SystemC-AMS modeling of photodiode based on PWL technique to be used in energy harvesting CMOS image sensor

Accurate modeling of the circuits at different steps of design process is an essential factor for having a cost effective and high throughput final circuit implementation. In this paper an efficient system level modeling of a Digital Pixel Sensor (DPS) architecture with the capability of harvesting the energy of illuminated light is presented. The SystemC and analog/mixed signal (AMS) extension is used for the modeling. The proposed model mainly focuses on the accuracy of the fabricated photodiode model, a 11×11μm2 n+/p-substrate photodiode which is fabricated in a standard CMOS 0.18μm technology. This accuracy is important for both light sensing and energy harvesting phases. The model uses electrical linear network (ELN) and timed data flow (TDF) models of computation (MoC) of SystemC-AMS. Simulation results of the proposed model are in good agreement with experimental results of the photodiodes. In order to evaluate the photodiode model in pixel architecture, a system level model of a DPS with capability of harvesting illuminated energy is also proposed and implemented in SystemC-AMS framework.