B What is a Clock? Syncing Technologies Explained

[Torog]

How do we decide what is a clock and how do clocks that rely on different technologies stay synchronized?

 

[Ibix]

Something that repeats regularly is a clock. You may wish to add a counter. They stay synched (leaving aside relativistic effects and noise) because that's what "repeats regularly" means.

 

[russ_watters]

Physicists seem to use the circular definitions:
-Clocks are devices that tell time.
-Time is what a clock measures.

I add the fact that "time" works similarly to other dimensions, so it is treated as a dimension...and describe that: Time is a non-spatial dimension that forms a continuum, separating events.
https://www.merriam-webster.com/dictionary/time

Regarding how we synchronize clocks of different technologies: At this point, the scale of time is defined to fit the caesium clock. The actual value of how many shakes equals 1 second is of course arbitrary.

 

[DrStupid]

How do we decide what is a clock

If the output of two or more devices comply with Einstein's clock synchronisation, the devices are a clocks and their output is a measure of time.

 

[PeterDonis]

If the output of two or more devices comply with Einstein's clock synchronisation, the devices are a clocks and their output is a measure of time.

This is certainly a sufficient condition, but it's much too strong to be a necessary condition. Two clocks in relative motion in flat spacetime don't meet the condition, but that does not mean they can't be considered clocks.

 

[hilbert2]

How do we decide what is a clock and how do clocks that rely on different technologies stay synchronized?

If you have two processes that are based on completely different mechanisms, and seem to be periodic with $n$ periods of process 1 taking the same time as $m$ periods of process 2, then it's an even stronger proof that both processes are actually periodic than two similar mechanisms behaving in that way. It's unlikely that the period of two different kinds of processes would happen to be changing at the same rate.

 

[Paul Colby]

A clock is just like a meter stick but for time intervals instead of distances.

 

[DrStupid]

Two clocks in relative motion in flat spacetime don't meet the condition, but that does not mean they can't be considered clocks.

Yes, they can be clocks. But how do you check that?

 

[Nugatory]

Yes, they can be clocks. But how do you check that?

First you use each clock to establish a length standard usable when that clock is at rest (for example, the meter is the distance traveled by light in 1/299792458 second, using the cesium clock definition of the second). Then you calculate the proper time between two consecutive readings of one clock using that clock and the other clock. If these calculations agree to the limits of your measurement accuracy then they are good clocks.

 

[MarekKuzmicki]

Something that counts any events is clock. Of course you want clock to be precise. You want to control what you count. For example: You need to construct "tick" - which is event to count. Then end of your event is starting new event in loop. Or, another example - you know atomic decay law. "Tick" is not too reliable here, but statistick over atomic decay is reliable, and you can compute timeframe from that.

 

[Torog]

I have a lovely pendulum clock on my wall. It is accurate to within a minute a week. The speed of the clock is regulated by the equivalence of inertial mass and gravitational mass. What about the other clocks? What do they depend on to be stable (lets assume we are not traveling at enormous speeds and we are all in the same gravitational field).

 

[russ_watters]

I have a lovely pendulum clock on my wall. It is accurate to within a minute a week. The speed of the clock is regulated by the equivalence of inertial mass and gravitational mass. What about the other clocks? What do they depend on to be stable (lets assume we are not traveling at enormous speeds and we are all in the same gravitational field).

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

Why do I feel like I'm providing research for a paper?

 

[Nugatory]

What do they depend on to be stable (lets assume we are not traveling at enormous speeds and we are all in the same gravitational field).

The "travelling at enormous speeds" part of this question is a red herring that you would do well to let go of. Everything is traveling at enormous speeds relative to something somewhere. For example, your pendulum clock works just fine (in the sense that I described in #9 above and to the limits of its accuracy) whether you or someone on Mars watching through a telescope and moving at a few kilometers per second relative to it is depending on it to measure time.

Where the stability comes from depends on the construction of the clock. Use an electronic LC circuit oscillator and you're depending on the laws of electricity and magnetism which govern the behavior of electronic circuits... Use the slowly increasing length of your hair and fingernails, and you're depending on the biological processes that control their growth... Use an ordinary bedside alarm clock plugged into a wall outlet and you're depending on your electrical utility's ability to provide 50 or 60 (depending on where you live) cycles per second, which in turn depends on classical physics applies to elextrical power generators... And long ago we used sandglasses based on some faiurly complex Newtonian mechanics, sundials based on the Earth's motion through space, candles that burned at a more or less constant rate because the laws of chemistry are stable.

 

[Torog]

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

Why do I feel like I'm providing research for a paper?

No not a research paper. Trying to get my own thoughts in some sort of a shape.

So in my clock one of the stabilizers is the inertial weight of the pendulum.

Does the inertial weight of the bits of the cesium atom provide stability to the atomic clock?

 

[russ_watters]

Does the inertial weight of the bits of the cesium atom provide stability to the atomic clock?

No, weight/mass have nothing to do with the operation of an atomic clock.

 

[Torog]

Thanks Russ,

I like this:

Physicists seem to use the circular definitions:
-Clocks are devices that tell time.
-Time is what a clock measures.

 

[Torog]

No, weight/mass have nothing to do with the operation of an atomic clock.

From: http://hyperphysics.phy-astr.gsu.edu/hbase/Nuclear/nspin.html#c1
"Nuclear Spin
It is common practice to represent the total angular momentum of a nucleus by the symbol I and to call it "nuclear spin". For electrons in atoms we make a clear distinction between electron spin and electron orbital angular momentum, and then combine them to give the total angular momentum. But nuclei often act as if they are a single entity with intrinsic angular momentum I. Associated with each nuclear spin is a nuclear magnetic moment which produces magnetic interactions with its environment. "

Surely the Spin or angular momentum - inertia - is an integral part of the mechanism to the hyperfine structure of the nucleus.

 

[russ_watters]

Surely the Spin or angular momentum - inertia - is an integral part of the mechanism to the hyperfine structure of the nucleus.

I don't know. I'm pretty thin on QM. What does this have to do with the topic?

 

[Paul Colby]

Surely the Spin or angular momentum - inertia - is an integral part of the mechanism to the hyperfine structure of the nucleus.

I think what @russ_watters was getting at is the ratio of weight to mass (weight/mass) has nothing to do with atomic clocks, which is correct. Certainly the mass of the particles involved will effect the transition frequencies in an atomic clock. Current theory is these masses are fixed.

 

[Torog]

I don't know. I'm pretty thin on QM. What does this have to do with the topic?

The question is: What is it that stabilizes the clock rate? And does the inertial weight/mass of the bits form part of the stabilizing mechanism?

 

[Grinkle]

The question is: What is it that stabilizes the clock rate? And does the inertial weight/mass of the bits form part of the stabilizing mechanism?

Its not fundamentally a clock rate, it takes a human to interpret it as a clock. Its a stable periodic thing that can be counted, and that is why it us useful as a clock. Why it is stable, and whether the mechanism for its stability has anything to do with why your pendulum clock is stable is not part of what makes it a handy clock. Some people may say "that is not a good clock" and have their own subjective value judgements on what makes a good clock. They may value ease of observing over accuracy, for instance.

I think that is what @russ_watters means by saying you are now asking a different question.

 

[jbriggs444]

The question is: What is it that stabilizes the clock rate? And does the inertial weight/mass of the bits form part of the stabilizing mechanism?

As I understand the mechanism, the energy requirement to make the hyperfine transition is quite precise. One tunes a microwave beam accurately to maximize the rate at which the hyperfine transition is made and uses the associated frequency as the basis for a clock.

The stabilising "mechanism" to the extent that there is one is the fact the the hyperfine transition has a well defined energy level.

Empirically, the fact that atomic clocks agree with one another is a demonstration that the stability exists regardless of why it does.

 

[Ibix]

The question is: What is it that stabilizes the clock rate?

What would you consider to be an acceptable answer? You say earlier that the equivalence of gravitational and inertial mass stabilises your pendulum clock, but you don't justify that statementand I'm not at all sure it's correct. For example, I would think that I could take the clock into zero g, give the pendulum bob an electric charge, place it in a uniform electric field and expect it to work (or at least, I could design a pendulum clock for which that was true - too much metal in an off-the-shelf model). I don't immediately see a reason why it would be any more or less stable operating in an electric or gravitational field.

 

[John M]

A layman's answer to the second part of the very first question - 'how do clocks that rely on different technologies stay synchronized?' It depends on the technology utilised to make the clock i.e. a mechanical clock, an electrical clock or an atomic clock. Each device has a built in regulator i.e. centrifugal weights, pendulums for mechanical clocks, pulses or crystals for electrical clocks and the movement of atoms for atomic clocks. Each can be adjusted until synchronised to a similar device. Devices of a dissimilar technology will not synchronise.

 

[Ibix]

Devices of a dissimilar technology will not synchronise.

The Earth is a sort of clock - it rotates at a steady rate. A quartz watch is a sort of clock which vibrates a crystal at a steady rate. These are dissimilar technologies. Are you telling me you can't tell the time of day with a quartz watch?

 

[russ_watters]

The Earth is a sort of clock - it rotates at a steady rate. A quartz watch is a sort of clock which vibrates a crystal at a steady rate. These are dissimilar technologies. Are you telling me you can't tell the time of day with a quartz watch?

Right: I would say it is a by-definition requirement that a "clock" be capable of synchronizing with other clocks, to within its capable accuracy.

 

[hilbert2]

Right: I would say it is a by-definition requirement that a "clock" be capable of synchronizing with other clocks, to within its capable accuracy.

Exactly. Similarly, if someone would have built the first "thermometer" hundreds of years ago by saying "the length of this metal rod multiplied by a unit conversion factor defines the temperature", there would have been no reason for anyone else to acknowledge the significance of this invention. Only when independently working people are able to make thermometers based on different physics (thermal expansion, pressure of a low density gas in closed system, thermal radiation intensity), and these devices consistently give mutually agreeing results when applied on the same sample system, it makes sense to say that the thing (temperature) that they're measuring is actually something objective and worthwhile to measure.

 

[r8chard]

I think the questioner does not understand what they are asking.
A clock or any similar function is what we arbitrarily use as an artificial measure of the span of time as we perceive that interlude.

What we use is a completely artificial and imaginary guesstimate of mechanical and biological functions. Check out the history of daylight savings time. And the struggles with determining longitude.

Our ancestors, a thousand years ago who were serfs determined seasons as plowing, growing and harvesting and starving. A priest would listen for the bells announcing primes. A merchant might use a burning candle to estimate time or if rich a waterclock. A noble would measure time by when the roads were to muddy to travel on and after the harvest, it an army could move to steal their neighbors harvest.

Their ancestors, a thousand years before them would go by wet or dry, cold or hot seasons. When the fish came upstream or the fur pelts were of quality to take. When to girdle a tree or set fire to thee scrub that grasses can grow for thee cattle.

And, to confuse the issue for historians. Every locality, village, tribe, cult festival would use a wild mishmash of calendars. Based on the whims of the pharaoh or king or primate or even the scribe making crap up instead of just admitting he didn't know!

 

[Paul Colby]

I guess I don't see a huge difference between clocks used to measure time being synchronized and rulers used to measure distances maintaining a stable length.

 

[russ_watters]

I guess I don't see a huge difference between clocks used to measure time being synchronized and rulers used to measure distances maintaining a stable length.

Nor do I...but the way you say it implies you think there should be a difference...?

 

[Paul Colby]

Nor do I...but the way you say it implies you think there should be a difference...?

Just hedging my bets. Time intervals and Space intervals are definitely different for some value of different. Their measurements have many issues in common though.

 

[jbriggs444]

I guess I don't see a huge difference between clocks used to measure time being synchronized and rulers used to measure distances maintaining a stable length.

Depends on what you mean by synchronized. If you mean that the elapsed time between one tick and the next on this clock over here should match the elapsed time between one tick and the next on that clock over there then I would count that as "syntonized". That would correspond to the centimeter markings on my ruler over here being separated by one centimeter, just like the centimeter markings on your ruler over there.

By contrast, two clocks are "synchronized" when they both tick off time zero at the "same time" (whatever that is chosen to mean). That would correspond to my ruler over here being placed on the same start line as your ruler over there. Just like "at the same time", "on the same start line" is a convention, not a fact of physics.

 

[Paul Colby]

Depends on what you mean by synchronized.

Independent of what I said, I had intervals of time and space in mind. An inch measured by ruler A and ruler B agree when I use them at my desk to measure things. Same with time, a second of lapsed time measured with clock A agrees with clock B when in the same room moving at sufficiently slow relative and so on, and so on.

 

[Herman Trivilino]

How do we decide what is a clock and how do clocks that rely on different technologies stay synchronized?

Metrologists decide, and they base their decision on precision, or as you've put it, synchronization.

You need two devices that repeat some kind of a pattern of behavior, you count the number of repetitions, and compare. For example, the hour hand on my watch makes 24 revolutions while the sun makes one. I do this kind of comparison over and over to see if that comparison stays the same every time I make it. If it does, the two are candidates for clocks. I can then repeat this process with other kinds of devices, and see if they stay synchronized, too. The ones that do are good candidates, the ones that don't get rejected.

By the way, don't confuse this type of synchronization between clocks that share the same location with the type that's done with separated clocks. When the clocks are separated synchronization becomes more complicated and the results counter-intuitive.

 

[Grinkle]

An engineer says clocks are synchronized if they are derived from the same source and hence by construction cannot drift away from each other. The only discrepancy is the accuracy of a single derived tick, one clock to the other. The per-tick error does not accumulate tick-by-tick because the construction of the clocks causes them to start each tick from the same source.

If clocks are not derived from the same source, they are not said to be synchronous with each other.

 

[jbriggs444]

If clocks are not derived from the same source, they are not said to be synchronous with each other.

The clocks used to maintain TAI do not derive from the same source. They derive from themselves and each other and stay synchronized that way.

 

[Grinkle]

@jbriggs444

Maybe they are, I don't know enough to say if they are synchronized in the stricter sense that I described. From here,

https://en.wikipedia.org/wiki/International_Atomic_Time

I see this -

TAI as a time scale is a weighted average of the time kept by over 400 atomic clocks[4] in over 50 national laboratories worldwide

That tells me they are probably not considered synchronized by the engineers responsible for maintaining TAI.

On the other hand, for all I know the values of the constants of nature do truly synchronize these atomic clocks and they should not / cannot in theory exhibit any drift one to the other.

 

[Torog]

I think the questioner does not understand what they are asking.

I just probably didn't word the question very well.

What I am really interested in is how clocks work. What is the process in one type of clock that makes it synchronize well with another type of clock? Is there a common factor in good clocks that allows the different technologies to offer a reasonable synchronization? ---( just don't tell me it is time! )---

1) My pendulum clock needs the inertial weight and the gravitational weight to be a stable ratio.
2) The world in it's orbit needs the same thing. Cited as a clock in post N. 25
3) An atomic clock. This seems to need the quantized angular momentum of the nucleus as part of the system to generate the necessary hyperfine frequency.
4) What about an electrical oscillator? Does the weight of the electrons affect the frequency?

 

[Grinkle]

Is there a common factor in good clocks that allows the different technologies to offer a reasonable synchronization?

Synchronizing clocks is an exercise of converting tick count on one clock to expected tick count on a different clock.

If two clocks are derived from the same source, for instance one set of gears runs two different clock faces on a clock tower, then one tick on the first clock must equal one tick on the second clock (or some known ratio between the two clocks exists by construction) and these clocks are synchronized to each other as well as any two clocks can be, regardless of how regular the tick on the two clocks is. If one can couple clocks in this manner, one is not as concerned with the quality of the clocks - they stay in synch by construction. If one clock is off by an hour, then both clocks are. They are in synch with each other.

If two clocks have a known period and there is zero error in either period then they needn't be derived from the same source to be synchronized, one just needs to know the periods of the clocks and calculate the ratio to convert from one clock to another. I don't know if atomic clocks fit this description or not, maybe they do. Barring that, I would say that this is just an ideal and no clock will have zero cycle-to-cycle variation in tick duration.

So, good clocks in the sense that you are asking have very low cycle-to-cycle variation in tick duration. The lower this variation, the less two not-coupled clocks will drift away from each other over time.

 

[Robert Albertson]

The way I see it is like this. Time is a measure of change. Hence what we call time is our way of quantifying the change of the universe. One may also use the entropy of the universe. In fact, this may be a better drfinition, i.e. time is a measure of increasing entropy of the universe. Hence, a clock is a numerical device that allows us to keep track of the changing state of the universe (increasing entropy).

 

[Herman Trivilino]

1) My pendulum clock needs the inertial weight and the gravitational weight to be a stable ratio.

How could you know that? You'd need another clock to compare it to, and some way to know that it's the pendulum clock's behavior that makes the ratio unstable, and not the behavior of that other clock.

2) The world in it's orbit needs the same thing. Cited as a clock in post N. 25

Same comment applies.

3) An atomic clock. This seems to need the quantized angular momentum of the nucleus as part of the system to generate the necessary hyperfine frequency.

Again, how would you know its oscillations occur with a constant frequency? You'd need another clock for comparison and some way of attributing any discrepancies to the other clock instead of to the atomic clock.

 

[Torog]

How could you know that? You'd need another clock to compare it to, and some way to know that it's the pendulum clock's behavior that makes the ratio unstable, and not the behavior of that other clock.
Same comment applies.
Again, how would you know its oscillations occur with a constant frequency? You'd need another clock for comparison and some way of attributing any discrepancies to the other clock instead of to the atomic clock.

These quoted are all "good" clocks. It is fairly easy to see them synchronized. I didn't say that we didn't need another clock. It is obvious that a clock cannot exist by itself. The existence of one clock implies a second clock.

 

[Paul Colby]

I love the sorry of John Harrison and the Longitude problem. Around the time of Newton people needed a means of navigation at sea, not a small issue. Harrison, basically considered a layperson at the time, approached the problem from the point of view of clocks that could function for months at a time and at sea. He's credited with several major technological innovations such as jeweled bearings. Those fancy looking pendulum clocks from that period are actually temperature compensated so they don't change length to first order as the room temperature changes. After years of work and many attempts he solved the problem. His final clock looks very much like a pocket watch with jeweled bearings. Sadly, the intellectual elite at the time (including Newton who was kind of a dick about it) tried to stiff Harrison out of the millions in prize money and the recognition. Newton failed to do so in the fullness of time.

 

[Herman Trivilino]

These quoted are all "good" clocks. It is fairly easy to see them synchronized.

Then it seems you've answered your original question:

How do we decide what is a clock and how do clocks that rely on different technologies stay synchronized?

Unless I'm missing something.

 

[Torog]

Then it seems you've answered your original question:
Unless I'm missing something.

Not in the least. Most people who contributed to this little discussion talked about synchronizing clocks. I'm not interested in that at all.

I love the sorry of John Harrison and the Longitude problem. Around the time of Newton people needed a means of navigation at sea, not a small issue. Harrison, basically considered a layperson at the time, approached the problem from the point of view of clocks that could function for months at a time and at sea. He's credited with several major technological innovations such as jeweled bearings. Those fancy looking pendulum clocks from that period are actually temperature compensated so they don't change length to first order as the room temperature changes. After years of work and many attempts he solved the problem. His final clock looks very much like a pocket watch with jeweled bearings. Sadly, the intellectual elite at the time (including Newton who was kind of a dick about it) tried to stiff Harrison out of the millions in prize money and the recognition. Newton failed to do so in the fullness of time.

This above is interesting to me - how clocks work - for example the balance wheel clock as perfected by Harrison shifts energy between the inertia of the balance wheel and the hair spring and this provides the stable rate and allows synchronization with the Earth clock. How do all the other clocks work?

 

[Herman Trivilino]

Most people who contributed to this little discussion talked about synchronizing clocks. I'm not interested in that at all.

It was part of your original question.

How do we decide what is a clock and how do clocks that rely on different technologies stay synchronized?

 

[russ_watters]

How could you know that? You'd need another clock to compare it to, and some way to know that it's the pendulum clock's behavior that makes the ratio unstable, and not the behavior of that other clock.

Same comment applies.

Again, how would you know its oscillations occur with a constant frequency? You'd need another clock for comparison and some way of attributing any discrepancies to the other clock instead of to the atomic clock.

I'm not sure if I'm accidentally arguing for The Empire here, but the difference between today and 500 years ago is that we know and use the physics by which clocks operate to accurately predict their accuracy. Before Newton (or perhaps Galieo) people didn't know how pendulums work, so they couldn't pre-calculate the oscillation rate. Hourglasses would have been even worse. And I doubt even Harrison in the 1700s used the underlying physics to calculate the performance of his chonometer. So all of these had to be calibrated against a master clock (the sun) to work at all. That just isn't the case today.

So we can take a clock up a mountain or fly it around the world in a plane (or GPS satellite) and have a pretty good idea of how it will behave.

 

[Herman Trivilino]

So all of these had to be calibrated against a master clock (the sun) to work at all.

But the only reason people knew that was that they saw that those primitive clocks couldn't stay in sync with each other or with the "master". It was only when the technology developed enough that people saw that they could do a better job of keeping them in sync with each other that they noticed that the master did a good job of joining them.

But even that didn't last as the technology advanced.

 

[russ_watters]

But the only reason people knew that was that they saw that those primitive clocks couldn't stay in sync with each other or with the "master".

I suspect that when designing those primitive clocks, the designers knew they were guessing and what it meant.

 

[Paul Colby]

This above is interesting to me - how clocks work - for example the balance wheel clock as perfected by Harrison shifts energy between the inertia of the balance wheel and the hair spring and this provides the stable rate and allows synchronization with the Earth clock. How do all the other clocks work?

I think your usage of the word "synchronized" is a bit off. A better term is clock rate for what you are discussing. For example, Harrison made pendulum (grandfather) clocks before he became interested in the Longitude issue. If all you have is a pendulum clock, how do you know it's quality? What Harrison did was build multiple clocks and compare them against one another. For example, he would run one clock with his front door open in winter while building a large fire in his living room. He would compare the clock by the front door with the one by the fire. Then he would switch the location of the clocks. This way he could make precision measurements of the thermal stability of his designs.