The method based on phase difference for shaft power measurement is used widely and supposed to be more accurate than traditional methods, like by strain gauge. Because the torque of shaft is calculated according to measure torsional angle between two coding wheels, there are some potential errors which probably have great influence to experiment result in practicle measurement. The sources of error are researched and analyzed in four aspects. The first one is from the quantification error in FPGA, which is caused by the counter's frequency. Lower counter frequency may leads to bigger error in logic algorithm for measurement of the width of pulse signal. The research in this problem is to find out a quantification frequency that is accurate enough but won't increase too much the amount of data. The second one is temperature drift in sensor. It is generated by the variation of environment temperature. The influence of temperature to sensor is universal in application. The error tolerance and treatment of temperature drift is discussed in the research work. The third and fourth error sources in measurement are decentered error and perpendicularity error. Both of them are the result of inexactitude installation of coding wheels on shaft. All the sources of error discussed may introduce a lot of uncertainty to the continuous torque and power monitoring. So the tolerances of the potential errors and the extent of improvement are discussed after the analysis of error sources. The research work will help the engineers to understand the processing of error generation and prevent unnecessary problems in measurement and monitoring of shaft torque and power, especially the error that may result in the missing of important signals for fault diagnosis.
展开▼