Windows Knowledge Base

[NOC Tutor]Fiber-Optic Tests Procedure

Just as installing fiber-optic cable is completely different from installing copper-cable, the testing
processes also differ greatly. Much of the copper-cable testing revolves around the various
types of interference that can affect the performance of a network. Fiber-optic cable is completely
immune from interference caused by crosstalk, EMI, and RFI, however, so tests for these are not needed.
What you do need for a fiber-optic installation is to ensure that the signals arrive at their destinations
with sufficient strength to be read and that the installation process has not degraded that strength.

Because of its superior signal-carrying capabilities, fiber-optic cable installations can include
various types of cable runs. The typical LAN arrangement consists of single-fiber links that connect
a patch panel in a wiring closet or data center to wall plates or other individual equipment sites
over relatively short distances, with patch cables at both ends to connect to a backbone network
and to computers or other devices. Because of the limited number of connections they use, testing
these types of links is fairly straightforward. However, fiber optic can also support extremely long
cable runs that require splices every two to four kilometers, which introduces a greater potential
for connection problems.
To completely test a fiber-optic installation, you should perform your battery of tests three
times. The first series of tests should be on the spooled cable before the installation to ensure
that no damage occurred during shipping. The installation costs for fiber-optic cable can be
high—often higher than the cost of the cable and other hardware—so it’s worthwhile to test
the cable before investing in its installation. Because excessive signal loss is caused mostly by the
connections, simply testing the continuity of the cable at this stage is usually sufficient. This
continuity testing is sometimes referred to as a “flashlight” test because it amounts to shining
light in one end of the fiber strand and seeing if there is light at the other end.
The second series of tests should be performed on each separate cable segment as you install
it, to ensure that the cable is not damaged during the installation and that each individual connector
is installed correctly. By testing at this stage, you can localize problems immediately,
rather than trying to track them down after the entire installation is completed.
Finally, you should test the entire end-to-end connection, including all patch cables and
other hardware, to ensure that cumulative loss is within certified parameters.
For fiber-optic LAN installations, only two tests are generally required: optical power and
signal loss. The following sections examine these tests and how you perform them. Other types
of tests are used on long-distance fiber-optic links and in troubleshooting, which are much
more complex and require more elaborate equipment.

Optical Power
The most fundamental test of any fiber-optic-cable plant is the optical-power test, as defined in the
EIA’s FOTP-95 standard, which determines the strength of the signal passing through a cable
run and is the basis for a loss-measurement (attenuation) test. The testing process involves connecting
a fiber-optic power meter to one end of the cable and a light source to the other. The
power meter uses a solid-state detector to measure the average optical power emanating from the
end of the cable, measured in decibels. For data networks using multimode cable, you should perform
optical-power tests at 850 and 1,300nm wavelengths; many testers run their tests at both
settings automatically. Single-mode cables require a 1,300nm test and sometimes 1,550nm, as
well. The 1,550nm test determines whether the cable will support wavelength division multiplexing
and can detect losses due to microbending, which are not apparent at 1,300nm.

Loss (Attenuation)
Loss testing, along with optical power, are the two most important tests for any fiber-optic
cable installation.Loss is the term commonly used in the fiber-optic world for attenuation; it is
the lessening of the signal as it travels through the cable. The physics of optical transmission
make it less susceptible to attenuation than any copper cable, which is why fiber cable segments
can usually be much longer than copper ones. However, even if your network does not have
extremely long fiber cable runs, there can be a significant amount of loss, not because of the
cable, but because of the connections created during the installation. Loss testing verifies that
the cables and connectors were installed correctly.
Measuring the loss on a cable run is similar in practice to measuring its optical power, except
that you use a calibrated light source to generate the signal and a fiber-optic power meter to measure
how much of that signal makes it to the other end. The combination of the light source and
the power meter into one unit is called an optical loss test set(OLTS). Because of the different
applications that use fiber-optic cable, you should be sure to use test equipment that is designed for
your particular type of network. For example, a light source might use either a laser or an LED
to create the signal, and the wavelengths it uses may vary as well. For a fiber-optic LAN, you
should choose a product that uses a light source at wavelengths the same as the ones your network
equipment will use so that your tests generate the most accurate results possible. The testing
procedure begins with connecting the light source to one end of a reference test cable (also called the
launch cable) and the power meter to the other end. The reference test cable functions as a baseline
against which you measure the loss on your installed cable runs and should use the same type of cable
as your network. After measuring the power of the light source over the reference test cable, you
disconnect the power meter, connect the reference cable to the end of the cable you want to test,
and connect the power meter to the other end. Some testers include a variety of adapters to accommodate
various connector types. By taking another power reading and comparing it to the first one, you can
calculate the loss for the cable run. As with the optical-power test, you should use both 850 and
1,300nm wavelengths for multimode fiber tests; you should also test the cable from the other direction
in the same way. When you have the results, compare them to the optical loss budget (OLB), which is
the maximum amount of signal loss permitted for your network and your application. (You may occasionally see
optical link budget; though optical loss budget is preferred, the two terms are synonymous.)

Depending on the capabilities of your equipment, the loss-testing process might be substantially
easier. Some power meters have a zero loss reference capability, meaning that you can set the meter
to read 0dB while measuring the reference test cable. Then, when you test the installed cable run,
the meter displays only the loss in decibels; no calculation is necessary.

Optical Loss Test Sets and Test Kits
In most cases, you need both an optical power meter and a light source in order to properly
install and troubleshoot a fiber-optic network, and you can usually save a good deal of money
and effort by purchasing the two together. You will thus be sure to purchase units that both
support the wavelengths and power levels you need and that are calibrated for use together.
You can purchase the devices together as a single combination unit called an optical loss test
set (OLTS) or as separate units in a fiber-optic test kit.
An OLTS is generally not recommended for field testing, because it is a single unit. While
useful in a lab or for testing patch cables, two separate devices would be needed to test a permanently
installed link because you have to connect the light source to one end of the cable and
the power meter to the other. However, for fiber-optic contractors involved in large installations,
it may be practical to give workers their own OLTS set so that they can work with a partner
and easily test each cable run in both directions.
Fiber-optic test kits are the preferable alternative for most fiber-optic technicians because
they include a power meter and light source that are designed to work together, usually at a
price that is lower than the cost of two separate products. Many test kits also include an assortment
of accessories needed to test a particular type of network, such as adapters for various
types of connectors, reference test cables, and a carrying case. Prices for test kits can range from
$500 to $600 for basic functionality to as much as $5,000 for a comprehensive kit that can test
virtually every type of fiber-optic cable.