In photovoltaic (PV) solar installations, the operative personnel are mainly concerned about energy production and safe operation. Undetected and un-located faults generate severe power losses and security risks; in this sense, the early detection and location of faults in the solar array is nowadays a necessity. However, many factors must be considered, i.e., failure modes per technology, the degree of fault severity, and the amount and disposition of solar panels, inverters, and other system equipment. This paper presents a fault detection and location technique (FDLT) that divides the PV array into groups located in the same string. The FDLT relies on differential voltage measurements across PV panels. It also presents a strategy to locate the IoT devices in different places of the PV array. The fault detection and localization algorithm compares the average voltages of each module against the averages of other photovoltaic modules. The algorithm is validated through numerical simulations with a PV array of \(9\times 4\) modules faulted with the following conditions: short-circuits, open-circuit, open-circuit inside the module, short-circuit to ground, permanent partial shadow, and internal degradation. The obtained results were quite satisfactory, with a hundred percent of faults being located under specified restrictions; moreover, the algorithm requires low computational power of only four samples per channel, making it suitable for embedded systems in real-time applications.