Some Causes and Cures of Frequency Instabilities (Drift & Noise) in Cesium Beam Frequency Standards

Abstract

Frequency drift of the order of several parts in 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">13</sup> per year is often observed in commercial cesium beam frequency standards, and on some occasions significant changes in the white noise frequency modulation level is also observed. Recently at the National Bureau of Standards some standards with these types of problems have been analyzed and their velocity distributions measured. A comparison of the changes in drift and noise performance with measurements of the velocity distribution leads to some interesting interpretations. Changes in focusing voltage at or in the vicinity of the detector may cause the finite surface area of the detector to act like a velocity selector; i.e., the detection efficiency of cesium atoms mapped across the surface of the detector changes with time as a result of changing electric fields which focus the cesium ion-beam. Changes in the microwave power cause changes in the most probable cesium atom velocity, which transduce via the Ramsey cavity phase shift into frequency changes. The magnitude of the cavity phase shifts in a cavity was inferred by reversing the current through the C-field inducing mixing of the m <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">F</sub> states.