measurement cycle, the ramping sequence of hold and up for positive and hold and down for negative current is repeated 20 times, and 25 data points were taken at every ramping step. In order to eliminate the voltage switching effect, the first five and the last five points were excluded in the evaluation of the mean value and standard deviation. The offset current caused by a device-under-test (DUT), voltage ramp and a capacitor due to an incomplete discharge can be eliminated by subtracting the null current measurement (hold step) from the measured current. The drift of the offset current can be averaged out by changing the polarity. However, the short term drift of the offset current caused by the discharge of the capacitor cannot be completely eliminated by the subtraction method, as shown in Fig. 2, where the generated current) can be expressed by the model equation.