The tip-tilt stabilisation system of the 1.8 m Auxiliary Telescopes of the Very Large Telescope Interferometer was never dimensioned for robust fringe tracking, except when atmospheric seeing conditions are excellent. Increasing the level of wavefront correction at the telescopes is expected to improve the coupling into the single-mode fibres of the instruments, and enable robust fringe tracking even in degraded conditions. We deployed a new adaptive optics module for interferometry (NAOMI) on the Auxiliary Telescopes. We present its design, performance, and effect on the observations that are carried out with the interferometric instruments.

We extend semi-analytical computations of excitation rates for solar oscillation modes to those of other solar-like oscillating stars to compare them with recent observations. Numerical 3D simulations of surface convective zones of several solar-type oscillating stars are used to characterize the turbulent spectra as well as to constrain the convective velocities and turbulent entropy fluctuations in the uppermost part of the convective zone of such stars. These constraints, coupled with a theoretical model for stochastic excitation, provide the rate 'P' at which energy is injected into the p-modes by turbulent convection. These energy rates are compared with those derived directly from the 3D simulations. The excitation rates obtained from the 3D simulations are systematically lower than those computed from the semi-analytical excitation model. We find that Pmax, the excitation rate maximum, scales as (L/M)^s where s is the slope of the power law and L and M are the mass and luminosity of the 1D stellar model built consistently with the associated 3D simulation. The slope is found to depend significantly on the adopted form of the eddy time-correlation ; using a Lorentzian form results in s=2.6, whereas a Gaussian one gives s=3.1. Finally, values of Vmax, the maximum in the mode velocity, are estimated from the computed power laws for Pmax and we find that Vmax increases as (L/M)^sv. Comparisons with the currently available ground-based observations show that the computations assuming a Lorentzian eddy time-correlation yield a slope, sv, closer to the observed one than the slope obtained when assuming a Gaussian. We show that the spatial resolution of the 3D simulations must be high enough to obtain accurate computed energy rates.