Temperature Notes FE Sem 2 Measurements and Sensors | MU
On this page
Temperature is the most frequently measured quantity in industry: furnaces, engines, chemical reactors, cold stores, hospitals and the battery inside your laptop all depend on it. This module of Physics of Measurements and Sensors (BSC2023, FE Semester 2) starts with the difference between heat and temperature and the temperature scales, then explains four practical thermometers: the bimetallic thermometer, the platinum resistance thermometer (Pt-100), the thermoelectric thermometer (thermocouple) and the NTC thermistor. It ends with how to select a thermometer, their ranges and a comparison, and the calibration of a Pt-100, which is also a lab experiment. It maps to CO5. Expect numericals on Pt-100 resistance, thermocouple emf and thermistor resistance.
Summary
Heat and temperature: Heat is energy in transit from one body to another because of a temperature difference; it is measured in joules and depends on the mass and material. Temperature is a measure of the degree of hotness of a body, related to the average kinetic energy of its particles; it decides the direction of heat flow (from higher to lower temperature). Two bodies in thermal equilibrium with a third are in equilibrium with each other (zeroth law), which is why a thermometer can measure temperature at all: it reaches equilibrium with the object and then a temperature-dependent property of the thermometer (length, pressure, resistance, emf) is read.
Temperature scales: The SI unit is the kelvin (K). Since 2019 the kelvin is defined by fixing the Boltzmann constant \(k = 1.380649\times10^{-23}\) J/K exactly; earlier it was defined as 1/273.16 of the triple point of water. Celsius: \(t(^\circ C) = T(K) - 273.15\); water freezes at 0 °C and boils at about 100 °C at standard pressure. Fahrenheit: \(t_F = \tfrac{9}{5}t_C + 32\). The International Temperature Scale of 1990 (ITS-90) defines practical fixed points (for example the triple point of water 273.16 K, the freezing points of zinc 419.527 °C, aluminium 660.323 °C and silver 961.78 °C) and specifies the standard platinum resistance thermometer as the interpolating instrument between them over a large range.
Thermometric property: Any property that changes regularly and reproducibly with temperature can be used: expansion of liquids or solids (liquid-in-glass and bimetallic thermometers), pressure of a gas (gas thermometer), electrical resistance of metals or semiconductors (RTD, thermistor), thermoelectric emf (thermocouple) and thermal radiation (pyrometer). A good thermometric property should vary linearly or predictably, be large enough to measure, be reproducible and have no hysteresis.
Bimetallic thermometer: Two metal strips with very different coefficients of linear expansion, typically invar (a nickel-iron alloy with very low expansion) and brass or steel (high expansion), are bonded (welded or riveted) together along their length. When heated, the high expansion metal tries to become longer, so the strip bends into an arc with the high expansion metal on the outside (convex side). The deflection is proportional to the temperature change. For a more sensitive dial instrument the strip is wound into a flat spiral or a helix inside a stem; one end is fixed and the other turns a pointer over a circular scale. Range about −50 to 500 °C, accuracy about ±1 percent of span. Advantages: simple, rugged, cheap, no power supply, easy to read. Disadvantages: not very accurate, slow, not suitable for remote reading, may drift after overheating. Uses: thermostats in irons, water heaters and refrigerators, circuit breakers and fire alarms, dial thermometers for ovens and pipes.
Platinum resistance thermometer (PRT): The resistance of a pure metal increases with temperature. Over a moderate range \(R_T = R_0(1 + \alpha T)\); more accurately (IEC 60751, for 0 to 850 °C) the Callendar-Van Dusen equation \(R_T = R_0(1 + AT + BT^2)\) with \(A = 3.9083\times10^{-3}\) per °C and \(B = -5.775\times10^{-7}\) per °C², giving a mean coefficient α = 0.00385 per °C between 0 and 100 °C. Platinum is chosen because it is chemically inert, does not oxidise, can be obtained very pure, has a stable and reproducible resistance-temperature relation and a high melting point. A Pt-100 has 100 Ω at 0 °C (and about 138.5 Ω at 100 °C). Construction: fine platinum wire wound non-inductively on a mica or ceramic former, or a thin platinum film deposited on a ceramic chip, enclosed in a protective stainless steel or ceramic sheath. Measurement: the element forms one arm of a Wheatstone bridge, or a constant current is passed and the voltage measured. Lead resistance adds error, so three-wire or four-wire connections are used to cancel it, and the measuring current is kept small (about 1 mA) to avoid self-heating. Range about −200 to 850 °C; accuracy can be better than ±0.1 °C. Advantages: high accuracy, stability, near-linear output, interchangeable. Disadvantages: costlier, slower than a thermocouple, needs excitation, lower maximum temperature than thermocouples.
Thermoelectric thermometer (thermocouple): Seebeck effect: when two dissimilar metals are joined to form a closed circuit and the two junctions are kept at different temperatures, an emf (thermo-emf) is produced and a current flows. The hot (measuring) junction is placed at the temperature to be measured and the cold (reference) junction is kept at a known temperature, ideally 0 °C (ice bath) or measured and compensated electronically (cold junction compensation). The emf varies with temperature difference approximately as \(E = aT + bT^2\) (T measured from the cold junction at 0 °C). It rises to a maximum at the neutral temperature \(T_n = -a/2b\) and falls to zero at the inversion temperature \(T_i\) (equal to \(2T_n - T_c\)). In practice standard pairs are used in the region where E is nearly linear. Laws: (1) law of intermediate metals: a third metal inserted in the circuit has no effect if both its junctions are at the same temperature, so a meter and copper leads can be connected; (2) law of intermediate temperatures: \(E_{T_1,T_3} = E_{T_1,T_2} + E_{T_2,T_3}\), which lets us correct for a reference junction not at 0 °C. Common types: type K (chromel-alumel, about −200 to 1250 °C, about 41 μV/°C), type J (iron-constantan, about −40 to 750 °C, about 52 μV/°C), type T (copper-constantan, about −200 to 350 °C), type E (chromel-constantan, highest output, about 68 μV/°C), type R and S (platinum-rhodium/platinum, up to about 1450 to 1600 °C, about 10 μV/°C), type B (up to about 1700 °C). Advantages: wide range, very small, fast response, cheap, self-powered, rugged. Disadvantages: small output needing amplification, non-linear, needs a reference junction, less accurate than a PRT.
NTC thermistor: A thermistor (thermal resistor) is a semiconductor resistor made by sintering mixtures of metal oxides such as manganese, nickel, cobalt, copper and iron oxides into beads, discs or rods. Negative temperature coefficient (NTC) thermistors decrease in resistance as temperature rises because more charge carriers are released. The relation is strongly non-linear: \(R_T = R_0\exp[\beta(1/T - 1/T_0)]\) with T in kelvin and β (typically 3000 to 5000 K) a material constant. The temperature coefficient \(\alpha = -\beta/T^2\) is about −3 to −5 percent per °C at room temperature, roughly ten times that of platinum, so thermistors are very sensitive. Range about −50 to 150 °C (some to 300 °C). Advantages: high sensitivity, small size and fast response, low cost, high resistance so lead resistance is unimportant. Disadvantages: non-linear, limited range, less interchangeable, self-heating, can drift at high temperature. Uses: digital clinical thermometers, battery packs, air conditioners, motor protection, inrush current limiters, temperature compensation.
Selection of a thermometer: Consider (1) the temperature range; (2) the accuracy and resolution required; (3) speed of response needed (small sensors respond faster); (4) the environment: corrosive, vibrating, high pressure, nuclear, electrically noisy; (5) whether contact is possible or non-contact (radiation) measurement is needed; (6) whether remote indication, recording or control is required (electrical output is then preferred); (7) size and mounting space; (8) stability, need for calibration and interchangeability; (9) cost of sensor and instrumentation.
Calibration of a Pt-100: Calibration establishes the actual R-T relation of a particular sensor. Method: place the Pt-100 and a reference thermometer of known accuracy (a standard PRT or a calibrated digital thermometer) close together in a stirred, well-insulated bath (ice-water bath for 0 °C, a temperature-controlled water or oil bath or a dry block calibrator for higher points). At each set temperature wait for equilibrium and record the reference temperature and the Pt-100 resistance (using a four-wire ohmmeter or a bridge with low current). Repeat at several points, both increasing and decreasing to check hysteresis. Plot R against T; fit a straight line (least squares) or the Callendar-Van Dusen equation. From the line find \(R_0\) and \(\alpha = (R_{100} - R_0)/(100R_0)\). The deviations from the standard table at each point form the calibration chart, and the fitted relation is used to convert later resistance readings into temperature.
Mind map
Temperature and its Measurement
- Basics
- Heat is energy, temperature is degree of hotness
- K = °C + 273.15, °F = 9/5 °C + 32
- ITS-90 fixed points
- Bimetallic
- Invar and brass, unequal expansion
- Spiral or helix, dial
- Thermostats
- Platinum RTD
- R = R0(1 + αT), α = 0.00385
- Pt-100, 3 or 4 wire
- −200 to 850 °C
- Thermocouple
- Seebeck emf, E = aT + bT²
- Neutral, inversion temperature
- Types K, J, T, R, S
- Cold junction compensation
- NTC thermistor
- Metal oxides, R falls with T
- R = R0 exp[β(1/T − 1/T0)]
- Very sensitive, non-linear
- Selection and calibration
- Range, accuracy, response, environment, cost
- Pt-100 calibration with bath and reference
Thermocouple emf, neutral and inversion temperatures
- Keep the cold junction at 0 °C and the hot junction at T °C. Experimentally \(E = aT + bT^2\) (for most pairs b is negative).
- Thermoelectric power (sensitivity) is \(dE/dT = a + 2bT\).
- Neutral temperature: the emf is maximum where \(dE/dT = 0\), so \(T_n = -a/2b\).
- Inversion temperature: the emf becomes zero again (and then reverses) where \(aT + bT^2 = 0\), so \(T_i = -a/b = 2T_n\) for a cold junction at 0 °C.
- If the cold junction is at \(T_c\), the neutral temperature does not change (it depends only on the pair), and \(T_i = 2T_n - T_c\).
- A thermocouple is used only well below its neutral temperature, where E rises steadily with T.
- Law of intermediate temperatures lets us correct a reading taken with the cold junction at room temperature: \(E(T,0) = E(T,T_r) + E(T_r,0)\).
Comparison of thermometers
| Thermometer | Principle | Typical range | Strengths | Weaknesses |
|---|---|---|---|---|
| Bimetallic | Differential expansion bends a strip | About −50 to 500 °C | Simple, rugged, no power | Low accuracy, local reading only |
| Platinum RTD (Pt-100) | Resistance of platinum rises with T | About −200 to 850 °C | Most accurate and stable, near linear | Costly, slower, needs excitation |
| Thermocouple | Seebeck emf between two metals | About −200 to 1700 °C (by type) | Widest range, fast, small, cheap | Small non-linear emf, needs cold junction compensation |
| NTC thermistor | Semiconductor resistance falls with T | About −50 to 150 °C (some to 300 °C) | Very sensitive, small, fast, cheap | Highly non-linear, narrow range |
| Liquid-in-glass | Expansion of mercury or alcohol | About −100 to 350 °C (mercury from −39 °C) | Simple, cheap | Fragile, no remote reading |
Important definitions
| Term | Meaning |
|---|---|
| Heat | Energy transferred because of a temperature difference. |
| Temperature | Degree of hotness; measure of the average kinetic energy of particles; decides direction of heat flow. |
| Thermometric property | A physical property that varies reproducibly with temperature and is used to measure it. |
| Temperature coefficient of resistance | Fractional change in resistance per degree, \(\alpha = (R_T - R_0)/(R_0T)\). |
| Seebeck effect | Generation of an emf in a circuit of two dissimilar metals whose junctions are at different temperatures. |
| Neutral temperature | Hot junction temperature at which the thermo-emf is maximum. |
| Inversion temperature | Hot junction temperature at which the thermo-emf becomes zero and then reverses. |
| Cold junction compensation | Correction for the reference junction not being at 0 °C. |
| Thermistor | Semiconductor resistor with a large temperature coefficient (NTC or PTC). |
| Self-heating | Rise in sensor temperature caused by the measuring current, which causes error. |
| Calibration | Establishing the actual relation between sensor output and temperature by comparison with a standard. |
Formulas
| Formula | Meaning |
|---|---|
| \( T(K) = t(^\circ C) + 273.15,\ t_F = \tfrac{9}{5}t_C + 32 \) | Temperature scale conversions. |
| \( R_T = R_0(1 + \alpha T) \) | Linear RTD relation (α = 0.00385 /°C for Pt-100). |
| \( R_T = R_0(1 + AT + BT^2) \) | Callendar-Van Dusen, 0 to 850 °C; A = 3.9083 x 10^-3, B = −5.775 x 10^-7. |
| \( \alpha = \dfrac{R_{100} - R_0}{100R_0} \) | Coefficient from two-point calibration. |
| \( E = aT + bT^2 \) | Thermo-emf with cold junction at 0 °C. |
| \( T_n = -\dfrac{a}{2b},\quad T_i = 2T_n - T_c \) | Neutral and inversion temperatures. |
| \( E_{T_1,T_3} = E_{T_1,T_2} + E_{T_2,T_3} \) | Law of intermediate temperatures. |
| \( R_T = R_0\,e^{\beta\left(\frac{1}{T} - \frac{1}{T_0}\right)} \) | NTC thermistor (T in kelvin). |
| \( \alpha_T = -\dfrac{\beta}{T^2} \) | Temperature coefficient of a thermistor. |
Key points to remember
- Heat is energy; temperature decides the direction of heat flow.
- Bimetallic strip bends towards the low expansion metal (high expansion metal on the convex side) when heated.
- Pt-100: 100 Ω at 0 °C, 138.5 Ω at 100 °C; use 3 or 4 wire connection and small current.
- Platinum: inert, stable, reproducible, nearly linear; the standard interpolating thermometer of ITS-90.
- Thermocouples are active sensors; RTDs and thermistors are passive.
- Thermocouple used below neutral temperature; inversion temperature is twice the neutral temperature for a 0 °C cold junction.
- NTC thermistor resistance falls exponentially with temperature; it is about ten times more sensitive than an RTD.
- Widest range: thermocouple. Best accuracy: platinum RTD. Highest sensitivity: thermistor. Simplest: bimetallic.
- Calibrate a Pt-100 against a reference thermometer in a stirred bath at several points and fit a line.
Solved examples
Example 1. A Pt-100 (α = 0.00385 /°C) reads 119.4 Ω. Find the temperature.
\(T = (R_T/R_0 - 1)/\alpha = (1.194 - 1)/0.00385 = 0.194/0.00385 = 50.4\) °C.
Example 2. Using the Callendar-Van Dusen equation, find the resistance of a Pt-100 at 200 °C. Compare with the linear formula.
\(R = 100[1 + 3.9083\times10^{-3}\times200 - 5.775\times10^{-7}\times200^2] = 100[1 + 0.78166 - 0.0231] = 175.86\) Ω. Linear: \(100(1 + 0.00385\times200) = 177.0\) Ω. The linear formula overestimates by about 1.1 Ω (about 3 °C), so the quadratic is used above 100 °C.
Example 3. In calibrating a Pt-100, the resistance is 100.2 Ω in melting ice and 138.9 Ω in steam at 100 °C. Find α. What temperature corresponds to 125.0 Ω?
\(\alpha = (138.9 - 100.2)/(100\times100.2) = 38.7/10020 = 3.862\times10^{-3}\) /°C. \(T = (125.0 - 100.2)/(\alpha R_0) = 24.8/(3.862\times10^{-3}\times100.2) = 24.8/0.387 = 64.1\) °C.
Example 4. The thermo-emf of a couple is \(E = aT + bT^2\) with a = 14 μV/°C and b = −0.02 μV/°C², cold junction at 0 °C. Find the neutral and inversion temperatures and the emf at 100 °C.
\(T_n = -a/2b = 14/(2\times0.02) = 350\) °C. \(T_i = 2T_n = 700\) °C. \(E(100) = 14\times100 - 0.02\times100^2 = 1400 - 200 = 1200\) μV = 1.2 mV.
Example 5. A type K thermocouple (take a constant 41 μV/°C) gives 12.3 mV with its cold junction at 25 °C. Estimate the hot junction temperature.
Temperature difference \(= 12.3\times10^{-3}/41\times10^{-6} = 300\) °C. By the law of intermediate temperatures, \(T = 300 + 25 = 325\) °C. (Accurate work uses the type K reference table instead of a constant sensitivity.)
Example 6. An NTC thermistor has 10 kΩ at 25 °C and β = 3950 K. Find its resistance at 50 °C, and the temperature at which it reads 2.5 kΩ.
\(T_0 = 298.15\) K, T = 323.15 K. \(R = 10000\exp[3950(1/323.15 - 1/298.15)] = 10000\exp(-1.0249) = 3588\) Ω, about 3.59 kΩ. For 2.5 kΩ: \(1/T = 1/298.15 + \ln(0.25)/3950 = 3.3540\times10^{-3} - 3.510\times10^{-4} = 3.0030\times10^{-3}\), T = 333.0 K = 59.8 °C. Coefficient at 25 °C: \(\alpha = -3950/298.15^2 = -0.0444\) per K, about −4.4 percent per °C, against +0.385 percent per °C for platinum.
Example 7. Convert normal body temperature 98.6 °F to °C and K.
\(t_C = (98.6 - 32)\times5/9 = 37.0\) °C; T = 310.15 K.
Important questions and answers
Q1. Distinguish between heat and temperature. (3 marks)
Heat is a form of energy that flows from a hotter to a colder body; it is measured in joules (or calories) and depends on mass, material and temperature change (\(Q = mc\Delta T\)). Temperature is the degree of hotness, a measure of the average kinetic energy of molecules, measured in kelvin or degrees Celsius; it does not depend on mass. Heat is the cause; temperature change is the effect. Two bodies at the same temperature exchange no net heat even if they contain very different amounts of internal energy.
Q2. Explain the construction and working of a bimetallic thermometer. (5 marks)
Construction: two strips of metals with very different expansion coefficients, such as invar and brass, bonded along their length; for a dial thermometer the strip is coiled into a spiral or helix, one end fixed to the stem and the other attached to a spindle carrying a pointer over a scale. Working: on heating, the brass expands more than invar; since they are bonded, the strip bends with brass on the outer (convex) side; in a coil this produces rotation of the free end, which moves the pointer in proportion to temperature. Range about −50 to 500 °C. Advantages: simple, robust, cheap, no power. Disadvantages: low accuracy, local reading only, slower response, calibration may shift. Applications: thermostats, fire alarms, circuit breakers, dial thermometers.
Q3. Explain the platinum resistance thermometer. Why is platinum preferred? (6 marks)
Principle: resistance of a metal increases with temperature, \(R_T = R_0(1 + \alpha T)\). Construction: fine pure platinum wire wound non-inductively on a mica or ceramic former (or thin film on ceramic), sealed in a protective sheath, connected by compensating leads. Measurement: element in a Wheatstone bridge; with three-wire connection the lead resistances appear in adjacent arms and cancel; or four-wire method with constant current and voltmeter; current kept small to avoid self-heating. Platinum preferred because it is chemically inert and oxidation-resistant, can be refined to high purity, has a stable reproducible and nearly linear R-T relation, wide range (−200 to 850 °C) and high melting point (1768 °C). Accuracy better than 0.1 °C is possible. Drawbacks: cost, slower response, small α compared with a thermistor.
Q4. What is a thermocouple? Explain the Seebeck effect, neutral temperature and inversion temperature, and the laws of thermocouples. (6 marks)
A thermocouple is a pair of dissimilar metal wires joined at a measuring (hot) junction, with the other ends at a reference (cold) junction connected to a millivoltmeter. Seebeck effect: a temperature difference between the junctions produces an emf, \(E = aT + bT^2\). Neutral temperature \(T_n = -a/2b\), where emf is maximum; inversion temperature \(T_i = 2T_n - T_c\), where emf is zero and reverses. Law of intermediate metals: introducing a third metal (meter, copper leads) does not change the emf if its junctions are at the same temperature. Law of intermediate temperatures: \(E_{T_1,T_3} = E_{T_1,T_2} + E_{T_2,T_3}\), used for cold junction correction. Types and ranges: K, J, T, R, S, B. Merits: wide range, fast, cheap, self-powered. Demerits: low output, non-linear, needs cold junction compensation.
Q5. Explain the NTC thermistor and its R-T characteristic. (5 marks)
An NTC thermistor is a sintered metal oxide semiconductor (Mn, Ni, Co oxides) in bead, disc or rod form, with leads, often glass coated. As temperature rises, more electrons are freed to conduct, so resistance falls exponentially: \(R_T = R_0\exp[\beta(1/T - 1/T_0)]\). The R-T curve is a steeply falling exponential; a 10 kΩ thermistor falls to about 3.6 kΩ at 50 °C. Coefficient about −4 percent per °C. Used in a bridge or voltage divider; linearised with a parallel resistor or by software. Applications: clinical and battery thermometers, temperature control in appliances, current inrush limiting, temperature compensation.
Q6. What factors are considered when selecting a thermometer for a particular use? (5 marks)
Temperature range of the process; accuracy and resolution required; speed of response (sensor mass and size); nature of the medium and environment (corrosion, pressure, vibration, electrical noise, hazardous area); contact or non-contact measurement; need for remote transmission, recording or control (favours electrical sensors); available space and mounting; long-term stability and ease of calibration; interchangeability; total cost including instrumentation. Examples: furnace at 1200 °C, type K or S thermocouple; laboratory precision, Pt-100; clinical thermometer, thermistor; oven dial, bimetallic.
Q7. Describe the calibration of a Pt-100 sensor. (5 marks)
Apparatus: Pt-100, reference thermometer (standard PRT or calibrated digital thermometer), ice bath and stirred water or oil bath with heater (or dry block calibrator), four-wire ohmmeter or bridge. Procedure: place sensor and reference close together in the bath; at 0 °C (ice point) record R; raise the bath to several steady temperatures (say every 20 °C to 100 °C), waiting for equilibrium each time, recording reference temperature and resistance; repeat while cooling. Analysis: plot R against T, fit the best straight line by least squares (or the Callendar-Van Dusen equation), find \(R_0\) and α, compare with the standard table, and prepare a calibration chart of corrections. Precautions: stir the bath, immerse sufficiently to avoid stem conduction, use low current to avoid self-heating, use four-wire leads.
Exam tips
- Write T in kelvin in thermistor formulas; this is the most common mistake.
- For thermocouples, state that the cold junction must be at a known temperature and how it is compensated.
- A comparison table of the four thermometers (principle, range, accuracy, merits, demerits) answers many 5-mark questions.
- Draw the three-wire bridge for the RTD to show lead compensation.
- Relate the Pt-100 calibration to the least squares method from Module 1.
FAQs
Why is the reading of a thermocouple wrong if the cold junction is at room temperature?
The emf depends on the temperature difference between the junctions. If the cold junction is at 30 °C instead of 0 °C, the emf is smaller, and the emf for 0 to 30 °C must be added (law of intermediate temperatures) to get the true reading.
Which thermometer is used in a digital clinical thermometer?
Usually a small NTC thermistor, because it is very sensitive over the narrow range of body temperatures, small and cheap.
Why does a Pt-100 use three or four wires?
The leads have resistance that changes with temperature. A three-wire bridge cancels the lead resistance, and a four-wire method measures voltage directly across the element, so lead resistance does not cause error.
What is a PTC thermistor?
A thermistor whose resistance increases with temperature, usually made from doped barium titanate. It shows a sharp rise above a switching temperature and is used as a resettable fuse and in motor protection.
These notes follow the University of Mumbai first-year engineering syllabus (NEP 2020, Physics of Measurements and Sensors) and were prepared with AI help and checked.
More in Physics of Measurements and Sensors
See all →- Nanotechnology Notes FE Sem 2 Measurements and Sensors | MU MU FE Sem 2 Measurements and Sensors Module 6: nano properties, surface to volume ratio, van der Waals bonding, SEM, TEM, AFM, gas sensing,...
- Measurement Errors Notes FE Sem 2 Measurements and Sensors | MU MU FE Sem 2 Measurements and Sensors Module 1: errors, calibration, instrument terms, accuracy, precision, standard deviation, least squares...
- Wave Interference Notes FE Sem 2 Measurements and Sensors | MU MU FE Sem 2 Measurements and Sensors Module 2: monochromatic light, optical flats, air wedge fringes, flatness testing and surface contour t...
- Transducers Notes FE Sem 2 Measurements and Sensors | MU MU FE Sem 2 Physics of Measurements and Sensors Module 3: transducer classification, potentiometer, strain gauge, RTD, LVDT and thermal phot...
- Solid State Sensors Notes FE Sem 2 Measurements and Sensors | MU MU FE Sem 2 Measurements and Sensors Module 4: Hall effect and Hall voltage, piezoelectric ultrasonics, ultrasonic flow and distance meters.