In Norman Ramsey's method (fig. 11) the atoms are perturbed by radio frequency radiation in two separate regions at a suitable distance from each other. The atoms can be excited in either one of the two regions. The resonance curve is changed by these two possibilities as shown in fig. 11b. The width of the central narrow peak is determined by the atoms time of flight between the two regions. Using perturbing radiation along the entire distance would also result in a narrow band width but the inhomogeneity of the magnetic field would then impair the result.
    The cesium clock, which is used as a time standard since 1967, is based on the above-mentioned principle (fig. 11a). A beam of cesium-133 atoms passes an inhomogeneous magnetic field. Some of the atoms are directed into the perturbing regions. Only those atoms that flip their magnetic moment during the passage of the two regions are directed by the second magnet to the detector. According to the adapted definition, the peak of the resonance curve corresponds to exactly 9 192 631 770 Hz.

The hydrogen maser (Ramsey) is another atomic clock (fig. 12). In this case the excited hydrogen atoms are selected by a hexapole magnet (Paul). These atoms are directed into a cavity that is part of an electric circuit tuned to the same resonant frequency as the radiation emitted by the excited hydrogen atoms. The radiation energy built up by the atoms causes the cavity to oscillate. The cavity can be connected to an antenna, the signal of which has a frequency stability of 1x10-15. When measuring continental drifts or checking Einstein's general relativity theory, it is in fact more important to have a clock with high stability than to know its exact frequency.

The Paul trap has been suggested by Dehmelt as a precision clock utilizing the so called quantum jumps (fig. 13) predicted by N. Bohr (Nobel Prize 1922). A single ion in the trap is illuminated by resonant laser light, thereby "lifting" an electron in the ion from an energy level 1 to an energy level 2. The electron rapidly returns to level 1 by emission of light (fluorescence). By using laser light of another frequency the electron can be excited to the more stable energy level 3. On level 3 no fluorescence occurs. When the electron has left level 3, it can again be excited to level 2 at which point the fluorescence starts again. The long-lived level 3 has a very narrow band width, a fact which may be exploited in defining an accurate clock frequency. Through experiments of this kind scientists have envisioned a clock with a stability of 1x10-18, i.e. during the entire lifetime of the universe it would only be wrong by 1 second.

Copyright © Nobel Media AB 2014