitivity of the cold-junction temperature. In this example, the measurement linearity is best from 40°C
to 50°C, to center the non-linearity closer to room temperature, raise the value of R1.
2.8.2.3
Measurement Conversion
The cold-junction compensation starts with a voltage measurement across R1 as shown in Equation 25
VR1 = (VREF • Code) / (215)
(25)
The measured voltage may be compared against the plot shown in Figure 22. Using the calculated result,
the temperature of the cold-junction is determined.
2.8.2.4
•
•
•
•
Generic Register Settings
Select multiplexer settings for AINP and AINN to measure the thermistor circuit
Disable the PGA
Select data rate and digital filter settings
Enable the internal reference for use as the ADC reference
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
33
Thermocouple Measurement Circuits
2.8.3
www.ti.com
Temperature Sensor Cold-Junction Compensation
Another option for cold-junction temperature measurement is to use a semiconductor device temperature
sensor. The following circuit shows a cold-junction measurement with an LMT70 device.
2.8.3.1
Schematic
AVDD
AVDD
Thermocouple
AIN0
AIN1
ADC
PGA
3.3 V
AIN2
LMT70
AIN3
Internal
Reference
Reference
Mux
Mux
AVSS
REFOUT
REFCOM
REFP0
REFN1
Figure 23. Thermocouple Measurement Circuit With Temperature Sensor Cold-Junction Compensation
2.8.3.2
Design Notes
Many semiconductor temperature sensors have temperature measurement accuracies below 1°C and can
be used for cold-junction compensation. Figure 23 shows a circuit that uses the LMT70 for measurement.
The measurement circuit requires:
• ADC input range that extends to ground
• AINP and AINN inputs
• Internal reference for voltage measurement
This ADC measurement is a directly measures voltage and requires a known reference voltage. The
output of the LMT70 gives an output voltage that can be used to calculate a temperature. Table 5 shows
voltages that can be used to convert the LMT70 output to temperature.
One important consideration for this measurement is the input range of the PGA. The output of the LMT70
extends to the ground node and an accurate temperature measurement of the temperature may not be
possible unless the PGA is disabled. Consult the ADC data sheet for PGA specifications and operation.
For more information about the LMT70, consult LMT70, LMT70A ±0.05°C Precision Analog Temperature
Sensor, RTD and Precision NTC Thermistor IC.
34
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
Thermocouple Measurement Circuits
www.ti.com
2.8.3.3
Measurement Conversion
The cold-junction compensation starts with a voltage measurement as shown in Equation 26.
VLMT70 = (VREF • Code) / 215
(26)
The resulting voltage is then converted to a temperature. Table 5 shows a table for temperature given the
output voltage of LMT70. Use this table be used to construct a piece-wise linear plot of temperature
versus output voltage of the device.
Table 5. Conversion From Voltage to Temperature for the LMT70
Temperature (°C)
2.8.3.4
•
•
•
•
VTAO (mV)
Local Slope (mV/°C)
MIN
TYP
MAX
–55
1373.576
1375.219
1376.862
–4.958
–50
1348.99
1350.441
1351.892
–4.976
–40
1299.27
1300.593
1301.917
–5.002
–30
1249.242
1250.398
1251.555
–5.036
–20
1198.858
1199.884
1200.91
–5.066
–10
1148.145
1149.07
1149.995
–5.108
0
1097.151
1097.987
1098.823
–5.121
10
1045.9
1046.647
1047.394
–5.134
20
994.367
995.05
995.734
–5.171
30
942.547
943.227
943.902
–5.194
40
890.5
891.178
891.857
–5.217
50
838.097
838.882
839.668
–5.241
60
785.509
786.36
787.21
–5.264
70
732.696
733.608
734.52
–5.285
80
679.672
680.654
681.636
–5.306
90
626.435
627.49
628.545
–5.327
100
572.94
574.117
575.293
–5.347
110
519.312
520.551
521.789
–5.368
120
465.41
466.76
468.11
–5.391
130
411.288
412.739
414.189
–5.43
140
356.458
358.164
359.871
–5.498
150
300.815
302.785
304.756
–5.538
Generic Register Settings
Select multiplexer settings for AINP and AINN to measure the LMT70
Disable the PGA, ensure that the ADC is able to use ground as AINN
Select data rate and digital filter settings
Enable the internal reference for use as the ADC reference
SBAA274 – September 2018
Submit Documentation Feedback
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
35
Summary
3
www.ti.com
Summary
Thermocouples are temperature sensors that are constructed from two dissimilar metals. The junction of
these metals is used as the temperature sensing element, while the remaining two leads are connected to
an isothermal block. Thermocouple measurements are made using precision ADCs, but still require
considerations for biasing, burn-out detection, and cold-junction measurement for the isothermal block.
The circuits shown in this application note are a simple guide to how thermocouple measurements are
made with precision ADCs. An overview was presented along with different thermocouple biasing
topologies and different methods used for burn-out sensing. Additional circuits for cold-junction
compensation were presented.
Topologies presented here are a sampling of different thermocouple topologies. Different methods of
thermocouple biasing and burn-out detection can be expanded and combined to create larger systems
with more channels. Alternate temperature measurement methods can be used for cold-junction
temperature measurement.
36
A Basic Guide to Thermocouple Measurements
Copyright © 2018, Texas Instruments Incorporated
SBAA274 – September 2018
Submit Documentation Feedback
IMPORTANT NOTICE AND DISCLAIMER
TI PROVIDES TECHNICAL AND RELIABILITY DATA (INCLUDING DATASHEETS), DESIGN RESOURCES (INCLUDING REFERENCE
DESIGNS), APPLICATION OR OTHER DESIGN ADVICE, WEB TOOLS, SAFETY INFORMATION, AND OTHER RESOURCES “AS IS”
AND WITH ALL FAULTS, AND DISCLAIMS ALL WARRANTIES, EXPRESS AND IMPLIED, INCLUDING WITHOUT LIMITATION ANY
IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT OF THIRD
PARTY INTELLECTUAL PROPERTY RIGHTS.
These resources are intended for skilled developers designing with TI products. You are solely responsible for (1) selecting the appropriate
TI products for your application, (2) designing, validating and testing your application, and (3) ensuring your application meets applicable
standards, and any other safety, security, or other requirements. These resources are subject to change without notice. TI grants you
permission to use these resources only for development of an application that uses the TI products described in the resource. Other
reproduction and display of these resources is prohibited. No license is granted to any other TI intellectual property right or to any third
party intellectual property right. TI disclaims responsibility for, and you will fully indemnify TI and its representatives against, any claims,
damages, costs, losses, and liabilities arising out of your use of these resources.
TI’s products are provided subject to TI’s Terms of Sale (www.ti.com/legal/termsofsale.html) or other applicable terms available either on
ti.com or provided in conjunction with such TI products. TI’s provision of these resources does not expand or otherwise alter TI’s applicable
warranties or warranty disclaimers for TI products.
Mailing Address: Texas Instruments, Post Office Box 655303, Dallas, Texas 75265
Copyright © 2018, Texas Instruments Incorporated
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertising