[time-nuts] Tour of METAS (Swiss Federal Institute of Metrology) time lab: any questions or requests?

[Magnus Danielson]

Hi,

On 05/07/2015 11:48 PM, Bob Camp wrote:
> Hi
>
>
>> On May 7, 2015, at 8:02 AM, Pete Stephenson <pete at heypete.com> wrote:
>>
>> On 5/7/2015 1:22 PM, Bob Camp wrote:
>>> Hi
>>>
>>> I guess on a general level:
>>>
>>> 1) How do they generate the Swiss version of UTC (fleet of hydrogen masers run into to a .. :)
>>
>> According to their website they use an ensemble of commercial cesium
>> clocks (they show some photos of HP 5071As) and a hydrogen maser. I'll
>> be sure to ask them for details.
>>
>>> 2) How do they link that version of UTC with the rest of the world? (BIH etc)
>
> BIH = Bureau International de l’Heure
> They are the people who (by treaty) decide what time it *really* is. In order to do this they need to
> be able to evaluate the various time scales around the world. That evaluation process requires some
> method of direct inter comparison of time scales over very long distances. Typical approaches range
> from “clock trips”  to TWSTT ( or TWSTFT) to GPS based techniques.

As I recall it, they do no longer check the time-scales alone, but 
report the time difference of all their clocks in relation to their 
time-scales, such that TAI gets calculated out of all available clocks.
Actually, they first produce another time-scale from all clocks for 
optimum stability and then frequency correct it into TAI and from that 
uses the leap-second to produce UTC. Typically each lab run their own 
ensemble algorithm and then use one of their clocks to feed a 
microstepper that then produces the UTC(lab) of that lab. Some also 
produces a TAI equivalent in the form of TA(lab). The algorithm steering 
the microstepper also get's fed corrections from the UTC calculations in 
order to steer the labs UTC to be inline. This is also used to steer 
slave-clocks such as the GPS clock to be inline with USNO (if they 
haven't changed anything in the details).

Knowing this overall process helps in understanding what you see and ask 
the questions.

> In addition to the comparison process it’s self, there is also a lot of internal cross checking and analysis
> that goes into a typical time scale. Often the associated equipment base is pretty substantial. It’s also a
> good thing to look at to pick up ideas for (lower cost) implementation in a basement lab.

Indeed.

Cheers,
Magnus