FIBEROPTICS BY JIM HAYES
But once the products are purchased
and installed, testing—usually performed
by the installer—verifies the installation
quality. Whenever questions about product performance arise after installation,
the initial judgment is based on that testing. If there is a problem, more often than
not, the dispute focuses on how the components were tested even more than how
they were installed.
Why the controversy?
Why is testing so controversial? Don’t
today’s instruments give us an easily
decipherable digital number we can
read, store and print for our customer?
Yes, they do. But what does that number
mean? How close is it to the actual value
we are trying to measure? What can
change the reading and cause errors?
Would anyone else making the same
measurement—e.g., the manufacturer’s
lab personnel—get the same reading?
The answers to those questions can
be found in the study called “metrology,”
which is the science of measurements.
Don’t stop reading for fear I’m getting
too technical, because I promise I’m not.
Metrology is quite simple and logical. Electrical contractors (ECs) should learn about
it because it is relevant to many of the things
they do beyond fiber optics installation. It
can be found in building construction, electrical installation or any part of business
that involves taking measurements.
For measurements, accuracy is the
topic of most concern. Accuracy refers
to how well the measured value agrees
with the actual value of what is being
measured. For example, when measuring the length of a fiber with an optical
time-domain reflectometer (OTDR) and
10.0 kilometers (km) is measured but the
actual length is 10. 1 km, the accuracy is
99 percent, or as we generally state it,
your error is 1 percent.
“Precision” is another term widely
used in measurements. A measurement
is precise when repeated measurements
produce nearly the same result. In the
OTDR example, a person might look at
the display, move the markers slightly and
measure 10.0 km, try again and get 9. 9 km,
again and get 10. 1 km and again to get 10. 2
km, caused by the uncertainty in placing
the cursors. Or, perhaps the trace is noisy
and several measurements are taken
with the same variations, or the OTDR is
allowed to automatically record the distance as an event in several tries as 10.0,
10. 1 and 9. 9. All of these measurements
have a variation of about 1 percent, which
we call the precision of the measurement.
The variations we get in our mea-
surement indicate the uncertainty of
the measurement or its accuracy. In the
1980s, when I worked with the scientists
at the U.S. National Bureau of Standards
(now the National Institute of Standards
and Technology) to develop a standard
for optical power to calibrate fiber optic
power meters, they always corrected me
when I said “accuracy.” They preferred
to say “measurement uncertainty”
because we generally refer to measure-
ment errors rather than how close we
are, which sounds strange. So I continue
There are two major contributions
to measurement uncertainty: random
errors and systematic errors. Using the
example of measuring length with an
OTDR, random errors would be varia-
tions caused by the operator variations
in setting the marker positions or noise
in the OTDR trace, affecting automatic
measurements. Random errors would
cause random variations in the length.
But the fiber length being measured
is a function of the speed of light in the
fiber, because the OTDR measures time
and converts it to distance. The OTDR
calibration uses the index of fiber refrac-
tion that determines the speed of light
in the fiber. What happens if we use the
wrong index of refraction? Every mea-
surement will be short if the index of
refraction is too high and long if it’s too
low. This mistake produces a systematic
error where all measurements are erro-
neous in the same way.
Next month, I’ll cover more about
fiber optic measurement uncertainty.
Let’s Talk About Precision
Fiber optic testing
WHEN MANUFACTURERS ARE ASKED what is the biggest problem with installing of fiber optic components and systems, they invariably say “testing.” Practically
everything they do depends on testing. During development, engineering and manufacturing in their labs, they test product performance. Testing reveals the most
effective installation processes that they document and teach to their customers.
HAYES is a VDV writer and educator and the president of the Fiber Optic Association. Find
JimHayes.com. I S T
The variations we get
in our measurement
indicate the uncertainty
of the measurement
or its accuracy.