Radio fluxes of Cyg OB2 #5 as a function of phase in the 2366 d cycle (with t_AB+C,0 = 34909) along with the best-fit model with s = 2 (see text). The crosses stand for the VLA data from Kennedy et al. (2010), whereas the filled dots indicate the VLBA fluxes from Dzib et al. (2013) multiplied by a factor 4/3 and with a 5 mJy offset added (see text). The VLBA data were not used in the fit but they nevertheless agree well with VLA results.