Context. Period variations have been detected in a number of eclipsing close compact binary subdwarf B stars (sdBs) and these have often been interpreted as caused by circumbinary massive planets or brown dwarfs. Various evolutionary scenarios have been proposed for these stars, but a definite mechanism remains to be established. Equally puzzling is the formation of these putative circumbinary objects which must have formed either from the remaining post common envelope circumbinary disk or survived its evolution. Aims. In this paper we review the eclipse time variations (ETVs) exhibited by seven such systems and explore if there is conclusive evidence that the ETVs observed over the last two decades can reliably predict the presence of circumbinary bodies. Methods. We report 246 new observations of the seven sdB systems made between 2013 September and 2017 July using a worldwide network of telescopes. We combined our new data with previously published measurements to analyse the ETVs of these systems. Results. Our data shows that period variations cannot be modelled simply on the basis of circumbinary objects. This implies that more complex processes may be taking place in these systems. From eclipse time variations, it has historically been suggested that five of the seven binary systems reported herein had circumbinary objects. Based on our recent observations and analysis only three systems remain serious contenders. We find agreement with other observers that at least a decade of observations is required to establish reliable ephemeris. With longer observational baselines it is quite conceivable that the data will support the circumbinary object hypothesis of these binary systems. Also we generally agree with other observers that larger values of (O-C) residuals are found with secondary companions of spectral type M5/6 or earlier as a result of an Applegate type mechanism