Charging Efficiency and Thermal Management Balance of Series Resonant Capacitor Charging Power Supply
Capacitor charging power supplies rapidly charge capacitor banks to high voltage for pulsed power applications including laser flashlamps, radar modulators, and particle accelerators. Series resonant charging circuits provide efficient, controlled charging with inherent current limiting. The charging efficiency and the thermal management of the power electronics must be balanced to achieve reliable operation at the required charging rates. The design optimization considers the tradeoffs between efficiency, thermal performance, size, and cost.
The series resonant charging circuit consists of a resonant inductor in series with the capacitor being charged, driven by a switching converter. The resonance between the inductor and the capacitor creates a sinusoidal current that charges the capacitor. The resonant frequency depends on the inductance and the capacitance. As the capacitor charges, the resonant frequency changes, affecting the charging characteristics. The switching converter controls the charging rate and can implement constant current or other charging profiles.
Charging efficiency is the ratio of energy delivered to the capacitor to the energy drawn from the input source. The efficiency is less than unity due to losses in the switching elements, the resonant inductor, and other circuit components. The losses depend on the charging current, the switching frequency, and the component characteristics. Higher efficiency reduces the input power requirement and the thermal load, but may require larger or more expensive components.
Switching losses occur in the semiconductor switches during the transitions between on and off states. During the transition, both voltage and current are simultaneously nonzero, causing power dissipation. The switching loss is proportional to the switching frequency, the voltage, and the current. Zero voltage or zero current switching techniques eliminate switching losses by ensuring that the switch transitions occur when the voltage or current is zero. Resonant circuits naturally provide zero current switching at the resonant frequency.
Conduction losses occur in the switches and other components while they are conducting current. The conduction loss equals the product of the current squared and the on resistance for MOSFETs, or the forward voltage drop times current for diodes and IGBTs. The conduction loss depends on the current waveform and the duty cycle. Component selection for low conduction loss may require larger die sizes or more expensive technologies.
Inductor losses include core loss from the alternating magnetic flux and winding loss from the current flow. The inductor carries the resonant current, which may have significant amplitude depending on the charging rate. The inductor design must minimize losses while providing the required inductance and current handling. Air core inductors avoid core losses but may be physically large for the required inductance. Magnetic core inductors are smaller but have core losses that increase with frequency.
Thermal management removes the heat generated by losses to maintain component temperatures within ratings. The thermal design includes heat sinks for power semiconductors, cooling for inductors and transformers, and overall enclosure cooling. The thermal resistance from components to the ambient determines the temperature rise for a given power dissipation. Active cooling with fans or liquid increases the heat removal capability but adds complexity and potential failure modes.
The balance between efficiency and thermal management involves tradeoffs at multiple levels. Higher efficiency reduces the thermal load, potentially allowing simpler cooling or smaller heat sinks. However, achieving higher efficiency may require more expensive components or larger magnetics, increasing cost and size. The optimal balance depends on the application requirements for cost, size, reliability, and operating environment.
Operating frequency selection affects both efficiency and size. Higher frequencies allow smaller inductors and capacitors, reducing size and weight. However, switching losses and core losses increase with frequency, potentially reducing efficiency. The optimal frequency balances size reduction against efficiency degradation. Modern wide bandgap semiconductors such as silicon carbide and gallium nitride enable higher frequency operation with lower switching losses, shifting the optimal frequency higher.
Reliability considerations in the thermal design include the impact of temperature on component lifetime. Semiconductor devices have lifetime that decreases exponentially with temperature. Capacitor lifetime also decreases with temperature, particularly for electrolytic capacitors. The thermal design must maintain temperatures that achieve the required reliability, which may require lower temperatures than the component ratings would suggest. Thermal cycling from the pulsed operation creates additional stress that affects reliability.
