RE: ATM Fiber light source - a burning issue

["Arjan te Marvelde (ELN)" <Arjan.te.Marvelde@eln.ericsson.se>]

Many years ago, i've done experiments with coupling light between mono-mode fiber and a 
thin optical waveguide. I've tried many methods, but the most stable one was to align the 
two properly and use polyurethane as intermediate medium (i.e. glue). The pur had a 
refractive index very well matched to that of the fiber and the waveguide.

: a small ball
: lens on the ends by quickly dipping in a clear setting thick solution.

Soo, what could be done is use a glass fiber (either mono- or multimode, doesn't really 
matter for the short distance needed), break it off in a controlled way to obtain a flat 
end (score and bend over a tube, while applying tension), and use PUR to eithe make the 
lense, or glue one on.

The PUR i used was fluid, it could be dissolved in methyl-acatate (smells like candy and 
gives a high too).

The monomode fiber may be a bit hard to feed with a laser, core diameter is typically 
less than 10 micron. Multimode glass is less critical, sometimes even unclad. 

: Communications fibres from some Telecom cable seemed to have 
: an outer low N
: plastic cladding whuich seems equally unuseable.

Depends on what kind of fiber is inside the plastic.

: Drawn fibres of simple glass are easy to make but without the optical
: cladding are not of much use.  Making a ball on the end is 
: easy but they do
: not conduct light efficiently without the low N cladding.
This is not neccessarily true, since air can be considered a low N cladding too. The lack 
of cladding is not realy an issue for the short distances in question, as long as the 
surface is unharmed light will not scatter out exessively.

: Finally.  From bitter experience. The plastic ones dont take kindly to
: solvents.  Even relatively mild solvents such as methylated 
: spirits damage
: the outer cladding causing millions of microcracks.  It may 
This type of fiber, often used for LANs, usually has no cladding at all. Any irregularity 
in the fiber surface will scatter light, due to the high contrast between fiber and air 
indexes. 
So using a solvent to taper a fiber is usually not a good idea, unless you apply a 
cladding of slightly lower index. You can taper glass fiber by heating and pulling it.

Why do you need a lense at all? If you want to have a tiny lightsource for a star test 
i'd go for the mono-mode fiber, and try to make a nice perpendiclar cut. You will then 
have a light source of 10 micron diameter, with high intensity and an aperture in the 
order of 10-30deg. If you place it a few tens of meters away it will give you a nearly 
flat wavefront with sufficient intensity.

Hope this helps,

/Arjan