[jeffreyH (OP)]

Is there any method that you can think of that could achieve this? Have a stab at it.

[Halc]

Roemer was the first to measure the speed of light, and he did it using a 1-way method.
Does that count?
There were clocks at both ends (the experiment was done shortly after the invention of a clock more accurate than a sundial), but the clocks were not 'in sync'.

[Bored chemist]

That's interesting. The Wiki article references the idea that the method can be considered a "doppler shift".
Well, maybe it can or not- I guess it depends on your definitions.

But there's a streetlight at the end of the road and it's not been replaced by LED yet, so it is a sodium lamp.
And I can put a sodium lamp in a car and drive towards the streetlight.
And, in principle, i can allow light from both curses to fall on a photodiode and I will get a voltage from that diode corresponding to the difference in frequencies - one of my local sodium lamp and the other of the doppler shifted streetlight.

Since I know my speed and the beat frequency and the wavelength of the light, I think that's enough to let me calculare teh speed of light.

Equivalently, I can use a radar "speed gun" and a vehicle with a known speed to measure the speed of light.


Does that count as "one way"?

[alancalverd]

Generate a radio signal at a convenient frequency and measure the wavelength at any point, by any means you wish (diffraction, quarter-wave dipole, whatever) . v = fλ by definition.

[CliffordK]

It should be fairly easy to calculate the one-way speed of light.  The problem is doing so with any reasonable amount of accuracy.

Here's how I would do it.


SpeedOfLight.gif (9.32 kB . 669x311 - viewed 46914 times)

Light (strong laser) passes through two shaft connected spinning discs, hits a cylindrical (conical) mirror, and is projected onto a wall to record.  Mostly interested in the trailing edge of the light spot.

Light speed  is: 299,792,458 m/s.

So, say the two discs are spaced 10 meters apart, then it will have traversed the gap in about 1/(29,979,245) s

Say the discs are 2 meters in diameter, and spinning at 60,000 RPM (1,000 RPS).  With a circumference of πD, or 6.28 m.

So, your discs are spinning at 6,280 m/s.

Multiply the two, and one gets about 1/4774 m, or .0002094 m  (2x10^-4)

Multiply by 1000, and one is 0.2094 mm

That is well within the range one can discern.

The problem is the accuracy of the reading.  If one uses UV light, one might be able to get down to 100nm accuracy, or about 1x10^-7 m. 

So that gives one about 3 digits of accuracy.  One may be able to add a 4th digit, or even 5th digit by going with a faster shaft speed, longer shaft, larger discs, etc.  Perhaps using X-Rays if they could be reflected on the cylindrical mirror (or mirror pair).

So, the question is would say 3 or 4 digits of accuracy matter?

Now, let's consider some kind of unknown aether or fabric of space(time). 

What is the fastest object known to man?  well, Earth is a good candidate.

The equator is rotating at about 460m/s.  And, even doubling it for the difference of forward rotation vs reverse rotation with respect to the fabric of space during the 24 hour day, and it still isn't fast enough.

Earth's orbital velocity around the sun is about 29,780,000 m/s, or about 1/10,000 the speed of light.  And, still not fast enough to pick up.  Earth's orbital speed around the Milky Way is about 220,000 m/s, and getting close to 1/1000 the speed of light.  And, the Milky Way is moving at about 600,000 m/s.

Now, if our aether or fabric of space is not distorted by gravity, electrical fields, matter, etc, then one can choose a place to orient one's device East/West with respect to the galactic rotation or motion, or North/South with respect to the motion.  This should give comparative velocities, but one is just barely on the cusp of being able to pick it up.

(oops, I got a conversion factor to mm off... but it means it is easier to snag the first digit of the speed of light with the theoretical machine, but still a difficult task to get multiple digits of accuracy).

Ok, thinking about this more, there are issues with the leading edge of the light from the first slot illuminating the trailing edge from the second slot making timing difficult.

One would likely fix that by either adding an adjustible flat mirror to the inbound light, or slightly changing the direction of the source.  Adjust to both maximize the intensity of the output beam as well as maximizing the beam width coming off of the cylindrical mirror.

A concave mirror rather than a convex mirror would also work, but would flip the image.  Parabolic?

[evan_au]

Roemer was the first to measure the speed of light, and he did it using a 1-way method.

He used the fact that the Earth's orbit takes it backwards and forwards over (part of) the distance to Jupiter's moons.
Does that make it partially 2-way?
See: https://en.wikipedia.org/wiki/R%C3%B8mer%27s_determination_of_the_speed_of_light

I can use a radar "speed gun" and a vehicle with a known speed

Radar/laser speed guns use the signal reflected from a moving object, and compare it with the source. So it's 2-way.
- But the Doppler method with a Sodium streetlight could be done moving away from the streetlight, so that would qualify. (Even though measuring the beat frequency of non-coherent light sources at optical wavelengths is a real challenge!)
See: https://en.wikipedia.org/wiki/Radar_speed_gun

Generate a radio signal at a convenient frequency and measure the wavelength at any point, by any means you wish (diffraction, quarter-wave dipole...

- Diffraction is measuring the difference between light taking two different paths; not back to the source, but it is 2 different paths. Maybe 1.5 directions?
- Measuring the wavelength by transmission-line effects in a line or a diploe (or in a laser cavity) depends on reflection from the ends of the transmission line producing nodes, so it is a 2-way method. It also produces the speed of light in a transmission line, not the speed in free space, which is what I think the OP was seeking.

SpeedOfLight.gif

The diagram shows light passing in both directions (but not reflected back to the source as in the Fizeau/Foucault experiment).
- But the description does not seem to make use of this two-way transmission...

See: https://en.wikipedia.org/wiki/Fizeau%E2%80%93Foucault_apparatus

My conclusion - measuring the speed of light using light and motion in one direction only is tricky.

[CliffordK]

The diagram shows light passing in both directions (but not reflected back to the source as in the Fizeau/Foucault experiment).

Right, two independent light sources, and essentially two independent experiments as independent controls.  One only needs one light source. 

If the slits are statically aligned, the lasers and mirrors for the reverse direction experiment would be independent, but the same slits would be used.

However, the reason to study the one-way speed of light is to prove or disprove that it is the same as the two way speed of light.  Thus, the interesting thing would be if one found discrepancies between the right to left speed of light vs the left to right speed of light.

One would likely mount the system on some kind of a circular rail or turntable, probably also reversible motor/gearing.  So, one could point it in different directions, and completely reverse it.  So, if it was calibrated at East/West, then rotate it to West/East and see if it gives the same results.  And compare to North/South or South/North. 

And, of course, run the experiment at least once a month for a year to see if there is an annual effect.

The thing would have to be run at a pretty high vacuum both because the speed of light is impacted by the atmosphere, and rotation speeds would be mighty high.  I was hoping the edge of the discs would get up to about mach 20 if not greater.  Magnetic bearings would be nice, but most are active systems which may or may not be as useful at high speeds.  One would likely have a mechanical gearbox, although a magnetic gearbox (or partly magnetic gearbox) might be possible.

The design is very similar to the single slitted wheel or mirror experiments, but has a mechanical connection between the two wheels rather than depending on a reflection back.  Thus, the reflection experiments use a much simpler apparatus and can cover large distances, but only get a 2-way result.

[Bored chemist]

Radar/laser speed guns use the signal reflected from a moving object, and compare it with the source. So it's 2-way.

OK, So, you point a radar gun at me and walk away.
I hook up a receiver to a frequency counter and observe the doppler shift. From that I can calculate your speed or, if I know your speed, I can calculate C.

But does that mean that the oscillators in the two instruments constitute "synchronised" clocks?

[evan_au]

does that mean that the oscillators in the two instruments constitute "synchronised" clocks?

I would thaht this assumes synchronised clocks.
- The frequency synthesiser has to be locked to the frequency when it is not moving, so it can measure the change in frequency when the source is moving.
- Einstein had some criteria that to truly synchronise two clocks, they needed to be colocated and at rest, relative to each other. As soon as you accelerate one, or put it at a different gravitational potential, they are no longer synchronised.
- We can relax some of those constraints if you don't want to synchronise the absolute time, but only the rate of the clocks.
- And you can "cheat" by using a clock like the yellow sodium vapour lamp wavelengths, where the synchronisation is done for you by the wavefunction of a sodium atom (provided they are at the same gravitational potential, if you want to be picky!).

[wolfekeeper]

Is there any method that you can think of that could achieve this? Have a stab at it.

No. There isn't any way to do this. All methods involve something having to go in two directions, either before, during or after the experiment and so are inherently subject to a delay due to the speed of light in both directions. If you even measure a distance light has to go in both directions.

[alancalverd]

All methods involve something having to go in two directions

Apart from reply #3, above.

[CliffordK]

All methods involve something having to go in two directions

Apart from reply #3, above.

Well, in #4, I show light going in both directions, but it is only being measured in one direction at a time.  The reverse direction is part of an internal control, and allowing measuring and comparing two independent directions at a time.

Generate a radio signal at a convenient frequency and measure the wavelength at any point, by any means you wish (diffraction, quarter-wave dipole, whatever) . v = fλ by definition.

Interesting concept.  Of course that requires v, f, & λ to all be measured independently.  Frequency does require some type of a clock, but not necessarily synchronized with the first clock.
 
Say one has a light bulb releasing light in all directions.  Is it guaranteed that the frequency and wavelength will be constant in all directions?  Relativity would seem to indicate that it is not constant.  Of course, any relativity distortion of matter could also be problematic even with a physical attachment device.

Astronomy has already taught us a few things.  Identical "Clocks" on other stars send us light at different wavelength/frequencies.  I.E.  red shift/blue shift.

I wonder if one could measure redshift/blueshift right here on Earth. So you have your oscillator creating a coherent light source projected against a fixed mirror, say a few miles away. If synchronized right, it should cancel it out perfectly along all locations along the path.  I.E.  near the mirror, as well as near the source. 

Any discrepancy in frequency/wavelength would indicate Earth being red-shifted/blue-shifted in relation to aether/fabric of space. 

[Bored chemist]

Astronomy has already taught us a few things.  Identical "Clocks" on other stars send us light at different wavelength/frequencies.  I.E.  red shift/blue shift.

Interesting point.
If you could measure the speed of those stars with a ruler and a stopwatch then you could use the blue or red shift to calculate the speed of light.
Ordinarily, we do it the other way round.
It's the same maths as my suggestion of looking at two sodium lamps.

[alancalverd]

Say one has a light bulb releasing light in all directions.  Is it guaranteed that the frequency and wavelength will be constant in all directions? 

On the one hand, who cares? The question was whether you could measure speed in one direction, to which the answer is obviously yes.

Using my method, you can measure speed in any direction because the dipole and frequency meter aren't associated with the source in any unique orientation. And has been pointed out, the earth is rotating and orbiting, so however you define direction, it is varying all the time. So the fact that you get the same answer twice is proof that speed is not direction-dependent.

[wolfekeeper]

All methods involve something having to go in two directions

Apart from reply #3, above.

Nope. There are no exceptions. How are you going to measure the distance? You can't. If you have a ruler, the molecular interactions along the ruler are affected by the speed of light in both directions; because they're electromagnetic. Hence the shape of the ruler is changed by any velocity the ruler has.

In fact because you can do a velocity boost in any direction, you can easily see that it doesn't matter if the speed of light is different in different directions; it can't affect the result.

You always measure the one-way velocity to be 'c', even if it isn't, because everything changes around it, in just the right way to make it 'c'.

And it's a good thing. If it wasn't like that, objects would probably fall apart if they started to move.

[CliffordK]

Nope. There are no exceptions. How are you going to measure the distance? You can't. If you have a ruler, the molecular interactions along the ruler are affected by the speed of light in both directions; because they're electromagnetic. Hence the shape of the ruler is changed by any velocity the ruler has.

On my apparatus above one measures the right/left and left/right speed of light simultaneously but independently on the same fixed apparatus.  So anything affecting one would affect both. 

If one twists it from East/West to North/South, or looks at diurnal drift, then you might be able to argue matter is expanding or contracting (not counting temperature which would need to be controlled, but a vacuum might aid with temperature control).

[wolfekeeper]

There's a hidden two directional interaction going on along the shaft in the chemical bonds which is synchronising the two ends.

[alancalverd]

How are you going to measure the distance?

You don't have to in #3. You can measure wavelength two ways - dispersion angle from a diffraction grating or peak amplitude of a dipole aerial - with the two instruments effectively at right angles, simultaneously.

[evan_au]

dispersion angle from a diffraction grating

Diffraction gratings work by the fact that light takes many paths from the light source to the diffraction grating, and many paths from the diffraction grating to the screen/meter where the pattern is observed.
- Not exactly a 2-way measurement
- But not exactly 1-way, either!

You could do a similar thing with light echoes from a planetary nebula puffed off by a star which subsequently goes supernova: Observe the supernova via the direct light path, and then look for the delayed light echoes from the planetary nebula (which is not on the direct path).
- By looking at the delay between the direct path and the indirect path, you could calculate the speed of light
- But this is still using multiple paths for the light, and not just 1-way travel
- And this assumes that you know the straight-line distance to the star, which is hard to calculate by 1-way means (even parallax uses 2 light paths to determine a stellar distance)

[Bored chemist]

peak amplitude of a dipole aerial

Only works because you set up standing waves which require a reflection.