Team : CIAN
Departure date : 02/18/2014
Supervision : Dimitri GALAYKO
Study, optimization and silicon implementation of a smart high-voltage conditioning circuit for electrostatic vibration energy harvesting system
This thesis studies vibration energy harvesters (VEHs) with electrostatic transducers. A typicalelectrostatic VEH consists of a transducer, conditioning circuit and a storage element. This work isfocused on investigations of the auto-synchronous conditioning circuit based on the charge pump andthe inductive flyback energy return (buck DC-DC converter). This architecture is very promising since iteliminates precise gate control of transistors employed in synchronous architectures.
This thesis addresses the theoretical analysis of the conditioning circuit. We developed an algorithmthat by proper switching of the flyback allows the optimal energy conversion strategy. By adding thecalibration function, the system became adaptive to the fluctuations in the environment.
Another contribution is the realization of the proposed algorithm on the circuit level. The major designdifficulties were related to the high-voltage (HV) requirement and the low-power design priority. Wedesigned a HV analog controller of the switch using AMS035HV process. Its power consumptionvaries between several hundreds nanowatts and a few microwatts, depending on voltage levels of thecharge pump, frequency of the flyback switching, frequency of calibration, etc.
We also implemented on silicon, fabricated and tested a HV switch with a novel low power level-shifting driver. By mounting on discrete components the charge pump and flyback circuit andemploying the proposed switch, we characterized the wideband HV operation of the MEMS transducerprototype fabricated alongside this thesis. When excited with stochastic vibrations having anacceleration level of 0.8 g rms distributed in the band 110-170 Hz, up to 0.75 µW of net electricalpower has been harvested.
Defence : 02/18/2014 - 14h - Site Jussieu 25-26/105 Jury members : Elie Lefeuvre, professeur, univ. Paris-Sud, IEF, [rapporteur]
Pascal Nouet, professeur, univ. Montpellier, LIRMM, [rapporteur]
Skandar Basrour, professeur, INPG, TIMA
Philippe Basset, enseignant-chercheur, ESIEE
Pascal Fornara, PhD, ST Microelectronics
Dimitri Galayko, HDR, UPMC, LIP6
François Pêcheux, professeur, UPMC, LIP6