Temperature Sensor Basics


This document goes into detail about the theory of operation, types, and considerations surrounding three types of temperature measurement sensors. Each sensor option—thermocouple, RTD, and thermistor—is unique in its design and theory of operation. Further, each sensor option comes in a variety of styles suited for different application needs. By understanding these variations and their potential impacts on measurement quality, you can select the appropriate sensor for your application. For assistance in choosing between these three types of temperature sensor, refer to the Overview of Temperature Sensors.



Theory of thermocouple operation

Thermocouples operate under the principle known as the Seebeck effect. When two wires made of dissimilar metals are joined and heated at one end, a thermoelectric circuit is formed that causes a measurable voltage differential known as the Seebeck voltage at the “cold” end. A given pairing of metals varies in temperature range, sensitivity, and error based on the properties of those metals.

Illustration of the Seebeck Effect

Figure 1: Illustration of the Seebeck Effect


Each type of thermocouple consists of a unique pairing of metals. You need to understand the operating specifications of the thermocouple type you select for your temperature measurement. Some thermocouples offer a wide temperature range at the expense of a very nonlinear voltage-temperature relationship, while others provide a smaller (but more linear) temperature range.

Thermocouple types

As mentioned above, you can choose from a variety of thermocouple types and constructions. Types are generally defined by a letter designation, for example, E, J, or K. The thermocouple type defines the metals used to create the thermocouple; therefore, it also defines the operating range, accuracy, and linearity of the thermocouple. The following graphs depict the voltage response of various thermocouple types over a range of temperatures.

Temperature Response of Different Thermocouple Types

Figure 2: Temperature Response of Different Thermocouple Types


In addition to the type of thermocouple, you must choose a sheathing configuration. Some of these options are shown in Figure 3, including grounding, isolated, sealed, and exposed .


Options for Thermocouple Sheathing 

Figure 3: Options for Thermocouple Sheathing

Each configuration has advantages and disadvantages regarding response time, noise immunity, and safety. Table 1 gives an overview of the impact of each configuration option.


Junction Configuration




Fastest response (~0.1 s to 2 s)

Ground loop and noise potential

no chemical  protection

most prone to physical damage

Exposed Bead

Fast response (~15 s)

Ground loop and noise potential

no chemical  protection

prone to physical damage

Sealed and Grounded

Physical and chemical protection

Slow response (~40 s)

Ground loop and noise potential

Sealed and Isolated

Physical and chemical protection

electrical protection (avoids ground loops and noise)

Slowest response (~75 s)

Table 1: Overview of Thermocouple Junction Configurations



Theory of RTD operation

RTDs, or resistance temperature detectors, are active measurement devices that operate by changing resistance with changes to ambient temperature. RTDs are typically constructed with a ceramic or glass core and a thin winding of metal that’s often platinum for its stability.

Basic Resistance Thermometer Components

Figure 4: Basic Resistance Thermometer Components


Alternative configurations use different insulation and/or winding materials, which results in different performances and temperature ranges. Another option, a thin-film RTD, consists of a thin layer of metal in between layers of insulating material. This style is best suited for surface temperature measurements because it provides more uniform contact across the surface of the RTD.

The key to an RTD’s ability to measure temperature is the thermal properties of the metal winding. If you understand these properties well, you can reliably predict the temperature at a measured resistance. The predictable resistance-temperature relationship leads to an accurate temperature measurement device.

Since RTDs are active sensors, they require external excitation to produce a measurable voltage drop that can be translated into resistance. Resistance values are generally very low, meaning lead-wire resistance can cause less accurate measurements. Because of this, RTDs often come in multiwire configurations. With all of these alternatives, you have to carefully select the measurement hardware to use with an RTD. For more information, refer to the Engineer's Guide for Accurate Sensor Measurements .

RTD types

The various types of RTDs have four primary defining attributes:

  1. Number of lead wires
    • Two-wire  RTDs are the simplest to use, but they can affect measurement accuracy if the resistance of the lead wires is significant relative to the measured resistance of the RTD element.

Two-Wire RTD

Figure 5: Two-Wire RTD


    • Three-wire RTDs are the most common in industrial applications. They allow the measurement hardware to characterize and correct for lead-wire resistance, which results in a more accurate measurement. Inconsistencies in the lead-wire resistance, however, can lead to unexpected measurement errors.

Three-Wire RTD

Figure 6: Three-Wire RTD


    • Four-wire RTDs provide separate paths for both excitation and resistance measurement. This allows for lead-wire resistance correction as well as the isolation of any noise in the excitation signal. Four-wire RTDs are also immune to mismatched lead-wire resistances, but they require measurement hardware capable of a four -wire resistance measurement.

Four-Wire RTD

Figure 7: Four-Wire RTD


  1. Wire material
    • Deployment environment and temperature range are the key factors to consider when selecting wire materials.
    • Select a material that can withstand expected water submersion, chemicals, and abrasion.
    • Wire material and sizing also impact lead resistance, depending on lead-wire configuration.
  1. Physical construction
    • This is application specific. Consider how the sensor must be mounted, any electrical isolation requirements, and the type of medium the sensor will be immersed in (liquid, gas, solid surface, and so on).
    • Available configurations include basic RTD elements, thin-film surface elements, and probes for which the RTD element is encased in a metal sheath and possibly electrically isolated.

RTD Construction Styles

Figure 8: RTD Construction Styles



Theory of thermistor operation

Thermistors, like RTDs, are active measurement devices that operate by changing resistance with changes to ambient temperature. They consist of a metal oxide semiconductor pressed into a small bead, disk, wafer, or other container and coated with epoxy or glass. Since thermistors are constructed of semiconductor materials, they provide the best sensitivity of any measurement device and are ideal for measuring smaller temperature changes. They also generally have a much higher resistance than an RTD.

Basic Resistance Thermometer Components

Figure 9: Basic Resistance Thermometer Components


Unlike an RTD, a thermistor is typically a negative temperature coefficient device. This means its resistance decreases with an increase in temperature.

Thermistor types

As with any temperature sensor, an important factor to consider with thermistors is material composition and its subsequent impact on temperature range, sensitivity, accuracy, and so on. Sensor vendors may provide different metal-oxide compositions and/or casing materials that affect how the sensor can be mounted as well as chemical/abrasive resistance capabilities. Thermistors also come in a variety of physical configurations suited for different applications. Figure 10 shows some of these options.

          Thermistor Configurations

Figure 10: Thermistor Configurations


Unlike RTDs, thermistors rarely require a configuration other than two-wire because their resistance is several orders of magnitude greater than any lead-wire resistance that may be present. This means the impact of lead-wire resistance on measured resistance is minimal and often negligible.