With the rise of Internet-of-Things and connected devices, the need for self-powered wireless sensor nodes is ever increasing. One promising technology for self-powered sensor nodes in noisy environments is acoustic energy harvesting: deriving energy from ambient sound. Existing acoustic energy harvesters are typically based on resonant structures, yielding narrowband, and therefore low-energy, collection from broadband noise sources. In addition, existing acoustic energy harvesters tend to exhibit MEMS-scale sizes, with consequently low power outputs. This two-part work addresses the size and bandwidth of such harvesters. A large-scale acoustic energy harvester is developed, based on piezoelectric PVDF (polyvinylidene fluoride) film 100 cm 2 in size. The harvester is designed to minimize reactive impedance, allowing for circuit loading for broadband energy harvesting in Part II of this paper. An energy-based dynamic analysis of the harvester driven by an acoustic source yields an equivalent circuit model and subsequently a Thévenin equivalent model of the harvester. The model is validated by experiments including acoustic and electric measurements, laser Doppler vibrometry, and finite element analysis. It is used to develop loadings in Part II.