Since the introduction of the pyramid wavefront sensor [P-WFS] concept (Ragazzoni), numerous investigations have clearly shown its ability to achieve better performance (sensitivity, dynamic range) than the standard Shack-Hartman [SH-WFS]. It has recently been successfully implemented on LBT and has already been provided very interesting results (Esposito et al). Then, most of the future adaptive optics [AO] systems, mainly for ELT instrumentation, will probably integrate one or several pyramidal sensors. However, the pyramid behavior still needs to be extensively studied in order to ensure its optimization in real conditions of operation. So, the coupling in an AO loop and the control of this type of sensor is fundamental for an efficient implementation in the future AO systems. At LAM, we recently carried out in labs demonstration of an extremely performant pyramid sensor (up to 60x60), using particularly an OCAM2 detector (1.5 kHz, RON close to zero). Both modulated and fixed configurations are investigated and compared with numerical models. The P-WFS is being coupled with a dedicated RTC and a 12×12 DM to achieve a first AO closed loop operation. For modulation, a fine control is needed: a specific electronic module, interfaced with the RTC, is being developed to drive the TT mirror (OCAM2 triggering). Then, various TT mirrors are under test to determine a suitable one. After tests of the pyramid specificities (optimiziation, calibration and operation procedures), the P-WFS will be tested on-sky and compared with an already existing SH-WFS (using the same OCAM²) on the ONERA bench.