In the design of
a fiber optic transmission system, the first step is to determine which
transmitters and receivers are best suited to the signal type. The best
way to find the right system is to compare data sheets and consult with
sales engineers to find which products best meet the system
specifications. Once this is done, the next consideration is the choice
of the fiber optic cable itself, the optical connectors to be used and
the method of attaching these connectors.
This portion of the system design is not so straightforward and is
shrouded in a great deal of misunderstandings and fear of complex “glass
grinding” techniques by the inexperienced. This booklet should clarify
several misconceptions about fiber cable and termination.
Categories of UTP
Category 3, data to 10 MHZ only (Ethernet)
Category 4, data to 20 MHZ (16 MHZ TokenRing)
Category 5, when supplier guarantee operation in bandwidth to the 100 MHZ abstractedly from using protocol (Ethernet, TokenRing, CDDI etc ..
- Extended category 5 (alternatively 5+), which works as well in bandwidth to the 100 MHZ, however requires new manners measuring parameters and in some of them is strictier. The goal is to run Gigabit Ethernet.
- Category 6 with bandwidth to the 200 MHZ.
- Category 7 with bandwidth to the 600 MHZ
Once exists categories 3 and 4. Cable distribution according to these categories is mostly necessary to remake today.
Fiber optic cable
Fiber optic cable physics
Optical Fiber: Thin strands of highly transparent glass or sometimes plastic that guide light.
Core: The center of the fiber where the light is transmitted.
Cladding: The outside optical layer of the fiber that traps the light in the core and guides it along - even through curves.
Buffer coating or primary coating: A hard plastic coating on the outside of the fiber that protects the glass from moisture or physical damage.
Mode: A single electromagnetic field pattern (think of a ray of light) that travels in fiber.
Multimode fiber: has a bigger core (almost always 62.5 microns - a micron is one one millionth of a meter - but sometimes 50 microns) and is used with LED sources at wavelengths of 850 and 1300 nm for short distance, lower speed networks like LANs.
Singlemode fiber: has a much smaller core, only about 9 microns, and is used for telephony and CATV with laser sources at 1300 and 1550 nm. It can go very long distances at very high speeds.
Both multimode and singlemode fiber have an outside diameter of 125 microns - about 5 thousandths of an inch - just slightly larger than a human hair.
Plastic optical fiber (POF): is a large core (about 1mm) multimode fiber that can be used for short, low speed networks. POF is used in consumer HiFi and starting to be used as part of a new standard for car communication systems called MOST
For more on optical fiber, go here.
Step index, Graded index optical fiber
Whether loose-buffer or tight-buffer, the actual glass fiber used in any fiber optic cable only comes in one of two basic types, multimode fiber for use over short to moderate transmission distances (up to about 10 Km) and single-mode fiber for use over distances that are generally greater than 10 Km. Communications grade multimode fiber normally comes in two sizes, 50 micron core and 62.5 micron core, the latter being the size most commonly available. The outer diameter of both is 125 microns and both use the same connector size. Single-mode fiber comes in only one size, 8-10 microns for the core diameter and 125 microns for the outer diameter. Connectors for single-mode fiber are not the same as those designed for multimode fiber but can look the same as we will soon discuss.
Figure 3 is a drawing of the construction of two types of optical
fiber, step index and graded index.
Losses in an optical fiber are the result of absorption and impurities within the glass as well as mechanical strains that bend the fiber at an angle that is so sharp that light is actually able to “leak out” through the cladding region. Losses are also dependent on the wavelength of the light employed in a system since the degree of light absorption by glass varies for different wavelengths. At 850 nanometers, the wavelength most commonly used in short-range transmission systems, typical fiber has a loss of 4 to 5 dB per kilometer of length. At 1300 nanometers this loss drops to under 3 dB per kilometer and at 1550 nanometers, the loss is a dB or so. The last two wavelengths are therefore obviously used for longer transmission distances.
The losses described above are independent of the frequency or data rate of the signals being transmitted. There is another loss factor however that is frequency (and wavelength) related and is due to the fact that light can have many paths through the fiber. Figure 4 shows the mechanism of this loss through step-index fiber.
A light path straighter through a fiber is shorter than a light path with maximum “bouncing”. This means that for a fast rise-time pulse of light, some paths will result in light reaching the end of the fiber sooner than through other paths. This causes a smearing or spreading effect on the output rise-time of the light pulse which limits the maximum speed of light changes that the fiber will allow. Since data is usually transmitted by pulses of light, this in essence limits the maximum data rate of the fiber. The spreading effect for a fiber is expressed in terms of MHz per kilometer. Standard 62.5 micron core multimode fiber usually has a bandwidth limitation of 160 MHz per kilometer at 850 nanometers and 500 MHz per kilometer at 1300 nanometers due to its large core size compared to the wavelength of the propagated light. Single mode fiber, because of its very small 8 micron core diameter has a bandwidth of thousands of MHz per kilometer at 1300 nanometers. For most low frequency applications however, the loss of light due to absorption will limit the transmission distance rather than the pulse spreading effect.
All of the common types of connectors are fairly simple to install, although you can expect a 10-percent loss until installers have a few days worth of experience. After that, figure on losses of 2 to 5 percent, depending on the cleanliness of the area in which the connections are made.
Before the installation of connectors onto a fiber cable, a breakout kit may have to be installed. This procedure will not be necessary on breakout cables having 2-mm buffered fibers, but will be required on 250-, 500-, and 900-micron tight-buffer cables. The breakout kit consists of a buffer tubing (usually 2 mm) over a 900-micron inner tubing. The bare fibers are inserted into these buffer tubes to provide handling protection and strength when mounted onto connectors.
Installing a fiber connector onto a pigtail or unbuffered fiber can be done in several ways. The three most common are epoxy glue with oven-cure, then polish; Hot Melt pre-glue, then polish; and cleave and crimp, no polish.
The epoxy-glue method is the oldest and is still widely used today. This process involves filling the connector with a mixed two-part epoxy, then insetting the prepared and cleaned fiber into the connector. After curing the epoxy in an oven for the specified period of time (usually 5 to 20 minutes) the fiber is scribed and cleaved nearly flush with the end of the connector. Finally, it's polished with a succession of finer and finer lapping papers (typically ranging from 3-micron grit down to 0.3- micron grit).
With the Hot Melt method (a trademark of 3M Co.), the connector come preloaded with glue and must be place into an oven to soften the glue. Clean, prepared fiber is then inserted into the connector, then left to cool. After cooling, fiber is scribed and polished in the same process as used in the epoxy method.
Cleave and crimp connectors do not require a polish procedure since these connectors already have a polished ferrule tip. Thus, installation simply involves inserting a properly cleaved fiber to butt against the connector's internal fiber "stub." The fiber connector is then crimped to hold the fiber in place. Each mounting method has advantages and disadvantages, varying from ease of installation to cost per connector to performance qualities.
Fiber Optic Splicing
What is Fiber Optic Splicing
of fiber optic splicing methods is vital to any company or fiber optic
technician involved in Telecommunications or LAN and networking projects.
Simply put, fiber optic splicing involves joining two fiber optic cables together. The other, more common, method of joining fibers is called termination or connectorization. Fiber splicing typically results in lower light loss and back reflection than termination making it the preferred method when the cable runs are too long for a single length of fiber or when joining two different types of cable together, such as a 48-fiber cable to four 12-fiber cables. Splicing is also used to restore fiber optic cables when a buried cable is accidentally severed.
There are two methods of fiber optic splicing, fusion splicing & mechanical splicing. If you are just beginning to splice fiber, you might want to look at your long-term goals in this field in order to chose which technique best fits your economic and performance objectives.
Mechanical Splicing vs. Fusion Splicing
hold the two fiber ends in a precisely aligned position thus enabling light to pass from one fiber into the other. (Typical loss: 0.3 dB)
Which method is better?
As for the performance of each splicing method, the decision is often based on what industry you are working in. Fusion splicing produces lower loss and less back reflection than mechanical splicing because the resulting fusion splice points are almost seamless. Fusion splices are used primarily with single mode fiber where as Mechanical splices work with both single and multi mode fiber.
Many Telecommunications and CATV companies invest in fusion splicing for their long haul singlemode networks, but will still use mechanical splicing for shorter, local cable runs. Since analog video signals require minimal reflection for optimal performance, fusion splicing is preferred for this application as well. The LAN industry has the choice of either method, as signal loss and reflection are minor concerns for most LAN applications.
Fusion Splicing Method
Four basic steps to completing a proper fusion splice:
Step 1: Preparing the fiber - Strip the protective coatings, jackets, tubes, strength members, etc. leaving only the bare fiber showing. The main concern here is cleanliness.
Step 3: Fuse the fiber - There are two steps within this step, alignment and heating. Alignment can be manual or automatic depending on what equipment you have. The higher priced equipment you use, the more accurate the alignment becomes. Once properly aligned the fusion splicer unit then uses an electrical arc to melt the fibers, permanently welding the two fiber ends together.
Step 4: Protect the fiber - Protecting the fiber from bending and tensile forces will ensure the splice not break during normal handling. A typical fusion splice has a tensile strength between 0.5 and 1.5 lbs and will not break during normal handling but it still requires protection from excessive bending and pulling forces. Using heat shrink tubing, silicone gel and/or mechanical crimp protectors will keep the splice protected from outside elements and breakage.
Mechanical Splicing Method
Four steps to performing a mechanical splice:
Step 1: Preparing the fiber - Strip the protective coatings, jackets, tubes, strength members, etc. leaving only the bare fiber showing. The main concern here is
Step 3: Mechanically join the fibers - There is no heat used in this method. Simply position the fiber ends together inside the mechanical splice unit. The index matching gel inside the mechanical splice apparatus will help couple the light from one fiber end to the other. Older apparatus will have an epoxy rather than the index matching gel holding the cores together.
Step 4: Protect the fiber - the completed mechanical splice provides its own protection for the splice.
More info at
Heat Shrink Tubing -
Heat Shrink Tubing - rainscreen cladding - printer ink