Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: [Phys-l] What think you all of this:

Hi All,

I'm in a hurry at the moment. But, a fast reading tells me that there is an error in the experiment.

1. Notice that in the first few frames there is an additional set of faint fringes from some other mode in the system either in the laser or in the interferometer. These fringes go across the entire pattern at a different speed than the main fringes. I have often seen these in my own interferometers. This tells me that there is something else unstable in the system.

2. Even high quality optical tables will flex with minute forces when you have an unequal arm Michelson interferometer at play. I have been able to detect the weight of a quarter on the corner of a 4 inch thick Newport optical table when I set up a very sensitive and accurate unequal arm Michelson interferometer.

3. Now, take this guy's system and hang a big honking arm off the table. I'm especially surprised that they suggested it could even be wood! Then rotate the table. The tiniest of errors in the balance and center of the pivot will cause the table to flex enough for the unequal arms to detect a change. No surprise, the change will always be in the same places too.

These interferometers are so dang sensitive that even if it was done on a 3 meter optical table with Invar rails, I'd still be suspect of the pivot. There is a good reason that Michelson and Morely floated the entire apparatus on a pool of mercury -- it supported the whole system uniformly with no flex.

My students and I are currently using an unequal arm Michelson interferometer to measure the wavelength of a highly stabilized laser. With a path difference of only a meter or two we are able to routinely detect sub-picometer changes in the wavelength. We go to great lengths to remove errors in these measurements including setting the whole system up in a room where the floor has been isolated from the building by a layer of cork and anchored to the bedrock. The table is eight inches thick and vibration isolated and we average our readings.

For what it is worth, it is adding problems by using long beams when you can keep the system compact and fold the beam repeatedly. You get a much stiffer system by folding the beams which removes a bunch of variables from the process.

This "data" does not fill my heart with dread for the demise of special relativity. Nope, what impresses me with this experiment is that he is getting such a tiny amount of fringe movement.

John

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
John E. Sohl, Ph.D.
Professor of Physics
Weber State University
2508 University Circle
Ogden, UT 84408-2508

voice: (801) 626-7907, fax: (801) 626-7445
e-mail: jsohl@weber.edu
web: http://physics.weber.edu/sohl/

Bernard Cleyet <bernardcleyet@redshift.com> 9/15/2007 9:31 AM >>>
For such discrimination the Michelson method is necessary; massive slab
floating on mercury.

bc, Panzer.

But, the experimenter did try doing the experiment over time using the
earth's rotation and revolution to effect the necessary movement.
However, being lazy, I didn't read the text carefully.

Brian Whatcott wrote:

At 03:05 PM 9/14/2007, you wrote:


http://www.sas.org/tcs/weeklyIssues_2007/2007-03-03/project1/index.html


bc, has no comment, yet.


p.s. some comments:

http://www.sas.org/tcs/weeklyIssues_2007/2007-04-06/project2/index.html




Good for the physician!

In speculating about error sources, I discounted differential bending of
the wooden beam with rotation when swung on a vertical axis because it was
reported that an iron tube provided much the same result.
However, swinging the beam on a horizontal axis appeared to give
much greater offsets. This would support the idea of greater bending
with beam horizontal than vertical. However, the effect would be more
sinusoidal, one supposes.
The next suspect is the accuracy of ball race at the rotational axis.
Suppose that comparing two angular positions provides a circumferential
offset from the horizontal plane in which the beam rotates
of 0.1 mm in 3 meters - an angular error of about 1 second of angle.
This is a plausible bearing error.
Both error candidates are avoided by fixed mounts oriented N-S
for laser and ccd

However,the experimenter reports no variation for this arrangement.



Brian Whatcott Altus OK Eureka!

_______________________________________________
Forum for Physics Educators
Phys-l@carnot.physics.buffalo.edu
https://carnot.physics.buffalo.edu/mailman/listinfo/phys-l