Measuring Temperature with RTDs

The dataTaker data loggers support several types of platinum, nickel and copper Resistance Temperature Detectors (RTDs).

These devices provide the most accurate and stable method for measuring temperature with the dataTaker data loggers.

The RTD element is a resistive device which changes resistance in an accurately known way with changes in temperature.

The dataTaker accurately measures the changes in resistance of the element, and linearizes the resistance to units of temperature.

RTDs have a defined resistance at 0.0 Deg C, which varies between manufacturers and types. The most common are PT100 sensors, which are Platinum RTDs that have a resistance of 100 Ohms at 0.0 Deg C. The dataTaker supports all RTDs types which have a resistance at 0.0 Deg C of 10 Ohms to 2000 Ohms.

The linearization errors for RTDs by the dataTaker are less than 0.1 Deg C over the range of -200 Deg C to +600 Deg C.

In practice the accuracy of temperature measurement using RTDs is limited by sensor imperfections such as impurities in the element, physical stress in the element, etc.

RTD Support by the dataTaker

The dataTaker measures RTDs directly as resistive devices. The resistance of the RTD element is measured by any of the 2 wire, 3 wire or 4 wire resistance measurement methods (See Section II ñ Measuring Resistance).

The 4 wire resistance measurement method for RTDs provides the highest accuracy, since this method fully compensates for cable wire resistances. However the 3 wire method also provides high accuracy.

The resistance of the RTD is measured, and the ratio of this resistance to the specified resistance at 0.0 Deg C is calculated. The ratio is then linearized to units of temperature using the standard RTD calibration tables.

Platinum RTD Types

The dataTaker supports two platinum RTD calibration standards, which are specified by the analog input type identifiers as follows

a=0.003850 Ohm/Ohm/Deg C dataTaker channel type PT385

a=0.003916 Ohm/Ohm/Deg C dataTaker channel type PT392

These platinum RTD temperature coefficients comply with the following standards

a=0.003850 Ohm/Ohm/Deg C DIN43760-1980, IEC751-1983

a=0.003916 Ohm/Ohm/Deg C US Standard JIS C1604-1981

The platinum RTD types supported by the dataTaker are all assumed to have a sensor resistance at 0.0 Deg C of 100.0 Ohms.

However Platinum RTDs with resistances of other than 100.0 Ohms at 0.0 Deg C can also be used. The zero temperature resistance of these RTDs must be declared as a channel option in the channel specification.

Using DeTransfer, the commands for example

1PT385
2PT385(120.0)
5PT392(200.0)

specifies three different types of platinum RTD sensor as follows

a platinum RTD sensor with a coefficient of a=0.003850 Ohm/Ohm/Deg C,
and a resistance of 100.0 Ohms at 0.0 Deg C (default)

a platinum RTD sensor with a coefficient of a=0.003850 Ohm/Ohm/Deg C,
and a resistance of 120.0 Ohms at 0.0 Deg C

a platinum RTD sensor with a coefficient of a=0.003916 Ohm/Ohm/Deg C,
and a resistance of 200.0 Ohms at 0.0 Deg C

Any number and type of RTD sensors which have different resistances at 0.0 Deg C can be used in the same application.

Using DeLogger, the zero temperature resistance for RTDs is entered into the RTD Wiring Configuration when the program is created in the Program Builder.

If the resistance at 0.0 Deg C for the platinum RTD sensors is not specified, then the default of 100.0 Ohms is applied. This is equivalent to

2PT385(100.0)
5PT392(100.0)

The resistance for the RTD sensor at 0.0 Deg C is used to calculate the ratio of the measured sensor resistance to the zero temperature resistance (Ohm/Ohm/Deg C).

Nickel RTD Types

The dataTaker also supports nickel RTDs, which have a temperature coefficient of 0.005001 Ohm/Ohm/Deg C.

The nickel RTD types supported by the dataTaker are all assumed to have a sensor resistance at 0.0 Deg C of 1000.0 Ohms.

However Nickel RTDs with resistances of other than 1000.0 Ohms at 0.0 Deg C can also be used. The zero temperature resistance of the RTD must be declared as a channel option in the channel specification.

Using DeTransfer, the commands for example

1NI
3NI(1200.0)
6NI(1500.0)

specifies three different types of nickel RTD sensor as follows

a nickel RTD sensor with a resistance of 1000.0 Ohms at 0.0 Deg C (default)

a nickel RTD sensor with a resistance of 1200.0 Ohms at 0.0 Deg C

a nickel RTD sensor with a resistance of 1500.0 Ohms at 0.0 Deg C

Any number and type of RTD sensors which have different resistances at 0.0 Deg C can be used in the same application.

Using DeLogger, the zero temperature resistance for RTDs is entered into the RTD Wiring Configuration when the program is created in the Program Builder.

If the resistance at 0.0 Deg C for the nickel RTD sensors is not specified, then the default of 1000.0 Ohms is applied. This is equivalent to

NI(1000.0)

The resistance for the RTD sensor at 0.0 Deg C is used to calculate the ratio of the measured sensor resistance to the zero temperature resistance (Ohm/Ohm/Deg C).

Copper RTD Types

The dataTaker also supports copper RTDs, which have a temperature coefficient of 0.0039 Ohm/Ohm/Deg C.

The copper RTD types supported by the dataTaker are all assumed to have a sensor resistance at 0.0 Deg C of 100.0 Ohms.

Copper RTDs with resistances of other than 100.0 Ohms at 0.0 Deg C can also be used. The zero temperature resistance of the RTD must be declared as a channel option in the channel specification.

The dataTaker support for copper RTDs allows the temperature of copper coils such as those in relays, solenoids, vibrating wire strain gauges, etc. to be measured. The resistance of the coils is declared as the channel option.

Using DeTransfer, the commands for example

1CU
3CU(120.0)
7CU(150.5)

specifies three different types of copper RTD sensor as follows

a copper RTD sensor with a resistance of 100.0 Ohms at 0.0 Deg C (default)

a copper RTD sensor with a resistance of 120.0 Ohms at 0.0 Deg C

a copper RTD sensor with a resistance of 150.5 Ohms at 0.0 Deg C

Any number and type of RTD sensors which have different resistances at 0.0 Deg C can be used in the same application.

Using DeLogger, the zero temperature resistance for RTDs is entered into the RTD Wiring Configuration when the program is created in the Program Builder.

If the resistance at 0.0 Deg C for the copper RTD sensors is not specified, then the default of 100.0 Ohms is applied. This is equivalent to

3CU(100.0)

The resistance for the RTD sensor at 0.0 Deg C is used to calculate the ratio of the measured sensor resistance to the zero temperature resistance (Ohm/Ohm/Deg C).

Excitation Current

The dataTaker outputs a precise constant current from the Excite terminal of the analog channel during measurement of the RTD.

This current is turned on when the channel is selected, and remains on until after the analog to digital conversion is completed. This is usually for a period of 30 mS, and so there is little self heating of the RTD.

The RTD temperature sensors are connected between the Excite terminal of the analog input channels and the Analog Return or Ground terminals, such that this excitation current passes through the sensor element.

The voltage produced across the RTD sensor element by the excitation current is connected by various methods to the analog input channels. The measured voltage drop is then used to calculate the value of the resistance of the RTD element.

The dataTaker can output either a 250.0 µA or a 2.500 mA excitation current from the Excite terminal (See Section II ñ Measuring Resistance).

The 2.500 mA excitation current is the most appropriate for RTD measurement, although the 250.0 µA excitation current can be used if desired.

While the temperature resolution obtained for the two currents is generally the same, the higher current is preferred because any input noise which may be present is reduced. However the trade-off is self heating, because the 30 mS pulse of the higher current during measurement can heat the sensing element.

If the non default 250.0 µA excitation current is to be used for RTD measurement, then this is specified as part of the channel specification in the schedule lists.

Using DeTransfer, the commands for example

3PT385(I)

specifies that the platinum RTD sensor connected to analog channel 3 is to be measured using a 250.0 µA (channel option I) excitation current from the Excite terminal.

Using DeLogger, click Channel Options:Excitation to open the Channel Properties dialog, and select the 250 µA radio button under the Excitation tab.

The 2.500 mA excitation current is selected by default when an RTD input type is specified, and does not need to be specifically selected.

Connection and Measurement of RTDs

RTDs can be connected to the analog channels in three different ways, which provide varying degrees of compensation for cable wire resistance. The connection methods are the 4 wire, 3 wire and 2 wire methods of resistance measurement, and are described in the following sections.

These RTD connection methods use all four terminals of the analog input channels. Therefore a total of 5 RTDs can be monitored by the dataTaker 50, and a total of 10 RTDs can be monitored by the dataTaker 500/600 series loggers and the Channel Expansion Module (CEM-AD).

A fourth RTD measurement method is also supported, where the RTDs are connected as single ended inputs to the analog input channels. The single ended input methods do not perform cable wire compensation, but allow up to twice as many RTDs to be measured.

Four Wire RTD Measurement

The 4 wire method of RTD measurement is the most accurate, since there is no current flowing in the signal cable wires.

In this method two cable wires are connected to each end of the RTD. One pair of cable wires carries the excitation current, the other pair of cable wires is used to measure the voltage across the RTD.

The 4 wire method provides accurate compensation for cable wire resistance, and should be used where cables are long or the individual cable wires are of unequal resistance

RTDs are directly connected to the analog input channels for 4 wire resistance measurement as follows

Figure 77 ñ Connection for Four Wire RTD Measurement

During RTD resistance measurement, the excitation current output from the Excite terminal passes through the RTD and returns into the Analog Return terminal.

The voltage produced across the RTD sensor is measured differentially between the +ve and ñve terminals.

The excitation current flows in the excitation circuit (shown by the red path), which is totally separate from the measurement circuit. Because of the high input impedance of the analog to digital converter, there is negligible current flow in the measurement circuit. Therefore there is a negligible cable resistance component in the voltage drop measured across the RTD.

Each RTD measured by the four wire method requires all terminals of the analog input channel.

Therefore a maximum of 5 RTDs can be measured using this method by the dataTaker 50, and a maximum of 10 RTDs can be measured using this method by the dataTaker 500/600 series loggers and the Channel Expansion Module.

RTDs connected to the analog input channels using the 4 wire method are sampled and the data is returned when a Schedule containing the channel is executed.

Using DeTransfer, the command for example

BEGIN
RA10M
1PT385(4W) 3NI(1200.0,4W)
END

instructs the dataTaker to measure the RTDs connected to analog input channel 1 and channel 3.

The excitation current channel option is not specified, and so the default 2.500 mA excitation current is used.

The platinum RTD connected to channel 1 is assumed to have a resistance of 100.0 Ohms at 0.0 Deg C. The nickel RTD connected to channel 3 has a declared resistance of 1200.0 Ohms at 0.0 Deg C

The PT385 indicates that the signal applied to these channels is from a platinum RTD, with a temperature coefficient of a=0.003850 Ohm/Ohm/Deg C. The resistance is linearized.

The NI indicates that the signal applied to this channels is from an nickel RTD. The resistance is linearized.

The 4W channel option indicates that the 4 wire method of resistance measurement is to be used.

Using DeLogger, RTDs connected by the 4 wire method can be measured by the following Program Builder program. The 4 wire connections are selected from the RTD Wiring Configurations dialog which opens when you have selected the analog input channel.

The data is returned in units of temperature.

The dataTaker will read the inputs every 10 minutes, and readings are stopped by entering a H (Halt) command.

Three Wire RTD Measurement

Most RTD measurements are made using the 3 wire resistance measurement method, which compensates for cable wire resistances in the measurement circuit (assuming that the two current carrying cable wires are of the same resistance).

In this method of measurement, one cable wire is connected to one end of the RTD, and two cable wires are connected to the other end.

The two cable wires connected to the lower end of the RTD are used to compensate the measured resistance for cable wire resistance.

RTDs are connected to the analog input channels for 3 wire measurement as follows

Figure 78 ñ Connection for Three Wire RTD Measurement

During RTD resistance measurement, the excitation current from the Excite terminal of the analog channel passes through the RTD and returns into the Analog Return terminal.

The voltage produced across the RTD sensor and the cable wire resistance is measured differentially between the +ve and ñve terminals.

Since all cable wires have resistance, the voltage produced across these is included in the total voltage measured. This introduces resistance offset errors, especially when long cables are used.

There is a much higher input impedance for the ñve input terminal than for the Analog Return terminal. This causes the excitation current to return in the cable wire connected to Analog Return. This ensures that there is no appreciable current flowing in the third cable wire which is connected to the ñve terminal.

The third cable wire is used to accurately measure the voltage due to the resistance of the return cable in the resistance measurement circuit. The dataTaker doubles this voltage, and subtracts it from the total voltage drop measured across the RTD and the cable wires.

This technique assumes that the two current carrying cable wires are of equal resistance. This in turn implies that cables used to connect RTDs to the dataTaker should be of equal length, gauge and type of wire.

Up to 5 RTDs can be measured using the 3 wire method by the dataTaker 50, and up to 10 RTDs can be measured by the dataTaker 500/600 series loggers and the Channel Expansion Module (CEM-AD).

RTDs connected using the 3 wire method are sampled and the data is returned when a Schedule containing the channel is executed.

Using DeTransfer, the command for example

BEGIN
RA10M
1..2PT392
END

instructs the dataTaker to measure the two platinum RTDs connected to analog channels 1 and 2.

The excitation current channel option is not specified, and so the default 2.500 mA excitation current is used. The platinum RTDs are all assumed to have a resistance of 100.0 Ohms at 0.0 Deg C.

The PT392 specifies that the signal connected to these channels is from a platinum RTD, with a temperature coefficient of a=0.003916 Ohm/Ohm/Deg C. The resistance is linearized.

Using DeLogger, RTDs connected by the 3 wire method can be measured by the following Program Builder program.

The 3 wire connections are selected from the RTD Wiring Configurations dialog which opens when you have selected the analog input channel.

The data is returned in units of temperature.

The dataTaker will read the inputs every 10 minutes, and readings are stopped by entering a H (Halt) command.

Two Wire RTD Measurement

The 2 wire RTD measurement method is the simplest method for measuring RTDs with the dataTaker. This method only requires a single pair cable to connect each RTD to the logger. The cable pair carries both the excitation current and signal voltage.

The 2 wire RTD measurement method has the advantage of less cabling, which may be a consideration in some applications.

However this method also has the disadvantage of including the cable wire resistances in the measured resistance, which can produce significant errors. Measurement errors will also be produced by temperature effects on the cable wires, particularly if the cable wires are subject to temperature variations.

RTDs are connected to the analog input channels for 2 wire measurement as illustrated below.

Figure 79a ñ Connection for Two Wire RTD Measurement

Figure 79b ñ Connection for Two Wire RTD Measurement

During RTD resistance measurement, the excitation current output from the Excite terminal passes through the RTD and cable wires, and returns into the Analog Return terminal.

The voltage produced by the excitation current across the RTD and the cable wires is measured between the +ve and ñve terminals.

Cable wire compensation is performed during measurement of the RTD, but is of little significance since the resistance of the links is very much less than that of the cable wires (Figure 79a).

Where cables with only two cable wires are already installed and must be used, cable wire resistance compensation can be obtained by replacing the link between the ñve terminal and Analog Return with a resistor (Figure 79b) of value equal to that of the total cable resistance.

However using a compensation resistor does not compensate for temperature effects on the cables.

RTDs connected using the 2 wire method are read and data returned when a Schedule containing the channel is executed.

Using DeTransfer, the command for example

BEGIN
RA10M
1..3PT385(120.0)
END

instructs the dataTaker to measure the three platinum RTDs connected to analog channels 1, 2 and 3.

The excitation current channel option is not specified, and the default 2.500 mA excitation current is used. The platinum RTDs all have a resistance of 120.0 Ohms at 0.0 Deg C.

The PT385 specifies that the signal applied to these channels is from a platinum RTD, with a temperature coefficient of a=0.003850 Ohm/Ohm/Deg C. The resistance is linearized.

Using DeLogger, RTDs connected by the 2 wire method can be measured by the following Program Builder program.

The 2 wire connections are selected from the RTD Wiring Configurations dialog which opens when you have selected the analog input channel.

The data is returned in units of temperature. The dataTaker will read the inputs every 10 minutes, and readings are stopped by entering a H (Halt) command.

Measuring RTDs Using Single Ended Inputs

Two RTDs can be measured using a combination of 4 wire input and a single ended input for each of the analog input channels. Note : This method does not provide cable wire compensation for the RTD connected as a single ended input.

Referenced to Analog Return

Pairs of RTDs are connected to the analog input channels with the second RTD referenced to Analog Return as follows

Figure 80 ñ Connecting RTDs for Single Ended Input

Up to 10 RTDs can be measured using this method by the dataTaker 50, and up to 20 RTDs can be measured by the dataTaker 500/600 series loggers and the Channel Expansion Module (CEM-AD) .

During resistance measurement, the excitation current output from the Excite terminal of the analog channel passes through the two RTDs and cable wires, and returns into the Analog Return terminal.

The voltage produced across the RTD1 is measured as a differential input between the +ve and ñve terminals.

The voltage produced across the RTD2 is measured as a single ended input between the ñve terminal and Analog Return.

The two RTDs connected to the analog input channels by this method are sampled and the data is returned when a Schedule containing the channel is executed.

Using DeTransfer, the command for example

BEGIN
RA10M
1PT385(4W) 1ñPT385
END

instructs the dataTaker to measure the two platinum RTDs

connected between the +ve and ñve terminals of analog input channel 1. The 4W channel option indicates that the 4 wire method of resistance measurement is to be used. This measurement will be cable compensated.

connected between the ñve terminal of analog input channel 1 and Analog Return. This measurement will not be cable compensated.

The excitation current channel option is not specified, and the default 2.500 mA excitation current is used. The two platinum RTDs are assumed to have a resistance of 100.0 Ohms at 0.0 Deg C.

The PT385 specifies that the signal applied to these channels is from an platinum RTD, with a temperature coefficient of a=0.003850 Ohm/Ohm/Deg C. The resistance is linearized, and the data is returned in units of temperature.

The dataTaker will read the inputs every 10 minutes, and readings are stopped by entering a H (Halt) command.

Measurement Ranges and Accuracy

The measurement accuracy for RTD sensors is the same as that for standard resistance measurement, and is discussed in Section II ñ Resistance Measurement.

Greater accuracy of RTD measurement is achieved when 3 wire or 4 wire methods providing cable wire compensation are used, and the 2.500 mA RTD excitation current is used.

The temperature resolution is approximately 0.025 Deg C for 100 Ohms RTD sensors. However the temperature resolution can be increased by a factor of ten if the RTDs are measured as a half bridge with constant current excitation (See Section II ñ Measuring Bridges and Strain Gauges).

The disadvantage of this method however is that the dataTaker will not convert the resistance to units of temperature. This must be done with a user defined polynomial, or later by the host computer.

Sources of Error

There are two sources of error in RTD measurements by the dataTaker.

The first is due to errors in resistance measurement (excluding cable wire effects), which is specified at ±0.1% over the logger temperature range of 10 to 40 Deg C. This systematic error can however be calibrated out.

The second source of error is linearization error. The graph below illustrates the dataTaker linearization error limits along with the various tolerances.

Figure 81 ñ RTD Linearization Error Limits

Error Messages

RTD resistances which fall outside of the linearization range of ñ200 to 600 Deg C are returned as temperature data, which is extrapolated from the linearization functions.

When this occurs, the dataTaker reports the error by returning the error message 'E16 - Linearization error' if the Messages Switch /M is enabled.