This paper presents a real-time simulator of a photovoltaic array based on a Growing Neural Gas (GNG) controlled DC-DC buck converter circuit. The DC-DC converter, driven by the current-voltage-irradiance-temperature mapping supplied by the GNG, is able to accurately describe the set of characteristics of a PV array, including their dependence on the solar irradiance and the cell temperature. A voltage linear closed loop control of the DC-DC converter has been done. The stability of the system in the discrete domain has been proven in all load conditions. A prototype of the GNG based PV emulator has been designed and set up both in its hardware and software part. The experimental implementation of the PV input-output mapping, based on the GNG network, has been done on a DSP-2 board. Results show a good agreement between the theoretical characteristics of the PV and the simulated and experimental ones in all tested working conditions. Particularly, the GNG network has correctly learnt the PV characteristics in the 4-dimensional space and is able to distinguish in production phase the different temperature and irradiance conditions. Also transient tests show a correct and stable transition between different points of the same characteristic or points of different characteristics.