Inexpensive Furnace to Fuse Two Glass Optical Fibers Together

This can't be much more complicated than welding saw blades. I have this butane torch that has a piezo-electric start that I just love. I bet it is hot enough to melt fiber-optic strands. The clever part is bringing the ends together during the weld. That's not much different from saw blade welding. If the weld needs tempering, it's just a matter of practice, I'm sure.

A good magnifier over the weld area would be necessary, with a light.

Having been involved in cable laying and laying of obstacle fibre cables, I can tell you that there is a "suitcase" solution.

any one of these will help you

http://www.globalspec.com/SpecSearch/Suppliers/Optics_Optical_Components/Fiber_Optics/Fiber_Optic_Termination_Splicing_Equipment?RegEvent=login

Are you splicing the fibers end to end, or are you trying for form a coupling region between two fibers side by side)? As referenced above, "suitcase" style fusion splicers for end to end splicing are quite common, available from a number of manufacturers, and very expensive. This is not something you are likely to accomplish with a home-built "kit", because one needs very tight tolerances on the alignment (measured in microns). If you are trying to form a beam splitter, it is unlikely you are going to find a device for accomplishing this- this operation generally requires clean room conditions...

I think a glassblower supplies store might have something to offer,maybe out of Venice or Italy. Have you tried looking on-line under furnaces .

Ceramic kilns bake on enamel.

Can you use a ready made 3dB splitter:-

http://www.fibersplitters.net/

Thank you for your advices.

Actually, the use of 3dB splitters seams the best choice.

With regards to this subject,I am reading "The Demon and the Quantum" by Marlan O. Scully, at Wiley editor.
He states that indeed you can split a beam by 3dB into two beams, but you can't combine 100% of two beams using the fiber 3dB connector.
There are two explainations:

1)passive optics such as glass is symetric versus time reversal.Thus, since it split's to 50%, it combines to 50% of 200%, i.e. the same as one.Because of time reversal symetry

2)alternatively,if time reversal symetry were violated, there would be a degeneracy breaking of the guided mode,and half the energy would be lost in the clading(or evanescent waves?)

3)perhaps there is some clever tricks, involving quarter wave phase shifter or so, to achieve time symetry breaking.This would provide an interesting time demon.

You might want to approach a local University. There is bound to be a fusion splicer available somewhere on campus that you might gain access for your small job.