Re: Asymmetrical Newton's Cradle

[Brian Whatcott <betwys1@SBCGLOBAL.NET>]

Air race organizers have commented more than once,
"Used to be, the race planes would arrive from over half the sky
to the finish line.  Any more, they come from one direction:
they all fly GPS (Loran was good too)."

I suggest that graded index fiber is rather like this:
the beam is shepherded to the center over a narrow cone of
angles, the stepped fiber has a wider cone, hence increased
 time scatter.

So it would seem sensible that the upper speed through a
fiber is limited by the central index and the axial length.
 But perhaps you were asking rather, about data rate?

Brian W


At 02:48 PM 11/8/2003, you wrote:
> The purpose of the graded-index fiber, according to
> Hecht (Optics, 3rd edition, page 202) is to reduce the
> intermodal dispersion (from about 20 ns/km, for the
> stepped-grade multimode fiber, to about 2 ns/km).
> Fibers are most often used at single wavelengths,
> for example, 1.3 microns. The intermodal dispersion
> limits the speed at which pulses (zeros and ones in
> the case of computers) can be transferred. It occurs
> because light traveling along the near axis paths arrives
> before the light traveling along the more zig-zag paths.
>
> Single mode fibers (diameter around 10 wavelengths)
> "esentially eliminate intramodal dispersion" writes
> Hecht. What is then responsible for the upper limit of
> the transmission speed in single mode fibers?
> Ludwik Kowalski
>
> On Saturday, Nov 8, 2003, at 08:17 US/Pacific, Brian Whatcott wrote:
>
> > At 12:50 PM 11/8/2003, Ludwik,  you wrote:
> >> On Saturday, Nov 8, 2003, Brian Whatcott wrote:
> >>
> >>> One next remembers that there is an alternative to a
> >>> quarter wave length of transmission line. This is a
> >>> device of specific length and impedance intermediate
> >>> between two line sections to be joined  of greatly
> >>> different characteristic impedances.
> >>>
> >>> One can instead,  intersperse a length of gradually
> >>> changing impedance, which matches the preceding
> >>> and succeeding sections exactly. This has the great
> >>> advantage of being broad band, so that one is not
> >>> limited to a frequency where a quarter wave has a
> >>> fixed length.
> >>
> >> Has this kind of approach been used to make a
> >> "broad band" antireflection coating on a lens?
> >
> > I am not familiar with examples of graded index
> > anti reflection coatings.
> > The graded index fiber light guide is ubiquitous.


Brian Whatcott    Altus OK    Eureka!