The patterns of signal that detected by using electric and acoustic emission (AE) methods can be used for the identification of partial discharge (PD) events in power cable. Because the transmission distance of sounds in the insulation layer of power cables is limited, when discharging occurs at continuous or joint positions, a high-frequency current transformer (HFCT) and/or AE sensor can be used to measure the discharge signal. However, when measuring on site, the signals are influenced by noise. Our solution for this problem is from two perspectives. First, we designed a low-cost simple PD detector and use multiple detectors to measure PD in power cables. Second, empirical mode decomposition (EMD) and fast Fourier transform (FFT) signal processing methods were used to analyze and compare the collected data.