An Energy-Efficient, Adaptive Voltage Scaling Output Stage for Neural Stimulators Based on Sepic Converter

Peng Wu,Songping Mai
DOI: https://doi.org/10.1109/ipemc.2016.7512680
2016-01-01
Abstract:This paper presents an energy-efficient output stage for neural simulators which generate needed bi-phasic pulses. Energy efficiency is improved by adopting a single-ended primary inductance converter (SEPIC) that has a wide range of output voltage varying with the stimulus current on tissue load for adaptive voltage scaling. Comparing with a typical current-controlled stimulator which has a fixed high voltage supply, this approach reduces power consumption on the current driver. On-chip compensation is achieved with a Type II compensator while the DC-DC converter is operating in the discontinuous conduction mode (DCM) for reducing power consumption. Voltage-mode control is adopted for its robustness and no need for slope compensation. Measured results indicate that the stimulus current could change from 0.6 mA to 11.8 mA with the step of 0.09 mA when the resistance of tissue is set as 1.5 kΩ and an output range of 0.9 V to 17.8 V with an input voltage of 3.7 V from a li-ion battery is achieved. The SEPIC converter achieves >60% power-conversion efficiency (PCE) when stimulus current was above 2.2 mA and this design can save power by up to 68% compared to the output stage of a typical current-controlled neural stimulator. The circuits have been implemented in a commercially-available 1 μm 40 V BCD process.
What problem does this paper attempt to address?