Testing of High Energy Density Capacitors

Abstract

Summary form only given. The U.S. Army Research Lab (ARL) is investigating the capabilities of high energy density capacitors at narrow pulse widths from two different manufacturers for high current pulsed power applications. The pulse widths of the shots were between thirteen and fourteen microseconds. These pulse widths were taken from the full wave half maximum pulse width (FWHM). Both capacitors are rated for 12.5 kJ of energy but have different volumes. Manufacturer A's capacitor has a total energy density of 1.3 J/cc while the manufacturer B's capacitor has an energy density of 0.64 J/cc. Both manufacturer A's and manufacturer B's capacitors are metallized film capacitors. While energy density is a major focus for our tests; attributes such as peak current, rise time, and high voltage lifetime were other factors to be considered. ARL is collaborating directly with a capacitor vendor to help provide data to improve all properties of their capacitors. Prior to any testing, the capacitors had their respective internal values checked with an LCR meter to find the internal resistance, capacitance, inductance and quality factor (Q). A mechanical closing switch and resistive load were used to dump the energy after the capacitor had been charged to 10 kV. Once the test circuit had been built, the total inductance and resistance of the system was found using the calibrated LCR meter. These values were then entered into Pspice for a more accurate simulation of the test circuit's total current and load voltage waveforms. After one hundred shots were taken on manufacturer B's capacitor as a reference, testing began on manufacturer A's capacitor. After 100 shots on manufacturer A's capacitor, the capacitance and Q values had dropped significantly. It was decided the capacitor should be taken to failure level (5% capacitance loss). This level of capacitance loss was achieved after 150 shots with each shot having an average peak current of 160 kA. The rise time of the capacitors reached upwards of 30 kA/mus. After 100 shots, the capacitance and Q values dropped slightly on manufacturer B's capacitor. From the peak Q value till the one hundredth shot, the Q dropped from 610 to 530. The capacitance only dropped 0.2 muF, which is a 0.1% loss of capacitance. The failed capacitor has been sent back to manufacturer B to identify the internal reasons behind failure. The capacitor was opened up so that an autopsy could be performed. This sequence of events is done so that each generation can be improved upon. This paper will discuss the methods and findings of the experiments on high energy density capacitors.