Electrically breaking a fiber connection?

[Pack34]

I'm looking to be able to terminate a fiber connection on-demand in order to simulate the fiber being disconnected and reconnected.

I was going to try to use some fiber media converters but they do not seem to work when converting fiber-copper-fiber. Only copper-fiber-copper.

[HB9EVI]

in theory it's working fiber-copper-fiber as well, but it depends...

what light are you running over fiber? are you running simplex or fullduplex?

in case you talk about a common FTTH access, it could be indeed tricky, since there are running two different wavelenght one in each direction at the same time; common mediaconverter normally work in a kind of CPE-mode; but you need then a converter like on provider side, which I don't know if they are available on the normal consumer market.

[coppercone2]

do they have optical mechanical relays?

maybe spin a prism or mirror?

would be nice to have a bullet proof component for breaking fiber lines in remote areas where you don't wanna put weird IC's and PSU shit.

[ejeffrey]

MEMS optical fiber switches are a thing.  Is that what you are looking for?

[TERRA Operative]

If you are talking standard LC or SC connectors, maybe a set of media converters, one to convert from fiber to copper and the other copper to fiber.

Then on the copper interconnect between the two converters, set up a set of transistors or optocouplers or relay(s) or something to make and break the connection.

It would allow you to use a manual switch or a microcontroller etc to simulate different faults as needed.

[coppercone2]

Mems is complicated. I imagined something optomechanical that uses a coil and somemirrors and optics to be really robust.

[ajb]

https://www.thorlabs.com/navigation.cfm?guide_id=2013

[nfmax]

Mems is complicated. I imagined something optomechanical that uses a coil and somemirrors and optics to be really robust.

Imagine away all you want! We used to make 1x8 MM fibre switches just like that - XY coils driven by DACs and individually calibrated for acceptable loss & return loss on each channel. It was wonderful when MEMS switches became available (after some teething troubles associated with the high power levels we use). And you can get SM switches too, and they are cheap.

[Ian.M]

It looks like electromecanical optical switches are still around.
e.g. here's a range of moving fibre ones:  http://luminos.com/products/switches/

Moving fibre is vastly preferable to lenses, mirrors and shutters etc. due to the much lower losses of the simpler optical path, and easier alignment requirments.

[helius]

Google "optical bypass switch". They were used in FDDI dual-ring networks, and are still being made for various switching or monitoring applications in modern networks.

[coppercone2]

does anyone have a teardown video? I want to see how it looks inside., 

[sd]

You could design a small board with two SFF/SFP modules just wired to each other.
To simulate the fiber being disconnected/reconnected you can toggle the TDIS signal of the SFF/SFP modules or enable/disable the 3V3 rail.

[TomS_]

The problem with active solutions is that they are protocol specific, and you cant cater for WDM.

There are electromechanical optical switches available, they are also commonly used in protection switches to select between 1 of n paths in transmission systems, e.g. in case of a fibre break to select an alternate path for either the receiver and/or the transmitter. They are purely optical, no OEO, so you can pass what ever protocols or as many wavelengths as you want through them.

[LapTop006]

does anyone have a teardown video? I want to see how it looks inside., 

Think a mirror on a solenoid. Combined with some lenses on the fiber ends to handle alignment.

[mikeselectricstuff]

I wonder if you could use a really bright source firing into the side of the fibre to saturate the receiver ?

[Kalvin]

Would bending the fiber very sharply give sufficient attenuation to emulate the loss of connection?

[LapTop006]

Would bending the fiber very sharply give sufficient attenuation to emulate the loss of connection?

Well yes, but the more critical question is can you straighten and get your light back, to which the answer is unlikely.

[Chalcogenide]

I wonder if you could move to free space optics (by collimating the light and then re-launching it into a fiber), and then use a mechanical shutter in between the free space path. The losses may be significant, but it should be possible, although maybe not too practical.

[ahbushnell]

do they have optical mechanical relays?

maybe spin a prism or mirror?

would be nice to have a bullet proof component for breaking fiber lines in remote areas where you don't wanna put weird IC's and PSU shit.

How about an Ax.   

[coppercone2]

I was thinking more like a damaged submarine cable or something

[nfmax]

I am amazed by this discussion. Electrically operated fibre switches are a commodity item. Just buy one and drive it with a logic-level input. Who cares what the actual mechanism of operation is? As an added bonus, you can get them with jacketed fibre leads with he connector of your choice ready fitted.

It's like people discussing how to switch an electric current by pumping the air out of a jam jar with some bits from broken light bulbs in it, to make a thermionic valve. Sure, in principle it is possible, but why, for heaven's sake?

[coppercone2]

uh because people will argue about relay choice here all day and quenching plasma arcs with jam jars is interesting?

and because the failure modes are different?