Search

Class $E^2$ DC–DC converter composed of Class $E$ power amplifier (PA) and rectifier is a promising candidate for MHz wireless power transfer. It is soft-switching based and able to achieve high efficiency at MHz frequency. However, the converter implemented through traditional static design is sensitive to the variations of operation condition. Its performance gets deteriorated when DC load and coil coupling deviate from their respective optimum values. This paper demonstrates that the degradation of system efficiency is mostly due to the mismatch of PA load, and the efficiency drop can be efficiently improved by adding an L-type impedance matching network (IMN) after PA. A fixed IMN is sufficient to maintain a high efficiency, while a tunable IMN is required to ensure stable output power when operation condition dramatically changes. Key techniques, particularly system-level optimization, are discussed in this paper that ensure high efficiency over a wide range of variation in operation condition and also with reduced capacitor/inductor tuning ranges in the IMN. The 6.78-MHz Class $E^2$ DC–DC converters with and without the fixed/tunable IMN are fabricated and measured for validation purposes. The experimental results show that both high efficiency ( $>$66%) and stable output power (around 9 W) are maintained for the tunable converter when there are variations in the DC load and coil coupling. [ABSTRACT FROM AUTHOR]

In this paper, an analytical model for the Class-F rectifier RF-dc conversion efficiency is presented. This model analyzes each kind of diode power losses in the rectifier due to the diode series resistor, junction capacitor, built-in potential, and breakdown voltage based on the time-domain diode voltage and current waveforms. The model provides a simple calculation routine to determine the threshold input power at which the peak reverse voltage across the diode starts to exceed the diode breakdown voltage and cause the diode efficiency to decrease. Two sets of closed-form equations are derived to calculate the Class-F rectifier efficiency in the input power region where the input power is either smaller or larger than the determined threshold input power, respectively. The implementation of a MATLAB code for the model calculation is also presented. Using this tool, the diode efficiency in Class-F rectifiers of different diode parameters can be easily determined, which provides a useful guideline for the optimal diode selection in various applications. The calculated diode efficiency using this model is compared with the Class-F rectifier ADS simulation at different load conditions, and all the results agree well with the model. To verify the model at different frequencies, two Class-F rectifiers working at 900 MHz and 5.8 GHz, respectively, are designed with the highest measured efficiency of 80.4% for the 900-MHz rectifier and 79.5% for the 5.8-GHz rectifier. Both of the measurement results agree well with the model prediction, which further verified the accuracy of the proposed model. [ABSTRACT FROM AUTHOR]