Understanding How Atomic Clocks Work: Uncovering the Science Behind Timekeeping

[biferi]

I just got an Atomic Clock and googled how they work.

It says it picks up the Atoms in the Air and sets the Clock from finding out the Energy they emit or something like this.

Can you tell me what I am missing or how they work?

 

[vk6kro]

If you are in the U.S., there is a radio transmitter in Boulder Colorado which sends time information from an accurate clock (using a Rhubidium or Cesium standard) and this is picked up by the radio receiver in a clock or watch.

This radio receiver works on a low frequency of about 60 KHz. Some other countries have similar services. Japan had two until the Fukishima disaster destroyed one of them.

 

[SteamKing]

It's Rubidium, Rb, atomic number 37.

 

[vk6kro]

Yes, Rubidium should not have a "H" in it.

I checked and it is a Cesium clock used as a standard in Boulder. These are better than Rubidium.

 

[etudiant]

Hi biferi,

Your 'atomic clock is built around a radio receiver, as noted earlier by vk6kro.
The radio is tuned to the time signals sent by the NIST from their Ft Collins radio site WWV.
If you check the site, they even have a list of the various receivers used, probably including yours, here:
http://www.nist.gov/pml/div688/grp40/receiverlist.cfm

The atomic part of this time keeping comes from the core instrument that provides the time. An introduction to this technology, which has many interesting wrinkles, is here:
http://en.wikipedia.org/wiki/Atomic_clock

Do note that your clock will be way off, by atomic clock standards, simply because the radio signal will be several milliseconds late in arriving at your site ;) .

 

[vk6kro]

Propagation delay is a problem for radio transmission of a time standard although a low transmitting frequency ensures negligible multi-path propagation.

So, you could add a constant correction to the indicated time to get the real time, if it mattered.

A domestic clock showing seconds would hardly notice a few millisecond error, while a seismology lab would need to do the correction.

The transmitting frequency is very accurate and unaffected by propagation so an alternative very accurate clock could be derived from just the carrier.

 

[f95toli]

Yes, Rubidium should not have a "H" in it.

I checked and it is a Cesium clock used as a standard in Boulder. These are better than Rubidium.

It is actually an ensemble of hydrogen masers (I think they have something like 11 of them), that is then synchronized to their cesium fountains. The reason you need both is that Cs clocks do not generate a continuous signal, it is more of a "pulsed operation". Hence, the actual time signal comes from masers which are "freewheeling"; steered by the signal from the Cs clock to avoid drift over longer timescales.

The time generated at NIST is also synchronized to the UTC, which is the "world time" (ensentially an "average" of all the Cs fountains in the world) which is administated by BIPM in Paris.

Rubidium is a reasonably good frequency standard (but about 2 orders of magnitude worse than a maser), but they are not very good clocks because of long-term drift (but they are cheap).

 

[biferi]

If Boulder Colorado Transmits the Time and my Clock is picking it up.

Then why is the Time right?

I am in PA, not Boulder Colorado.

 

[f95toli]

If Boulder Colorado Transmits the Time and my Clock is picking it up.

Then why is the Time right?

I am in PA, not Boulder Colorado.

It is only approximately right. It takes about 10 millisecond for the radio signal to travel from Boulder to you.
This is of no consequence for an ordinary clock, but there are plenty of applications where this is not accurate enough; in which case you need to start using various tricks to improve the accuracy.
Time transfer is quite literally a science in its own right.

 

[russ_watters]

If Boulder Colorado Transmits the Time and my Clock is picking it up.

Then why is the Time right?

I am in PA, not Boulder Colorado.

Could you tell us what the make and model of your clock is?

 

[vk6kro]

The time signal is meant to hold your clock to within a second or so, which is more than accurate enough if the display is in seconds.

The one hour time difference between you and the transmitter is taken care of in the clock.

When you set it up, it asked you where you are or how many hours behind UTC (London England) time you are.
It then corrects the "hours" reading of the clock to allow for your location, and also for daylight saving if it applies in your location.

 

[Baluncore]

A cesium clock has higher short term phase noise but drifts less than a rubidium clock. Using the two together results in a better reference than either.

The transmitting frequency is very accurate and unaffected by propagation so an alternative very accurate clock could be derived from just the carrier.

Not quite true. The propagation path changes as the ionosphere changes. Multipath signals can cause phase reversals.

As ionospheric ionisation changes, the virtual point of reflection rises and falls during the night and day. It is the rate of change of path length that determines the apparent frequency of the Boulder Colorado reference signals. You can estimate the ionospheric conditions from the beat frequency.

If you need an accurate clock then you should consider a GPS locked receiver. It knows where you are and so corrects for path length. However, it is still slightly effected by variation in ionospheric transit time.