With the growing need for disturbance-free power distribution, there has been significant focus under the ‘Smart Grid’ initiative on fault and disturbance analysis. Compared to the high voltage (HV) transmission systems where high-end relays can be used, medium- and low-voltage (MV, LV) distribution systems employ low-end relays for cost reason. Therefore, computationally inexpensive signal processing algorithms are needed for disturbance identification at the MV and the LV distribution systems. In this paper, a novel amplitude tracking square wave concept is presented, which can effectively represent the complex sinusoidal signals under disturbance. The amplitude tracking square wave is very sensitive to any degradation introduced in the sinusoidal signals due to disturbances. These are typically reflected in the time-domain as spikes in the square wave, without requiring any frequency analysis. This makes it particularly attractive for embedding into the MV and the LV relays with limited computational resources. Successful application of the algorithm on real disturbance records substantiate its potential.