RTD Basics—An Introduction to Resistance Temperature Detectors

Resistance temperature detectors, or RTDs, are perhaps the simplest type of temperature sensor. These devices work on the principle that the electrical resistance of a metal changes with temperature. Pure metals typically have a positive temperature coefficient of resistance, meaning that their resistance increases with temperature. RTDs can operate over a large temperature range from -200 °C to +850 °C, as well as offer high accuracy, excellent long-term stability, and repeatability.

In this article, we'll go over the trade-offs of using RTDs, the metals used in them, two types of RTDs, and how RTDs compare to thermocouples. 

Before diving in too far, let's take a look at an example application diagram to better understand RTD basics. 

 

Example RTD Application Diagram

An RTD is a passive device and doesn’t produce an output signal on its own. Figure 1 shows a simplified RTD application diagram.

 

Figure 1. Example RTD application diagram. Image used courtesy of TI

 

An excitation current, I1, is passed through the temperature-dependent resistance of the sensor. This produces a voltage signal that is proportional to both the excitation current and the RTD resistance. The voltage across the RTD is then amplified and delivered to the ADC (analog-to-digital converter) to produce a digital output code that can be used to calculate the RTD temperature.

 

Trade-offs for Using RTD Sensors—RTD Sensor Pros and Cons

Before delving too far, it's important to note that the details of RTD signal conditioning will be addressed in a future article. As for this article, I’d like to highlight some basic trade-offs when working with RTD circuits.

First off, note that the excitation current is normally limited to less than 1 mA or so to minimize the self-heating effect. As the excitation current flows through the RTD, it produces I²R or Joule heating. The self-heating effect can raise the sensor temperature to a value higher than that of its surrounding environment, which is actually being measured. Reducing the excitation current can reduce the self-heating effect. It’s also worth mentioning that the self-heating effect depends on the medium in which the RTD is immersed. For example, the self-heating effect can be more pronounced for an RTD placed in still air than for an RTD immersed in moving water.

For a given minimum detectable change in temperature, the change in the RTD voltage should be large enough to overcome the system noise as well as the offset and drift of different system parameters. With the excitation current limited due to the self-heating effect, we need to use an RTD with sufficiently large resistance, so a relatively large voltage is produced for the downstream signal processing blocks. While a larger RTD resistance is desired to reduce the measurement error, we cannot increase the resistance arbitrarily, as a large RTD resistance can lead to a slower response time.

 

RTD Metals: Differences Between a Platinum RTD, Gold RTD, and Copper RTD

In theory, any kind of metal can be used to construct RTDs. The first-ever RTD, which was invented by C. W. Siemens in 1860, used a piece of copper wire. However, Siemens soon discovered that platinum RTDs produced more accurate results over a wider range of temperatures.

Today, platinum RTDs are the most widely used temperature sensor in precision thermometry. Platinum offers a linear resistance-temperature relationship and high repeatability over a large temperature range. Additionally, platinum does not react with most contaminant gasses in the air.

Aside from platinum, two other common RTD materials are nickel and copper. Table 1 provides the temperature coefficient and relative conductivity of some common RTD metals.

 

Table 1. Temperature coefficients and relative conductivity of common RTD metals. Data used courtesy of BAPI

 

In the last section, we discussed that a larger RTD resistance could reduce measurement error. When compared to platinum and nickel, copper has higher conductivity (or, equivalently, a lower resistance). For a given sensor size and excitation current, a copper RTD can produce a relatively smaller voltage. As a result, it can be more challenging for a copper RTD to measure small temperature changes. Besides, copper oxidizes at higher temperatures and is also limited to a measurement range of -200 to +260 °C. Despite these limitations, copper is still used in some applications because of its linearity and low cost. As depicted below in Figure 2, copper has the most linear resistance-temperature characteristic among the three common RTD metals.

 

Figure 2. Resistance vs temperature characteristics of nickel, copper, and platinum RTDs. Image used courtesy of TE Connectivity

 

Gold and silver also have relatively low resistance and are rarely used as RTD elements. Nickel has a conductivity close to that of platinum. As you can see from Figure 2, nickel provides the largest change in resistance for a given change in temperature.

However, compared with platinum, nickel offers a lower temperature range, a greater amount of nonlinearity, and a larger long-term drift. Besides, the resistance of nickel varies considerably from batch to batch. Due to these limitations, nickel is mainly used in low-cost applications such as consumer goods.

Common platinum RTDs are Pt100 and Pt1000. These designations describe the metal type used (platinum or Pt) in the sensor construction and the nominal resistance at 0 °C, which is 100 Ω and 1000 Ω for Pt100 and Pt1000 types, respectively. The Pt100 type used to be more popular; however, today, the trend is toward higher-resistance RTDs because a higher resistance can provide higher sensitivity and resolution at little or no extra cost. RTDs made of copper and nickel also use a similar naming convention. Table 2 lists some common types.

 

Table 2. RTD types, materials, and temperature ranges. Data used courtesy of Analog Devices

 

In addition to the type of metal used, the mechanical construction of the RTD can also affect the sensor performance. RTDs can be categorized into two basic types: thin film and wire-wound types. These two types are discussed in the following sections.

 

Thin Film vs Wire-wound RTD

To further our conversation on RTDs, let's explore two types: thin film and wire-wound. 

 

Thin Film RTD Basics

The structure of the thin film type is shown in Figure 3(a).  

 

Figure 3. Example thin film RTDs where (a) shows the structure and (b) shows different overall types. Image (modified) used courtesy of Evosensors

 

In a thin film RTD, a thin layer of platinum is deposited onto a ceramic substrate. This is followed by very high-temperature annealing and stabilization, as well as a thin protective glass layer to cover the entire element. The trimming area shown in Figure 3(a) is used to adjust the fabricated resistance to the specified target value.

The thin film RTD relies on relatively newer technology and enables a substantial reduction in assembly time and production cost. Compared to the wire-wound type, which we'll go into depth on in the next section, thin film RTDs are more resistant to damage from shock or vibration. In addition, thin film RTDs can accommodate a larger resistance in a relatively smaller area. For example, a 1.6 mm ✕ 2.6 mm sensor provides enough area to create a 1000 Ω resistance. Due to their small size, thin film RTDs can quickly respond to changes in temperature. These devices suit many general-purpose applications. The disadvantages of this type are relatively poor long-term stability and a narrower temperature range.

 

Wire-wound RTDs

Figure 4 below shows the construction of a basic wire-wound RTD.

 

Figure 4. Overview structure of a basic wire-wound RTD. Image used courtesy of PR Electronics

 

This type of RTD is built by winding a length of platinum around a ceramic or glass core. The entire element is normally packaged inside a ceramic or glass tube for protection purposes. RTDs with a ceramic core is suitable for measuring very high temperatures. Wire-wound RTDs are typically more accurate than the thin film type. However, they are more expensive and more likely to be damaged by vibrations.

To minimize any strain on the platinum wire, the coefficient of thermal expansion of the materials employed in the sensor structure should match that of platinum. Identical coefficients of thermal expansion minimize the long-term stress-induced change in the resistance of the RTD element and consequently improve the repeatability and stability of the sensor.

 

RTD vs Thermocouple Attributes

To wrap up this conversation on RTD temperature sensors, here’s a short comparison between RTDs and thermocouple sensors.

A thermocouple produces a voltage that is proportional to the temperature difference between its two junctions. While a thermocouple is self-powered and doesn’t need external excitation, RTD-based temperature measurement requires an excitation current or voltage. The thermocouple output specifies the temperature difference between the cold and hot junctions, and thus, cold junction compensation is required in thermocouple applications. On the other hand, cold junction compensation is not required in RTD applications, which leads to a simpler measurement system.

While thermocouples are normally used in the -184 °C to 2300 °C range, RTDs can measure the -200 °C to +850 °C range. Though RTDs are typically more accurate than thermocouples, they are, however, about two or three times more expensive than thermocouples. Another difference is that RTDs are more linear than thermocouples and exhibit excellent long-term stability. With thermocouples, chemical changes in the sensor material reduce the long-term stability and lead to a drift in the sensor readings.

 

---

 

To see a complete list of my articles, please visit this page.