28 Apr
2007
28 Apr
'07
9:39 p.m.
On Apr 28, 2007, at 3:13 PM, Warren Wolfe wrote:
EVERY frequency drifts. The question is: How
much, and how
repeatably? (or, maybe that's two questions....)
As far as I know, there's no way to adjust the Rb beam. A beam of
electrons is deflected by electronics, and hits a target. The
frequency
of the adjusting signal is "tuned" to keep the beam spot on the
target.
The Rb standards were a couple of orders of magnitude less stable than
the Cs standards. Again, I don't know why, I just know WHAT.
That is nowhere near how a rubidium oscillator works. Actually,
it sorta describes how a cesium oscillator works, but not quite...the
beam in a Cs oscillator is a beam of Cs-133 atoms, not electrons, and
the beam is steered magnetically, not electrically.
A rubidium oscillator servo-controls its frequency around the
opacity/transparency of a glass cell containing Rb-87 in a gaseous
state. Drift in this system is primarily a function of impurities
from the glass envelope contaminating the cell over
very long periods
of time, and to a much greater extent, degradation of the lamp
used
to illuminate the sensor on the other side of the cell. Fortunately
the lamps are easily replaced in most designs.
In practice, Rb oscillators are very long-lived devices, in terms
of both functionality and stability.
-Dave
--
Dave McGuire
Port Charlotte, FL