Raspberry pie based thermocouple measurement HAT module
The use of thermocouples to measure temperature is a common method because of its low cost, ease of use, and wide measurement range. Measurement Computing Corp (MCC) has a long history of designing and building accurate data acquisition devices for measuring thermocouples. In an uncontrollable environment, it is very difficult to design a device that can accurately measure thermocouples on the Raspberry Pi. This article explains the challenges encountered in accurately measuring thermocouples, how the MCC 134 HAT does its job, and how the MCC 134 user should minimize measurement errors.
MCC 134
How does a thermocouple work?
Thermocouples are a type of sensor used to measure temperature. It works by converting a temperature gradient into a potential difference, a phenomenon known as the Seebeck effect. A thermocouple consists of two different metal wires, and their ends are connected to each other to form a node. Since the two wires produce different potentials in the temperature gradient, the voltage can be measured in the circuit induction. This potential difference is obtained by measuring the voltage in the loop.
In different types of thermocouples, the connection of the wires is not the same, so they can be measured in different temperature ranges. For example, a J-type thermocouple consists of iron and constantan (copper-nickel alloy), suitable for measurement between -210 ° C and 1200 ° C; and T-type thermocouple consists of copper and constantan, suitable for -270 ° C and 400 ° C Measure between.
The above temperature gradient refers to the temperature difference between two nodes -- the measuring point, the hot end we are concerned with; the reference point, which is the cold end at the measuring device.
Note: The hot end refers to the measuring end, regardless of the temperature of the end; its true temperature may be higher or lower than the reference point, ie the cold end temperature.
The basic principle of thermocouple measurement
The thermocouple produces a voltage corresponding to the temperature gradient, ie the potential difference between the hot and cold ends. The only way to determine the absolute temperature of the hot end is to obtain the absolute temperature of the cold end.
While older systems rely on ice baths to achieve cold junction reference temperatures, modern thermocouple measurement devices use one or more sensors to measure the temperature of the terminal (cold end) to which they are connected.
Source of error in thermocouple measurement
Thermocouple measurements have many sources of error, including noise, linearity and offset errors, thermocouples themselves, and temperature measurements at the reference or cold junction. Modern 24-bit measurement equipment uses high-precision analog-to-digital converters (ADCs) and uses design practices to minimize noise, linearity, and offset errors.
Thermocouple measurement errors are unavoidable, but they can be minimized. This error is due to alloy defects used because they vary slightly from batch to batch. Some thermocouples have small errors in their own right. Standard K-type and J-type thermocouples can have errors of up to ±2.2 °C, while T-type thermocouples can have errors of up to ±1 °C. More expensive thermocouples (SLE-special error limits) consist of better quality wires, which can reduce these errors by half.
Accurate measurement of the cold end, the junction of the thermocouple and the device, is very difficult. More expensive instruments, such as the DT MEASURpoint series, use a thermally insulating metal plate to keep the cold end temperature constant and make measurement at high precision easier. Insulated metal sheets are too expensive for lower cost equipment. However, if there is no insulating metal plate, it is impossible to measure the temperature of the contact point between the thermocouple and the copper connector. This results in the measurement of the cold junction temperature being highly susceptible to rapid changes in temperature or power conditions in its vicinity.
MCC 134 design challenges
To better understand the design challenges of the MCC 134, we need to compare it to MCC's popular E-TC series -- a high-precision Ethernet thermocouple measurement device. The cold junction temperature of the E-TC series is measured by the analog device ADT7310 IC temperature sensor.
Due to the controllable and stable measurement environment, IC sensors & CODECs work well in the MCC E-TC series. An external plastic box is used to control airflow, circuit components, and a processor that operates at a constant load. In the controlled environment of the E-TC, the IC sensor can accurately measure the cold junction temperature.
However, when the MCC 134 was first designed with an IC sensor to measure the cold junction temperature, the problem of insufficient accuracy was particularly prominent during the equipment calibration process. The IC sensor cannot be placed near the connector module because the large and uncontrollable temperature gradients caused by the Raspberry Pi and the external environment can cause poor measurement repeatability.
The MCC redesigned the MCC 134 with an improved solution that provides better accuracy and repeatability while maintaining low cost. Unlike using one IC sensor and one terminal module, the MCC redesigned the board using two terminal modules and three thermistors -- each of the thermistors is placed on one side of the terminal module ( As shown below). Although this increases the difficulty of the design, the thermistor can capture the change in cold junction temperature more accurately even when the load of the processor and the ambient temperature change.
This design approach makes measurement results virtually unaffected by the uncontrolled environment of the Raspberry Pi. But even with new designs, certain factors can affect measurement accuracy. However, the user can improve the measurement by reducing the rapid change in temperature gradient on the MCC 134.
Best Practices for MCC 134 Accurate Measurement of Thermocouples
When operating in a standard environment, the MCC 134 maximizes the accuracy of the thermocouple. Intense temperature changes and changes in airflow can affect the results. In most cases, the MCC 134 will complete its typical specifications. In order to achieve the highest precision thermocouple readings, the MCC has the following practical recommendations:
Reduce the load on the Raspberry Pi processor. When the load of the running program occupies 4 cores of the Raspberry Pi processor, its temperature will rise above 70 °C. Running a program that loads only 1 core will run at a temperature of approximately 20 °C.
Minimize the ambient temperature change value. The MCC 134 is moved away from the recirculating heat source or cooling source. Instantaneous temperature changes can cause an increase in error.
Provides continuous air flow, such as a fan. Stable air flow dissipates heat and reduces errors.
When several MCC HATs are configured on the stack, the MCC 134 is placed at the far end of the Raspberry Pi. Since the Raspberry Pi is a non-negligible heat source, placing the MCC 134 farthest away from it will increase accuracy.
in conclusion
Thermocouples offer a low-cost and flexible method of measuring temperature, but accurate measurement of thermocouples is difficult. Through innovative design and extensive testing, MCC overcomes the challenge of accurately measuring thermocouples in an uncontrolled environment when using the Raspberry Pi. The MCC 134 DAQ HAT is able to combine standard thermocouples with the needs of a fast-growing, low-cost computing platform.