Long-period binary systems containing a B-type hot subdwarf (sdB) and a main-sequence (MS) companion are thought to originate from binary interactions involving stable mass transfer from the red giant, the progenitor of the sdB, to the MS companion. However, despite the recent progress in modelling their population, some of their observed properties are not entirely understood. Because determining their orbits requires extended campaigns of high-resolution spectroscopic observations, only a limited number of long-period sdB binaries have been studied with completely determined orbital parameters. We aim to expand the current sample of long-period sdB binaries with fully determined orbital parameters through the analysis of high-resolution spectroscopic data. In addition, the atmospheric parameters of the cool companions are analysed. Increasing the number of well-characterised systems will provide valuable insights into their formation channels and main characteristics. A sample of 32 wide binary systems containing sdB stars was selected for the analysis of the radial velocity (RV) curves of both companions. The dataset consisted of high-resolution spectra obtained with the HERMES and UVES spectrographs. The orbital parameters were derived by simultaneously fitting Keplerian orbits to the RVs of the sdB and its companion. The atmospheric parameters of the cool companions were determined using the GSSP code, which analyses the master spectra of the systems with a grid of LTE atmospheric models. An additional sample of wide sdB binaries was built by cross-matching the Gaia NSS catalogue with literature catalogues of sdB candidates and spectroscopically confirmed sdB systems. The outcomes from both samples were compared with existing theoretical models to assess their consistency with current formation and evolutionary scenarios. We obtained complete orbital solutions for 32 wide sdB binaries. The orbital period distribution of the ground-based spectroscopic sample is in reasonable agreement with population-synthesis predictions, except for two outliers. The CMD further suggests that current models overpredict systems with the coolest companions, since the observed systems with BP-RP>0.3 are associated with companions hotter than 6000K. The observed period-mass ratio distribution is consistent with recent population synthesis predictions, and suggests that the unexplained second branch found in these models is mainly populated by old systems. Excluding the two long-period outliers, the data do not support a clear increase in eccentricity with orbital period, whereas the Gaia-based candidate sample shows discrepant behaviour, owing to selection effects and larger uncertainties.