Atomic Clock Running Rate: Adjustment Possibilities?

[exmarine]

Pendulum clock running rates can be adjusted by changing the length of the pendulums. Likewise the running rates of mechanical and electronic crystal clocks can be changed and adjusted. Can the running rate of an atomic clock also be adjusted? It doesn’t seem like there could be any way to change the quantum properties of the cesium 133 atoms, or whatever they are using now. So when they set the GPS clocks to run about 37 μs slow before launch, is that just some offset between what the cesium atoms are really doing and what the physics package is broadcasting to the outside world?

 

[jtbell]

There are intermediate mechanisms between the "clock atoms" in an atomic clock, and the clock's time display. When you "set the clocks to run slow", you adjust those intermediate mechanisms.

I don't know how those intermediate mechanisms actually work, but in the end, they must include some way to adjust, in effect, the number of "atomic cycles" that correspond to one "display second".

 

[PeroK]

Since one second represents about 9 billion transitions of the Caesium atom, I imagine you simply count a different number of transitions per "second" to alter your timekeeping.

 

[Mister T]

Pendulum clock running rates can be adjusted by changing the length of the pendulums. Likewise the running rates of mechanical and electronic crystal clocks can be changed and adjusted. Can the running rate of an atomic clock also be adjusted? It doesn’t seem like there could be any way to change the quantum properties of the cesium 133 atoms, or whatever they are using now. So when they set the GPS clocks to run about 37 μs slow before launch, is that just some offset between what the cesium atoms are really doing and what the physics package is broadcasting to the outside world?

Which type of adjustment are you talking about? There are two. One is synchronization, that is, two clocks can be running at the same rate, but might not be synchronized. The other is the rate. You can synchronize two clocks that are running at different rates, but of course they won't stay synchronized.

In the first part of your question you talk about rate adjustment, but then at the end your question seems to refer to a synchronization adjustment.

Anyway, I don't think that anything is done before launch that can't be undone or otherwise adjusted after launch. The atomic clock just puts out a signal that oscillates with some frequency and that signal is received at antennas. What the engineer does with that signal after it's received is a matter of choice, so it can be used to tell time only if it's interpreted correctly. Note that if the transmitter is moving relative to the receiver the emitted frequency won't match the received frequency anyway! (Doppler effect). Likewise if they are at different gravitational potentials.

 

[PeterDonis]

when they set the GPS clocks to run about 37 μs slow before launch, is that just some offset between what the cesium atoms are really doing and what the physics package is broadcasting to the outside world?

Yes. There is a separate adjustment applied on board the satellites to the output of the atomic clocks before the time stamps that go into the emitted signals are generated.

For more information on this, and on the use of relativity in GPS in general, see the classic review article by Ashby:

https://link.springer.com/article/10.12942/lrr-2003-1

 

[pervect]

Yes. There is a separate adjustment applied on board the satellites to the output of the atomic clocks before the time stamps that go into the emitted signals are generated.

For more information on this, and on the use of relativity in GPS in general, see the classic review article by Ashby:

https://link.springer.com/article/10.12942/lrr-2003-1

Note that this uses the technique of counting a different number of cycles, it doesn't change the behavior of the atoms, but you can change the number of cycles you count to output a second.

 

[nitsuj]

Since one second represents about 9 billion transitions of the Caesium atom, I imagine you simply count a different number of transitions per "second" to alter your timekeeping.

An EXTREMELY precise clock that only measures in seconds lol