[time-nuts] Caveats on Allan Deviation with ultra stable oscillators
magnus at rubidium.dyndns.org
Sat May 31 05:30:41 EDT 2014
On 05/31/2014 12:24 AM, Poul-Henning Kamp wrote:
> In message <email@example.com>, "John Miles" writes:
>> people have even reported similar behavior with cesium standards, although I
>> don't see how that could happen. There aren't supposed to be any
>> first-order temperature effects in a CBT, and I'd think that any lower-order
>> effects would be way beneath the tube's flicker floor...)
> One important trick in this area:
> Always locate your DUTs physically orthogonal to each other.
> Almost none of our DUTs have 3 axis of symmetry, and therefore
> most environmental effects are not symmetric with respect to
> I noticed this by accident comparing three "identical" OCXO's
> because I had put one of them in a different orientation than
> the other two: The environmental noise were much larger between
> that one and the two other, than between the two co-aligned
> I'm not entirely sure this is relevant for Rb/Cs/H, their environmentals
> should be attenuated enough for it to not matter.
Their physical packages do have a sensitivity to magnetic fields. The
mymetal shield should handle most of it. Preferably you should let it
stay in the same orientation not to see any shift.
Hydrogen masers have a sensitivity to barometric pressure and
temperature, as it will deform and thus detune the resonant cavity. This
shifts with design, so some is sensitive and others is less sensitive.
Rubidium and Cesiums is passive clocks, in that you need to insert a
frequency generated from a fly-wheel crystal oscillator. Hydrogen masers
exist in both passive and active form. In the active maser it actually
outputs a signal. I believe that many of the active masers still use a
crystal oscillator for the mix-down and locking.
Regardless, while *most* of the environmental effects of that crystal
oscillator is being canceled by the loop to the atomic reference, for
all passive masers it's traditionally a frequency comparison and phase
stability depends on how accurate phase deviation is being captured.
Most time it is a bit crude detection which works better for frequency
Cesium was selected because it had the second best insensitivity to
magnetic fields, but was believed to be easier to work with and thus
easier to reproduce. If the choice would have been different we would
all be comparing our Thallium-beams and then some of us would be poor
enough only to have cesium beams. The basis of selection has shifted,
because since we have invented C-field servo to reduce the C-field
effect. The means of detection now include lasers, which allows good
efficiency. We can also do selective laser pumping which means we don't
have to dump half the beam, increasing the signal to noise right there.
We can do laser cooling and significantly remove a whole bunch of shifts
due to temperature, doppler etc. We can then bounce a ball of atoms up
and down a tube, removing the two-cavity systematic shift as well as
making the observation time much longer. By cooling down the tube we can
then remove the black body temperature shift of frequency.
Yeah, there is a whole bunch of environmental effects there. I haven't
mentioned the Stark effect, which is the electrostatic field effect.
See, the list grows longer.
The closer you look, the more effects you will find.
More information about the time-nuts