Bit Resolution Improvement for Continuous Data Acquisition of Electrical Waveforms in Multiphase Energy Measurement Systems

Abstract

A data acquisition platform has been previously developed that enables extraction of information about an electrical machine’s health and energy conversion efficiency by monitoring its electrical signals. It is desired to extend the functionality of this platform to enable multi-phase signal acquisition with higher bit resolution while providing continuous waveform sampling that does not present any gaps in information acquisition. The previously designed platform samples signals at 8 kHz and can achieve about 15 bits resolution following down sampling to 1920 Hz. To get higher bit resolution without altering the hardware, more oversampling is required. Thus, the hardware is set to sample at 64 kHz which is the maximum sampling frequency the 6-channel simultaneous sampling analog to digital converter (ADC) can provide. Sampling at higher frequencies results in larger raw data sizes while the sampling time window remains same, and this increases the data transfer time reducing the information available for analysis in a given period of time. To address this issue, in this thesis, several proposed approaches are explored and a final hybrid solution is used. The final solution achieves about 1.3 bits improvement in signal resolution compared to the original firmware, while also performing continuous waveform acquisition if the end-to-end network delay is within expected ranges.