Last issues about Building Integrated Photovoltaic Systems (BIPV) still show average Performance Ratio (PR) values in the range of 0.75-0.80. The main causes well known: partial shadows, temperature effects, PV inverter losses, thermal losses, etc. and mismatching losses. Ideally, all the modules work in the same conditions, but differences between modules really exist due to differences in the working temperature, the inclination or orientation angles, differences in the I-V characteristic coming from the manufacturing process, etc. The effect is that the output power of the complete PV system is lower than the addition of the power of each PV module. These mismatching losses can be decreased by means of suitable electronics. This paper presents the experimental results obtained over PV systems equipped with controlled PV modules, PV modules with low cost and high efficiency DC-DC converters, including MPPT algorithm and other functions, such as power control and Power Line Communications (PLC). Tests have been divided into two great categories: tests on the electronic performance of the DC-DC converter and tests on grid-connected PV systems with multiple DC-DC converters. Many of these tests have been carried out taking advantage of the PV System Test Platform, a powerful tool especially designed by Robotiker to evaluate all kind of PV systems, especially systems with differences between modules. Aspects of the DC-DC converter performance have been detailed and among the most important experiments, the paper analyses different situations such as partial shadows, different inclined planes, PV systems with different PV modules, and finally a comparison between a conventional system and a system composed by controlled PV modules have been described. To sum up, the importance of a good system dimensioning is analysed, with very interesting results. (C) 2007 Elsevier Ltd. All rights reserved.