This paper describes a one year work started at CSEM in 2009. At the end of the project, a system level prototype of an Extra Small Atomic Reference (XSAR) should be realized by teaming up different know-hows available at CSEM. The work followed a top down systematic. A breadboard level coherent population trapping (CPT) atomic clock working on Rubidium-87 atoms confined in a glass cell and controlled by laboratory electronics was built at first. The different building blocks of this breadboard CPT atomic clock were characterized in order to select them for the XSAR prototype. The CPT signal and the corresponding operating parameters were accordingly optimized. The knowledge acquired in the first phase was used to define the goals to be re ached with the first XSAR prototype. Priorities were settled on the miniaturization of the optical assembly unit (laser and optics), the atomic vapor cell unit and finally the RF electronic unit. After one year, the first XSAR prototype showed promising short term frequency stability reaching 1.8·10-11?-1/2 when fully driven by laboratory RF electronics and 2·10-10?-1/2 when driven by an integrated RF synthesizer.
EFTF-2010 24th European Frequency and Time Forum, Noordwijk (Netherlands), pp. 1-6