Types of RTDs/resistance thermometers

For the sake of understanding, the temperature-dependent change in resistance can be approximated as a linear equation for restricted temperature ranges:

            ∆R = k(T) ⋅ ∆T

Resistance thermometers are divided into PTC and NTC sensors:

• PTC: factor k is positive, with increasing temperature the electrical resistance increases.
• NTC: factor is negative, as the temperature drops, the electrical resistance increases.

Resistance thermometers exist in the various accuracy classes AA, A, B and C, each of which has its own limit deviations and validity ranges, with class AA having the highest measuring accuracy. The various accuracy classes, including the temperature-related tolerances, can be seen in the following table.

The accuracy class of a sensor, should be selected based on the desired target accuracy of the measurement. Resistance thermometers can be made of different temperature sensitive materials. The choice of material and the dimensioning of the sensor determine the possible resistance value and also the temperature range for which the sensor is suitable. Thus, resistance thermometers can be specially designed for different temperature and resistance ranges. KTY resistance sensors, for example, are used as a low-cost alternative to platinum sensors. However, due to the wide tolerance range of 1% - 5%, KTY sensors are more commonly used in applications that do not require precise measurement. Resistance thermometers made of nickel are also suitable for detecting smaller temperature differences. Ni sensors, due to their higher sensitivity than platinum sensors, achieve a greater relative change in resistance for the same temperature difference.

For classification in the following table typical materials and PTC/NTC properties:

The user must check whether a sensor is suitable for the intended measuring purposes and measuring devices. Factors to consider here include:

• Temperature range: is the sensor suitable for the intended temperature measuring range?
• Resistance measuring range: can the sensor resistance be measured in the intended temperature measuring range?
• Characteristic curve: how can the sensor resistance be converted into a temperature value? (if this is not done automatically via the measuring device, a vendor-specific sensor characteristic curve or table is required)
• Sensitivity/slope: how large is the change in resistance for an available change in temperature (ideally, the largest possible resistance range) - and how large is the digital sensitivity of the measuring device in digits/Ω.
• Noise: both sensor and measuring device introduce a noise component into the measurement, which becomes visible as temperature noise depending on sensor and measuring device sensitivity.
• Velocity: how often is the sensor resistance measured?

Sensor exchange Please note that 1:1 exchangeability is not always guaranteed, especially in the case of manufacturer-specified sensors. If necessary the new sensor must be recalibrated in the system.

Resistance measurement To determine the resistance, it is common to let a small measuring current I in the mA range (I < 5 mA) flow through the sensor and measure the resulting voltage. Three effects must be taken into consideration when doing this: The measuring current can lead to self-heating of the sensor. However, this usually has only a minimal effect on the measuring accuracy. But this can play a significant role in high-precision measurements; see the comments in the following section: "Self-heating of RTD sensors". The sensor supply lines also have a resistance and add a (usually) constant additional resistance to the measurement. Compensation can be made by ◦ 3- or 4-wire connection of the sensor, ◦ manual consideration of the known line resistance in the calculation, or ◦ use of a sensor with higher nominal resistance - then the line effects are less significant. Insulation faults or thermovoltages can affect the measurement.