The cm-wavelength radio flares on Cygnus X-3 have been studied for many years. Our recent paper [1] looked again at the minor flares (flux density 𝑆 of a few 100 mJy) and compared their properties with those of a sample of major flares (𝑆 > 1 Jy). We find that the minor flares have risetimes and duration of ∼ 1 hour, as opposed to ∼ days for the major flares. Minor flares show more
rapid expansion of the synchrotron radiation emitting material than in the strong flares. They also appear closer to the binary, whereas the large flares form a more developed jet, i.e. the jets formed in minor flares are short and wide, those in major flares are long and thin. We used the results of Fender & Bright [2] to calculate the magnetic field and expansion velocity as a fraction 𝛽 of the speed of light under minimum energy conditions when the source is optically thick for samples of minor and major flares. The minimum power in the source was found using the rise time of the flares. The minor flares have lower minimum power but have larger velocities and energy densities than the major flares. Minor flares can occur while a major flare is in progress, suggesting an indirect coupling between them. The spectral evolution of the minor flares can be explained by either an expanding synchrotron source or a shock model. Further investigation requires high resolution VLBI observations at the 1 mas level if we wish to understand the development of the
source. The problem is that Cygnus X-3 is strongly scattered by the interstellar medium so high frequencies in the several 10s of GHz are required for the resolution needed. The minor flares
are rapid and require high cadence observations to follow the flux density behaviour and hence only short snapshot VLBI observations can capture the structure. Large numbers of telescopes
are required which is a problem at the highest frequencies. We discuss the VLBI possibilities and trade-offs for this awkward object.