[time-nuts] Lifetime of Cesium tube
dgcarlson at sprynet.com
Thu May 12 12:56:34 EDT 2005
From the Agilent perspective, the expected lifetime of a new
Performance CBT is an average of 5 to 6 years of continuous operation. We
don't speculate on non-operating useful life. Lifetime is not a
specification for a number of reasons. The 5061A/B electronics are
not microprocessor-controlled as in the 5071A so there are fewer guards
against the possibility of a CBT being operated outside of its nominal range
of Cs oven temperature, EMult voltage, HWI voltage, etc. Any of
these could have an effect on CBT lifetime.
Agilent is still manufacturing and selling the replacement CBT for the
as products with model numbers 05061-6101 and 05061-6077
(standard-performance). N.B. that in order for the CBT warranty to be valid
for these models it is stipulated that it be operated only in conjunction
with the newer A11 Cs Oven Controller, identified as HP P/N 05061-6144.
CBTs manufactured after December 2002 now have more Caesium to extend their
expected operating life. This is true for all four of our CBT models, the
aforementioned two above as well as the 10890A and 10891A models for the
5071A Primary Frequency Standard. Although there is little or no empirical
data for the improvement in the 05061-6xxx models, initial testing results
on the 10891A model give expectations that the lifetime will average 9 to 10
years instead of the previous 6 years. To back up that
improvement Agilent also increased the warranty on that model 10891A to 5
years. These newer CBTs with increased quantity of Caesium are denoted by
the following serial number prefixes:
05061-6077US4247 5-year warranty
05061-6101US4248 3-year warranty
10890A US4239 10-year warranty
10891A US4240 5-year warranty
HP presented a paper at the Frequency Control Symposium in 1999 that
described lifetime and end-of-life symptoms, geared toward the 5071A. Here
are the first
several lines, to help in searching for the document:
"LONG-TERM EXPERIENCE WITH CESIUM BEAM FREQUENCY STANDARDS
John A. Kusters, Hewlett-Packard Santa Clara, USA
Leonard S. Cutler, Hewlett-Packard Laboratories, USA
Edward D. Powers*, United States Naval Observatory, USA
Since its introduction 7 years ago, the Hewlett-Packard cesium beam
frequency standard has been installed in many national laboratories. At the
As to pumping down a CBT in non-operating storage, the recommendation in the
5071A Operating manual is that it be done every 6 months. This applies a
conservative approach because in addition to time causing eventual increase
in internal vapor pressure, temperature also has an impact. A CBT
or Cs standard stored un-powered in a very cool environment will have less
likelihood of this than one stored in a galvanized-metal
warehouse in full sunshine.
Every 6 months is a good rule of thumb.
As to pump-down procedures, Appendix B in the 5061B Operating/service manual
tells how to use an external HV supply to pump the CBT by itself, should the
HV supply installed in the instrument chassis fail to do the job.
---- Original Message -----
From: "Chuck Harris" <cfharris at erols.com>
To: "Discussion of precise time and frequency measurement"
<time-nuts at febo.com>
Sent: Thursday, 12 May, 2005 5:36 AM
Subject: Re: [time-nuts] Lifetime of Cesium tube
> glenn.tracy wrote:
> > What kind of lifetime can one expect to see from an HP 5061-6101 cesium
> > tube?
> It is my understanding that you can expect about 5 years in continuous
> or 2 years on the shelf.
> > What typical symptoms do these exhibit near the end of their useable
> > lifetime?
> The end of life can be due to several things, but most notable is a
> of cesium ions to send hurling down the length of the tube, and an excess
> of other molecules for the cesium ions to run into. Either condition
> in a shortage of cesium molecules for the detector to detect.
> A c-beam is essentially a notch filter. The beam is modulated (mixed?)
> a 9Ghz microwave signal that is derived from a very stable and low noise
> reference. When the modulation signal is at the right frequency, the beam
> current will peak. There is usually a servo mechanism that keeps the
> reference tuned so that the beam current stays on top of this peak.
> At end of life, the beam current peak becomes indistinct. There isn't
> distance between the peaks and the valleys, and the servo has trouble
> out where the peak is, so it ends up bouncing all over the place. This
> up as increased phase variations. Ultimately, no peaks can be found and
> crystal oscillator ends up slammed against its maximum, or minimum
> time-nuts mailing list
> time-nuts at febo.com
More information about the time-nuts