Hi Martin, I think that the difficulties you mention with fiber are mostly with high grade singlemode fiber used for long distance links or very high data rates. I interned with a (now defunct) company which packaged and sold various optical assemblies for telecoms via fiber. We bought the bare die parts (e.g., laser diodes, photodiodes, etc.) and put them into hermetically-sealed packages with a pre-aligned singlemode fiber pigtail with a connector on the end. I worked on the part of the system which did the internal alignment of the fiber to laser diodes but I saw the benches where people would terminate fiber in connectors and perform fusion splicing in other instances. Anyway, I think for short-range digital audio links you can use cheap plastic fiber and hand tools which cleave the ends. You then use enough transmit power that the reflections at each interface, along with the fiber attenuation, don't reduce the signal too low at the receiver end. One interesting tidbit that I learned at this company was that you could make a quick-and-dirty optical attenuator by just winding fiber around an object which had a radius tight enough that total internal reflection can no longer happen inside the fiber and light is lost at a constant rate per length (which depends on the internal indices of refraction of the core and cladding, as well as the bend radius). You could then vary the attenuation by changing the number of turns. Sean On Thu, Nov 15, 2018 at 12:20 PM Martin McCormick wrote: > The thing that impressed me about fiber optics is that > the fiber isn't that terribly expensive but terminating it is. > > You've got to polish the cut end of the fiber until it is > optically flat and free of deformations or one must use an > electric arc to melt two ends of fiber such that they fuse > together creating one continuous fiber. Equipment that will do > that reliably is quite expensive. Some of the consumer HiFi gear > that uses fiber-optic audio cables may use plastic fibers for all > I know but still terminating them is not trivial. > > Basically, if the fiber optic connection is meant to be > easily connected and disconnected, you polish and lose 2 or 3 DB > per interface or you fuse if you are permanently joining a fiber > to something that you don't plan to remove under normal > circumstances. Those fusion splices don't have any loss since > there is no boundary. > > --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .