This paper proposed new transient power quality disturbances (PQDs) detection using strong trace filter (STF). By appropriate filter model design, when there are stationary PQDs the STF works as the same as Kalman filter, while when there are transient disturbances the STF indicates each sudden change of the distorted waveform by the fading factor (FF). The FF can also reveal which parameter of the signal component is changing and its sensitivity to sudden change can be tuned by the soften factor easily to avoid noise influence. Besides, the STF is a small algorithm, which can be easily implemented on embedded system for real-time and time-varying detection. Both simulation and experiment suggest that the STF is a good solution for transient PQDs detection.