Glowing optical fiber for identification

[bitwelder]

Some time ago I saw that during installation of an optical fiber that was passing by our side of the office building the serviceman must have put on the remote end of the fiber a testing/detection device that was making the whole fiber glow.

How is that accomplished? I thought that the regular optical fiber for networking is only capable to sending light axially.
Is is only about shooting light at the right angle into the fiber (outside the core?), or does also the installed fiber itself need to have some kind of side-glowing cladding capability?

[coppice]

Inject light into the core of the fibre and you will see nothing, as the loss is extremely low. Inject the light into the outer layer of glass, and you will see some glow.

[Marck]

If you have access to the cores you can connect a visual fault locator at one end that injects visible light.  Then all you need to do is look for the glow at the end of the fibre.  You can also put a bend in the fibre and visible light will then show up through the sides of the core. 

Fibre cores are not as fragile as people might have you believe in our training we tied a knot in the core which showed huge losses but once you released the knot results returned to normal.  The problem is if they are mistreated and you break a core its a massive pain to terminate or splice all the cores at that point to fix it. 

Keep in mind that these VFL’s have an output up around 20mw so looking directly at the ends of the cores is not recommended.

Its been 10 years since i did my training so things may have changed a bit since then.

M

[TERRA Operative]

Fibre cores are not as fragile as people might have you believe in our training we tied a knot in the core which showed huge losses but once you released the knot results returned to normal.

This is actually a workable way to make a poor man's attenuator.

Grab a pen or pencil and stick a light tester on the fiber then wrap it around until the light level drops to the correct level then tape it in place (usually done so you don't blow out the Rx sensor with a too powerful signal).

Also, you can get light testers that will look for the small amount of light leakage through the walls of the fiber so you can get a go/no-go confirmation if a fiber is dark or not.

[Marck]

Throwing a multimode fibre in the single mode port is another trick you can use to get enough attenuation if you want to run gear up on the bench and don’t have 5 km of fibre laying around.

[bitwelder]

Yeah, I have/had a torch with an adapter to push onto the terminated ends for that sort of job.

Ok. Any suggestion for what could be the name of that device? Not planning to buy, but just to see how it looks like.

Worth noting that it may not be the fibre that glows because, IIRC, a lot of fibres have a plastic coating on the glass as well as the aramid and plastic/nylon sheath.

Hmm... I'm not 100% sure about my recollection, but I think what I saw was the whole fiber from the patchpanel to the free end was glowing, while being illuminated remotely, maybe about 10-20 m away in the building basement, jumping though several cables and patchpanels. So whatever method they use, it seems to be able to cross through optical fiber connectors.

[madires]

Fibre cores are not as fragile as people might have you believe in our training we tied a knot in the core which showed huge losses but once you released the knot results returned to normal.  The problem is if they are mistreated and you break a core its a massive pain to terminate or splice all the cores at that point to fix it. 

Yep, but it depends on the fiber type. One type is optimized for a small bend radius and is therefore used for patch cables.

[madires]

This is actually a workable way to make a poor man's attenuator.

Grab a pen or pencil and stick a light tester on the fiber then wrap it around until the light level drops to the correct level then tape it in place (usually done so you don't blow out the Rx sensor with a too powerful signal).

That attenuates the signal but also increases some optical side effects considerably which might break the link, especially for high speed links. Better use proper attenuaters.

[madires]

Throwing a multimode fibre in the single mode port is another trick you can use to get enough attenuation if you want to run gear up on the bench and don’t have 5 km of fibre laying around.

But that can damage the port's connector, i.e. the polished fiber, causing an increased loss. And that might break a long haul link when each dBm counts. The connectors typically used create a physical contact between fibers. MM connectors have poorer tolerances and the fiber has a rougher polish than SM, besides the obvious difference in core sizes.

[TERRA Operative]

Yeah, I have/had a torch with an adapter to push onto the terminated ends for that sort of job.

Ok. Any suggestion for what could be the name of that device? Not planning to buy, but just to see how it looks like.

It's called a VFL or Visual Fault Locator.
You can find them on ebay and the common ones look like a standard aluminium flashlight, except it has a laser diose and a (usually) SC fiber connection in the end instead of a globe and reflector.

[Berni]

Its a thing that looks like a flashlight, except that on the end you have a small port to plug the fiber into.



I'm not quite sure how it works. The one that the technicians from my ISP used when connecting me up to fiber internet seamed to be using laser light because i could see the fiber glowing with that weird speckily self interference pattern that you see when looking at coherent light sources. They used it to locate the correct fiber going to my house at the other end.

My guess is that this works by shining a massive amount of non colimated light into the fiber. Since a lot of the light comes in at weird angles means the fiber doesn't work so well like a fiber anymore and exceeds the total internal reflection angles so the light starts escaping, especially in tight bends. Fiberoptic lines also tend to be designed for infrared light so red light also might have a easier time escaping. I'm also guessing a lot of this light is traveling in the outer layers of the fiber.

[TERRA Operative]

That's the VFL I have, it's basically a laser pointer with the SC 'nozzle' fitted. (I bought a separate SC-LC adapter too).
It's saved me a few hours troubleshooting bad fibers in the data centers already, already paid for itself a few times over.

[NiHaoMike]

Once, someone asked for help identifying a single fiber out of an unused bundle. I suggested holding one end up to the power light of a nearby PC so it would be a distinct color compared to ambient light. Worked great.

[bitwelder]

Thanks. I did look up at the VFLs before asking here, but I thought that they work so that either you see light at the other end (if the fiber is ok) or a little bit on the side locally where the fiber has snapped/broken.
Or perhaps there is a different model that blasts a lot of un-collimated light to make the whole fiber visual identification easier (although probably it won't work on the long distances) ?

[ConKbot]

Most of the light in a VFL isn't launched correctly, so you can end up with a few meters of fiber glowing, as the incorrectly launched light bleeds out.

There is something similar for fiber power/loss meters. The LED or simple laser diode sources don't fill the fiber correctly, so if you take two 1m patch cables, zero the meter, then hook to your DUT, your DUT loss will read high as the "bad" light bleeds out in the DUT.

The solution is a controlled bend which encourages all but the correctly launched light to please exit the fiber, so the power meter can be zeroed correctly.

https://www.fiberoptics4sale.com/products/f1-ccm04

Not to be confused with the launch cable for an OTDR which is primarily a delay line for the dead-time of the port.

[I4get]

You can also use your camera phone to see the light coming out of the end of the fiber from the transceiver.  While it is high enough intensity that you wouldn't want to point it at your eye, and if it is 10gig or higher very likely out of the visible spectrum, it will show up as a blue light on your smartphone camera.  This usually works for IR in general (point it at the end of your TV remote for testing).

[TERRA Operative]

10G and 100G are both IR in the 10's of mW range max.
But being IR, there is no blink reflex, so best to not stare into the fiber with your remaining eye.