The conventional algorithm of perturb and observe (P&O) is widely applied due to its simplicity, low cost and easy implementation. However, it suffers from instabilities during rapid changes of weather and/or oscillation around maximum power point (MPP) at steady state. Instabilities occur due to the incorrect decision taken by the conventional P&O algorithm at the first step change in duty cycle during the rapid change in radiation. The reason for the steadystate oscillation is the continuous perturbation and tradeoff between step sizes and the convergence time. This study presents a modified P&O algorithm to overcome such drawbacks. It uses a constant load technique to help the conventional P&O algorithm for recognising the cause of power change and to enable it in taking the right decision at first step change in duty cycle during rapid change of weather. The proposed algorithm is simulated using a single solar photovoltaic module of 80 W and a DC/DC boost converter. It is validated experimentally and implemented within an embedded microcontroller. The experimental setup presents a proposed model-based design methodology that uses measurements’ data for MPP tracking systems’ design. It combines hardware-in-the-loop simulation and prototype testing using actual weather measurements. Simulation and experiments show excellent results.