Table of contents

The Thompson-Lampard calculable cross-capacitor is used at the Physics and Engineering Laboratory (PEL) for the derivation of the ohm. A description of a calculable capacitor capable of determining the value of a 0.5 pF capacitor with an uncertainty of 2.2 parts in 10⁸ is given. A detailed model of the frequency dependence of calculable capacitors has been developed. This model is used to specifically derive the frequency correction for the PEL calculable capacitor. The model predicts a departure from a simple square-law frequency dependence of the correction and explains the observed frequency-dependent non-linearity of the capacitor. The model is also used to investigate the effect of changing the length and diameter of the capacitor and the significance of the use of ferromagnetic materials is also investigated.

The reflection coefficient and complex discontinuity admittance of a coaxial open circuit with a hollow center conductor are derived from fields correct to first order in the skin depth. Results show an admittance terminating the line at the plane of the discontinuity and consisting of a resistance in parallel with a capacitive reactance. The first-order fields are also used to derive equations for the characteristic admittance, series impedance and shunt admittance of the line. These equations include terms neglected in the well-known expressions used to calculate the line parameters, enabling error limits to be assigned to the latter.

A critical review of all beam-reversal phase-shift measurements carried out on NBS-6 (NBS-5) since 1973 is presented. The analysis takes into account a recent study of frequency shifts in atomic-beam frequency standards resulting from overlapping tails of adjacent lines. A strongly correlated variation of the cavity phase difference with time is shown, which appears to fit nicely on an exponential decay curve with a time constant of 3.7 years. The suggestion is made that some mechanical relaxation in the microwave structure may be responsible for this phenomenon. Furthermore, the line-pulling model seems to suggest a better explanation for observations previously attributed to distributed cavity phase shift. Verification of these interpretations through more involved evaluations may allow substantial reductions in the error budget for the standard.

An international intercomparison of neutron fluence-rate at 144 and 565 keV was organized under the auspices of the Bureau International des Poids et Mesures. Small indium foils were used as the transfer device, the induced radioactivity from the ¹¹⁵In(n, γ)^116mIn reaction being assayed with a 4πβ counter. The results of the intercomparison showed very good agreement between four standards laboratories at 144 keV, but there was a wider spread between six laboratories at 565 keV. The suitability of the indium-foil transfer device is also discussed, and it is concluded that in the future a more accurate system will be required.

As a result of this intercomparison, it has been found that neutron scatter in the target backing, when using the ⁷Li (p,n) reaction as a source of neutrons in the energy range 100 to 600 keV, has apparently been underestimated in the past, which can have a large effect on activation capture cross-section measurements.

An intercomparison has been made of the frequencies of He-Ne lasers stabilized using intracavity saturated absorption in ¹²⁷I₂ at 633 nm. Lasers at ten standards laboratories (PEL (New Zealand), NRC (Canada), NBS (USA), NPL (UK), PTB (FRG), NPL (India), NIM (China), NRLM (Japan), KSRI (South Korea), and NML (Australia)) were each compared in turn with a portable laser from PEL during March to May 1984.

The optical frequencies of all participating lasers exhibited a standard deviation of 23 kHz (5 parts in 10¹¹) with a range of ± 40 kHz. This standard deviation is a factor of seven smaller than the standard deviation of the absolute value recommended for the optical frequency of these lasers. If allowance is made for one iodine cell believed to be contaminated, the standard deviation is reduced to 18 kHz. Power-shift coefficients were measured, and ranged from + 1.4 kHz/mW to - 6 kHz/mW, referred to intracavity power.

It is concluded from the measured frequency offsets and power shifts that the best frequency reproducibility amongst different designs of laser is likely to be obtained at low levels of intracavity beam power (5 - 10 mW). The frequency reproducibility was not improved by extrapolating to zero power.