The power supply
AD7124-4/AD7124-8 has separate analog and digital power supplies. The digital supply IOV _DD is independent of the analog supply and can have values ​​in the range 1.65 V to 3.6 V (referenced to DGND). Analog power AV _DD is referenced to AV _SS and ranges from 2.7 V to 3.6 V (low and medium power modes) or 2.9 V to 3.6 V (full power mode). The circuit shown in Figure 1 operates from a single supply, so AVSS is connected to DGND, using only one ground plane. The AV _DD and IOV _DD voltages are generated separately using the ADP1720 low dropout regulator. The AV _DD voltage is set to 3.3 V and the IOV _DD voltage is set to 1.8 V using an ADP1720 voltage regulator. Using a separate voltage regulator ensures minimal noise.

Serial Peripheral Interface (SPI)
SPI communication with the AD7124-4/AD7124-8 is handled by the Blackfin® ADSP-BF527 on the EVAL-SDP-CB1Z board , as shown in Figure 1. To access the AD7124-4/AD7124-8 registers, use the AD7124-4/AD7124-8 EVAL+ software . Figure 4 shows the main window of the software. Click THERMOCOUPLE to configure the software for T-type thermocouple measurements.

The AD7124-4/AD7124-8 have on-chip diagnostics that can be used to detect SPI communication failures. These diagnostics include checking SPI read and write operations to ensure that only valid registers are accessed. The SCLK counter ensures that the correct number of SCLK pulses is used, while the CRC function checks whether the bit value changed during the transfer. When any SPI communication diagnostic function is enabled and a related error occurs, the corresponding flag in the error register will be set to 1. All enabled flags are ORed together to control the ERR flag in the status register. This feature is particularly useful when appending status bits to ADC conversion results.

The analog input
AD7124-4 can be configured as 4 differential or 7 pseudo-differential input channels, while the AD7124-8 can be configured as 8 differential or 15 pseudo-differential input channels.

The AD7124-4/AD7124-8 have on-chip diagnostics that can be used to check whether the levels on the analog pins are within the specified operating range. The positive (AINP) and negative (AINM) analog inputs can be independently checked for overvoltage and undervoltage occurrences, as well as for ADC saturation. When the voltage on the analog input exceeds AV _DDSS , the undervoltage flag is set.

For the circuit shown in Figure 1, two analog input pins are used to connect the thermocouples (AIN2, AIN3) and three analog pins are required for cold junction compensation (AIN1, AIN6, AIN7). AIN2 and AIN3 are configured as fully differential input channels for measuring the voltage generated by the thermocouple. For this circuit, as shown in Figure 1, the thermocouple is floating. To bias a thermocouple to a known level, enable the V _BIAS voltage generator on AIN2, biasing the thermocouple to the following values:



Thermocouple measurements are absolute measurements and therefore require a voltage reference, use the AD7124-4/ The AD7124-8 has a built-in 2.5 V reference.

For cold junction compensation, an excitation current source is used to excite the RTD. This current is generated from AV _DD and flows to AIN1. Figure 5 shows the analog pins and their configuration in detail.


For this circuit, the cold junction circuit uses the reference input REFIN1(±). The current flowing through the 4-wire RTD (used for cold junction measurements) also flows through the precision reference resistor, producing the reference voltage. The voltage developed across this precision reference resistor is proportional to the voltage across the RTD, so fluctuations in the excitation current are canceled out. Since the reference buffer is enabled, it is important to meet the required headroom for proper operation (AV _DD − 0.1 V and AV _SS + 0.1 V). A margin of 0.125 V (500 μA × 250 Ω) is provided by the 250 Ω resistor to ground, as shown in Figure 5.


Digital and analog filtering
Differential filters (cutoff frequency approx. 800 Hz) and common-mode filters (cutoff frequency approx. 16 kHz) are implemented at the analog inputs and reference inputs. This filtering is necessary to suppress interference at the modulator frequency and its multiples.

The AD7124-4/AD7124-8 offer great flexibility in on-chip digital filtering. A variety of filter options are available, and the selected filter affects the output data rate, settling time, and 50 Hz/60 Hz rejection performance. For this circuit note, the circuit implements a sinc4 filter and a post filter. The sinc ⁴ filter is used because of its excellent noise performance over the entire output data rate range, in addition to its excellent 50 Hz/60 Hz rejection. The post filter is used to provide simultaneous 50 Hz and 60 Hz rejection with a settling time of 40 ms. Calibrating the AD7124-4/AD7124-8 provides different calibration modes to eliminate offset and gain errors. For this circuit note, the circuit uses internal zero-scale calibration and internal full-scale calibration. Thermocouple Configuration The circuit shown in Figure 1 uses the AD7124-4/AD7124-8 for precision T-type thermocouple measurements. Thermocouple measurements require cold junction compensation. As shown in Figure 1, a 4-wire Pt100 RTD is used for this purpose. Using the configuration shown in Figure 1, a precision excitation current source is required to excite the RTD as part of the cold junction compensation measurement. This RTD is connected to analog inputs AIN6, AIN7. The bottom end of the RTD is connected to a precision reference resistor, which then applies an external reference voltage to the device. This precision reference resistor is connected between the reference input pins REFIN1(±). This configuration represents a ratiometric configuration, and any deviation in the excitation current affects both the RTD and the reference resistors and does not appear in the measurement results. The thermocouple itself is connected to analog inputs AIN2, AIN3. One of the inputs is biased using the ADC's internal bias voltage generator. The thermocouple voltage ranges from −8 mV to +17.2 mV, representing a temperature range of −200°C to +400°C. This low-level voltage is amplified by the AD7124-4/AD7124-8's on-chip PGA and then converted to a precision digital signal by a 24-bit Σ-Δ ADC. To ensure that the full range of the ADC is utilized, the PGA gain is set to 128. This thermocouple measurement is performed relative to an internal low-drift 2.5 V reference. A 4-wire Pt100 Class B RTD is used for cold junction measurements. The excitation current of the Pt100 RTD was set to 500 μA. Choose an appropriate external precision resistor value so that the maximum voltage developed across the RTD is equal to the reference voltage divided by the selected gain. Circuit Note CN-0381 discusses the following required steps in detail:

• Select a precision reference resistor
• Selecting the appropriate PGA gain for RTD measurements
• Margin resistor selection
• Excitation current output compliance voltage

The AD7124-4/AD7124-8 full system configuration for thermocouple measurement is as follows:

• Thermocouple measurement (T type)
  • Differential input (AINP = AIN2, AINM = AIN3)
  • Gain = 128
  • 2.5 V internal voltage reference
  • Digital filtering (sinc ⁴ and post filter)
• Cold junction compensation measurement (4-wire RTD)
  • Differential input (AINP = AIN6, AINM = AIN7)
  • Excitation current: IOUT1 = AIN1= 500 μA
  • Gain = 16
  • 5.11 kΩ precision reference resistor
  • Digital filtering (sinc ⁴ and post filter)

Thermocouple Temperature Calculation

After implementing the above procedure, the next step is to solve the thermocouple and cold junction calculations. Different methods can be used for linearization and compensation, including:

• Lookup table: Requires memory for storage, but also provides fast, accurate calculations.
• Software linear approximation method: only needs to store the conversion polynomial coefficients, no other memory is required. Processing time is required to solve multi-order polynomials. However, it can also produce very precise results. This is the method used in this circuit.

The software linear approximation method requires two inputs: the voltage measured on the thermocouple and the cold junction temperature.

The analog input channels (AIN2, AIN3) are used to measure the voltage on the thermocouple. Equation 1 is used to convert codes to voltages, and it assumes the ADC is in a bipolar configuration. The AD7124-4/AD7124-8 software automatically converts codes to voltages based on the implemented configuration.



Where:
V _TC is the thermocouple (TC) voltage.
CODE _TC is the thermocouple (TC) code.
N is the resolution of the ADC (24 bits).
V _REF is the reference voltage used for measurement. For this circuit, the internal reference voltage is used for thermocouple measurements.
Gain is the gain selected for TC mode (128).

A 4-wire RTD for cold junction needs to be linearized on its own. When the ADC is operating in bipolar mode, the general expression for calculating the RTD resistance (R) is as follows:



where:
R _RTD is the resistance of the RTD.
CODE is the ADC code.
N is the resolution of the ADC (24 bits).
R _REF is the reference resistor.
G is the selected gain (16).

The steps involved in converting RTD voltage to temperature and linearizing are described in Circuit Note CN-0381 .

Calculating the thermocouple temperature requires the following steps:

• Convert cold junction temperature to voltage
• Calculate thermoelectric voltage
• Convert thermoelectric voltage to temperature representation.

Cold junction temperature must be converted to voltage. The cold junction temperature is converted using a polynomial provided by the National Institute of Standards and Technology (NIST), as shown in Equation 3.



Where:
V _CJ is the thermoelectric voltage.
a _x are the polynomial coefficients related to the thermocouple type.
T is the cold junction temperature (℃).
n is the polynomial order.

Increasing the order of the polynomial can improve the conversion accuracy of cold junction temperature to voltage. However, the higher the order, the more processing required. Therefore, there are trade-offs when performing this conversion. This circuit is calculated using an eighth-order polynomial.

The cold junction temperature voltage must be added to the differential voltage measured across the thermocouple. The final voltage is an approximation of the thermoelectric voltage produced by the thermocouple temperature sensing junction.

This thermoelectric voltage can then be used to calculate the overall thermocouple temperature. This step involves the power series polynomial given by Equation 4. This circuit uses a sixth-order polynomial, and the polynomial coefficients of the T-type thermocouple are obtained from the NIST website.



where:
V is the thermoelectric voltage (μV)
a _x are polynomial coefficients related to the thermocouple type.
T is temperature (℃).
n is the polynomial order.


Thermocouple Measurements and Results

For the circuit shown in Figure 1, data was collected for the AD7124-4/AD7124-8 in different digital filter and power mode configurations.

The first configuration uses a sinc4 filter, full power mode, and an output data rate of 50 SPS. Under these conditions, the AD7124-4/AD7124-8 provide optimal speed and noise performance. Figure 6 shows the noise distribution at room temperature when a thermocouple is connected between the AIN2 and AIN3 input channels as shown in Figure 1. The corresponding rms noise is typically 70 nV, equivalent to approximately 16.4 bits of noise-free resolution. Under the same conditions, the noise performance of the AD7124-4/AD7124-8 is typically 48 nV rms with the input shorted, which is equivalent to 17 bits of noise-free resolution. The increase in noise comes directly from the thermocouples connected on the input channels (AIN2, AIN3).


For the thermocouple configuration with the sinc ⁴ filter and full power mode selected, the thermocouple temperature is swept from −50°C to +200°C while the cold junction is held at −40°C, +25°C, and +105°C. For each set thermocouple temperature, measure the corresponding voltage on the thermocouple using the AD7124-4/AD7124-8 as described above. Also recorded is the cold junction temperature measured using a 4-wire RTD. Calculate the thermocouple temperature using the thermocouple voltage and the voltage representation of the cold junction temperature. Figure 7 shows the error between the set temperature value and the measured temperature of the thermocouple after linearization when the cold junction temperature is −40°C, +25°C, and +105°C. Internal zero-scale and full-scale calibrations are performed at each cold junction temperature. As shown in Figure 7, the error between the calculated temperature and the thermocouple set temperature is within the combined root mean square error window of the T-type thermocouple and the Pt100 RTD. The maximum error of T-type thermocouple is 1℃ or 0.75%; according to IEC751 standard, the error of Pt100 is ±(0.3 + 0.005 × |T|).


The second configuration tested, using a post-filter, low-power mode, and 25 SPS output data rate, provides simultaneous 50 Hz and 60 Hz rejection and allows the user to trade settling time for rejection performance. Figure 8 shows the noise distribution at room temperature when a thermocouple is connected between the AIN2 and AIN3 input channels as shown in Figure 1. The corresponding root mean square noise is typically 220 nV rms, which is equivalent to approximately 14.7 bits of noise-free resolution. Selecting the same filter, gain, and output data rate, but with the input shorted, the noise performance of the AD7124-4/AD7124-8 is typically 170 nV rms, equivalent to 15.1 bits of noise-free resolution. The increase in noise comes directly from the thermocouples connected on the input channels (AIN2, AIN3).


For the AD7124-4/AD7124-8 configuration with post filter and low power mode selected, the RTD temperature is swept from −50°C to +200°C. For each set thermocouple temperature, measure the corresponding voltage on the thermocouple using the AD7124-4/AD7124-8 as described above. Also recorded is the cold junction temperature measured using a 4-wire RTD. Calculate the thermocouple temperature using the thermocouple voltage and the voltage representation of the cold junction temperature.

Figure 9 shows the error between the set temperature and the measured temperature of the thermocouple after linearization when the cold junction temperature is −40°C, +25°C, and +105°C. As shown in Figure 9, the error between the calculated temperature and the thermocouple set temperature is within the combined root mean square error window of the T-type thermocouple and the Pt100 RTD. The maximum error of T-type thermocouple is 1℃ or 0.75%; according to IEC751 standard, the error of Pt100 is ±(0.3 + 0.005 × |T|).