Platinum resistance thermometers (Pt100/Pt1000)
Building blocks of the Pt100 sensor
Properties and sources of error
Wiring of Pt100 sensors
Platinum resistance thermometers (Pt100/Pt1000)
Resistance thermometers work on the principle that the resistance of a metal varies with temperature. When accurate laboratory measurements are required, only standard platinum resistance thermometers (SPRTs) are used.
There are two reasons for this:
Industrial platinum resistance thermometers are known as IPRTs, the most widely used of which is the Pt100, which has a resistance of 100 ohms at 0 °C.
Standard platinum resistance thermometers (SPRTs) are used for calibration under ITS-90. These are made of extremely pure platinum and therefore respond more faithfully to the laws of physics (thermodynamics) andare more predictable than any other type of temperature sensor.
The temperature scale is created using a number of fixed points - phase-transition points for different materials. The PRT serves as an inter-polation instrument for the intervals between the fixed points on a scale from approx. -259 °C (14 K) to 962 °C.
The SPRT has the following properties:
This results in an extremely accurate detector, although it is unfortunately too delicate for use outside the calibration laboratory.
The reference detector for calibration is the SPRT, which is a
IPRTs—for industrial use
The industrial platinum resistance thermometer (IPRT) is of more robust design than the SPRT and capable of withstanding greater physical stresses. As far as possible, it corresponds to the laboratory standard.
Nonetheless, industrial platinum probes always constitute a compromise in order to optimize certain required properties. Common requirements include:
When accuracy is crucial Pentronic usually uses detectors made by TDI-Isotech in Britain. The construction is as close to that of the SPRT as it is possible to get at present.
The mixing of different metals along the leads from the resistor to the sensor terminals can give rise to Seebeck voltages that cause reading errors.
Pentronic mainly uses the Pt100 detector,
Photo showing IPRTs. Above two thin film Pt100 detectors and below one wire wound device which measures diameter 2,8 x 25 mm.
Bobbin wound Pt100 detector in principle.
Film detectors (Pt100/Pt1000)
Film detectors first had their market in whiteware applications. Film detectors are made in automated processes and fit the measurement needs of household equipments very well.
Now these detectors are used in industrial applications where low cost takes priority over accuracy and high temperature measurements.
Be aware of the fact that film detectors have limited temperature ranges compared to the wire wound resistors. For example class A -30 to +300 °C for assembled sensors.
Film detectors have a platinum pattern fixed onto its surface that gives 100 or 1000 ohms resistance at 0 °C. They are known as Pt100 and Pt1000 respectively.
Building blocks of the Pt100 sensor
An industrial Pt100 sensor usually comprises three main components: a sensing element or detector, a protective tube and a connection. These can be combined in numerous ways to achieve the desired properties.
Pentronic stocks a wide range of Pt100 wire wounded detectors, (also called resistors) the models varying in size and tolerance class. The diameter of the resistor element ranges between 0.9 and 2.8 mm, and the length between 6 and 50 mm.
There are two principal sizes of industrial sensor for applications in which priority is given to accuracy and stability: 2.8-mm dia. x 25 mm and 1.5-mm dia. x 15 mm. If you need to specify a different length, we will ensure that the diameter is the size best suited to the protective tube being used.
Tolerances as per IEC 60751:2008
Industrial Pt100 detectors are divided into four tolerance
The new standard makes a difference between wire wound resistors and film resistors. Experience shows that film resistors can't handle as wide temperature ranges as wire wound resistors under the different tolerance classes.
The sensitivity of Pt1000 is ten times greater than the one of Pt100 in ohms/°C. Tolerances are 10 times greater as well in ohms. In degrees C both sensor's tolerances are equal.
Pt100 tolerances less than class AA
Most manufacturers employ closer tolerances, but
these apply to just one temperature: 0 °C. This means,
for instance, that a tolerance of ± 0.03 °C applying to
“1/10 DIN” (Class B/10) is valid only at 0 °C. For other temperatures, the tolerances will follow the respective slope of curve for Class A or B, depending on which materials have been used. See diagramme to the right.
Thus, the Class-B value cannot be divided by ten: it is impossible to achieve this tolerance curve using the IEC platinum alloy.
Pentronic uses wire-wound Pt100 detectors normally made of Class-A material. Unless we have been given specific instructions to the contrary, during final inspection of the temperature sensor we check the tolerance against the Class-B specification, so that other types of error are accommodated as well. Details of the measured values are included with the delivery. Results from the final inspection testing are available under Test certificates in this web site.
Film resistors are gradually used more often in industrial applications. The test limits are suited to the customer's requirements.
Closer tolerances at other temperatures can only be achieved through calibration that determines the specific properties of the individual sensor.
This is because the quality of platinum specified in IEC 60751 is reduced by the material being an alloy with added palladium - done for compliance with the traditional DIN standard. The alloy gives rise to departures from the ideal curve, necessitating a safety margin corresponding to the slope of curve A or, at worst, of curve B.
Some Japanese and American standards specify a platinum alloy of higher purity, which gives closer tolerances over a wider temperature range. However, these detectors generate a different output signal, which is not compatible with European instruments. For competetive reasons American and Japanese instrument manufacturers usually produce their indicators including the IEC 60751 scale.
Temperature-resistance relationship as per IEC 60751:2008
IEC 60751:2008 describes resistance as a function of the temperature of a Pt100 which relationship was established in 1983. See Tables and polynomials.The amendments made in 1995 have little significance to measurements made in practice. Pentronic will be pleased to explain the effect of these on calibration.
IEC 60751:2008 is not at all changed compared to the 1955 version concerning resistance as a function of temperature.
Tolerances for assembled Pt100 sensors as per IEC 60751:2008
Detectors usually need to be enclosed in a protective tube before being used. Tubes are of two types, depending on the temperature range. Pentronic uses seamless tubes of stainless, acidproof steel unless otherwise stated. Up to 250–300°C. A steel tube with PTFE or polyimid-insulated leads is generally used. The temperature
Detectors and tube diameters are chosen to obviate the formation of problematic air gaps. As regards largebore tubes, we usually prevent the formation of air gaps by packing the tube with metallic material.
This results in the detector being centrally located in the tube, giving it greater protection against vibration. The heat-transfer properties are also greatly enhanced, which means shorter response times and smaller reading errors caused by heat losses through the protective tube.
Other manufacturers pack the Pt100 sensing element in powder, which has inferior heat-transfer properties and can be dispersed by vibration, allowing air gaps to form below.
(1). Non-terminated wires
(2). Extension fitting (sleeve connector) with lead and optional
(3). Fitted connector.
Output signal connections
Sensors are equipped with one of the following types of connection:
Pt100 sensors can usually be used to measure temperatures up to approximately 250 °C. For higher temperatures, protective tubes, eg, with mineral-oxide insulation, are required.
There is little point in measuring temperatures in excess of about 600 C using a Pt100. This is because the resistance of the platinum wire is shunted unpredictably by the surrounding material. Special detectors are available for measuring temperatures up to approximately 700 °C. In this temperature range, type-N and type-K thermocouples constitute a good alternative.
(4). Terminal block.
(5). Mounted transmitter.
(6). For retrofitting terminal block or transmitter.
Properties and sources of error
All IPRTs exhibit hysteresis, ie, give different readings depending on whether the temperature is rising or falling. In 1982, D.J. Curtis at Rosemount investigated different designs.
He found that the best design was an expensive special sensor element, closely followed by the wire-wound Pt100 (the model used by Pentronic). Film-type thermometer detectors and bobbin-wound elements exhibited errors 5–10 times higher.
The following error values are percentages of the measuring range:
Theoretical example of hysteresis in Pt100 detector elements on cycling between low and high temperatures. The deviation (dT) depends on different expansion coefficients between platinum and supporting material which are more or less locked to each other physically.
Film and bobbin wound detectors follow graph A while the 80% free wire detector will be affected by a factor down to 0,1 less (graph B). SPRTs with very free wires are subject to hysteresis in the order of < 1 mK.
Platinum detectors are highly stable over time but flaws introduced in the design and during manufacture can adversely affect the properties.
Detectors need to be heat treated to homogenize the crystal structure and remove any oxides that may have formed.
Wires secured to the base are stressed during heating: the freer the wires, the smaller the drift of the sensor on temperature changes.
The resistance value changes, eg, if a kink is introduced during manufacture, if the sensor is subjected to impact or vibration, or if it cycles between high and low temperatures.
A typical drift value for a Pt100 detector is 0.05°C per annum.
High-quality detectors exhibit maximum drift of 0.01°C. If the temperature range is confined to 25–150°C, drift is as low as 0.005°C a year.
Because it takes longer for the components of a Pt100 detector
For the response time to be short, the sensor must have good thermal-conductivity properties and a low mass. Pentronic’s design is unique in that there is continuous metallic contact between the surrounding medium and the sensor. This gives a faster response, a smaller measuring error due to heat dissipation through the protective tube, and greater tolerance to vibration.
A Pt100 measures across the entire length of the wire. The temperature reading is therefore the average for the wire length.
It is important to remember this if you are measuring
A Pt100 measures across the entire length of the wire. However, some very small film detectors can make measuring locations almost as short as thermocouple ones possible.
Faults introduced during manufacture
Pt100 detectors are delicate instruments and need to be treated with the utmost care during manufacture. Impurities can create defects that cannot be detected during final inspection but manifest themselves only after some time in use.
If the tube to the sensor insert is moist or has traces of oil on it, it will adversely affect the insulation of the sensor. At room temperature, this will probably not matter but, at higher temperatures, the contaminants may be vaporized and able to penetrate to the platinum wires.
Shortcomings in the manufacture can also lead to the platinum wires being contaminated by other metals, eg, iron.
Detectors from the “1/10 DIN” range increase the influence of the lead connections. To make allowance for these and other variations, Pentronic usually permits a deviation of 0.05 °C during the final inspection of sensors at 0 °C, which corresponds to “1/6 DIN”.
The surface of platinum becomes oxidized. Since the sensor wires are thin (approximately 0.02 mm), the coating of oxide will have a measurable effect on the resistance.
In normal environments, the process is a slow one and the effects can usually be minimized through regular calibration.
Wiring of Pt100 sensors
Two-wire connections make life easy for the technician but at the expense of a much greater likelihood of measuring errors arising when long extension leads are used.
The problem results from the resistance of the lead being added straight into the measured value. At 20°C, the resistance of a 10-m, 2 x 0.25 mm2 copper lead is 1.4 ohms. In terms of temperature, this translates into a measuring error of 3.6 °C.
It is possible to eliminate the error through calibration, but every time the ambient temperature around the lead changes, the resistance will also change, producing a new error. A Pt1000 film detector would reduce the magnitude of the error to a tenth. If this is acceptable still there would then be a greater risk of self heating.
A three-wire connection will eliminate most of the effect of the lead resistance on the measured value, although this is conditional on all three wires having the same resistance. This is almost impossible to achieve in practice. In fact only two wires have to have equal resistances but as two of them are colour coded red you cannot easily know which is the critical one. This is the reason we say all three wires should hav equal resistances.
In extreme cases, there may be a combination of these factors at play. For example: a ten-metre length ofthree-wire lead that has a resistance difference across the wires of 10% will give a reading error of 0.18°C. However, a three-wire lead will solve the problem of the effect of the temperature range on the wires.
All instruments designed for optimum measuring accuracy have four-wire connections. The current and the signal are separated into two circuits, which renders the unbalance in the resistances of the wires insignificant.
This is true provided that the difference is not of an excessive magnitude - up to 100 ohms is often acceptable with modern instruments. The four-wire system was previously confined to
Colour coding for extension lead to two-, three- and four-wire single Pt100 as per IEC 60751. One side of the Pt100 detector is connected to red wires while the other side is connected to white ones.
Because the former IEC standard offered no recommendations of colour code for dual Pt100s, Pentronic employs its own standard (above) whenever possible. The extra circuits are connected to blue and yellow wires. The IEC:2008 recommends black or grey colour where we chose blue.
The IEC recommends that different colour codes be used for Pt100 leads, in order to ease connection work. Pentronic uses the recommended codes whenever possible. See the picture and comments in the right-hand column.
Depending on market demand probably cable manufacturers will follow the IEC colour recommendations sooner or later.
We can also supply cable of the same colour code for additional runs. See illustrations in the righthand column.
What the Pt100 indicator really measures
Resistances included in the measured value for different
connections. Calibration of 2-wire and 3-wire connections
assumes that the lead resistances and the differences in the
resistances in the current winding are known.
|2-wire||Sensor + (R1 + R2)||All lead resistance included|
|3-wire||Sensor + (R1 - R2)||Difference in lead resistance included|
|4-wire||Sensor||Unaffected by lead resistance|
Synopsis of error sources
Typical error sources and ranges for the Pt100 detector.
The table does not include any allowance for environmental
effects, such as contamination by metal migration, platinum
oxides and the like.
|Source of error||Error contribution interval °C|
|Sensor construction/installation||0,1 - 3|
|Pt100 sensor||Tolerance||0,03 - 0,3 (at 0 °C)|
|Lead configuration||2 wire||0,1 - 5|
|3 wire||0,01 - 0,5|
|Instrumentation||0,02 - 3|
Comparison - thermocouple vs Pt100
The following table describes briefly the main differences between thermocouples and Pt100s in general terms:
|Measuring range||Large: -200 < T < 1000 + °C||Limited: -200 < T < 600 °C|
|Stability||Not good, especially in high temperature||Excellent. Annual drift < 0.01 °C|
|Measuring location||From hot junction to reference location||Across entire Pt-wire length|
|Ageing||Significant in high temperature||Insignificant, see Stability.|
|Response time||Very short possible < 1 s||Not as short as corresponding TC|
|Excitation power||None||Significant < 1 mW|
|Physical strength||Very good||Limited|
|Pricing||Slightly lower than corresponding Pt100s||Slightly higher than corresponding TC|