An Input Power-Aware Efficiency Tracking Technique for Energy Harvesters in IoT

This paper presents a two-dimensional maximum efficiency tracking technique for switched-capacitor based energy harvesting circuits in IoT applications. Conventional maximum power point trackers that sweep switching frequency (fSW) achieve high power conversion efficiency (PCE) for a limited range of input power. The purpose of this technique is to extend the input power harvesting range at which high power conversion efficiency can be achieved. To realize that, the proposed input-power aware system adds the capability of tuning the average load current to minimize converter power losses based on the sensed input power. This is fundamentally based on a discontinuous charging technique that delivers current to the load only during a controllable clock non-overlap time. Hence, the system can dynamically optimize the maximum PCE to actual ambient power levels. The system also tunes fSW to deliver the maximum output power. To demonstrate this technique, the system is designed and simulated in 90nm CMOS technology. For the same 40kΩ load conditions, the proposed technique achieves at least 45% PCE at a 5uW input power, as opposed to a conventional system which requires at least 18uW of input power to maintain the same PCE. The harvesting range is therefore extended by 72%.