Self-powered light sensor for simultaneous intensity-and-direction sensing and maximum-energy harvesting with shared photodiodes

Abstract

Self-powered light sensors are ubiquitous in internet-of-things applications for sustained operation without battery replacement. Figure 1 right shows an example, in which a museum needs thousands of light sensors to sense light intensity and direction for lighting control. Photodiodes (PDs) are inherently suited to both light sensing (LS) and light energy harvesting (LEH). Figure 1 top-left shows a PD’s P-V curve. A PD must be forward-biased for simultaneous LS and LEH, and be biased at its maximum powerpoint (MPP) voltage VMP for maximum LEH. Recent PD-based light sensors can sense light intensity [1]–[4] and/or direction [4] and harvest light energy [1], [3]. However, [1] performs LS and LEH with separated PDs. The PDs for LS in [1]–2, [4] are zero/reverse-biased and thus cannot perform LEH. Although the PD in [3] is forward-biased, the PD harvests little energy since it operates near the open-circuit voltage V <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">oC</inf> and consumes all for proper LS function. Light direction sensors [4] commonly use masks on PDs to increase light-direction sensitivity, but the incident light at some angles is largely blocked. Thus, the incorporation of LEH into these light sensors [1]–[4] is inefficient. This paper proposes a self-powered light sensor for simultaneous LS and maximum-LEH with the same PDs, as shown in Fig. 1 bottom-left. Instead of minimizing the power consumption of the sensor operating under high dynamic range, the proposed sensor can harvest more energy while detecting higher light intensity. This sensor includes a light-to-frequency converter (LFC) and a frequency-to-digital converter (FDC). The LFC uses a dc-dc converter with input (output) connected to a PD (an energy storage unit, ESU) to regulate the PD’s voltage VP around VMP with pulse-skipping modulation (PSM). Thus, the average frequency f <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ILo</inf> of the inductor current I <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Lo</inf> charge packet is linear to the light intensity, while the packets deliver energy from a PD to an ESU. The dc-dc converter is extended to multiple inputs (dual outputs) to sense light direction (provide a regulated voltage V <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">L</inf> for the FDC and RF TX). The FDC simply uses a counter and a look-up table for readouts. Combining LEH and LS alleviates the need for reducing FDC power in high light intensity. In addition, a low-cost plastic Fresnel lens is adopted to concentrate incident light to increase light-direction sensitivity [5], which can be further increased by placing the PDs on different surfaces of a polyhedron [6].