Cable Testing 101 There's No Gain with Gainers

1
Cable Testing 101 There's No Gain with
"Gainers"-Mark Mullins
www.flukenetworks.com 2006-2017 Fluke
Corporation
2
Cable Testing 101 There's No Gain with "Gainers"
Akin to water flowing from a small pipe into a
large pipe, gainers are essentially perceived
increases in optical power that occur at splice
points due to variations in fiber
characteristics, including core diameter,
numerical apertures, mode field diameters and
backscatter coefficients. Since gainers can be a
significant source of confusion for new optical
time domain reflectometer (OTDR) users, let's
take a look at why these occur, what impact they
have and how to avoid them.
3
Cable Testing 101 There's No Gain with "Gainers"
What are the Causes?
Gainers can show up when using an OTDR to measure
loss from one end of a fiber link, and they occur
due to the way in which an OTDR measures
reflected light along the length of the fiber. An
OTDR assumes that fiber characteristics such as
core and cladding size are consistent along the
length with no variations, and it calculates
signal loss based on the amount of reflected
light, or backscatter, that it detects. But even
if two connected fibers are the same type of
fiber, they may not necessarily be cut from the
same length of fiber so variations can still
occur, including different backscatter
coefficients (a fancy term for information about
the relative backscatter level of the fiber). And
that means that the fiber used for the launch and
receive cords of the OTDR may also have a
different backscatter coefficient than the fiber
under test. Different backscatter coefficients
can cause more light to be backscattered after a
connection than before the connection, causing
the OTDR to show a loss value that is less than
it actually is--a gainer.
4
Cable Testing 101 There's No Gain with "Gainers"
Why Is It a Problem?
The term "gainer" makes it seem like you are
actually gaining something, and you might think
that ending up with a lower than actual loss
value is a good thing. Think again. Gainers
ultimately don't gain you anything but headaches
and increased cost. When loss results are lower
than they actually are, you might be under the
misconception that there is plenty of headroom to
add another connection point, extend the distance
or simply guarantee performance. But gainers are
false positives that when taken as true, can
result in the fiber link ultimately not
supporting the application. For example, an OM4
150-meter channel has a maximum channel loss of
1.5dB to support 40 gigabit per second speeds
(40GBASE-SR4). If you're measured loss comes in
at 1.3dB, you might think it's okay to add
another 0.2dB connector. But what if your
measured loss includes a gainer, and the actual
loss of the channel is really closer to 1.4dB?
Now you end up with a customer asking you to come
back and troubleshoot the installation to
determine why they're not getting the data rate
they should.
5
Cable Testing 101 There's No Gain with "Gainers"
How Do You Prevent It?
Preventing gainers from impacting the quality of
your installation is actually quite simple.
That's because wherever there's a gainer, there's
a loser. That's right. While transmitting in one
direction may cause the gainer, when the
measurement is taken in the other direction where
less light is backscattered after the connection,
the measured loss is greater than the actual
loss. The simple solution, and the one required
by industry standards, is to measure in both
directions--otherwise known as bidirectional
testing. As shown here, when the OTDR trace with
the gainer is averaged with the trace of the
loser, the result is the actual loss. Fluke
Networks makes this even easier. To reduce the
cost and time involved in measuring from both
ends, our OptiFiber Pro or DSX-5000
CableAnalyzer Pro feature a built-in SmartLoop
Assistant that uses a loop at the remote end of a
duplex fiber link, allowing you to test in both
directions from one end. It also features
on-board averaging of the two measurements to
provide an accurate final loss measurement.
6
(No Transcript)