# How accurate are atomic clocks??

Discussion in 'Physics & Math' started by Prosoothus, May 26, 2002.

1. ### ProsoothusRegistered Senior Member

Messages:
1,973
It is believed that atomic clocks are accurate, even when they are moving.

Scientists claim that results of the measurements of moving atomic clocks indicate that Einsein was correct, and that time really does slow down the faster an object is moving.

However, after reading about the construction of caesium atomic clocks, I found that there are three factors that would influence a moving atomic clock, independently of the time that it is measuring:

a) The motion and direction of the magnetic fields which are used to seperate the different caesium atoms.

b) The motion and direction of the caesium atoms themselves.

c) The motion and direction of the microwaves used to excite the caesium atoms.

As the motion of the atomic clock increases, one or more of the above factors would influence the clock to give a false reading. This would mean that atomic clocks are only accurate if there at rest(at least relatively).This would also mean that time is constant, but the speed of the atomic clock changes based on its motion.

Tom

3. ### c'est moiall is energy and entropyRegistered Senior Member

Messages:
583
""The motion and direction of the microwaves used to excite the caesium atoms.""

--> microwaves are light and isn't it so that its motion is absolute?

5. ### Merlijncurious catRegistered Senior Member

Messages:
1,014
I believe that the answer to the question is that the motion of the clock (or of the atoms for that matter) do not at influence the accuracy because the velocity of photons (that make up the facors (a) and (c) in your post) are indepemdent of the velocity of the clock.

edited to add: oh, C'est moi, you beat me again.

7. ### ProsoothusRegistered Senior Member

Messages:
1,973
Merlijin and c'est moi,

What about the red or blue shift of the microwaves. Doe these shifted microwaves have the same effect on caesium atoms as unshifted microwaves?

What about the magnetic fields that separate the caesium atoms. Since the magnetic fields travel light speed, do they have less of an effect on the caesium atoms the faster the clock is traveling???

Tom

8. ### ProsoothusRegistered Senior Member

Messages:
1,973
Just to correct my previous post, when I said red and blue shifted microwaves, I meant microwaves that have frequencies shifted above or below their normal frequencies.

Tom

9. ### ProsoothusRegistered Senior Member

Messages:
1,973
Merlijn,

That is the problem, the velocitiy of the photons are independent. If the velocities of the photons were dependent on the clock, everything would be synchronised regardless of the clock's speed.

But since the velocity of the photons are independent of the velocity of the clock, the faster the clock travels, the less the photons are synchronised with the clock which, in turn, results in faulty readings.

Tom

10. ### James RJust this guy, you know?Staff Member

Messages:
30,353
The atoms in the clock experience no Doppler shifts, since the microwave source is moving with the rest of the clock.

11. ### ProsoothusRegistered Senior Member

Messages:
1,973
James R,

Good one!!! You got me there. I'll have to think about it a little more.

I'm also considering something c'est moi told me. He claims that increased velocity causes the Bohr radius of atoms to shrink. This reduction in the radius could cause a change the frequency between orbitals. Since atomic clocks count the frequency emitted by atoms, this change in frequency could give a false impression that time is slowing down(or speeding up).

But I still get the feeling that there are too many factors in atomic clocks that can lead to false readings.

Tom

12. ### thedIT GopherRegistered Senior Member

Messages:
1,105
Prosoothus

The photon frequencies emitted by orbital electrons is a quantum mechanical affect. What matters is the total spin, angular momentum and energy of the electron. If the source was moving with respect to the detector the frequency would indeed then be shifted, by doppler shifting, by an exact amount that can be determined. I refer you also to <a href="http://rd11.web.cern.ch/RD11/rkb/PH14pp/node16.html">Bremsstrahlung radiation and synchrotron radiation</a> so you can see that these affects are well known and accounted for. But as the emitter and detector are at rest relative to each other this should not be required.

The Bohr Radius of an atom is a classical approximation of what atoms really are and should not be used in this case. You will get misleading results from using it.

Also, as the entire experiment is moving it is in the same frame of reference and so special relativistic corrections do not apply between each element of the apparatus. What does matter and needs accounting for, is a another clock moving relative to your 'stationary' clock and general relativistic corrections which affect both.

13. ### ProsoothusRegistered Senior Member

Messages:
1,973
Thed,

You have to take into consideration that the magnetic fields, and the microwaves, in the atomic clock travel independently of the frame of reference of the clock. Both the magnetic fields and the microwaves travel in their own frame of reference, which may cause problems as the clock travels faster.

Tom

14. ### thedIT GopherRegistered Senior Member

Messages:
1,105
Tom

Give me a reference that clearly shows the magnets and lasers are moving independant of the source. That is, all moving on differnet vectors and not together.

15. ### ProsoothusRegistered Senior Member

Messages:
1,973
Thed,

The magnets and lasers are not moving independent of the source, but the magnetic fields and microwaves that they produce are.

This is one of the characteristics of light: It travels independently of any frame of reference. Since microwaves are photons as well, this characteristic applies to microwaves as well. I also believe that this characteristic applies to magnetic fields, but I'm not sure.

Tom

16. ### thedIT GopherRegistered Senior Member

Messages:
1,105
Tom

You seriously misunderstand what has been said. The speed of light is invariant between inertial frames of reference. That is not the same as saying that light in independant of the frame.

You would do well to understand the difference and what invariance means. Clue, it is not affected by rotations (changes) between inetial frames of reference.

17. ### ProsoothusRegistered Senior Member

Messages:
1,973
Thed,

Are you saying that light's speed and direction is dependent on the frame of reference of the object producing it, or not???

As far as I know, the moment a beam of light is produced, that beam of light's characteristics become independent of the source producing it. It's speed and direction become constant, regardless of the speed and direction of the source that produced it.

Tom

18. ### thedIT GopherRegistered Senior Member

Messages:
1,105
Tom

Let's look at what you said above

And

The point here is that the entire apparatus is in the same frame of reference. There is no relativistic correction between the emitter (the photons) and the detector because there is no relative motion between them.

The important thing to realise, if I see where you are going wrong, is that even if the clock was moving at 0.99c all elements of the clock are travelling at the same speed. The photons are still emitted at c and traverse the distance to the emitter in the same time as if the thing where stationary.

Only to an observor outside that frame (point of view or whatever) would you need to worry about the relative velocities.

Going back to something I said I while back. From the frame of reference of a distant observor we are travelling at near c. We can't tell that and apply no corrections. If that observor could measure our atomic clock they should be able to correctly use it to measure time, after applying the Lorentz transform.

19. ### Adam§Þ@ç€ MØnk€¥Registered Senior Member

Messages:
7,415
On topic: How accurate are atomic clocks?

My wrist-watch is atomic, runs off a thirty-pound nuclear reactor backpack. The crappy piece of so-and-so loses about three minutes each day; have to reset it all when I get up in the morning.

20. ### ProsoothusRegistered Senior Member

Messages:
1,973
Thed,

Imagine that you had two fixed plates, one an emitter and the other a detector, and they were one meter apart. While the two are at rest, if a laser beam shoots out of the emitter towards the detector the the time it would take the laser beam to hit the detector would be:

Time = 1/c

However if the two plates were traveling at .99c away from the laser beam, it would take much longer for the laser beam to hit the detector because the speed of light is always constant regardless of any frame of reference.

Now if it were possible for the two plates to travel at c, the laser beam would NEVER reach the detector. In order for the laser beam to reach the detector in this case, the laser beam would have to travel faster than light. And as you know, this is impossible.

Tom

21. ### James RJust this guy, you know?Staff Member

Messages:
30,353
Tom,

<i>Imagine that you had two fixed plates, one an emitter and the other a detector, and they were one meter apart. While the two are at rest, if a laser beam shoots out of the emitter towards the detector the the time it would take the laser beam to hit the detector would be:

Time = 1/c</i>

Yes. I agree.

<i>However if the two plates were traveling at .99c away from the laser beam, it would take much longer for the laser beam to hit the detector because the speed of light is always constant regardless of any frame of reference.</i>

Wrong. In the reference frame of the plates, the light would still take time 1/c to travel between them. The fact that the whole apparatus is moving is completely irrelevant.

BUT (and this is a big "but") according to a "stationary" observer watching the plates go past, the distance between the plates would be length contracted, and the time frame of the plates would appear to run slower. The net effect of this is that this observer would measure the speed of light (distance between plates as seen by him divided by time taken as seen by him) to be c, still. That is what is meant by the invariance of the speed of light.

<i>Now if it were possible for the two plates to travel at c, the laser beam would NEVER reach the detector.</i>

Yes, but that isn't possible.

22. ### ProsoothusRegistered Senior Member

Messages:
1,973
James R and Thed,

Here is a link I posted on the "Galileo & Einstein - second thoughts" thread:

http://www.physics.wustl.edu/~visser/physics-216/notes-light-clock.html

It illustrates one of the problems a moving atomic clock can experience. As you can tell from the link, the vertical motion of light can change in an object that is moving. This change may result in an atomic clock giving false readings.

Tom

23. ### thedIT GopherRegistered Senior Member

Messages:
1,105
Tom

Read the last line of the third paragraph of that link. The analysis is done from an external frame of reference. That makes a world of difference and is the point you are consistently or deliberately missing.

Nothing affects the rate the clock 'ticks' from the frame of reference of the clock. Only to another observor, in another frame of reference, would the clock appear to run slower. That is what they measured.