1. How do RTD resistance thermometers work?
RTDs are a type of temperature sensitive resistive device in which the resistive element is made of a pure metal. This sets them apart from thermistors, which contain a semiconductor resistive element and achieve higher accuracy but have a very limited measurement range. The RTD resistive element is made of either platinum, copper or nickel (or occasionally tungsten). The choice of metal, discussed in more detail within the purchasing tips, effects cost, temperature measurement range, corrosion resistance, linearity and sensitivity. The sensors are designated according to their resistance at 0 °C (because this temperature can be recreated accurately). A platinum element resistor with a reference resistance of 100 ohms at a reference temperature of 0 °C is known as a PT100 sensor and is the most common type of RTD. Other commonly available Platinum RTDs include PT50, PT200, PT500 and PT1000. Copper and Nickle RTDs are designated by the abbreviations Cu and N e.g. Cu100 and Ni100.
RTD sensors are available in one of three geometrical forms, wire wound, thin film and coiled element. In wire wound RTDs, a length of metal resistance wire is wound around an insulating glass or ceramic core and placed inside a protective metal sheath. Wire wound RTDs are extremely robust but incur long response times. Furthermore, strain resulting from thermal expansion or contraction effects wire resistance, thereby reducing accuracy. Thin film RTDs (also known as surface mount RTDs) consist of a resistive metal film printed onto a ceramic of flexible polymer substrate. They have fast response times but are subject to errors due to thermal expansion much like wire wound RTDs. Coiled element RTDs consist of a coiled resistive wire similar to wire wound RTDs but without an insulating core. Instead, the resistive coil is placed into a metal sheath and packed with sand. This configuration prevents the occurrence of thermally induced strain and therefore improves accuracy. Coiled element RTDs are delicate and therefore used predominantly in high accuracy laboratory applications.
To measure RTD resistance, we could place it in series with a fixed resistor of known resistance and apply a known voltage. Such a circuit is known as a voltage divider and causes the potential difference across the RTD to shift as a function of its resistance. This enables resistance to be calculated from the measured voltage. In practice a Wheatstone bridge circuit is the preferred method for detecting the resistance of a RTD. The Wheatstone bridge performs a similar function to a voltage divider but provides improved sensitivity. RTD resistance thermometers are available with either 2, 3 or 4 wires. When using the 2 wire configuration, the resistance of the lead wires is included in the resistance measurement. This introduces measurement error (which can be cancelled through post-processing), particularly when long lead wires or low resistance RTDs are used. The 3 wire configuration enables lead wire resistance to be subtracted from the resistance reading by the Wheatstone bridge. This improves accuracy, but only works when all 3 leads are of identical resistance. In contrast, 4 wire RTDs enable subtraction of the lead wire resistance irrespective of their resistances but are less commonly used because 3 wire RTDs are sufficient for the vast majority of applications.
The equations for calculating RTD resistance of 2 wire (with cancellation of lead wire resistance) and 3 wire RTDs are presented in the illustration above. Once resistance is known, the equation below is used to determine temperature. α, the temperature coefficient, is the change in resistance per unit temperature, i.e. the sensitivity of the sensor. The equation assumes that the RTD resistance is a linear function of temperature. The assumption of linearity is very accurate at temperatures between 0 – 200 °C. Outside of this range, if high accuracy is desired, then temperature can be calculated using the Callendar Van Dusen equation (Platinum RTDs only), which accounts for non-linearity. In practice, most post processing electronics will perform the calculations automatically and so the equations are provided here for academic benefit.
2. General characteristics of RTD resistance thermometers
RTDs are reasonably low cost, high accuracy temperature sensors, used across almost every industry. They have a large temperature measurement range (up to -200 to 850 °C, depending on RTD type) and high long term stability. They are active sensors, meaning that they require an external voltage supply and are therefore susceptible to self-heating. Their output is in Volts, from which the RTD resistance and therefore the temperature can be determined. They are available in three geometries; wire wound, thin film or coiled. Furthermore, they are available with different electrical insulation and protective braiding materials, the choice of which effects their allowable operating temperature.
3. Input and output signals
RTD resistance thermometers are active sensors and therefore require a supply voltage for their Wheatstone bridge circuit. The Wheatstone bridge outputs a voltage as a function of RTD resistance, from which temperature can be determined. RTD resistance thermometers rarely contain built in electronics and therefore require external electronics to implement the Wheatstone bridge. The external electronics will also amplify the bridge output and convert it into a standard output signal (e.g. 0-10 V) that is proportional to temperature.
4. Applications of RTD resistance thermometers
RTDs are popular temperature sensors, used across most industries, with the exception of application requiring either extremely high or cryogenic temperatures. Most of their applications overlap with those of thermocouples e.g. within the process manufacturing industry. RTD sensors are generally favored where high accuracy is required, and a fast response time is not crucial. Whilst most of their applications are for measuring the temperature of liquid or gas, surface mount RTDs are popular for measuring surface temperature e.g. of a heating vat.
5. Typical specification
|Cost||Low (depends on metal)|
|Measurement range||See chart below|
|Accuracy||0.1-1°C for class A, 0.2-2°C for class B*|
|Long term stability||High|
|Supply voltage||5-20 V|
|Passive / active||Active|
|Contact / non-contact||Contact|
|Sheath diameter||3-8 mm typical|
|Sheath length||25-1,000 mm typical|
*Absolute accuracy is lower away from the reference temperature. Furthermore, there are two common accuracy classes, A and B. For Platinum RTDs these are defined in IEC 60751.
A comparison of metal types
6. Purchasing tips
- Metal type: The RTD resistive element is made of either platinum, copper or nickel. The choice of metal effects cost, temperature range, corrosion resistance, linearity and sensitivity, as shown in the table above. Platinum is by far the most popular choice due to its large measurement range, excellent corrosion resistance and long term stability. Nickle and copper are used where cost is a significant consideration.
- Reference resistance: The reference resistance is the RTD resistance at the reference temperature, normally 0 °C. Common resistances include 50, 100, 200, 500 and 1000 ohms, of which 100 ohms is the most common. Higher resistance RTDs are less sensitive to errors due to lead resistance and less prone to self-heating because the higher resistance reduces current. However, the reduced current makes them more sensitive to electrical noise.
- Number of wires: RTDs are available with 2, 3 or 4 wires. The difference between these is in the ability to cancel out the resistance of the lead wires, as explained in section 1.
- Response time: Due to their thermal mass, RTDs require a finite time to arrive at the temperature of the fluid. Thin film RTDs have shorter response times than wire wound RTDs. However, both have longer response times than thin wire, exposed thermocouples. Response time depends not only on the sensor but also the fluid’s thermal conductivity and velocity e.g. response times in water are shorter than in air. Response times are defined in terms of time constants. The time constant is the time taken for the temperature difference between the thermocouple and the fluid to decay by 63.2% (e.g. from 100 °C to 46.8 °C). The 2nd time constant is the time taken for temperature difference to decay by a further 63.2%, i.e. from 63.2% to 86.5%. As such, the time taken for the temperature difference to decay from 0% to 63.2% is equal to the time taken from 63.2% to 86.5%. The first 5 time constants are illustrated in the graph below, where 0 represents the initial sensor temperature and 100% represents the fluid/process temperature.
- Sheath length and diameter: RTD resistance thermometer sheaths are typically 3-8 mm in diameter and 25-1,000 mm in length, though other sizes are available.
7. Advantages of RTD resistance thermometers
RTD resistance thermometers:
- Are low cost temperature sensors, particularly if copper or nickel wires are used.
- Are of high accuracy, particularly when compared to thermocouples.
- Have a higher output stability than thermocouples.
- Have very good long term stability.
8. Disadvantages of RTD resistance thermometers
RTD resistance thermometers:
- Have a smaller measurement range than thermocouples.
- May suffer from self-heating, which reduces accuracy.
- Are not capable of achieving extremely fast response times.
9. Application tips
- Calibration for improved accuracy: The Accuracy of RTD resistance thermometers can be improved by calibrating the RTD every 12 months, to account for long term drift.
- Avoid thermal shunting: Thermal shunting occurs when the presence of the RTD effects the measured temperature by absorbing heat. The susceptibility to thermal shunting is application dependent. Thermal shunting can be minimized by using smaller diameter sheaths.
- Self-heating: The current passing through the RTD generates heat which can increase the temperature of the RTD, causing measurement error. Self-heating can be reduced by using a high resistance RTD so that the current is limited. The heating power is equal to the RTD resistance (in ohms) multiplied by the current squared (in amps) and has the units of Watts. Unfortunately, there is no simple method to calculate the temperature rise from the dissipated heat. Some manufacturers provide the self-heating temperature rise in still air. The self-heating effect in water should be negligible because of the improved thermal conductivity.
- Thermowells: A thermowell is type of protective sheath which performs the same function as a regular RTD sheath but is not permanently attached to the sensor. It enables the RTD to be removed from a pipe or vessel without any down time or leakage of fluid because the thermowell remains in place providing a reliable fluid tight seal.