Time Scales

The accuracy of the National Research Council 2.1 m cesium beam frequency standard has been re-evaluated using the two NRC hydrogen masers as stable comparison standards. At the time of re-evaluation, the various uncertainties gave rise to an error limit of :!:3.8 x 10-12, with a most probable value of :!:1.5 x 10-12 if they were assumed to be random. Mter application of all corrections, the agreement with several other cesium standards, as determined from monthly VLF comparisons during January to March, 1968, appears generally to be within the narrower limits of :!:1.5 x 10-12. this, and in the synthesizer settings, are used to calculate the corresponding changes in the apparent cesium resonant frequency as determined by the cesium standard. The frequency ix of the 5 MHz crystal oscillator locked to the cesium resonance can be expressed as follows : 9192632000 .fw = 9192631770 + ~iH + L~i, (1) where 9192632000 refers to the nominal frequency synthesized, 9192631770 Hz is the defined value of the cesium hyperfine transition frequency, L~i, represents the sum of systematic errors ~i" i = 1, ..., n, and ~iH is the O field correction. In the NRC standard, ~iH is determined from measurements of the (4, -4) +-* (4, -3) transition frequencies, 'Vk, k = 1, ...8, over eight adjacent sections of the drift space between the two oscillating field regions. Introduction With the completion of a pair of hydrogen masers at the National Research Council [1] it has become possible to re-evaluate the accuracy of the NRC 2.1 m cesium beam frequency standard. This standard, in operation since 1965 [2], has exhibited much closer agreement [3,4,5] with other cesium frequency standards than the original estimate of accuracy:!: 1.9 x 10-11, might indicate. This initial estimate was based on experimental comparisons with a stable crystal oscillator, and appears to have been influenced mainly by the limitation of frequency stability imposed by this crystal rather than that of the standard itself. Use of the hydrogen masers, which possess frequency stabilities over short periods far in excess of that of the cesium standard, has eliminated this primary source of frequency instability, and has appreciably improved the precision of estimating the accuracy of the cesium standard. is Method of Re-Evaluation The technique used has been to observe changes in the relative frequency of the hydrogen and cesium standards when certain modifications to the cesium standard alone are made. The output of the 5 MHz crystal locked to the cesium resonance is used to drive a multiplier and synthesizer system used to detect the maser signal. The beat frequency between these two signals at 1420 MHz is then monitored, and changes in 24 Metrologla ~ = 230-<5IH}:;<511 (4) I 9192632000 and any change in this offset, d (7~ due to changes in I.5fH, or in :L:l.5f" is