Program

Abstract:

Since decades magnetically-deflected cesium beam tubes have been used as primary reference frequency standards for various industrial applications (metrology, time scale, navigation, communication …). When an increased frequency stability and a better accuracy are targeted while keeping a compact size product, one of the single possibilities is to increase the atomic flux by overheating the atomic beam oven. An improvement by a factor three is achievable (Allan deviation = 8.5E-12 / sqrt(tau) instead of 2.7E-11 / sqrt(tau) ) at the expense of significant lifetime reduction (5 years instead of 10 years for a long-life tube). By replacing the traditional magnetic deflection by an optical pumping process [1], the useful atomic beam flux can be significantly increased (factor 100), without reducing the tube lifetime as the atomic beam oven is kept at a reasonable temperature (100°C). Frequency stability improvements by an order of magnitude (1 to 3E-12 / sqrt(tau)) have been demonstrated at laboratory prototype level [2, 3]. Moreover a commercial high performance clock (7E-12 / sqrt(tau)) is commercialized in China [4]. As for almost all frequency standards using atom-light interaction processes, the AC Stark shift (light shift) limits the long-term stability and accuracy, including atomic beam standards [5]. One of the main challenge of continuous optically-pumped beam standards is to improve the short-term stability without degrading its long-term stability in particular for the light shift influence. Oscilloquartz has developed an industrial optically-pumped cesium beam clock. It fits in a standard 19’’ wide rack (450 mm), 3U high (133 mm) and 450 mm deep. It weighs 23 kg and consumes 35 W in standard lab conditions. In order to maximize the clock reliability, the simplest optical scheme has been chosen: a single laser wavelength is used (no acousto-optic modulator) for the atomic state selection and the atomic state detection (pumping transition D2:3->4’). This transition has the advantage to have a minimal power light shift coefficient. In addition, to further reduce this sensitivity, a special algorithm has been developed and implemented (patent pending). Its benefits are a reduced clock thermal sensitivity and an improved long-term stability. The frequency stability of the optically-pumped cesium beam clock has been measured against a reference active hydrogen maser for 50 days. The overlapped Allan deviation perfectly averages as ADEV = 4E-12 / sqrt(tau) down to 4E-15 @ 1 Ms, without showing any frequency flicker noise floor. In term of the time interval error (TIE), this clocks has wandered by +/- 6 ns over 50 days in free running mode. Such holdover timing performance for an industrial and long-life product represents a significant improvement compared to commercially available magnetic deflected cesium beam clocks. It can find place in numerous high performance and high resiliency applications. In particulier Positioning, Navigation and Timing (PNT) become of significant and strategic importance. In case of Global Navigation Satellite System (GNSS) outages, such optically-pumped frequency standard can be an interesting, high performance and commercially available backup solution. ? [1] G. Avila et al.. State selection in a cesium beam by laser diode optical pumping, Phys. Rev. A , Vol. 36, Iss. 8, 1987, pp. 3719-3728. [2] R. Lutwak. Optically pumped cesium-beam frequency standard for GPS-III, in Proc. of “Precise Time and Time Interval (PTTI 2001)”, Long Beach, CA, USA, 27-29 November 2001, pp. 19-32 [3] C. Sallot et al. 3E-12 ?-1/2 on industrial prototype optically pumped cesium beam frequency standard, in Proc. of the “Joint meeting IEEE-IFCS / EFTF 2003”, St Petersburg, FL, USA, 4-8 May 2003, pp. 100-104 [4] Y. Cao et al. A merchandized optically pumped cesium atomic clock, in Proc. of the “Joint meeting EFTF / IEEE IFCS 2017”, Besançon, France, 9-13 July 2017, pp. 618-621 [5] A. Brillet. Evaluation of the light shifts in an optically pumped cesium beam frequency standard, Metrologia, Vol. 17, 1981, pp. 147-150