We present a code which can be used to simulate the light curve of a binary system harbouring a non-circular accretion disk, and show that such a system geometry can produce light curves with asymmetric minima and different heights of the maxima. We have applied this code to check for the possible existence of a non-circular accretion disk, as suggested by Smak & Plavec (1997), in the Algol-type binary KU Cyg. We compare the best solutions for two sets of multicolour light curves, obtained within circular and non-circular disk models.

We present and discuss the Hα (λ = 6563Å) observations of the contact (W UMa type) binaries AE Phoenicis and YY Eridani, obtained in 1989, 1990 and 1995 with the CAT/CES telescope of the Southern European Observatory (ESO). In particular, we compare the intrinsic equivalent widths of both components with the NextGen theoretical models and the saturation limit. We find that the average Hα equivalent widths are close to the saturation border and that the primary components have excess Hα-emission, indicating enhanced chromospheric activity. This is compatible with both theoretical and observational suggestions that the primary is the more magnetically active component and is filled with (mostly unresolvable) dark spots and associated chromospheric plages.

We discuss the impact that the space experiment CoRoT, whose launch is scheduled for late 2006, will have on the field of close binary.