Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Can You Build a Fiber Test Kit On a Budget?

timothy posted more than 6 years ago | from the un-un-explained-acronyms dept.

Networking 53

An anonymous reader writes "Have any Slashdot readers hacked together cheap test kits for fiber optic cable? More and more IT infrastructure is using multimode and single mode fiber optic cabling. Commercial test equipment is extremely expensive, running the gamut from a few hundred dollars for a basic light source, to tens of thousands for an OTDR. What equipment do you consider essential to your fiber kit? Is there a way to save costs when it comes to fiber test equipment? It is worth it to do so?"

Sorry! There are no comments related to the filter you selected.

Just get.... (3, Informative)

b96miata (620163) | more than 6 years ago | (#24658149)

Two Media converters. If you can run ethernet over it, it's good.

Re:Just get.... (5, Informative)

juiceboxfan (990017) | more than 6 years ago | (#24658783)

Two Media converters. If you can run ethernet over it, it's good.

That's true but only to a certain extent.
It is possible that 1Gig ethernet will work fine on a cable that will not work or has a high error rate at 10Gig.

You also will not be able to determine if a long run of single mode fiber has a multi mode patch in the middle of it (OTDR is the only thing that might give you that info).
An Optical Power Meter [] is your best investment [] . That along with a good cleaning kit will give you the best results. A lot of "bad" fibers can be traced back to dirty connectors. _Always_ clean the fiber before plugging it into an interface - it is much easier to clean the fiber than it is to clean interface optics that have been contaminated by a dirty connector end.

Beyond that if this is for in-house work just plan on using your fastest interface as a tester during downtime - setup the interface to expect loopback and put a short loopback patch at the far end. Run data through the cable and check for errors.

Or as the parent said get a couple of cheap O/E converters for field work - not as good as a fast interface but better than a power meter alone.

Re:Just get.... (1)

ArsonSmith (13997) | more than 6 years ago | (#24666941)

hold one end to your eye, shine a flashlight in the other.

Re:Just get.... (3, Informative)

mbeckman (645148) | more than 6 years ago | (#24673819)

Two Media converters. If you can run ethernet over it, it's good.

Alas, this can give you very misleading results. Ethernet is such an error-tolerant protocol that you can "run it" over even a very poor fiber link. You'll get rotten performance, but most people won't notice right away, and won't know how to isolate the problem to the fiber link even if they do notice.

One particularly insidious performance degradation occurs when only one fiber in a send-receive pair has high power loss: random spanning-tree packet storms that can take down an entire network. Even pros have trouble curing this kind of systemic problem.

If you verify the optical integrity of your fiber network, then you'll be able to troubleshoot other components -- such as transceivers and pach cables -- much more easily (since these parts can be readily swapped out to check for failures). By far the single most common cause of campus LAN problems I see as a network engineer are defective but working fiber links.

Sometimes it's just dirty connectors (cleaning fiber cables consistently every time you manipulate them is the best defense against this), but often it's poor installation quality, remedied by re-terminating high-loss connectors.

no (0, Insightful)

Anonymous Coward | more than 6 years ago | (#24658157)

No. You get what you pay for.

That would depend on the size of the budget (1, Funny)

Anonymous Coward | more than 6 years ago | (#24658229)

If your budget is $1 million, no problem. If your budge is $5, problem.

Re:That would depend on the size of the budget (1)

Tubal-Cain (1289912) | more than 6 years ago | (#24658291)

If he had a million-dollar budget, I don't think he'd be asking.

Re:That would depend on the size of the budget (1)

Bastard of Subhumani (827601) | more than 6 years ago | (#24658949)

If he had a million-dollar budget, I don't think he'd be asking.

Perhaps he spent 999,995 bucks on a feasibility study from McKinsey?

The conclusion being, to cut a long story with lots of barcharts short: "maybe".

Re:That would depend on the size of the budget (1)

Bloke down the pub (861787) | more than 6 years ago | (#24672053)

Recession must be biting, looks like they've lowered their rates.

Re:That would depend on the size of the budget (1)

Bandman (86149) | more than 6 years ago | (#24660205)

You might be surprised, but money doesn't buy expertise or knowledge.

Re:That would depend on the size of the budget (2, Funny)

Anonymous Coward | more than 6 years ago | (#24659135)

Even at $5, you have options. You can use you eyes up to 2 times to perform a rudimentary check of the light source.

Cheap Kit == Cheap Results (4, Insightful)

Anonymous Coward | more than 6 years ago | (#24658267)

Depends on what you need to do.

There are many different tests you can do on fiber. A basic college course will teach you that.

Do you just want to know if the fiber is not "broken"? a LED and a phototransister can do that.

Want to know if the fiber is the correct mode, optimized for the wavelength of the led, or can support multimode and run parametric tests on the fiber, gee, that gets more expensive.

Its the same for the DIY ethernet cable testers too.

Want to test continuity? a 5$ multimeter will do that.

Want to determine if the cable meets spec? You need better equipment.

Better Equipment GENERALLY equals better [test] results. [ yes, a lot can depend on the skill of the tech operating said equipment too].

Re:Cheap Kit == Cheap Results (2, Insightful)

Creepy Crawler (680178) | more than 6 years ago | (#24658517)

And that means a Fluke with full fiber gear.

Re:Cheap Kit == Cheap Results (0)

Anonymous Coward | more than 6 years ago | (#24658701)

Cheap Kit == Cheap Results

(Psst, I think you mean that "cheap kit" leads to "cheap results", not that they are the same thing, which is what "equals" mean.)

Depends on what you need to do.

There are many different tests you can do on fiber. A basic college course will teach you that.

Yes, you must absolutely take a college course in every subject you intend to approach.

Do you just want to know if the fiber is not "broken"? a LED and a phototransister can do that.

Yeah, my sister does it all the time.

Better Equipment GENERALLY equals better [test] results.

You know, "generally" means that something holds for all considered cases, as opposed to just some special case. An example is "general relativity" versus "special relativity". That's what it means in mathematics, and other meanings are of course of no interest.

Re:Cheap Kit == Cheap Results (1)

wsanders (114993) | more than 6 years ago | (#24680249)

There is tons of this stuff for sale cheap on Ebay.

If your maintenance budget is only $50, though, maybe you should have stuck with Cat-V. The equipment for terminating and connecting fiver is much more expensive than for UTP, and considerable skill is required to do it right.

I've got a better idea (1, Funny)

neokushan (932374) | more than 6 years ago | (#24658337)

Nick one.

Heh (2, Informative)

Xtense (1075847) | more than 6 years ago | (#24658383)

Try a flashlight!

Re:Heh (3, Informative)

phreakngeek (1250360) | more than 6 years ago | (#24661089)

Bad moderator! No cookie! The parent is NOT offtopic. A flashlight works for a basic transmission test. I was a fiber optic technician in a former life and after terminating the ends with connectors we would shine a maglite on one end and turn out the lights on the other and if we saw the "light at the end of the tunnel" we'd call it good. If you want to get fancy you're going to need an optical time domain reflectometer. Probably not the cheapest thing.

Re:Heh (1)

Tmack (593755) | more than 6 years ago | (#24666647)

"Do not look into fiber with remaining eye"


Re:Heh (1)

phreakngeek (1250360) | more than 6 years ago | (#24667213)

My eyes are fine. There was nothing dangerous about looking at the maglite beam from the other end of the fiber. This wasn't some kind of fancy magnified light that attached to the connector - it was just a regular flashlight that we held up to the end. But, we also use a cheap laser pointer for longer runs - here the guy on the "dumb" end of the test in the dark (often me) would just hold them up and look for the glowing red end of the fiber without looking down the tube. We'd only order out the OTDR as a last resort.

1 step up from a flashlight (0)

Anonymous Coward | more than 6 years ago | (#24658437)

Everybody has used the flashlight (or desk lamp) approach to testing fibre. One step up is a small project box with a battery and LED in it, drill a small hole in line with the LED that you can put the test cable into. The other side is a project box with battery and phototransistor hooked to a beeper (or alternatively to a small current meter ). The light from the transmitter box runs through the cable and hits the detector box and makes noise or moves the needle on the current meter. While its by no means accurate for a measurement its good enough to tell you if the cable is cut or badly damaged. In general I'd expect similar results from each strand in a multi-strand bundle.

You use this word, 'gambit...' (4, Informative)

Khakionion (544166) | more than 6 years ago | (#24658441)

"running the gambit" I don't think it means what you think it means.

Re:You use this word, 'gambit...' (2, Funny)

JeffSh (71237) | more than 6 years ago | (#24658673)

indeed. it is "gamut".. which is ironic since he is talking about multi mode fibre..

Re:You use this word, 'gambit...' (1)

Hatta (162192) | more than 6 years ago | (#24664941)

For all intensive purposes it's the same.

Re:You use this word, 'gambit...' (0)

Anonymous Coward | more than 6 years ago | (#24666929)

intensive purposes? do you mean intents and purposes?

Re:You use this word, 'gambit...' (1)

Khakionion (544166) | more than 6 years ago | (#24667361)

intensive purposes? do you mean intents and purposes?

O <---- Joke

~=~ <----Earth's Atmosphere

X <-----Your head

Running the GAMUT. GAMUT. Not "gambit". (1, Funny)

blach (25515) | more than 6 years ago | (#24658465)

Please, for god's sake editors, do some editing.

It's "running the GAMUT," which means "the full range or compass of something; a range from one extreme to the other"

GAMBIT, on the other hand, means:
1. Chess. an opening in which a player seeks to obtain some advantage by sacrificing a pawn or piece.
2. any maneuver by which one seeks to gain an advantage.
3. a remark made to open or redirect a conversation.

But I could see how they could be confused, they both start with a G and have vowels *rolling eyes*

Re:Running the GAMUT. GAMUT. Not "gambit". (1)

spazghost (1346955) | more than 6 years ago | (#24658793)

I thought it was running the Gumbo... If we're following the 'starts with G and has vowels' rule than that'd work perfectly.

Re:Running the GAMUT. GAMUT. Not "gambit". (1)

unitron (5733) | more than 6 years ago | (#24668555)

I thought it was running the Gumbo...

Nah, any halfway decent gumbo will be too thick to run.

Although I suppose gumbo could be used in a different kind of "fiber" test. :-)

Even though it doesn't (1)

Bastard of Subhumani (827601) | more than 6 years ago | (#24659037)

That begs the question of how long it will be before a large number of repliess along the lines of "languages evolve" "if people use it to mean that then that's what it means" or "so what, I understood it" will arrive.

Re:Even though it doesn't (1, Interesting)

Anonymous Coward | more than 6 years ago | (#24659523)

languages evolve

"Begs the question" was a terrible translation of petitio principii in the first place, and the language never meant what people asserts it means. A proper translation from Latin would be "seeking the principles", which actually DOES describe the fallacy wherein assuming X (your principle) you make a long winded argument that eventually proves X (seeking it), which has nothing at all to do with "begging" or a "question" in the first place.

I hope people continue to use "begs the question" in its proper English meaning as you have, and that we can hold back the tide of linguistic Humpty Dumptys [] who continue to insist on "evolving" the language to mean whatever they wish it to be.

Evolve? Boy does it ever. (1)

hey! (33014) | more than 6 years ago | (#24659893)

Hasn't "language evolves" come to mean "I can't wait until the ignorant take over the world?"

Oh, wait.

Running the "gambit". (1)

camperdave (969942) | more than 6 years ago | (#24659069)

Ah, but it is correct. "Running the gambit" would mean, essentially, "playing the game". The commercial test equipment manufacturers are maneuvering price points and feature sets in order to gain the advantage of making a sale. So, they are "running the gambit", ie. playing the marketing game.

Okay, I can't keep a straight face anymore. It is gamut, not gambit. But I held it together pretty good there, for a while.

Re:Running the GAMUT. GAMUT. Not "gambit". (1)

HTH NE1 (675604) | more than 6 years ago | (#24659125)

Please, for god's sake editors, do some editing.

Ostensibly they are quoting the anonymous user. Which is most proper: to correct such mistakes in quotations, to call them out with "[sic]", or let them stand?

For my part, I tagged with typoinsummary, runningthegamut, and !gambit.

Fiber doesn't cost much here.... (3, Funny)

Barsteward (969998) | more than 6 years ago | (#24658779) []

Re:Fiber doesn't cost much here.... (0)

Anonymous Coward | more than 6 years ago | (#24659015)

That's fibre not fiber ;-)

For some reason Weetabix brings to mind X-Ray Spex []

Confucius say (1)

Profane MuthaFucka (574406) | more than 6 years ago | (#24658873)

Confucius say "cheapest way to test fiber is to eat cow pie."

Had to answer this question once... (3, Interesting)

caffiend666 (598633) | more than 6 years ago | (#24659333)

Had to answer this question once ourselves. Never got to the implementation phase, but the plan was to have one or two high-end test sets, and build the others out of the prototypes for the equipment we were building, and Lucent OptiStar NICs on Linux systems. IE, each protype would test the next, and we would have racks full of 1U Linux Boxen with Lucen OC-12 NICs. Can't find out of Lucent or Alcatel-Lucent is still making those things, but you can find used/refurbed OptiStar (?) SONET IP Nics for Linux (PCI) fairely cheaply. No idea if they work, but as low as $60 bucks. New, they were $10k to $6k. Searching for Optistar on Ebay or Google returns a lot of info. Or, searching for SONET PCI NIC. Or, just FIBER PCI NIC. You can probably set up a laptop with Gig-E, if you just want connection tests, regardless of which protocol you're going to use. Honestly, eBay/Amazon are your friends here. At times, you can find million-dollar test sets for $100 bucks (Adtech, Omniber, Cerjac...) Most are left-overs from when startups fail. Also depends on what you're testing.

I would use ... (1, Interesting)

Anonymous Coward | more than 6 years ago | (#24659339)

I used to build fibre optic cables and every cable that didnt have glowing bits when you pushed a laser down them seemed to test perfectly. So I would use a 5mW green laser and if any bits on the cable glow you know that cable is broken.

Re:I would use ... (0)

Anonymous Coward | more than 6 years ago | (#24661235)

I've only been able to push a quarter of the laser into the cable. Were you using a Vise-Grip instead of duct tape to hold the cable, or did you have some sort of cable lubricant so the laser could be inserted more easily?

A power meter, swabs, and spare cables... (0, Troll)

sirwired (27582) | more than 6 years ago | (#24659431)

Really, the cheapest solution is a simple power meter, some cable-cleaning swabs, and a nice long spool of pre-terminated spare cable.

There are full-fledged cable certification devices out there, and you cannot afford one, nor do you need one.

Yes, there are a lot of problems that can happen with a cable that won't be caught with a power meter, but those problems can be prevented by only using factory-terminated cables from a quality supplier.


The basic kit bag is cheap and easy to get (4, Interesting)

mbeckman (645148) | more than 6 years ago | (#24659591)

The primary tools you need are a fiberscope for visual connector inspection and a power meter with laser source, a set of laser goggles (if you plan to work on single-mode or long-haul multimode fiber), an assortment of fiber patch cables, and a mini-maglite with fiber adapter (for simple continuity tests).

You can get these used on eBay all day long for a few hundred dollars for the entire set (search terms: "fiber test*" "fiber patch*" "fiber splice*"). You may have to watch a few days for a bargain, but I'm constantly amazed at how often high-quality fiber test gear goes for a song on eBay. Purchase only name-brand gear, such as Fluke, Noyes, Microtest, TestUM, etc. (You can learn the name brands by looking at new-equipment listings at places like CDW).

There are many tutorials on the web showing how to use these tools, most from equipment vendors themselves, and some are even high quality video presentations. A useful starting point is [] (although it's become less useful since Fluke took it over).

The mini-maglite will instantly identify any fiber -- the light will be clearly visible at the far end of even a thousand-foot run. If high-power IR lasers are in use on the network, be careful to be wearing fiber goggles whenever looking at fiber ends, even at your own white light. You can't tell when a fiber is energized with high-power IR, since it's invisible.

The power meter lets you measure the light loss through a fiber path, which when correctly interpreted will give you performance information. Get one that reads tenths of a microwatt, and that also directly displays dB loss from a reference signal. If you know the installed fiber specifications (you can read these off the fiber jacket), you can compute the available bandwidth based on fiber length.

The fiberscope reveals otherwise-invisible defects in a connector so that you don't spend hours trying to make an unworkable connection work. Again, be sure you're wearing laser goggles if you don't have both ends of a fiber in your hands. Magnifying harmful IR radiation is very dangerous.

For routine work you don't need an OTDR. Besides measuring the length of a fiber, an OTDR will locate defects along a fiber span so you can locate and repair them. If you're not repairing cable, or splicing it, an OTDR is overkill. Fiber installers have such gear, and they'll be happy to use theirs on your network for a fee when you need that capability.

Do it right, spend some money (0)

Anonymous Coward | more than 6 years ago | (#24659885)

Spend your money on the commerical stuff. It's just not worth it to do anything else. As previous posters have noted, use ebay or some reseller. There are some that specialize just for fiber optical equipment.

You need to have

1) An optical power meter that handles multiple wavelengths and can measure at least 0 dBm.

2) A video optical scope. Do not get a passive scope. You'll burn your eye out. Westover scientific sells a great hand-held, USB scope for about $2k-$3K. It's worth it. Get it.

3) A cletop, optical pads, "soft" and "hard" swaps, and some alcohol (for the ferrule, not your body)

If you are going to use DWDM you need an Optical Spectrum Analyser. Aglient is the best but there are others. There is a real cheap handheld that interfaced with LabView (or something) that someone was trying to market for about $2k-$3K but I can't find the page.

If you are doing long distance fiber work, you need an OTDR. No alternatives. Buy an Aglient. (I like them, can't you tell) If you just want to see if you have a fiber break, buy a light source from NOYES or anybody. That with the power meter will tell you if something is wrong with a short piece of fiber.

cheap "laser" detector... (3, Informative)

ak_hepcat (468765) | more than 6 years ago | (#24661069)

There's a really cheap "laser" detector that nearly every tech carries, although you may have forgotten it.

It's that camera on your cell-phone.

Yup, in a pinch, you'll see a nice little purple dot appear on your screen if you've got IR coming down the fiber. Works well enough to identify active cables.

You can also pick up a mag-lite->fiber adapter that'll shine visible down the line fairly cheap.

Re:cheap "laser" detector... (0)

Anonymous Coward | more than 6 years ago | (#24720063)

YMMV, just tested with 5 remotes and a Motorola Razor. 3 worked, 2 can control the device, but show no light on camera. Test with a working fiber first.

military stuff that i use (1, Insightful)

Anonymous Coward | more than 6 years ago | (#24664019)

Im active duty, and in my shop we use a LS/PM model TS-4358 made by EXFO AMERICA. For an OTDR we use an ACTURNA MTS 5000 measurses 850nm and 1300nm wavelengths. The machines are simple to use but have many options.

Multimode fiber reels that I repair come in 500m lengths. I use my led Maglight for simple testing of it passing light or not. So easy a caveman can do it.

Im sure the civilian market has higher standards of min dB loss then the military.

Re: (1)

clint999 (1277046) | more than 6 years ago | (#24670703)

I friggin love this site. And here's about as cheap as you're gonna get. Though it won't exactly test your landlines.

Legos (1)

dontPanik (1296779) | more than 6 years ago | (#24673979)

I remember when I was a kid I had a lego spaceship that had fiber optics on it.
They didn't do anything but they sure looked cool!

Wait, what were we talking about?

Don't put the fiber up to your eyes! (0)

Anonymous Coward | more than 6 years ago | (#24675165)

You should never hold fiber to your eye, as you should never hold a gun towards you.

Buy 2 fiber collimators (they're cheap and you can get them from anywhere that sells laser optics like scientific atlanta). Shine a flashlight down one columator, then...decollumate the beam and shine it on a white surface. You should see even illumination all over the circle. If there are little shooting stars of light around the've got a problem with your termination.

You can compare new fiber to the fiber you want to test to get a sense of how bright the light should be.

hope that fiber to the eyes. The pulses that are used in Fiber Channel and Ethernet fiber can damage your eyes!!!!

An old ancient FCAL hub (0)

Anonymous Coward | more than 6 years ago | (#24683009)

You can get a 1gb FCAL fibre channel HUB for next to nothing. Most people throw them out. But with just two GBIC's installed you can put a fibre cable on it and it will go through its Loop Init and link give you link lights. With some SC/LC adapters you could at least test multimode fibre.

Depends on how much and what type of fibre work... (1)

uncledrax (112438) | more than 6 years ago | (#24692445)

As some pointed out.. what is defined as 'Cheap'? And also what level of work are you doing?

If you're just a cable/server/network guy (like me) then you really just need a cleaner and something basic to check for breaks.. (see comments about mag-lites; etc). I mostly deal with pre-made patch MM/SM.
A decent DB tester on 1300/850 is helpful too just so you can check for light-loss assuming your optics don't tell you (some do).

If you are a 'Fibre Guy', or make your own SM/MM; then you'll want to look at things like Fusion Splicers and OTDRs and other (generally expensive) test kit.

Check for New Comments
Slashdot Login

Need an Account?

Forgot your password?