2001 July 13
In response to the NIST press release the question came up of
what good are more ticks of a clock if the measuring reference
can not keep track of them. Well, on the average, they can be kept
track of, even with a poor reference.
Today I visited a staff member, Tom Parker, of the Time and Frequency
Division of NIST here in Boulder. I had to find out what is remembered
by the electron in the atom during the drift time (about one second) in
the fountain of F1.
F1 is the frequency standard now in use by NIST. The standards are not
time standards. Time is made by an ensemble of "about a dozen" clocks
across the hall from the fountain. These are averaged to create the time.
In the fountain, what is remembered is the hyperfine state of which
two for Cesium are used. The electron is in one of the two states and
stays there during the drift time, however long that is. Part of the art
is to keep down collisions so the electron does indeed stay in the one
state, that is, it remembers.
The frequency is measured, tracked, by sensing with 9 192 631 770 Hz
microwaves from a synthesizer driven by a hydrogen maser at, as I recall,
5 MHz. The maser is not stable. It has, for the frequency standard
business, a lot of drift and other noises. The digital tracker finds the
peak of the resonance of the electron and records, by computer, whatever
frequency count it has (of course, near 9 192 631 770). A lot of readings
are made and averaged to remove the noise.
The computer corrects the average output of the fountain tracker by
knowing that whatever the measured number is, it is in fact,
9 192 631 770 Hz.
Robert Bushnell
