What is a thermocouple?
A thermocouple (TC) is a passive sensing device that responds to the temperature that it is exposed to based on the Seebeck effect, which is the production of a differential voltage when there is a temperature differential between junctions of dissimilar metals, as shown in Figure 1. A thermocouple is made of a pair of dissimilar metallized electric cables connected together to form two junctions. The first junction is connected to the object whose temperature is to be measured. This is known as the hot or measuring junction. The other junction is connected to a known temperature, such as a resistance temperature detector (RTD). This is known as the cold or reference junction. The measured temperature indicated by the voltage VM in Figure 1 is a differential measurement. Hence, measuring the absolute thermocouple temperature depends on both the hot and cold junction temperatures and requires reference to a known temperature. Unlike other direct methods using RTDs and other devices, temperature measurement using a thermocouple is a calibrated measurement. Thermocouples provide a wide range of cold and hot temperatures in laboratory, industrial, and automotive applications, as well as consumer applications, because they are very fast, accurate, repeatable, reliable, and stable.
Figure 1. Thermocouple circuit.
The junction between metal A and metal B is the hot junction. This is the connected end of the thermocouple. The open end of the thermocouple is the cold junction and is connected to the measuring device such as a voltmeter or the MAX11410 ADC inputs, AIN1 and AIN2.
Table 1 lists some of the common type of thermocouples.
Table 1. Some Common Types of Thermocouples
|Sensitivity (°C)||Material Type|
|K||-180 to +1250||41||
|J||-180 to +800||55||
|N||-270 to +1300||39||
|R||-50 to +1700||10||
|S||-50 to +1750||10||
|B||0 to +1820||10||
|T||-250 to +400||43||
|E||-40 to +900||68||
In this application note, a K-type thermocouple is used for the temperature measurement.
The MAX11410 is a low-power, 10-channel, 24-bit delta-sigma (Δ-Σ) ADC with features and specifications optimized for precision sensor measurements. The device includes a low-noise, programmable gain amplifier (PGA) with very high input impedance and available gains from 1x to 128x to optimize the overall dynamic range. This internal PGA is perfect for applications measuring a wide temperature range using thermocouples, which generate a very low thermoelectric voltage in the millivolt range. The programmable matched-current sources provide excitation for RTD sensors, which can be used as the reference (cold) junction for the thermocouple measurement. An additional current source/sink aids in detecting broken sensor wires. The 10-channel input multiplexer provides the flexibility needed for complex, multisensor measurements, and GPIOs ease control of external switches or other circuitry. In addition, the MAX11410 has an internal 50Hz/60Hz filter to provide improved common-mode rejection, and self- and system-calibration options are available to reduce software development time.
The first step in a thermocouple measurement with the MAX11410 is to establish the reference (cold) junction. Figure 2 shows the thermocouple temperature measurement using the MAX11410 with the 2-wire RTD as the reference junction connected at the AIN8 and AIN9 inputs of the MAX11410. To accurately measure the temperature based on the RTD resistance, a voltage is generated across the 2-wire RTD by the MAX11410 internally programmable matched-current sources from 10µA to 1.6mA through the AIN8 pin. This current first goes through the RTD cable parasitic resistance RC. It then passes through the 2-wire RTD and the bottom cable parasitic resistance RC. Finally, it travels through the 4kΩ reference resistor (RREF) to establish the reference voltage at the REF1P pin. Pins AIN8 and AIN9 are used to measure the voltage drop across the RTD and VRTD, and the RTD value is calculated as follows:
VAIN8 - VAIN9 = VRC + VRTD... + VRC = (2 × RC + RTD) × I (EQ. 1)
Because the RTD is connected directly to the input pins AIN8 and AIN9 of the MAX11410, the RTD leads are very short. Therefore, the parasitic resistance RC is negligible.
Therefore, RC = 0 in equation 1.
VAIN8 - VAIN9 = RTD × I (EQ. 2)
VREF = RREF × I (EQ. 3)
The output code, which is a digital binary number produced by the MAX111410 when it compares the analog input voltage to the reference voltage, is obtained as follows:
Where N is the number of bits of the ADC, which is 24 for the MAX11410.
So, from equation 4:
Also, since Code(ADC Fullscale) = 2N:
Equation 7 shows that the calculated RTD is independent of the excitation current from the device in this ratiometric measurement. Therefore, any error in the current is canceled, and the measured voltage across the RTD depends solely on the precision of RREF.
Note that, in this 2-wire RTD topology, the voltage drop across the RTD is measured differentially between the AIN8 and AIN9 pins. This voltage includes the voltage drop across the two very short cable parasitic resistance RCs. Therefore, the voltage error caused by the RTD cable is negligible. If the parasitic cables are long, the 3-wire RTD measurement method should be implemented, as discussed in the Maxim application note 6793, “How to Measure Multiple Temperatures Using a Single MAX11410 and Resistance Temperature Detectors.”
Figure 2. Simplified block diagram of 2-wire RTD/thermocouple temperature measurement using the MAX11410.
Once the reference (cold) junction is established, the thermocouple measurement may take place. As depicted in Figure 2, a K-type thermocouple is connected to pins AIN1 and AIN2 of the MAX11410 through the 1kΩ overvoltage protection resistors. Unlike in the RTD case where ratiometric measurement can be carried out using the internal excitation current of the MAX11410 and any error in the current is canceled out, this thermocouple temperature measurement is implemented by measuring the differential voltage across the thermocouple device at the AIN1 and AIN2 pins. So, a precision external reference voltage such as the MAX6070 is required, which is connected to the REF2P and REF2N pins as shown in Figure 2. The voltage across AIN1 and AIN2 is then amplified by the internal PGA with a gain of 32 and converted to the measured temperature as follows.
First, a reference or cold junction temperature is established using the 2-wire RTD measurement technique, as previously discussed. As shown in the MAX11410EVKIT example in Figure 3, a cold junction PT1000 RTD measures and generates an ambient temperature of +24.8°C.
Figure 3. Cold junction temperature measurement using the MAX11410 ADC and PT1000 RTD.
The thermocouple temperature measurement uses +24.8°C as the reference temperature to generate the calibrated thermocouple temperature indicated by VTC as follows.
Referring to Figure 1:
VM = VTC - VCJ (EQ. 8)
Where VTC and VCJ are thermocouple voltage and cold junction voltage, respectively.
Rearranging equation 8 yields:
VTC = VM + VCJ (EQ. 9)
To measure the thermocouple temperature indicated by VTC, measure the differential voltage (VM) at inputs AIN1 and AIN2 of the MAX11410 and convert it to the temperature according to the Thermocouple Reference Table published by Omega™. A simplified version is provided in Table 2. Then add the cold junction temperature (VCJ) of +24.8°C obtained from the RTD measurement. This VCJ of +24.8°C is automatically saved in the MAX11410 software graphic user interface (GUI) and is added to the temperature corresponding to VM, as shown in equation 9.
Table 2. K-Type Thermocouple Temperature and the Corresponding Voltage
For example, the K-type thermocouple is now immersed in the Fluke® 7341 high precision oil bath for accurate and stable temperature, which is set to +28°C as shown in Figure 4. The MAX11410 measures VM across AIN1 and AIN2, which is further amplified by the internal PGA set at 32X. Because the voltage produced by the thermocouple is only approximately 51mV at +1250°C, after the PGA of 32, the maximum amplified voltage is approximately 1.6V, which is ideal for use with the MAX6070 2.5V voltage reference.The Hart Scientific 1502A thermometer is used to monitor the oil bath temperature for establishing the accuracy of the MAX11410 measurement. The calibrated temperature reading from the MAX11410 is shown in Figure 5.
Figure 4. Simplified block diagram of thermocouple temperature measurement using the MAX11410 setup.
Figure 5. K-type thermocouple temperature measurement using the MAX11410 and oil bath.
The oil bath temperature was varied from approximately -35°C to +125°C of its operating range. The thermocouple temperature was measured and is summarized in Table 3. Figure 6 shows the accuracy of the temperature measured with the MAX11410 ADC.
Table 3. Temperature Measurement Using the MAX11410 with the Thermocouple in the Oil Bath
|TC (°C) (Hart Scientific 1502A Thermometer)||MAX11410 Measured Temperature (°C)||Error (°C)||Error (%)|
Figure 6. Temperature measurement accuracy using the MAX11410 with the thermocouple in the oil bath.
Due to the limited temperature range of the oil bath from approximately -35°C to +125°C, the Krohn-Hite Model 523 (KH523) precision DC source voltage calibrator was used to evaluate the thermocouple temperature measurement accuracy of the MAX11410, since it can generate precision low voltage in millivolts as specified in Table 2.
Based on Table 2, the KH523 was set to generate the voltage which corresponds to the K-type thermocouple temperature at AIN1 and AIN2. The MAX11410 then displayed the temperature on the GUI based on the measured voltage. Table 4 shows the measured thermocouple temperature from -200°C to +1250°C. Figure 7 plots the temperature accuracy error produced by the MAX11410.
Table 4. Temperature Measurement Using the MAX11410 with the KH523
|TC (°C)||V at TC (V)*||MAX11410 PGAx||VREF (V)||MAX11410 Measured Voltage (V)||MAX11410 Measured Temperature (°C)||Error (°C)||Error (%)|
Figure 7. Thermocouple temperature measurement accuracy using the MAX11410 with the Krohn-Hite Model 523 (KH523) precision DC source voltage.
With the implementation of the internally programmable matched excitation current sources, ratiometric feature, low-noise programmable gain amplifier (PGA) with very high input impedance, and available gains from 1x to 128x, the MAX11410 is optimized for precision temperature measurements using both RTDs and thermocouples. Thermocouple temperature measurement is a two-step process. The first step is to establish the reference temperature, and the RTD feature of the MAX11410 is perfect for this purpose. The second step requires precision low-voltage measurement in millivolts generated by thermocouples. The MAX11410 is an ideal ADC for this low voltage measurement since it is designed with internal variable gains up to 128x. The results obtained from the K-type thermocouple measurement using the MAX11410 show that the accuracy error is approximately only ±0.04% over the entire temperature range from -200°C to +1250°C.