Investigation of sampling frequency and jitter effects on transient current signal evaluation for insulation condition monitoring

The proposed online monitoring method detects insulation defects in a machine by evaluating the current response after a certain voltage step excitation is applied. The transient current response is analyzed to determine the status of an electrical machine insulation regarding aging, in a longer perspective. If insulation degradation occurs the parasitic winding components (e.g. winding capacitances) have changed and as a consequence the high frequency characteristic. A deviation in the frequency components can be interpreted as insulation deterioration of the insulation state. This work includes a strategy to improve the performance of the monitoring method with limited hardware and software resources (e.g. quite low sampling frequency of ADC units, usage of standard current sensors only). Different test scenarios conducted on a 1.4MW induction machine have been made to prove the applicability.