By its simplicity, sphere gaps method for the measuring of peak voltage is preferred to be used as standard for calibration purposes. The limits of accuracy of the standard sphere-gaps depends on the ratio of the gap spacing S to the sphere diameter D, as follows. For AC, switching and lightning impulse voltage, S ?0.5D, uncertainty is ?3%, but for gap spacing 0.75D > S > 0.5D, the uncertainty is larger than ?3%. However, the U50% disruptive discharge of some standard spheres at the same gap spacing showed the different value from each other. The significant case occurred between 5cm and 10cm sphere-gap for negative polarity.
Hence, the objective of this thesis is to study the characteristic of the curve of U50% disruptive discharge voltage against the gap spacing S for 5cm and 10cm diameter sphere. The standard lightning impulse voltage (T1=1.2 ?s and T2=50 ?s) of negative and positive polarities for S= 0.5, 1.0, 1.5, 2.0 and 2.4cm, was taken for this study by taking account the influence of humidity and the application of external irradiation. In order to control the atmospheric condition around the sphere-gap during the breakdown phenomena, both sphere-gap was enclosed in a cylindrical shape test chamber. This chamber was designed by using the finite element method (FEM) to ensure that there is an acceptable influence on the electrical field distribution across the gap of spheres. Two different measuring devices were chosen according to the applied voltage level for the purpose of measuring accuracy. The experiment results showed that the difference between U50% of this present study and the one of the standard decrease as the gap spacing increase, especially for S?1.0cm. And the value of U50% for D=5cm and D=10cm from this experiment tended to be close to each other.