2006MNRAS.372.1129G

We have measured the orbital light curve of dwarf nova OY Carinae on eight separate nights between 1997 September and 2005 December. The measurements were made in white light using CCD photometers on the Mt Canopus 1-m telescope. The time of eclipse in 2005 December was 168±5s earlier than that predicted by the Wood et al. ephemeris. Using the times of eclipse from our measurements and the compilation of published measurements by Pratt et al., we find that the observational data are inconsistent with a constant period and indicate that the orbital period is decreasing by 5±1x10^–12 s/s. This is too fast to be explained by gravitational radiation emission alone. It is possible that the change is cyclic with a period of ∼35 yr and a fractional period change ΔP/P = 2.6x10^–7. This is probably due to solar-cycle magnetic activity in the secondary. There are also large systematic deviations, with a time-scale of years, from a sinusoidal modulation.