In this video within the implementing isolated shunt sensors using isolated delta sigma modulators and digital filters section of the training series, we will cover the different calculations used to calculate the metrology parameters in this design. This slide shows the formulas used by the foreground process to calculate the metrology readings. For the r-miss readings, a running sum of squares samples for a full frame of data, approximately a second's worth of data, is divided by the number of samples. An offset can then be subtracted from the resulting quality to track the effect of Gaussian noise for more accurate RMS measurements.

Subsequently, the square root operations performed and the resulting quantity is multiplied by a scaling factor to convert the results into real-world units like Fulton amps. For calculating the active power, the sum of the product of voltage and current samples for a full frame of data is then divided by the count of samples within the current frame and then multiplied by a scaling factor. An offset, which is used to compensate for voltage-to-current crosstalk is then subtracted.

For reactive power, a similar process is followed, except a 90-degree shift inversion of voltage is used for the multiplications instead of the original voltage samples, and a different reactive power offset is also used. An FIR filter and sample shifts are used to generate the 90-degree shift in voltage waveform used for calculating reactive power. The reactive power scaling factor is calculated internally based on the active power scaling factor and the characteristics of the selected FIR filter.

Apparent power is calculated by squaring the calculated active power, adding it to the squared reactive power, and then taking the square root of the results. The active power is divided by the apparent power to calculate the power factor. Using the individual power readings, the MSP432 calculates the cumulative power readings by summing their respective powers. The calculation of frequency is done by estimating the number of samples between two zero-crossings and dividing the sample rates of the ADC, 5,859.4 hertz in this case, by this number.

Metrology calculations done by the MSP432 can be broken down into two portions, the foreground process and the background process. The background process mainly deals with timing critical events in software. It is called whenever there is an interrupt from the AMC1210's ACK pin. The background process triggers the DMA to automatically read the newly generated current samples from the AMC1210. While the DMA is requesting the new current samples from the AMC1210, per-sample processing is done on the previously obtained voltage and current samples by calling the per_sample_dsp function.

When the per_sample_dsp function is called, it removes the residual DC content from the previous voltage and current samples, and updates the running sum of the product quantities, also called the dot products, needed for metrology parameter values. The background process is also responsible for counting the number of samples between zero-crossings, which is later used by the foreground process to calculate the frequency in units of hertz.

After approximately one second of data that is synchronized to zero-crossings, the background process alerts the foreground process that enough data has been captured for the foreground process to take the data captured in the background process to calculate metrology readings in real-world units. In addition to calling the per_sample_dsp function, the background process also calls the per_sample_energy_pulse_processing function to output energy.

The per_sample_energy_pulse_processing function is used to output pulses with the frequencies that are proportional to the amount of energy that is consumed. These pulses are used to calibrate the meter, as well as to determine the accuracy of the meter itself. This function accumulates the average powering reading of the previous frame into a buffer. When the value of this buffer is greater than the energy that corresponds to one pulse, then the MSP432 would output a pulse and subtract the amount of energy that corresponds to this pulse from the buffer, so that the leftover energy would be used as a starting point for accumulation the next time this function is called.

The foreground process in the code is responsible for setting up the MSP432 hardware, and initializing the AMC1210 registers. After this, the code gets into a loop where it checks and responds to any new messages from the GUI. Subsequently, the foreground process checks to see if it has been alerted by the background process that new metrology parameters are ready to be calculated. If the background process alerts the foreground process that there are new metrology parameters ready to be calculated, the foreground process uses the dot products from the background process to calculate the metrology readings in real-world units using the formulas in the last slide.

LCD management function is also called within the foreground process to change the displayed parameter on the LCD once every two seconds. In addition, a signal is outputted once a second in the foreground process, and fed to the TPS3850 to prevent the device from resetting the MSP432.