[time-nuts] Zeeman frequency and cesium tube interchange

Tom Van Baak tvb at leapsecond.com
Fri Apr 22 13:32:35 EDT 2005

Hi Brian,

The SI second is defined for mean sea level and no
external fields.

If there were no magnetic field, in theory, a cesium
tube would show a resonance when its synthesizer
generated exactly 9192.631770 MHz; the definition
of the second.

But in practice, a weak uniform magnetic field is
necessary for the beam apparatus to operate; to
isolate the center peak from the other peaks. This
DC field also has the side-effect of slightly shifting
the frequency of the center line.

Fortunately the frequency shift is a calculatable
amount (a function of magnetic field strength) so
the trick is that the synthesizer must be designed
to generate a slightly higher frequency to exactly
compensate for the shift that will be induced by
the field.

Thus the synthesizer for a 5061A does not actually
generate 9192 631 770 Hz as one might expect,
but because of the nominal 61 milligauss C-field,
the synthesizer must generate 9192 631 771.6 Hz
in order to lock onto the Cs peak precisely. The
Zeeman frequency for 61 mG is 42.82 kHz (info
from an old 5061A manual).

Other 5061A/B use a 76 mG field, corresponding
to a 53.53 kHz Zeeman frequency, and require the
synthesizer to generate 9192 631 772.5 Hz (info
from a new 5061B manual).

The short HP 5062C runs at 9192 631 774.3 Hz
with a Zeeman of 70.40 kHz.

Below is a great reply from Corby Dawson about
the problems this can cause when mixing FTS
4050, 4060, HP 5060A, 5061A, 5061B parts.


----- Original Message -----
> Tom,
> The zeeman frequency required depends on two things, the magnitude of the
> C-field current and the synthesizer frequency.
> HP 5061A and B units that have a synthesizer freq. of 12.6317725 Mhz are
> configured for a lower value of C-field current by selecting a higher
> value series resistor on the A15 board. In this case any tube installed
> (5061A/B 5060 4050 4060) will operate at the 53.53Khz zeeman frequency.
> HP 5061A and B units and 5060A units that have a synthesizer frequency of
> 12.6317716 Mhz are configured for a higher C-field current due to a lower
> value series resistor in the C-field circuit. In this case any installed
> tube will operate with a 42.82Khz zeeman frequency.
> Problems arise when synthesizer and or A15 modules are swapped around
> indiscriminantly leaving a unit with modules that do not match!
> Since the HP and FTS tubes C-field windings are designed to provide the
> same field for a given input they are interchangeable. You can operate
> any of these tubes at a 42.82 or 53.53Khz zeeman. I usually just stick
> with how the mainframe came configured.
> I don't remember what I ended up with as far as the zeeman freq. was
> concerned when I installed a Frequency Electronics tube into a 5061A, but
> do remember the line width was quite broad in keeping with the reduced
> accuracy spec. of the FE tubes. (Same spec. as the 5062C tube)
> I have installed FTS tubes into the 5062C giving it the +-7Z10-12th
> accuracy spec. and did have to modify the synthesizer for 42.82Khz and
> set the C-field accordingly.
> Hope this helps!
> Corby_

----- Original Message -----
> Brooke-
> I have a question for you. I know this topic has come-up before
> on the Time-Nuts reflector, but what Zeeman frequency did you use
> when you initially set the C-field on your 4060?
> I have heard different answers from different people about this
> question. Some claim the exact freq follows the tube while others say
> it follows the model of the standard.
> In my case, I put a "used" 5061 series tube into an HP-5060 and find that
> a Zeeman freq of 53.53KHz results in a C-field setting that is not
> correct (ie: the resulting freq against GPS and my opt004 5061A is
> offset).
> The 5060 manual calls for 42.83KHz and at first glance this appears to
> result a "better" C-field setting even though the tube is from a 5061.
> Tom -  Any experience here?
> Brian, WA1ZMS

More information about the time-nuts mailing list