HCMC University of Technology DEPT.

OF BIOMEDICAL ENGINEERING

Chapter 2+6
Temperature sensors
V Nht Quang

Objectives
Discuss various types of temperature sensor Discuss their characteristics Look at circuits incorporating temperature sensors

Chapter 3

Temperature is sensed by a variety of methods such as:

1.

Thermoresistor effects: Resistance thermometers Thermistors Thermoelectric effects Thermocouples Radiation sensing pyroelectric sensors
3

2.

3.

Chapter 3

2.3 RESISTIVE TEMPERATURE DETECTORS (RTDs)

Chapter 3

An RTD is a temperature detector based upon a variation in electric resistance. The commonest metal for this application is platinum, which is sometimes designated PRT (platinum resistance thermometer)

Chapter 3

RTDs rely on the positive temperature coefficient for a conductors resistance. In a conductor the number of electrons available to conduct electricity does not significantly change with temperature. But when the temperature increases, the vibrations of the atoms around their equilibrium positions increase in amplitude.
Chapter 3 6

This results in a greater dispersion of electrons, which reduces their average speed. Hence, the resistance increases when the temperature rises. This relationship can be written as:

R0: the resistance at the reference temperature T0 The coefficients can be determined from resistance measurements at fixed-point temperatures
Chapter 3 7

Restrictions First, it is not possible to measure temperatures near the melting point of the conductor. Second, we must avoid any self-heating due to the measurement circuits. Otherwise, the sensor temperature would be higher than that of the surrounding medium.
Chapter 3 8

For a conductor in a given environment, the heat dissipation capability is given by the heat dissipation constant or heat dissipation factor (mW/K), which depends on the surrounding fluid and its velocity, because heat loss increases by convection. As for other sensors, RTD must be stable. Time and thermal drifts, particularly at high temperature, limit temperature resolution.
Chapter 3 9

Chapter 3

10

Chapter 3

11

Advantages High sensitive (ten times that of thermocouples) High repeatability Long-term stability and accuracy for platinum Low cost for copper and nickel

Chapter 3

12

For metals used as RTD probes:

is the temperature coefficient of resistance, calculated from the resistance measured at two reference temperatures (e.g. 0C and 100C) is sometimes termed

relative sensitivity
Chapter 3 13

Chapter 3

14

Chapter 3

15

2.4 THERMISTORS

Chapter 3

16

2.4.1 Models Thermistor comes from thermally sensitive resistor and applies to temperature-dependent resistors that are based not on conductors as the RTD but on semiconductors. There are two types:
NTC: negative temperature coefficient PTC: positive temperature coefficient

Chapter 3

17

Chapter 3

18

Thermistors are small semiconducting sensors made from sintered oxides of cobalt, nickel and manganese. Can write temperature dependence as:

RT0: resistance at 25C or other reference temperature; and T0: this temperature in K : characteristic temperature of the material (2000K 4000K)
Chapter 3 19

Temperature Coefficient of Resistance (TCR)

This is the TCR (or relative sensitivity) which is usually quoted by manufacturers.
Chapter 3 20

10

Characteristics Thermistors usually have a negative temperature coefficient, compared to positive temperature coefficient of Resistance thermometers. Resolution is theoretically infinite. Linearity depends on range of operation. Typical value of Temp. coefficient: -0.04 to +0.14 Range is typically: -100C to +200C
Chapter 3 21

2.4.2 Thermistor types and applications

Disc type

Bead type

Probe type
Chapter 3

Rod type
22

11

Chapter 3

23

2.4.3 Linearization

NTC thermistors exhibit a distinctly nonlinear R/T characteristic. If a fairly linear curve is required for measurements over a temperature range, e.g. for a scale, series-connected or paralleled resistors are quite useful.

Chapter 3

24

12

Chapter 3

25

The parallel combination of both resistors is then

and its sensitivity with temperature is:

Chapter 3

26

13

Rp is not linear, yet its change with temperature is smaller than that of RT because the factor multiplying dRT/dT is smaller than 1. The equivalent TCR is:

Chapter 3

27

Resistor R, or alternatively the NTC thermistor, can be chosen to improve linearity in the measurement range. An analytical method to calculate R is by forcing three equidistant points in the resulting resistance temperature curve to coincide with a straight line. If T1 T2 = T2 T3, the condition is: RP1 RP2 = RP2 PP3
Chapter 3 28

14

Solving for R, we obtain:

This expression does not depend on any mathematical model for RT. Thus this method can also be applied to PTC thermistors and other nonlinear resistive sensors.
Chapter 3 29

With both thermistors and resistance thermometers, we must be careful to limit the current flowing in device to reduce self-heating effects. Typically: Resistance thermometers: 1 to 2 mA Thermistors: 5 20 mA

Chapter 3

30

15

6.1 THERMOCOUPLES

Chapter 3

31

6.1.1 Reversible thermoelectric effects Thermoelectric sensors are based on two effects that are reversible as contrasted with the irreversible Joule effect: The Seebeck effect The Peltier effect The Thomson effect A pair of different metals with a fixed junction at a point or zone constitutes a thermocouple.
Chapter 3 32

16

The Seebeck effect Thomas J. Seebeck discovered in 1822

Chapter 3

33

The relationship between the emf EAB and the difference in temperature between both junctions T defines the Seebeck coefficient SAB:

Where SA hay SB are, respectively, the absolute thermoelectric power for A and B
Chapter 3 34

17

The Peltier effect Jean C.A. Peltier discovered in 1834 The heating or cooling of a junction of two different metals when an electric current flows through it.

Chapter 3

35

When the current direction reverses, so does the heat flow. That is, if a junction heats (liberates heat), then when the current is reversed, it cools (absorbs heat), and if it cools, then when the current is reversed, it heats. This effect is reversible and does not depend on the contact, namely, on the shape or dimensions of the conductors. This effect depends on the junction composition and temperature.
Chapter 3 36

18

The Thomson effect Discovered by William Thomson in 1847 1854 Consists of heat absorption or liberation in a homogeneous conductor with a nonhomogeneous temperature when there is a current along it. The heat liberated is proportional to the current , not to its square, and therefore changes its sign for a reversed current.
Chapter 3 37

Heat is absorbed when charges flow from the colder to the hotter points, and it is liberated when they flow from the hotter to the colder one. In other words, heat is absorbed when charge and heat flow in opposite directions, and heat is liberated when they flow in the same direction.

Chapter 3

38

19

Chapter 3

39

6.1.2 Common thermocouples

In thermocouple junctions, there is a simultaneous requirement for 1. A low resistivity temperature coefficient 2. Resistance to becoming oxidized at high temperatures, in order to withstand the working environment 3. A linearity as high as possible
Chapter 3 40

20

Chapter 3

41

Chapter 3

42

21

6.1.3 Practical thermocouple laws

In addition to the advantages and disadvantages, there are several experimental laws for temperature measurement using thermocouples that greatly simplify the analysis of thermocouple circuits.

Chapter 3

43

6.1.3.1 Law of homogeneous circuits

An electric current cannot be sustained in a circuit of a single homogeneous metal, however varying in section, by the application of heat alone. What this means for thermocouples, is that if there is a temperature distribution along the wires between the hot and cold junctions, the total thermal EMF will be unaffected. Only the temperature at the junction between the two dissimilar metals will have an effect on the EMF produced.
Chapter 3 44

22

Chapter 3

45

6.1.3.2 Law of intermediate metals

If two dissimilar metals A and B with their junctions at T1 and T2 are joined to a third metal C at one leg, if C is kept at a uniform temperature along its entire length, the total EMF in the circuit will be unaffected.

Chapter 3

46

23

Chapter 3

47

6.1.3.3 Law of successive or intermediate temperatures If you have one thermocouple with it's junction at 32F and some reference temperature and another thermocouple at the same reference temperature and the measured temperature. This is equivalent to a single thermocouple with it's junction at 32F and the measured temperature.

Chapter 3

48

24

Chapter 3

49

6.1.4 Cold junction compensation in thermocouple circuits In order to apply the Seebeck effect to temperature measurement, one junction must remain at a fixed reference temperature. Placing the reference junction into melting ice is easy and highly accurate, but it requires frequent maintenance and has a high cost.
Chapter 3 50

25

Standard thermocouple arrangement

Chapter 3

51

Practical arrangement

This is the more usual configuration using cold junction compensation. It relies on the junctions between metal a and metal c and metal b and metal c being at the same known temperature.
Chapter 3 52

26

Cold junction compensation

Leaving the reference junction to undergo the ambient temperature fluctuations but at the same time measuring these by another temperature sensor placed near the reference junction. Then a voltage equal to that generated at the cold junction is subtracted from the one produced by the circuit.
Chapter 3 53

VOLTAGE OUTPUT ANALOG TEMPERATURE SENSORS

Chapter 3

54

27

LM135/235/335 Kelvin sensors Develop an output voltage proportional to absolute temperature with a nominal temperature coefficient of 10 mV/K. The nominal output voltage is therefore 2.73 V at 0C and 3.73 V at 100C.

Chapter 3

55

Chapter 3

56

28

The error of an untrimmed LM135A over the full -55C to +150C range is less than 2.7C. Using an external trimpot to adjust accuracy reduces error to less than 1C over the same temperature range.

Chapter 3

57

LM35/LM45 Celsius sensors

Three - terminal devices that produce output voltages proportional to C (10 mV/C). The output voltage is 0 mV at 0C and 1000 mV at 100C. These sensors can measure temperatures below 0C by using a pull down resistor from the output pin to a voltage below the ground pin
Chapter 3 58

29

The LM35 is more accurate (1C from 55C to +150C vs. 3C from -20C to +100C)
Chapter 3 59

Chapter 3

60

30

CURRENT OUTPUT ANALOG SENSORS

Chapter 3

61

LM134/234/334 current output temperature sensors A current output temperature sensor with an output current proportional to absolute temperature. The sensitivity is set using a single external resistor. Typical sensitivities are in the 1A/C to 3A/C range, with 1A/C being a good nominal value.
Chapter 3 62

31

By adjusting the value of the external resistor, the sensitivity can be trimmed for good accuracy over the full operating temperature range (-55C to +125C for the LM134, -25C to +100C for the LM234, and 0C to +70C for the LM334) LM134 typically needs only 1.2V supply voltage, so it can be useful in applications with very limited voltage headroom.
Chapter 3 63

Chapter 3

64

32

References
Ramon Pallas Areny and John G. Webster, Sensors and signal conditioning, John Wiley & Son Inc, 2001 Dr. Paul W Nutter, Sensors and sensing principles Lecture 6, University of Manchester, 2008 National semiconductors temperature sensor handbook
Chapter 3 65

33