Industrial

Control &
Automation
Lecture 06
Field Instrumentation and
Sensors
Temperature Sensors
Spring 2019
 Field Instrumentation and
Sensors
 Temperature
 Pressure
 Level/Position
 Proximity

12-Feb-19 Industrial Control and Automation - Spring 2019 2

Temperature Sensors

 Thermo-resistive Sensors
 Resistance Temperature Detector (RTD)
 Thermistors
 Thermoelectric Sensors
 Thermoelectric Effect
 Thermocouple

12-Feb-19 Industrial Control and Automation - Spring 2019 3

Thermo-resistivity - Basics
 One of the primary methods for electrical measurement of
temperature involves changes in the electrical resistance of
certain materials.
 A measure of the material resistance indicates the temperature
of the device and the environment.
 Time response becomes very important in these cases because
the measurement must wait until the device comes into thermal
equilibrium with the environment.
 The two basic devices based on this principle are:
 Resistance-Temperature Detector (RTD) - based on the
variation of metal resistance with temperature
 Thermistor - based on the variation of semiconductor
resistance with temperature

12-Feb-19 Industrial Control and Automation - Spring 2019 4

RTD - Working Principle
Metal Resistance vs Temperature
When a thermal
energy is present in
the material, the
atoms vibrate.
The conduction
electrons tend to
collide with the
vibrating atoms, which
impedes the
movement of electrons
and absorbs some of
their energy, hence
causing resistance to
electrical current flow.

12-Feb-19 Industrial Control and Automation - Spring 2019 5

RTD - Working Principle
Metal Resistance vs Temperature
 The resistance of various
metals increases almost
linearly with temperature.
 For a specific metal of high
purity, the curve of relative
resistance versus
temperature is highly
repeatable, and thus either
tables or graphs can be
used to determine the
temperature from a
resistance measurement
using that material.

12-Feb-19 Industrial Control and Automation - Spring 2019 6

RTD - Working Principle
Metal Resistance vs Temperature

If the resistivity (𝜌) of some metal

is known as a function of
temperature, then this equation
can be used to determine the
resistance of any particular
sample of that material at the
same temperature.
12-Feb-19 Industrial Control and Automation - Spring 2019 7
RTD - Working Principle
Metal Resistance vs Temperature
In fact, curves such as these are
curves of resistivity versus
temperature because they are
proportional, for example:

The use of resistance versus

temperature graphs/tables is
practical only when high accuracy
desired. For many applications,
we can use an analytical
approximation of the curves.

12-Feb-19 Industrial Control and Automation - Spring 2019 8

Linear Approximation
An equation for a straight line
can be defined over some
specified span, that
approximates the resistance
versus temperature (R-T) curve.
Mathematically, it is given as:

12-Feb-19 Industrial Control and Automation - Spring 2019 9

Linear Approximation

The value of 𝛼0 can be found

from values of resistance and
temperature taken either
from a graph/table of resistance
vs temperature. In general,

The units of 𝛼0 are typically 1/℃ or 1/℉

12-Feb-19 Industrial Control and Automation - Spring 2019 10
Example – Linear Approximation

12-Feb-19 Industrial Control and Automation - Spring 2019 11

Quadratic Approximation
 A quadratic approximation is a more accurate representation of
the R-T curve.
 It includes both a linear term, as before, and a term that varies
as the square of the temperature.

 Values of 𝛼1 and 𝛼2 are found from tables or graphs using

values of resistance and temperature at three points.
12-Feb-19 Industrial Control and Automation - Spring 2019 12
Example – Quadratic
Approximation

12-Feb-19 Industrial Control and Automation - Spring 2019 13

RTD Properties
Signal Conditioning
 An RTD does not produce any
voltage by itself. A source of
voltage and an excitation resistor
can be used in voltage divider
configuration.
 Another frequently used circuit is a
four-resistor Wheatstone bridge.
 The bridge output voltage is
measured.
 The output voltage depends only
on the bridge resistance unbalance
– not much effect of the resistance
of the lead wires.

12-Feb-19 Industrial Control and Automation - Spring 2019 17

RTD Measurement Circuits

12-Feb-19 Industrial Control and Automation - Spring 2019 18

RTD Properties
Signal Conditioning Issue
The compensation line in the R3 leg of the bridge is required when the lead
lengths are so long that thermal gradients along the RTD leg may cause
changes in line resistance.

These changes show up as

false information. By using
the compensation line, the
same resistance changes
also appear on the R3 side
of the bridge and cause no
net shift in the bridge null.

12-Feb-19 Industrial Control and Automation - Spring 2019 19

RTD Properties
Dissipation Constant
 Because the RTD is a resistance, there is an I2R power dissipated by
the device itself that causes a slight heating effect, a self-heating.
 Thus, the current through the RTD must be kept sufficiently low and
constant to avoid erroneous reading due to self-heating.
 Typically, a dissipation constant is provided in RTD specifications that
relates the power required to raise the RTD temperature by one degree
of temperature.
 Thus, a 25-mW/°C dissipation constant (PD) shows that if I2R power
losses in the RTD equal 25 mW, the RTD will be heated by 1°C.
 The self-heating temperature rise can be found as:

12-Feb-19 Industrial Control and Automation - Spring 2019 20

Example - RTD

12-Feb-19 Industrial Control and Automation - Spring 2019 21

Example - RTD

12-Feb-19 Industrial Control and Automation - Spring 2019 22

Example - RTD

12-Feb-19 Industrial Control and Automation - Spring 2019 23

Thermistor
The thermistor represents another class of temperature
sensor that measures temperature through changes of
semiconductor resistance.

12-Feb-19 Industrial Control and Automation - Spring 2019 24

Thermistor – Working Principle
 In semiconductors, a small gap of energy ∆WG, typically in the
range 0.01 − 4 𝑒𝑉 (1𝑒𝑉 = 1.6 × 10−19 J) exists between valence
and conduction electrons.
 When the temperature of the material is increased, the
molecules begin to vibrate providing additional energy to the
valence electrons.
 When such energy equals or
exceeds the gap energy, ∆WG,
some of these electrons become
free of the molecules and enter the
conduction band to carry current
through the bulk of the material.

12-Feb-19 Industrial Control and Automation - Spring 2019 25

Thermistor – Working Principle
Contrary to metals,
 The semiconductor
becomes a better
conductor of current as its
temperature is increased,
i.e., its resistance
decreases.
 The change in
semiconductor resistance
is highly nonlinear.

12-Feb-19 Industrial Control and Automation - Spring 2019 26

Thermistor – Example

12-Feb-19 Industrial Control and Automation - Spring 2019 27

Thermistor – Example

• The real answer to the problem involves a

new design that reduces the thermistor
current so that self-heating temperature <=
0.1 °C.
• Alternatively, the calculations can be made
to cater for the self-heating effects.

12-Feb-19 Industrial Control and Automation - Spring 2019 28

Thermocouples and
Thermoelectric Effect
 The Thermoelectric effect is characterized by the different electrical
and thermal transport properties of different metals.
 Analogy: A potential difference in electrons in the hotter end of the
material have more thermal energy than those in the cooler end, and
thus tend to drift toward the cooler end.
 Similarly, if a circuit is closed by connecting the ends of two metals
through another conductor at a certain temperature, a voltage-
generating sensor is formed due to the potential difference.
 This causes an electromotive force (emf) that is proportional to
temperature almost linearly and very repeatable for constant materials.
 Devices that measure temperature on the basis of this thermoelectric
principle are called Thermocouples (TCs).

12-Feb-19 Industrial Control and Automation - Spring 2019 29

Seebeck Effect
 The most common representation of a Thermoelectric effect, called
Seebeck effect is formed by two different metals, A and B, used to close
the loop with the connecting junctions at temperatures T1 and T2.
 We could not close the loop with the same metal because the potential
differences across each leg would be the same, and thus no net emf
would be present.
 The emf produced is proportional to the difference in temperature
between the two junctions.

12-Feb-19 Industrial Control and Automation - Spring 2019 30

Peltier Effect
 An extension of the same thermoelectric properties occurs when the
reverse of the Seebeck effect is considered, called the Peltier Effect.
 In this case, we construct a closed loop of two different metals, A and
B, as before.
 Now, however, an external voltage is applied to the system to cause a
current to flow in the circuit.
 Because of the different electrothermal
transport properties of the metals, it is
found that one of the junctions will be
heated and the other cooled; that is,
the device acts as a refrigerator!
 Some practical applications of such a
device, such as cooling small
electronic parts, have been employed.

12-Feb-19 Industrial Control and Automation - Spring 2019 31

Thermocouple Measurement
 A definite relationship between measured emf of thermocouple and the
unknown temperature is established using an arrangement as shown.
 The measurement junction, TM is exposed to the environment whose
temperature is to be measured.
 This junction is formed of
metals A and B as shown.
 Two other junctions are then
formed to a common metal, C,
which then connects to the
measurement apparatus.
 The “reference” junctions are
held at a common, known
temperature TR, the reference
junction temperature.

12-Feb-19 Industrial Control and Automation - Spring 2019 32

Thermocouple Measurement
 In this arrangement, an open-circuit voltage is measured (at high
impedance) that is then a function of only the temperature difference
(TM - TR) and the type of metals A and B.
 The voltage produced has a magnitude dependent on the absolute
magnitude of the temperature difference and a polarity dependent on
which temperature is larger, reference or measurement junction.
 Thus, both magnitude and sign
of the measured voltage must
be noted.
 The reference junctions must
be held at a known, constant
temperature.

12-Feb-19 Industrial Control and Automation - Spring 2019 33

Thermocouple Types
 Certain standard configurations of thermocouples using specific metals
(or alloys of metals) have been adopted and given letter designations.
 The voltage produced by a TC is differential in the sense that it is
measured between the two metal wires, hence polarity is important.
 As noted in the footnote (a) to Table, by convention the description of a
TC identifies how the polarity is interpreted.

12-Feb-19 Industrial Control and Automation - Spring 2019 34

Thermocouple Types
 A type J thermocouple is called iron-constantan, with a 0°C reference:
 It will produce +5.27 mV for a measurement junction of 100°C, meaning
that the iron is more positive than the constantan.
 For a measurement junction of -100°C, polarity changes, and voltage
will be -4.63 mV, meaning that iron is less positive than the constantan.

12-Feb-19 Industrial Control and Automation - Spring 2019 35

Thermocouple Types
 The type J and K
thermocouples are noted
for their rather large
slope, that is, high
sensitivity.
 The type S thermocouple
is much less sensitive,
but has higher
temperature range.
 These curves are not
exactly linear.

12-Feb-19 Industrial Control and Automation - Spring 2019 36

Thermocouple Tables
 The thermocouple tables simply give the voltage that results for a
particular type of thermocouple when the reference junctions are at a
particular reference temperature, and the measurement junction is at a
temperature of interest.
 In most cases, the measured voltage does not fall exactly on a table
value. When this happens, it is necessary to interpolate between table
values that bracket the desired value.

 The measured voltage VM lies between a higher voltage VH and a lower

voltage VL which are in the tables.
 The temperatures corresponding to these voltages are TH and TL
respectively.
 Similarly, for voltage interpolation,

12-Feb-19 Industrial Control and Automation - Spring 2019 37

Thermocouple Tables - Example

12-Feb-19 Industrial Control and Automation - Spring 2019 38

Change of Table Reference
 Tables of TC voltage versus temperature are given for a
specific reference temperature, but can be used to relate
voltage and temperature for a different reference
temperature by using a voltage correction factor.
 The key point to remember is that the voltage is
proportional to the difference between the reference and
measurement junction temperature.
 This correction factor is simply the voltage that the new
reference would produce from the tables.
 If the new reference is greater than the table reference, subtract
the correction factor from table values.
 If the new reference is less than the table reference, Add the
correction factor to table values.

12-Feb-19 Industrial Control and Automation - Spring 2019 39

Change of Table Reference
 The correction process is graphically illustrated in the figure.
 The curve for a 0°C reference is assumed to be the table values.
 For a reference of 20°C, the curve is simply reduced everywhere by the
correction voltage of 1.02 mV.
 In effect, the original curve
slides down by 1.02 mV.
 Note that you cannot simply
add or subtract the new
reference temperature as a
correction factor.
 Correction is always
applied to voltages.

12-Feb-19 Industrial Control and Automation - Spring 2019 40

Change of Table Reference
Example

12-Feb-19 Industrial Control and Automation - Spring 2019 41

Comparison - Construction

RTD Thermistor Thermocouple

The construction is A bulk semiconductor A thermocouple by itself

typically such that the that can be fabricated in is simply a welded or
wire is wound on a form the form of discs, even twisted junction
(in a coil) to achieve beads, and rods, between two metals.
small size and improve varying in size from a
thermal conductivity to bead 1 mm in diameter
decrease response to a disc several
time. centimeters in diameter
and several centimeters
thick.

12-Feb-19 Industrial Control and Automation - Spring 2019 42

Comparison - Sensitivity
RTD
For platinum, this
number is typically on
the order of 0.004/°C,
and for Nickel a typical
value is 0.005/°C. Thus, noted. For example, a
with platinum, a change
of only 0.4Ω would be
expected for 100Ω RTD
if the temperature is
changed by 1℃.
Thermistor Thermocouple
Sensitivity is much The actual sensitivity
higher than RTD and strongly depends on the
changes in resistance type of TC and also the
about 10% per °C are signal conditioning
employed.
10k thermistor at some Worst and best case of
temperature may sensitivity:
change by 1 kΩ) for a Type J: 0.05 mV/°C
1°C change in (typical)
temperature. Type S: 0.006 mV/°C
(typical)

12-Feb-19 Industrial Control and Automation - Spring 2019 43

Comparison - Range
RTD
The effective range of
RTDs principally
depends on the type of
wire used as the active
element
• Platinum RTD:
-100° to 650°C
• Nickel RTD: -180°
to 300°C
12-Feb-19 Industrial Control and Automation - Spring 2019
Thermistor Thermocouple
There are three range The thermocouple
limitation factors: temperature sensor has
(1) melting or the greatest range of all
deterioration of the thermal sensors.
semiconductor, (2) For Type J: -150°C to
deterioration of 745 °C.
encapsulation material, For type S: 0°C to
and (3) insensitivity 1765°C.
at higher temperatures.
The typical range varies
from -50°C to 300°C.
44
Comparison – Response Time
RTD
In general, RTD has a
response time of 0.5 to
5s or more.
Generally, time
constants are specified
either for a “free air”
condition (poor thermal
contact - slower) or an
“oil bath” condition
(good thermal contact -
faster).
12-Feb-19 Industrial Control and Automation - Spring 2019
Thermistor Thermocouple
For the smallest bead Thermocouple time
thermistors: response is simply
• 0.5s in an oil bath related to the size of the
(good thermal wire and any protective
contact); material used with the
• 10s in free air. sensor.
Large disc or rod Large, industrial TC:
thermistors may have 10 to 20s.
response times of 10s Small-gauge wire TC:
or more, even with 10 to 20ms.
good thermal contact.
45