This paper considers thermodynamic conversion of solar energy into electric energy (up to maximum 50 kWe), presenting a very brief review of the possible systems: the ‘Dish/Stirling’ technology, which relies on high temperature Stirling engines and requires high solar energy; low temperature differential thermal engine using direct solar energy without any concentration but with very low power per unit volume or unit mass of the system; and the intermediate solar energy concentration ratio.A theoretical investigation on the coupling of a two-stage parabolic trough concentrator with a reciprocating Joule cycle air engine (i.e. an Ericsson hot air engine in open cycle) is presented. It is shown that there is an optimal operating point that maximises the mechanical power produced by the thermal engine. The interest of coupling a simple, low cost parabolic trough and a simple, low technology, mid-DT Ericsson engine is confirmed.