The objective of this study to develop Partial Discharge (PD) analysis system to assess its use as potential PD measurement in high voltage equipment. Longer service period of insulation system in high voltage has cause major problem to the high voltage equipment which typically due to ageing and deterioration. The insulation system is the most important part of high voltage equipment to prevent any discharge adjacent at the protected area. The cable that effected due to the aging process may interrupt the ability of insulation property in the power cable. In principle, the aging mechanism lead to the process of PD that cause by voids from inside the solid insulator, bubble in liquid insulator, contamination by particles on the surface of insulating material, floating particles in gas insulator, corona activity which is discharges around an electrode in gas and mechanical failure or damage of insulation materials. The effects of PD are, insulation erode, losses and temperature increased and immediate failure of the system. Nowadays, PD measurement is applied to detect the occurrences of discharge and online measurement is more reliable and can effectively diagnose insulation problem compared to the offline measurement. However, online measurement face problem of disturbance during industrial onsite measurement. Thus, to overcome this problem, PD analysis system that equipped with denoising technique has being created. LabVIEW software is being used to model the system with Discrete Wavelet Transform (DWT) denoising technique. In this work, different decomposition level and type of mother wavelet in DWT denoising technique were investigated. The performance of denoising result is evaluated by calculation of Mean Square Error (MSE) and Signal-to-Noise Ratio (SNR). Research is divided into three phases, which are data acquisition, data process and data analysis. The first phase is data acquisition which involved modelling of PD and noise signal using the mathematical equation. Next, data process phase involved noise rejection using wavelet transform. The result is being manipulated with different type of mother wavelet, such as Haar, Daubechies, Coiflets, Symlets and Biorthogonal. Lastly, data analysis is being calculated the MSE and SNR value of denoising signal to select the best mother wavelet in denoising process.
