Title
Neuchâtel Observatory thallium beam
Authors
Neuchâtel Observatory
Date of Issue
1962-07
Linked Datasets
Device Type
frequency
Operating Principle
beam
Reference Species
thallium
Quantum Transition
hyperfine
(F=0,mF=0) ↔ (F=1,mF=0)
Reference Frequency
21 310.833 945 1 MHz
Fabrication Source
noncommercial
Host Institution
Neuchâtel Observatory
Status Timeline
Development: 1961–1962 (Rapport 1964)
Operational: 1962-07 (Rapport 1963)
Public demonstration: 1963-01-17, to some 100 manufacturers, watch adjusters, and press, radio, television and Ciné suisse representatives (Rapport 1963)
Visit: 1963-02-18, guided visit by the Council of State of the Canton of Neuchâtel (Rapport 1963)
Dismantled for reconstruction: 1963-05 (Rapport 1963)
Rebuilt: 1964–1967 (Rapport 1964)
Resonator complete, electronics unfinished: 1967 (Rapport 1967)
Programme definitively terminated: 1967 (Rapport 1968)
Technical Design aspects: size, design etc.
Overall length: 2240 mm, oven to detector
Oven temperature: 950 K (677 °C)
Cavity: Ramsey separated oscillating fields, 940 mm arm separation
State selection: Alnico dipole A- and B-magnets, 14 000 G, pole pieces 250 mm long, 3 mm gap
Detector: oxidised tungsten ribbon 1 mm wide, surface ionisation, 0.5 s dwell time
Technical specifications: stability, precision
Uncertainty: 5E-11 (1 Hz), scale A1, quartz reference (Bonanomi 1962); 2E-11 (0,4 Hz), scale A2, rubidium reference (Rentsch & Bonanomi 1964)
Stability: 2.2E-11 @ 28 sec, quartz reference (Bonanomi 1962); 7E-12 @ 10 sec, 2E-12 @ 100 sec, rubidium reference, May 1963 only (Rentsch & Bonanomi 1964)
Precision: 1E-11 @ 150 sec
Line width: 135 Hz
Q: 1.6E8
The frequency of the thallium resonance was determined against the LSRH cesium standard between July 1962 and May 1963. Two values were published. The first, from the opening month of operation, is 21 310 833 945.1 ± 1 Hz on the A1 scale (Bonanomi 1962); the second, from the completed campaign, is 21 310 833 945.1 ± 0.4 Hz on the A2 scale (Rentsch & Bonanomi 1964). The uncertainty applies to the measurement of thallium against cesium, not to the thallium standard itself: in 1962 it combined, in quadrature, 2E-11 from the absolute value of the cesium beam standard, 2E-11 from the comparison chain cesium → quartz clock → thallium, and 4E-11 from possible asymmetric sidebands in the klystron power spectrum, and no contribution attributable to the thallium standard was found — the apparatus systematics tested, cavity phase (reversal of cavity II gave no observable shift; cavity I gave (2.5 ± 2)E-11) and the C-field (a correction of only 2E-12 at 50 mG), lay at or below the 2E-11 resolution of the comparison. By 1964 the factors still limiting absolute accuracy were named as the symmetry of the cavity and the imperfections of the electronics — not the reference oscillator. The National Bureau of Standards measured the same resonance from September 1963 and obtained 21 310 833 946.4 ± 0.4 Hz, 6E-11 higher; the divergence was attributed largely to the inaccuracy of the cesium standards on both sides rather than to either thallium machine. Short-term stability was set by the interpolation oscillator, of which two were used. Against the Observatory's quartz clock (aging 1.5E-11 per day, about 1E-11 over periods of hours), which served for most of the campaign, the dispersion of a single measurement was 2.2E-11 in 28 s. For the May 1963 measurements a Varian rubidium standard replaced it, and the dispersion fell to 7E-12 in 10 s and 2E-12 in 100 s — an improvement in the measurement chain, which is why the two figures belong to different reference oscillators rather than to different states of the thallium standard. The beam itself limited the rate rather than the accuracy: the oxidised tungsten detector's dwell time — 0.5 s as first reported, about 0.1 s at 700 °C in the 1964 account — dictated an averaging time of 35 s per point and about three minutes per comparison at the 1E-11 level.
Modification History
Rebuilt from 1963 around the original vacuum vessel. The ribbon-shaped beam and two-pole permanent deflection magnets were replaced by six-pole electromagnets, raising the detected beam intensity by a factor of 30 and making velocity selection possible, so the spectral line width could be modulated and any width-dependent perturbation detected. The six-pole system produces a cylindrical rather than ribbon-shaped beam, which required a new oven, detector and cavity; the electronics were rebuilt entirely, the originals not having been designed for precisions of 1E-12. Only the high-vacuum vessel and general measuring instruments were to be retained (Rapport 1964). Completion was expected during 1965, but from 1965 the work became marginal as priority went to the HBG time-signal service, though most mechanical components were manufactured that year (Rapport 1965). From 1966 the objective was a prototype suitable for industrial manufacture (Rapport 1966). By the end of 1967 the resonator was complete, the electronics still under construction and first frequency measurements under way (Rapport 1967); work stopped definitively at the end of that year (Rapport 1968).
Usage History
An experimental standard (exécution de laboratoire) was put into service in July 1962 and used as a frequency reference in the Observatory's time determination, alongside the LSRH cesium resonator (Rapport 1962). Its principal use was the determination of the thallium frequency relative to cesium and the comparison of that value with a more recent American measurement; a Varian rubidium standard on deposit at the Observatory from February to November 1963 provided the short-term reference for those measurements (Rapport 1963; Rentsch & Bonanomi 1964). The campaign was completed in May 1963 and the instrument was dismantled; from 1964 the Observatory's quartz clocks were calibrated about twice a month against the LSRH cesium standard instead (Rapport 1964). Because a laboratory apparatus did not lend itself to routine measurement, the intention was to rebuild it as an operational standard (étalon d'exploitation). That work continued from 1964 to the end of 1967, by which point the resonator was complete, the electronics still under construction and the first frequency measurements under way (Rapport 1967); it was then stopped definitively.
Product
bonanomi1962
RapportDirecteur1962
RapportDirecteur1963
RapportDirecteur1964
RapportDirecteur1967
RapportDirecteur1968
RentschBonanomi1964
Identifiers
https://libra.unine.ch/handle/20.500.14713/100460