Energy trapping and adaptive clocking innovations applied to capacitor charging series resonant inverters

Abstract

The charging of a capacitive energy store is commonly accomplished by means of a high frequency series resonant switching inverter. A small amount of energy is switched each cycle but only a fraction is actually transferred to the store during the first portion of the charging process. The series resonant inverter behaves as a current source; and a constant amount of charge is delivered to the capacitor store each switching cycle. During the first portion of the charging process, the voltage on the store is low and the power or energy transfer is small compared to the constant volt-ampere capacity of the circuit. That is, there is a low power factor or impedance mismatch. During the switching cycle, there is an initial transfer of energy from the power source through the switching circuit; however, only a small portion of the energy stays in the capacitor store. The remainder flows back to the source and is then recycled again during the next switching cycle. Therefore, to transfer a given energy to the store requires the processing of a large amount of reactive power. By introducing an "energy trap" innovation to the circuit, the initial energy is forced to transfer to the store. The time required for this forced transfer to occur depends upon the voltage on the store. The transfer time is long when the voltage is low and decreases as the voltage increases. Since a new switching cycle cannot be initiated until the energy transfer of the present cycle is completed, the switching frequency must be chirped to match the circuit state as the charging process proceeds. An "adaptive clocking innovation" is used to monitor the energy transfer process and to clock the beginning of the next switching cycle when the energy transfer is completed. The result is a lower over-all switching frequency with lower losses for the same total energy transfer to the load. Since the circuit operates at a lower frequency at the beginning of the charging process one may be concerned that it would force the transformer size to increase. This is not the case because the transformer is working at a lower voltage during the low frequency part of the charging process and this combination of frequency and voltage results in a constant transformer flux and a lower over-all core loss.