[time-nuts] Updating the unit of,time: the second.

[Attila Kinali]

On Tue, 28 May 2019 22:56:35 +0200
Mike Cook <michael.cook at sfr.fr> wrote:

> a. There is no need for a new definition.

There is. Current optical clocks deliver a lower uncertainty than
Cs fountain clocks. Ie the reference we have is less precise than
the measurement tools we have. Hence a redefinition of the second
is needed.

> b. Any new definition would have to be realizable and easily verifiable. 

That's one of the main concerns and this is also the main reason why
nobody is actively pursuing a redefinition just yet. But there are people
out there who are already working on this topic and gathering all the
requirements to a successful redefinition of the second. My guess,
based on the current speed of things, is that we will have a new
definition of the second within 10-15 years. 

> c. The first commercial cesium clocks were available in 1956, but the second did not get redefined until 1967.  There is no rush.

Which caesium beam standards were available in 1956? AFAIK the first one
was the HP5061 and that came much later. Essen and Parry built their
clock in the 1950s and published the results in 1955. The picture of the
beam tube is only a small fraction of the clock itself. There are multiple
racks full of RF equipment not shown. I would be very surprised if there
was any company that was able to commercialize this contraption within
only a year. Even in this large size.

>     I believe that commercial optical clocks are available but:

No. As far as I am aware of, there are no commercial optical clocks
available. There are a few optically pumped microwave clocks out there
(e.g. by Oscilloquartz) and even cold atom clocks (by Muquans and SDI)
but no optical clocks.

The main problem with optical clocks is the frequency division of the
optical signal down to something that can be used in electronics.
This is usually done using an optical comb. But the commercially
available ones are big, and according to Michael Wouters also quite
expensive. There are efforts to use non-linear optical rings to
generate these combs, but there is no commercial version available
yet (it's a very new technique, which has been around just a few years)

The closest I know to a commercial product is what NIST reported
in Optica just a few days ago[1] (based on two-photon absorption
in a Rb vapor cell and using two optical combs to divide the
778nm down to 22GHz).

> d. There are too many flavors of optical clocks around on lab benches. So despite their increased precision and stability which flavor would get the vote?

This is another issue. Of course, a redefinition will use one atomic species
only (with the others probably becoming secondary definitions). So far
the jury is still out which of the atoms and which method is the best one.
As there are not yet enough optical clocks out there, we don't have enough
data to decide yet. And it doesn't help that an optical atomic clock takes
several years and a quite large team to build.

			Attila Kinali

[1] "Architecture for the photonic integration of an optical atomic clock",
by Newman et al., 2019
https://doi.org/10.1364/OPTICA.6.000680
-- 
It is upon moral qualities that a society is ultimately founded. All 
the prosperity and technological sophistication in the world is of no 
use without that foundation.
                 -- Miss Matheson, The Diamond Age, Neal Stephenson