On Wed, 29 Mar 2006 20:23:25 +0100, "George Dishman" <[email protected]>
>"Henri Wilson" <HW@..> wrote in message
>> On Mon, 27 Mar 2006 19:53:16 +0100, "George Dishman"
>> <[email protected]>
>>>"Henri Wilson" <HW@..> wrote in message
>>>> On Sat, 25 Mar 2006 14:59:37 -0000, "George Dishman"
>>>> <[email protected]>
>>>> I could easily come up with a few equations that would support the
>>>> would have to speculate about some pretty odd values though, eg, for
>>>> mass and 'drag coefficient' of light moving at grazing angle around a
>>>You start by writing down hte equations so the
>>>'drag coefficient' would be "mu". Once you have
>>>those, you don't even need to speculate about
>>>the values, as I said before, you work them out
>>>from what is needed to match the Sganac result.
>> I suppose I could do something like that.
>No 'suppose' about it, you have to do it if you are
>going to perform the experiment so that you know how
>much fibre and how many turns you need.
>> It must be assumed that the drag due
>> to 'c' alone is balanced in each direction and cancels out, like a
>> bridge circuit. Sagnac only senses the effect due to the 'v' bit.
>Not quite, that's the "deliberate mistake" I mentioned
>before. Sagnac is sensitive to the difference in arrival
>time of the beams, not the difference in the speed. The
>two are equivalent if the speed change is small compared
>to c but you need to take it into account if the slowing
>is significant. Since the distance is the same, Sagnac
>is sensitive to the difference in '1/v' rather than the
>difference in 'v', and you have to integrate that along
but the slowing due to 'c' cancels. It is equal in both directions.
>> I think it would be much better to first perform the 'light coil'
>> It would be rather sensational if light could be slowed simply by sending
>> around a solenoid. It would also be interesting to vary the materials or
>> in its core. For instance would a strong magnetic field affect it in any
>> Another variation would be to use IR light and adjust the phasing so that
>> turn was an exact mulptiple of a wavelength.
>IR would be too short, you would need microwaves and
>that means waveguides, not fibres. We know the speed
>is the usual value at the end of a waveguide since
>many micrwave components depend on the speed having
>the right value for their operation. Consider resonant
>cavities for example, for a given size they would work
>at the wrong frequency if the speed changed.
The type of speed changes we are talking about wouldn't be noticeable
>> I'm surprised no bright young enthusiast has thought if this and already
>> performed such an experiment.
>Most students in that field will have to do experiments
>on waveguides, there's a huge industry reliant on such
Not waveguides coiled into tight circles with many turns, George,
>>>>>> The coil is fixed at one end.
>>>>>> Light from a single source is split into two beams that go in opposite
>>>>>> directions around the loop. The beam from S that passes through the
>>>>>> will move at reduced speed for the rest of the trip. That from D will
>>>>>> at c
>>>>>> for most of the trip.
>>>>>> The return of the two beams can be detected with one device capable of
>>>>>> resolving any phase diference between the two.
>>>>>How do you remove any systematic phase difference
>>>>>in the measuring instrument?
>>>> There is only ONE detecting instrument. How can it have a phase
>>>You are comparing the phase of light from one end
>>>of the fibre with light from the other. At least
>>>you would need a two-channel scope and you then
>>>need to address the difference in propagation
>>>times between the channels within the instrument.
>> Yes. .and you can simply swap the detectors to check for response time
>> differences. (make one pulse smaller than the other for recognition)
>So you swap the leads and the difference changes. Was
>that due to differences inside the scope or in the
>fibres from the end of the experiment to the scope?
>Then you change them back and the result isn't the
>same as when you started. Been there, done that :-(
>Don't touch the leads, change the procedure!
If you suspect the scope leads, just connect them both to the same signal
I can see you are searching desperately for ways to escape from this George.
>> If the fibres are 3000 m long, (with say 1500 m in the coil. There would
>> be an
>> optimum value for this proportion), the travel time is about 1 us.
>> We would be looking for nanosecond differences or less.
>Proof? You haven't produced your equations yet so
>these guesses are unsupported.
Do you have a better suggestion?
>> ....that might be difficult depending on the size of the effect if any. We
>> would need thousands of turns in a small coil...not impossible.
>Good scopes can work faster than that but they
>are expensive, what do you have?
I only have an old scope that isn't very fast at all.
>> So let's look at interferometric methods again.
>> You suggest moving the coil, but that introduces mechanical stability
>That's why I said you need to clamp the sensitive
>parts, the directional couplers and so on. The rest of
>the loop doesn't matter, if conventional theory is
>right any change to the configuration should have no
>effect. You might get some transients during changes
>though if the fibre is being accelerated but microphony
>coupled to the interferometer will probably mask that
>> How about this for a start.
>> The thing is set up as an interferometer, with the coil at one end and
>> to be moved. Instead, we try varying the conditions inside the core eg. by
>> moving large objects in and out or passing electric or magnetic fields
>How does that test your explanation for Sagnac?
It doesn't. It is just an interesting experiemnt.
>> Any positive effect would surely amount to a major discovery.
>Fields are already know to produce interesting effects,
>consider Pockel cells and so on.
This would be something new.
>> If that didn't produce anything then you might try moving the coil ..but
>> would have to be would very carefully.
>> There might be a way around this that we haven't yet thought of...eg,
>> using an
>> AC signal or modulated light.
>That is how I would measure the speed of the light
>but the benefit of the interferometer setup is that
>you get a sensitivity dependent on the wavelength.
>The modulation wavelength would be shorter so you
>lose the benefit.
How about this.
The light signal intensity is modulated and sent in both directions. If the
speeds are the same in both, the travel times will be the same and a steady
interference will be obtainable. If the speeds are different, a change in
frequency should see a change in the interference pattern.
>>>>>The simple way is to
>>>>>move the loop along one arm and measure the slope
>>>>>of the phase difference as a function of distance.
>>>> Very funny.
>>>> You obviously don't have a clue.
>>>I am speaking from experience.
>> Our friend Jerry claimed to have done this and achieved a positive result.
>> because Einstein said it couldn't happen, she concluded that the effect
>> was due
>> to mechanical distortion only.
>Mechanical distortion might produce anisotropic
>speed though it's unlikely depending on the
>material. What is contrary to SR is that the
>changed speed would persist after the region
>affected which is why moving the loop removes
>most of the possible objections to your result.
It isn't very easy to MOVE a coiled optic fibre.
>> Being a DHR yourself George, you would probably come to the same
>Nope, I would repeat the experiment until I could
>prove it had a known cause :-( or I could patent a
>device based on the effect :-)
You're smarter than Jerry then.
>>>> You are looking for extremely small differences in time. You would
>>>> coil if you moved it.
>>>Your equations will tell you how much, and you
>>>don't know if the coil is 'distorted' when you
>>>first make it. Move it from end to end in small
>>>steps then move it back again in small steps and
>>>do that whole process a dozen times. Produce a
>>>scatter plot and see what the shape is (hopefully
>>>from your point of view a straight line) then do
>>>a best fit of a line. The residuals (rms error
>>>from the line) gives you a way of assessing
>>>random errors without which you have no way to
>> Maybe you could get something out of it but I don't like your chances.
>> A variation on mine would be to use a distant mirror rather than sending
>> light through the fibre for the whole length. It would require a very
>> laser pulse though and a lens would be required to focus the return beam
>> on the
>> fibre end. That introduces RI problems.
>> We probably need hollow fibres for this type of experiment.
>You haven't explained what effect a mirror would
>have that could explain Sagnac's result.
I thought we were looking at the possibility of light's slowing in a coiled
>you work that out, then an experiment could be
>designed but what you are missing is that the
>design depends on the nature of the effect.
>>>Anyway, you need to do that to eliminate basic
>>>systematic factors like differences between your
>>>instrument channels and more esoteric
>>>considerations like anisotropic refractive index.
>>>Your results have to stand up to scrutiny and
>>>unless you can show you have thought about these
>>>problems and designed the experiment to get round
>>>them you will waste your time.
>> ...and you were the one who said it was fairly simple......
>> My version IS quite simple...but it mightn't have the resolution.
>In my version, you bring the two ends close together
>and parallel and put a pice of card in front to see
>a fringe pattern, that's all. In yours you need a
>stable high frequency oscillator, high frequency low
>drift photodetectors, high bandwidth scope and you
>have yet to work out how to calibrate it.
But your version requires moving the coil. You seem to think that is easy. I
don't....but you are welcome to try.
>Add a piece of thin glass and two mirrors with mine
>and I think you can also measure the speed of each
>of the beams to see if they are half the usual value
>as your theory predicts.
What's this 'half'?
>>>>>> I have also introduced my photon
>>>>>> axis, idea.
>>>>>Which also has no effect on the time taken for
>>>>>the light to get round the loop. You can introduce
>>>>>all the ideas you want, unless they affect the
>>>>>measurement they are irrelevant.
>>>> You have this funny idea that light behaves like frictionless ball
>>>> rolling around a toroidal tube.
>>>You have this funny idea that light behaves like
>>>ball bearings. I am just saying that if they did
>>>then the direction of the spin axis of the bearing
>>>doesn't affect the time it takes to travel from one
>>>end of the tube to the other.
>> It certainly would if the spin axis was not parallel to the coil axis.
>No, only the speed of the ball changes the time
>taken for a given distance.
There would be friction.
>>>>>You'll never know whether there is one or
>>>>>if I'm pulling your leg unless you do the
>>>>>calculation yourself. Bottom line Henry,
>>>>>unless you can show that your effect would
>>>>>precisely explain the Sagnac Experiment
>>>>>to within about 2 parts in 10^12 (IIRC the
>>>>>paper I cited) then ballistic theory is
>>>>>falsified, and to do both Paul and I have
>>>>>shown the speed of the light would fall to
>>>>>a stupidly low level, easily detected in
>>>>>your experiment above.
>>>> How would it have been detected.
>>>With enough gentle kinks in the fibre, it
>>>would take light hours to get through
>> Don't be silly. There are almost straight.
>> I should imagine the effect would be small even in a coil a thousand tight
>> turns .
>Sagnac's table used one turn. A dozen bends
>of 30 degrees would be the same, and your
>theory implies the speed would probably be
>halved. You really need to get that equation
>sorted out and work out what the real loss
>of speed would be.
George, we are talking now abouit fibre gyros. Why suddenly revert to four
>>>>>Remember I don't have to prove you wrong,
>>>>>you are claiming a previously unrecognised
>>>>>effect and it is for you to prove it exists
>>>>>at exactly the necessary level to give the
>>>>>result in every Sagnac experiment ever
>>>>>performed regardless of the details of the
>>>>>setup. Until you achieve that, ballistic
>>>>>theory remains falsified.
>>>> Sure. ...If the RI of glass is ignored :)
>>>Still making that mistake? I corrected you
>>>twice, the RI cancels out in the Sagnac
>>>setup. You'll find the proof on the web.
>> So should any light speed variation then.
>Did you look up the proof? I can't remember
>the details but they are in one of the sites
>we looked at last year.
Yes I recall reading it. I wasn't impressed.