Abstract:
In industrialized countries, motors consume between 40 to 50% of all the generated capacity. Induction motors are prone to faults and require special attention. Few systems exist that can detect faults in motor through vibrations, thermal and current analysis but they are expensive, invasive and not suitable for small industries. The solution for the above problem is to develop a system that will detect most of the incipient faults using only non-invasive current sensor and should be cost effective. Most of the motor faults can be detected analyzing stator current. In this system, non-invasive current sensor is wrapped around input cable and the obtained signal is fed to ADC after conditioning. The digital data from ADC is transmitted wirelessly to the receiver. At receiver end, data is fed to the computer through USART. Data is passed through low-pass filter and Fast Fourier Transform is applied. Signal spectrum analysis helps the software to detect different types of fault in motor e.g. broken rotor bar and end ring faults, Eccentricity related faults, Bearing Faults and Stator or Armature Faults via detecting undesirable peaks other than the one at principal frequency. Our system provides the capability to monitor more than ten motors at same time. And different systems installed in same facility will not interfere with each other. In a nutshell our system detects incipient faults and saves users’ money, time and space by preventing motors from complete damage.
