Measuring Speed of Light with One Clock

[lowemack]

I know there is a problem with measuring the one way speed of light, but why can we not use just one clock. If somebody could explain the problems with this method, then it might help me to understand.

Can we measure the one way speed of light the following way:- Have a long sealed tube with a light sensor at the front and a light sensor at the rear. The front sensor starts a clock (very accurate atomic type), the one at the rear stops the clock. If we then open the front of the tube (like a camera shutter perhaps), light will enter the tube and some of the photons will start the clock, some will carry on and some time later, when they reach the end of the tube, stop the clock. We will have a time, we will know the length of the tube, then we can calculate the speed of light. We can make the tube as long as we like, the length of cable between sensors the same size and the clock as accurate as possible.

We cannot measure the speed of a single photon* but if we assume that all photons from the same light source are traveling at the same speed then some of the photons from the front of the beam will start the clock and some from the front of the beam will stop it.

*How can we say we can measure the speed of a single photon? As soon as a photon is sensed somehow, it has been absorbed by an orbiting electron, and then re-emitted, (maybe after shifting the electron to a new energy level briefly). It is not the same photon, and the re-emitted speed may not be dependent on the original photon’s speed.

 

[pervect]

I know there is a problem with measuring the one way speed of light, but why can we not use just one clock. If somebody could explain the problems with this method, then it might help me to understand.

Can we measure the one way speed of light the following way:- Have a long sealed tube with a light sensor at the front and a light sensor at the rear. The front sensor starts a clock (very accurate atomic type), the one at the rear stops the clock.

How does it do that? For instance, if you have a wire or coaxial transmission line, how do you account for the propagation delay through the wire/cable? Realistically you'd expect propagation delays on the order of - the speed of light in the reverse direction. But then you are doing a two-way measurement.I

 

[lowemack]

If both sensor cables were the same length, and the start and stop signals took 1 second each to reach the clock and the time taken for light to travel the distance between the sensor was 0.2 sec, then at T=0 start sensor activated, T=0.2, stop sensor activated, T=1, clock started, T=1.2 clock stopped, clock reads 0.2 seconds.

 

[pervect]

If both sensor cables were the same length, and the start and stop signals took 1 second each to reach the clock and the time taken for light to travel the distance between the sensor was 0.2 sec, then at T=0 start sensor activated, T=0.2, stop sensor activated, T=1, clock started, T=1.2 clock stopped, clock reads 0.2 seconds.

For this to work, you need to assume that the propagation speed of the signal is the same in both directions on the cable. This assumption is known as isotropy, which is equivalent to explicitly assuming the Einstein clock synchronization method and using two clocks and subtracting the results.

 

[harrylin]

To expand on that: according to reference systems relative to which your system is in motion, the propagation speed of the signals in the cables is not the same in both directions.