Temperature Sensors
While the International Temperature Scale of 1990 (ITS-90) defines the theoretical realization of temperature, practical applications in industry and laboratories depend on secondary and tertiary measurement devices. The three most common electrical temperature sensors are Resistance Temperature Detectors (RTDs), Thermocouples, and Thermistors.
Each of these technologies exploits a different physical phenomenon related to heat and electrical properties. Selecting the appropriate sensor involves balancing accuracy, temperature range, environmental robustness, and cost. This section explores the fundamental physics, construction, and calibration of these indispensable instruments.
Resistance Temperature Detectors (RTDs)
Resistance Temperature Detectors operate on the principle that the electrical resistance of pure metals increases predictably with temperature. The most common and accurate RTDs are made of platinum and are often referred to as PRTs (Platinum Resistance Thermometers).
The Pt100 Standard
The most prevalent industrial RTD is the Pt100. The "Pt" denotes platinum, and "100" signifies its nominal resistance of exactly $100\,\Omega$ at $0^{\circ}\text{C}$. Their behavior is largely governed by the Callendar-Van Dusen equation, and standardized by norms such as IEC 60751.
For temperatures above $0^{\circ}\text{C}$, the resistance $R_t$ at temperature $t$ is closely approximated by a quadratic equation:
Where $R_0$ is the resistance at $0^{\circ}\text{C}$, and $A$ and $B$ are empirical constants derived during the material characterization.
Advantages: RTDs offer exceptional stability, repeatability, and accuracy, making them the standard for reference measurements in many calibration laboratories. They provide a nearly linear response over a wide range (typically $-200^{\circ}\text{C}$ to $850^{\circ}\text{C}$).
Disadvantages: They are relatively fragile, especially wire-wound types, and are susceptible to self-heating errors if the measurement current is too high. Furthermore, their response time is slower compared to thermocouples due to their physical mass. Four-wire measurement configurations are strictly necessary for high-accuracy applications to eliminate lead wire resistance errors.
Thermocouples
Thermocouples rely on the Seebeck effect. When two dissimilar metal wires are joined at one end (the measuring junction) and exposed to a temperature gradient, a small thermoelectric voltage is generated between the open ends (the reference junction).
The Seebeck Effect and Cold Junction Compensation
The voltage $V$ generated by a thermocouple is proportional to the temperature difference between the measuring junction ($T_m$) and the reference junction ($T_{ref}$):
Where $S_{AB}(T)$ is the Seebeck coefficient of the two metals (A and B) as a function of temperature. Because the voltage depends on the temperature difference, knowing the exact temperature of the reference junction is critical. Modern instruments use Cold Junction Compensation (CJC), employing a secondary sensor (like a thermistor) at the instrument's terminal block to measure $T_{ref}$ and computationally compensate for it.
Advantages: Thermocouples are incredibly robust, inexpensive, and have extremely fast response times due to their small thermal mass. They can measure over a vast temperature range, with some specialized types (like Type C or D) capable of measuring above $2000^{\circ}\text{C}$.
Disadvantages: They are the least accurate of the three main sensor types, typically offering uncertainties in the range of $\pm 1^{\circ}\text{C}$ to $\pm 2^{\circ}\text{C}$ for standard limits of error. Their output is non-linear, requiring complex polynomial characterization in the measurement device, and they are susceptible to electromagnetic interference (EMI).
Thermistors
Thermistors (Thermal Resistors) are semiconductor devices made from ceramic or polymer materials. Unlike metallic RTDs, most common thermistors have a Negative Temperature Coefficient (NTC), meaning their resistance drops drastically as temperature increases.
The Steinhart-Hart Equation
Thermistor behavior is highly non-linear. To achieve precise temperature measurements over a usable range, the Steinhart-Hart equation is commonly utilized:
Where $T$ is the temperature in Kelvin, $R$ is the resistance in ohms, and $A$, $B$, and $C$ are the Steinhart-Hart coefficients provided by the manufacturer or derived from calibration.
Advantages: Thermistors possess extreme sensitivity. A small temperature change yields a massive resistance change, allowing for high-resolution measurements (often better than $0.01^{\circ}\text{C}$ over a narrow span). They are compact, fast-responding, and relatively inexpensive.
Disadvantages: Their primary limitation is their narrow usable temperature range (typically $-50^{\circ}\text{C}$ to $150^{\circ}\text{C}$). At higher temperatures, they can permanently drift or degrade. Their profound non-linearity requires more complex signal conditioning or software algorithms to convert resistance to temperature accurately.