This paper is aimed to present an experimental criterion that allows researchers and designers to evaluate the performance of DC micro-grids dedicated to charging operations of full EV. The laboratory methodology is explained through tests on a demonstrator of power architecture, specifically designed as simplified case study of a DC charging station for fully-electrified low-power two-wheeler, such as: electric scooters and bikes. This experimental prototype is composed by a 20. kW AC/DC bidirectional grid-tied converter, which realizes the DC conversion stage, and two DC/DC power converters, interconnected with the micro-grid through a DC bus. The power architecture provides on one hand the charging operations of electric two-wheeler battery packs and on the other hand the integration of the micro-grid with an ES buffer, which has the main function of supporting the main grid while an electric vehicle is on charge. The performance of the considered architecture is characterized and analyzed in different operative conditions, through a specific management of the energy fluxes. The laboratory tests evaluate efficiency, charging times and impact on the main grid, with specific reference to the DC charging operations of electric scooters. The obtained experimental results show the advantages of adopting the DC buffer architecture, compared with an AC commercial battery charger, considered as reference in this work. Finally, the numerical data of this paper, related to the single power components and to the experimental results evaluated for the whole demonstrator, effectively support the lack of knowledge in the literature about charging stations for EVs. In fact, these pieces of information, based on experimental tests, are expected to support the building of simulation models and the identification of the best energy management and control strategies to be adopted in a smart-grid scenario, characterized by distributed generation systems including renewable energy sources.