B Defining 1 Second: The History and Science Behind Time Measurement

[Pedroski55]

Hi,

I'm wondering about time. Time is very important in physics.

It seems we inherited the idea of 60 minutes in an hour from the Babylonians. I don't know if the Babylonians were even interested in minutes, let alone seconds, probably not.

Later on, we needed more accuracy, I suppose it seemed natural to divide 1 minute into 60 seconds, rather than 100 or 10.

Mechanical clocks came and went, then came the quartz crystal clock. Shock a quartz crystal and it will vibrate. A quartz crystal tuning fork is skimmed until it vibrates at 32768 oscillations a second. 32768 = 2 to the power of 15. Run through a flip-flop array, this may be reduced to 1 pulse per second.

Now, it appears to me , I could easily be wrong, that this is a chicken or egg situation: frequency is vibrations per second.

If I know what a second is, I can skim my tuning fork to vibrate at exactly this frequency. Or I define 32768 vibrations of my tuning fork as 1 second. (Accurate enough for most purposes, but prone to environmental influences.)

I hear even atomic clocks use a quartz crystal as initial impulse giver, then 9 192 631 770 oscillations of a Caesium 133 atom = 1 second.

Which came first, the chicken or the egg? The second or the quartz crystal tuning fork?

 

[Dale]

The first quartz clock was built in 1927. The first clock measuring seconds was built in 1570 or so. Even the SI second was first defined in 1889.

 

[PeroK]

Which came first, the chicken or the egg? The second or the quartz crystal tuning fork?

The second came first, then the clock (caesium), then the second was redefined based on the clock.

 

[Ibix]

Isn't this one of those things that only seems like a chicken and egg if you mix unit systems?

Old style, a second is 1/86400 days. This turned out not to be a precise enough definition for modern purposes because day length varies, so we found a more stable basic time unit, the emissions of Caesium, and defined a new style second in terms of those. So that nobody except metrology nerds needed to care about the redefinition, the new style second is defined so that (for as many practical purposes as possible) it's the same size as an old style second within tolerance.

So this is not circular. There are two completely separate definitions of the second, one based on fractions of the rotation period of the Earth and one on multiples of the vibration period of Caesium. The modern one was chosen to match the old one, but if we'd been willing to pay the price in confusion we could have defined it otherwise.

 

[Mayhem]

So this is not circular. There are two completely separate definitions of the second, one based on fractions of the rotation period of the Earth and one on multiples of the vibration period of Caesium.

I don't know, seems very circular to me?

 

[Pedroski55]

Thank you all very much! I am the wiser!

Wouldn't happen to have a link or info on how, in practice, one catches, measures or reads the oscillations of caesium 133?? That's a very high frequency!

 

[PeroK]

Thank you all very much! I am the wiser!

Wouldn't happen to have a link or info on how, in practice, one catches, measures or reads the oscillations of caesium 133?? That's a very high frequency!

I typed "how does a caesium clock work" into Google and found this:

https://science.howstuffworks.com/atomic-clock3.htm

 

[Dale]

Typically it is done using an absorption spectroscopy type of experiment. It is explained here in some detail:

https://en.m.wikipedia.org/wiki/Atomic_clock

 

[jbriggs444]

I don't know, seems very circular to me?

As of 1960 the definition was elliptical, not circular.

 

[libertyk]

Prior to 1967 one second of time was defined by international agreement as 1/31,556,925.9747 of the time it took Earth to orbit the Sun.

In 1967 a second was redefined as 9,192,631,770 oscillations of the Cesium-133 atom.

 

[phinds]

I don't know, seems very circular to me?

So you believe that there is a relationship between the time it takes for the Earth to circle the sun and the time between the oscillations of cesium 133 ?

 

[Mayhem]

So you believe that there is a relationship between the time it takes for the Earth to circle the sun and the time between the oscillations of cesium 133 ?

No, but for n = 4, I can conclude that 1 in 2 people got the joke.

 

[libertyk]

So you believe that there is a relationship between the time it takes for the Earth to circle the sun and the time between the oscillations of cesium 133 ?

There used to be in 1967. The two would have diverged slightly in the 54 years that have passed since they redefined the definition of a second.

 

[phinds]

There used to be in 1967. The two would have diverged slightly in the 54 years that have passed since they redefined the definition of a second.

You completely missed the point.

 

[Psnarf]

"this is a chicken or egg situation"
Obquote: "What is time to a chicken?" -Roy Blunt, Jr.

GPS satellites have to keep a different time. Is the difference in time on Earth vs time in Earth orbit due to relativity or gravity? I read that one second to an observer near the center of a black hole, is a billion years to an observer on Earth, a different frame of reference. Does suggests a relationship between time and gravity ?

 

[jbriggs444]

General relativity describes how gravity is entirely a feature of space-time geometry. It accurately describes the relationship between time and gravity.

At a popular science level, the effect of relativity on GPS satellites can be split into two parts. There is a "special relativity" part which is due to the velocity of the moving satellite in a reference frame where the Earth is stationary -- fast clocks run slow. Then there is a "general relativity" part which is due to gravitational time dilation -- low clocks run slow.

[At a more advanced level, both the "special relativity" and "general relativity" effects are part and parcel of general relativity. The two can be split out and computed separately when gravity is weak and linear approximations are valid. When gravity is stronger, the non-linear nature of general relativity becomes important and more difficult calculations are required]

The two effects act in opposite directions. The gravitational effect wins by 38 microseconds per day. You can Google up tons of articles on this at various levels of sophistication.

For an observer near the center (the "singularity") of a black hole, the notion of time dilation is not really sensible because time dilation depends on a synchronization definition and there is no single "correct" one to choose. For hovering observers outside the event horizon, there is a natural synchronization standard to use. Using that standard, arbitrarily great time dilation ratios can exist as the horizon is approached more and more closely.

None of this has anything much to do with the definition of the second. Further discussion belongs in the Relativity forum instead.

 

[sophiecentaur]

That's a very high frequency!

It wasn't chosen on the basis of its frequency. It was chosen for repeatability. The atomic resonance is the same wherever you are or whoever is 'making' the reference clock. Some atoms are, comparatively 'all over the place' Some may be very consistent but hard to engineer into a reliable time reference.

We could be having a similar conversation about the metre. It's no longer defined in terms of a standard length of platinum in Paris but in terms of how far light will travel in a particular timed. Time is such a reliable quantity to measure - your digital watch is pretty good and you can get timing accuracy to many orders of magnitude better than that. You could cry 'circular argument' again and get the same reply here.

Afaik, they are still trying to get a better way of defining the kilogram than a standard lump of platinum in France. One approach is based on collecting a large (known) number of atoms of one isotope of an element and calling that the kilo but there's a problem with getting the reference mass pure enough. You need a substantial number of atoms to give better accuracy than the platinum kgs. When I last heard about it they were still working on it.

 

[Dale]

Afaik, they are still trying to get a better way of defining the kilogram than a standard lump of platinum in France.

Actually, they redefined the kilogram in 2019. Now it is defined in terms of Planck's constant, just like the meter is defined in terms of the speed of light.