Académique Documents
Professionnel Documents
Culture Documents
Date: 27-03-2008
Version No.: 2.1
1 Introduction
When the fiber section/span was first bought, BOL (Beginning Of Life), the
values of this fiber from end to end, most likely was near perfect or close to
it. Over time the cable, with the fiber inside, may be stressed for numerous
reasons. The stress imposed to the cable could be digging, the ground
moving, rocks pressing on the cable or other changes to the cable’s
environment. These changes are called aging, and change the fibers
possibilities and capacity to transport light with close to no loss to a little
more loss every time a change occurs.
This is why a fiber is said to have an EOL (End Of Life) as well as a BOL. This
is perfectly normal behavior during a fibers lifetime. Reaching EOL however,
does not mean the end of that particular fiber, just a change of purpose to
something less demanding, but this is entirely up to the owner of the fiber to
decide.
Unwanted factors that also play a role in the fibers lifetime is, direct damage
to the cable, in most cases a fiber cut. The fiber needs to be fixed as soon as
possible. To help find the location to dig up the cable in order to splice the
fibers together again, an OTDR (Optical Time-Domain Reflectometer) is used.
The OTDR instrument will produce a report of every fiber measured, and by
looking at those, it can be established what happened to the fiber and what
to do about it.
2 Objective
The objective of this document is to unify and standardize any received OTDR
report from any of NORDUnets fiber providers. Second objective is to make
sure all staff within NORDUnet has the same understanding of OTDR reports.
3 Target group
The target group is NORDUnet staff and fiber providers to the NORDUnet
optical network.
Dark Fiber Evaluation with OTDR Ver. 2.1
Date: Version: All Rights Reserved:
27-03-2008 1.0 NORDUnet A/S
4 Introduction
This document concentrates on defining how and why the OTDR reports
should look like they should in order to get the best possible results. Looking
into the future it is vital that the measurements follow the exact same
structure for easy comparison between old and the most recent OTDR
measurements.
The OTDR measurements physical elaboration is vital for the results, since
the results are used to maintain and further develop the existing network.
5 OTDR techniques
This section explains in short the techniques behind any OTDR measurement;
this is done to point out why it is very important to understand how to
perform an OTDR measurement and what to consider when doing so.
5.1 Backscatter
The OTDR makes measurements on optical fibers by sending a very high
intensity pulse of light into the fiber, minute reflections occur along the
length of the fiber as well as at all discontinuities at such places as splices
and connections and reflect the light back to the transmitter (this is known as
backscattering, see Fig. 1).
Light Particle
Backscatter
Fig. 1 Light reflected in a particle backscatters light.
Increasing the pulse width to 10 microseconds (10 us) will allow more light to
travel down the fiber and consequently more light to be reflected back. In
effect, increasing the pulse width increases the signal to noise ratio and thus
allows for a longer measurement.
When characterizing a fiber link using an OTDR, it is the reflected light that is
used to calculate the attenuation of the link, the characteristics of loss and
the length of the fiber span. The OTDR software displays obvious faults and
terminations (typically patch cables) on a generated graph, known as a trace,
and provides the loss value in dB as a function of distance. The faults, called
events, are listed in a table of events. Fig. 2 illustrates a typical OTDR
acquisition, displaying a trace and its corresponding table of events.
Receiver saturated
Recovery
period
OTDR
receiving pulse
reflection
Why not bury this unusable measurement time in a piece of fiber that is not
part of the fiber under test? That is exactly what the test fiber (or so called
Dead Zone Eliminator, hence DZE) accomplishes.
By placing the DZE in front of the fiber to be tested and shooting the OTDR
through this device, the receiver can be in saturation while the light is still in
the DZE and has not yet traveled into the fiber under test. This lets the
receiver “recover” while the launch pulse is still traveling in the DZE. The
attenuation measurements are then starting at a point before the beginning
of the fiber under test – but still not right at a point inches from the start of
the fiber under test.
Why? Pulse suppressors still need to be connected to the fiber under test and
this interface will cause a reflection that may result in another dead zone.
Loss measurements cannot be made within these reflections.
So how does the Dead Zone Eliminator (DZE) allow the OTDR to measure
through this dead zone event? By allowing the OTDR to place a measurement
cursor in the linear (non-saturated) portion of the trace, prior to the start of
the fiber under test, and within the length of the DZE. This measures through
the dead zone events giving an indication of the loss of this section. When
measuring loss of fiber through these events it is important to realize that the
OTDR is measuring not only the fiber under test, but also a small length of
the DZE and the DZE to fiber connector pair. This method gives a good
indication of the quality of the initial section of the fiber under test, a loss
Dark Fiber Evaluation with OTDR Ver. 2.1
Date: Version: All Rights Reserved:
27-03-2008 1.0 NORDUnet A/S
indication that may not be available without the use of the DZE. For the same
technical explanation the DZE is used at the end of the fiber under test.
The influence of the receiver’s bandwidth resides in the fact that low-
bandwidth transmission smoothes down curves and attenuates sharp
transitions. Although this reduces noise and improves SNR at the far end of
long-range measurements, it increases dead zones. A high receiver
bandwidth, on the other hand, transmits sharp transitions from the
backscattered and reflected signals, but also lets through noise. This shortens
dead zones, but may require more averaging in order to reduce noise level.
A high receiver bandwidth has a limited range mainly due to the random
noise it contains. However, it improves front-end fiber link event
measurement accuracy and spatial resolution, especially when measuring
closely spaced events.
Sampling points have a more complex influence on the end results. Without
going into a detailed description of sampling rates and phases, it is safe to
say that the main influences of sampling points may be described as follows:
Parameters such as the pulse width, the number of sampling points, the
measurement distance, the receiver bandwidth and averaging all affect the
SNR (ratio of back reflected signal-to-noise level). As a whole, these
parameters have a considerable impact on the end results and the influence
of these factors on event detection measurement and spatial resolution is
very difficult to predict as they interact differently in every OTDR. Changing
one parameter will modify others, which is something the user might not be
aware of or might not want to do.
6 Inspecting events
The OTDR techniques have been explained, this section uncovers what is
interesting to analyze from the trace result.
The fiber changes characteristics over time; however the most common
change to the fiber is a direct cut.
What an OTDR measurement uncovers is how well the fiber is spliced back
together again, but this is just in an event of a fiber cut. There are various
different events to look for in every fiber.
Dark Fiber Evaluation with OTDR Ver. 2.1
Date: Version: All Rights Reserved:
27-03-2008 1.0 NORDUnet A/S
If the span of the fiber is more than 20-25 km long, the span will also be
having changes in the core diameter, not much, and this too will have a
visual impact on the loss in the fiber span. The reason for the core change is
because a new roll of fiber is used to extend the first roll, each roll is typically
20-25 km long when arriving from the factory.
When the fiber is laid it can be bent excessively and fixed in this position, this
is known as a macro bend and can in some circumstances be accounted for.
All these small changes to the fiber can be identified when looking at a
correctly performed OTDR measurement. Fig. 5, crudely shows these
occurrences.
Light Rejected
Perfect Dust
scratch
Fig. 6 The perfect connector far left and far right a connector that needs replacement
The perfect connector can have as little as 0.2 dB loss. A dusty connector
needs to be cleaned, which in most cases is enough, but it is important that
all dust is removed since it can burn and stick to the connector causing it to
fail over time.
Visual inspection should be carried out at x400 time’s magnification, which
will reveal even the tiniest scratch the connector might have. A light scratch
on the connector on the ferrule part can be accepted since the power of light
is fairly low, but a scratch in or near the core of the fiber must be rejected at
once.
A scratch at the core could mirror the laser in a small space, accumulating
higher and higher energy which will make the connector explode when the
energy gets too high.
At some point a fiber needs to be spliced with another fiber, this could be
when installing the fiber between point A and B, or it could be when the fiber
needs repair.
All splices are done with a fusion splicer, which align the two fibers cores to
each other and then splices the fibers with an electric arc. Depending on the
fusion splicer the result could be near zero loss or 0.1dB but normally a splice
should be no more than 0.16dB.
As with fiber events and connectors a splice event can also be identified with
an OTDR measurement.
7 Measurement requirement
When taking the technical and surrounding events into consideration this
implies strict requirements when measurements are performed.
NORDUnet requirements are outlined as values for events and total loss pr.
km.; this is a repetition of the part of the Telenor and Global Crossing
requirements. Offering.
A fibre connection of a span shall consist of an uninterrupted glass fibre end-
to-end; if this requirement is impossible to fulfil then every fibre span must
be considered on a case by case basis.
All of the requirements must be met within the 1530-1570nm wavelength
spectrum if not stated elsewhere.
The optical fibre itself (not only the cable) shall have been manufactured
after 1992.
The Dark Fibre shall be specified according to ITU-T G.652 or ITU-T G.655.
For ITU-T G.655 the True Wave Classic shall not be accepted.
The attenuation at 1550 nm should not exceed 23 dB and shall not exceed 35
dB.
The attenuation at 1550 nm should not exceed 23 dB and shall not exceed
26 dB.
Dark Fiber Evaluation with OTDR Ver. 2.1
Date: Version: All Rights Reserved:
27-03-2008 1.0 NORDUnet A/S
The Dark Fibre parameters of the installed cable shall during the whole
contract period comply with the values in 1-7 below. The values include
effects of possible future splicing, repair, aging, etc.
7.1.7 Documentation
If PMDQ exceeds 0.20 ps/ km then the Supplier shall provide measurement
reports or manufacturing data for each span in an electronic readable format
stating the Polarisation Mode Dispersion (PMDQ) value.
7.1.8 Repairs
Each repair loss in each direction shall be less than 0.5 dB at 1550nm.
1. Measuring with a DZE on each end of the fiber under test (FUT).
a. The DZE should be 200 meters long as a minimum and not more
than 2.5 km long matching the pulse width approximately.
b. A standard 2.5 km DZE is acceptable.
See Fig. 7 and 8 for correct physical setup before performing the OTDR
measurement.
Fig. 8 Physical setup for measuring a pair of fibers, not longer than 80 km long in total
A single fiber measurement takes longer time, but allows greater detail and
resolution to the measurement and it allows the fiber to be up to 160 km
long.
The dual fiber measurement is only suitable for short fiber distances up to 40
km pr. fiber, but it decreases measurement time dramatically.
2. And one set done after the actual fiber work (splicing of the fiber).
8 Example files
From this document laboratory OTDR traces can be requested at any time.
9 Document relations
10 Document changes
2008.02.25
Firs official release of this document.
END OF DOCUMENT