The USNO Cesium Fountain Project

We are pursuing a research project to build and study atomic fountain frequency standards. This work initially dealt with the processes of laser cooling and trapping of cesium atoms.


Cesium atoms are widely used in atomic clocks. One of the transitions in Cesium has an oscillation frequency of 9 192 631 770 Hz, which is used to define the second. In a standard atomic clock, the cesium atoms in a hot beam are interrogated twice by microwave radiation. The first pulse starts the oscillation between two hyperfine states and a second, later pulse stops the oscillation. The information about the frequency of the microwaves is encoded in the population of the two states of the cesium. This type of clock is a passive device.


The Atomic Fountain

The fountain geometry increases the time between the two interrogations by gently tossing the atoms up and letting them fall back down under the influence of gravity, all under high vacuum. Atoms are collected and then launched through a single microwave cavity, which interrogates the atoms both on the way up and again on the way down. The atoms are then detected optically to determine the information about the microwave frequency. This cycle is then repeated. The longer time between interrogations improves the precision of the measurement, as does the use of a single microwave cavity.

We use light pressure to manipulate the atoms: collect them, launch them at a controllable velocity, and keep them cold to minimize the expansion of the atom cloud during its ballistic flight.


The USNO Fountain Design Goals

The USNO fountain is designed to be a reference device, not a primary frequency standard. In other words, our device does not have to "tick" at precisely one second, (the job of defining the second in the US belongs to NIST) but must be as stable as possible. This means that we must minimize fluctuations in factors that might affect the fountain, such as temperature, magnetic field, and number of atoms in the signal.

Short- and long-term performance goals

The short-term performance of the fountain is a combination of two factors: the interrogation time, which determines the spacing of the interference fringes generated by scanning the external microwave frequency, and the resolution of an individual fringe, which depends on the signal-to-noise ratio.

Our launching allows a one-half second interrogation of the atoms. This means that the interference fringe peaks will be spaced 2 Hz apart. If our microwaves are on the cesium resonance, this means we know the microwave frequency to better than 1 Hz, or about 1 part in 1010 (ten billion). We can generate signal-to-noise sufficient to resolve the fringe to about one part in a thousand, so our overall short-term performance is between one and two parts in 1013 (ten trillion) at one second.

As the clock runs over a period of time, we are able to average out any random noise that is present, so the performance improves until non-random (systematic) noise sources begin to dominate. This systematic floor depends on how well we limit fluctuations in the systematic noise terms, and a long-term performance of a few parts in 1016 should be achievable.

In a recent data run we were able to steer out the drift of the Hydrogen Maser with which the fountain is compared, revealing the underlying fountain's medium-term performance.

email questions or comments to theanswerguy at


Recent Fountain papers

Take a look at the Fountain!

A schematic and photo of the fountain

A closeup of the atom trapping chamber

The laser system

The fountain without the magnetic shields

The fountain and the web page are still under construction


Info on the new Rubidium Fountain

New pictures! (3/15/04)

Last updated 8/13/04

Please read this
Privacy and Security Notice and this Disclaimer for External Links