Guide to Fiber Optic Measurement

Reference: 901GFOM/00

Reprinted: September 2001

 2001 Acterna

The information contained in this document is the

property of Acterna. It is only provided for the operation
and maintenance of the instrument. It must not be
duplicated without the prior written permission of
Acterna.

Acterna Saint-Etienne
34 rue Necker
42000 Saint-Etienne
Tel. +33 (0) 4 77 47 89 00
Fax +33 (0) 4 77 47 89 70
Web www.acterna.com
Technical Specifications OFI 2000

11-8 USER MANUAL 720000992/03

Otdr.bk : otdr.pdg Page ii Mercredi, Mars 4, 1998 3:31 PM

Acterna shall not be liable for errors contained herein.

This document must not be photocopied, reproduced, or translated into
another language without the written consent of Wavetek.

Printed in France

Authors J. Laferrière
R. Taws
S. Wolszczak

ii Guide to Fiber Optic Measurements

Otdr.bk : otdr.tdm Page iii Mercredi, Mars 4, 1998 3:31 PM

1 Table of contents

Fiber Principles ................................................................................ 1-1

Types of fibers....................................................................................... 1-1
Multimode fiber ........................................................................... 1-7
Singlemode fiber .......................................................................... 1-8
Fiber standards and recommendations .................................... 1-10
Optical Testing ................................................................................ 2-1
Families of optical fiber tests ............................................................... 2-1
Transmission tests................................................................................. 2-2
Field tests ..................................................................................... 2-3
Different families of optical testers ..................................................... 2-7
Sources, Power meters and Attenuators .................................... 2-7
Mini-OTDR ............................................................................... 2-11
Mainframe or full-featured OTDR ..........................................2-12
Monitoring systems ................................................................... 2-13
Other general test equipment ..................................................2-16
Principles of an OTDR ..................................................................... 3-1
Fiber Phenomena.................................................................................. 3-1
Rayleigh scattering ..................................................................... 3-2
Fresnel reflection ......................................................................... 3-4
OTDR block diagram ........................................................................... 3-5
Laser diodes ................................................................................. 3-6
Pulse generator with laser diode ................................................. 3-6
Photodiode ................................................................................... 3-7
Time base and control unit ......................................................... 3-7
OTDR specifications ............................................................................ 3-8
Dynamic range ............................................................................. 3-8
Dead Zone .................................................................................. 3-11
Resolution .................................................................................. 3-14
Accuracy ..................................................................................... 3-15
Wavelength ................................................................................ 3-16
Using an OTDR ................................................................................. 4-1
Acquisition ............................................................................................. 4-1

Guide to Fiber Optic Measurements iii

Otdr.bk : otdr.tdm Page iv Mercredi, Mars 4, 1998 3:31 PM

Injection level .............................................................................. 4-2

OTDR wavelength ..................................................................... 4-3
Pulse width .................................................................................. 4-4
Range ............................................................................................ 4-6
Averaging ..................................................................................... 4-6
Smoothing .................................................................................... 4-8
Fiber parameters ......................................................................... 4-8
Measurement ..................................................................................... 4-10
Slope or fiber section loss ......................................................... 4-14
Event loss ................................................................................... 4-14
Reflectance and Optical Return Loss ...................................... 4-17
Measurement artifacts and anomalies ............................................... 4-19
Ghosts ......................................................................................... 4-19
Splice "Gain" .............................................................................. 4-21
Getting the most out of your OTDR ................................................ 4-26
Using launch cables ................................................................... 4-26
Verifying continuity to the fiber end ....................................... 4-28
Fault location ............................................................................. 4-29
Effective refractive index ......................................................... 4-30
Glossary ........................................................................................... A-1
Notes ................................................................................................ N-1
Index ................................................................................................... I-1

iv Guide to Fiber Optic Measurements

Otdr.bk : OTDR_C1.doc Page 1 Mercredi, Mars 4, 1998 3:31 PM

Chapter
Fiber Principles
1
1
1.1 Types of fibers
An optical fiber is made of very thin glass rods composed of two parts:
the inner portion of the rod or core and the surrounding layer or cladding.
Light injected into the core of a glass fiber will follow the physical path
of that fiber due to the total internal reflection of the light between the
core and the cladding. A plastic sheathing around the fiber provides the
mechanical protection.
Fibers are classified into different categories based on the way in which
the light travels in them, which is closely related to the diameter of the
core and cladding.

Principle of the transmission (simplified version):

• a ray of light enters into the fiber at a small angle α.
• the capability (maximum acceptable value) of the fiber cable to
receive light on its core is determined by its numerical aperture NA:
2 2 where: α0: maximum angle of acceptance
NA = sin α 0 = n1 – n 2
(i.e limit between reflection and
refraction)
2 2 n1: core refractive index
α 0 = arc sin n 1 – n 2
n2: cladding refractive index
Note : 2 α0 is the full acceptance angle.

Cladding
Full
acceptance n1
angle α0 Core

n2

Guide to Fiber Optic Measurements 1-1

Otdr.bk : OTDR_C1.doc Page 2 Mercredi, Mars 4, 1998 3:31 PM

Fiber Principles

Light propagation
• If α > α0: the ray is fully refracted and not captured by the core.
n2 Refraction :
α0
αi n1 n1 sin αi = n2 sinαr
αr n2

• If α < α0: the ray is reflected and remains in the core

n2 Reflection :
α0 αi αr αi = αr
n1

n2

Velocity
The velocity at which light travels through a medium is determined by the
refractive index of the medium. The refractive index (n) is a unitless number
which represents the ratio of the velocity of light in a vacuum to the velocity
of the light in the medium.

c
n = ----
V

where:n: Refractive Index

c: Speed of light in a vacuum (approximately 3 x 108 m/s)
V: Speed of light in the transmission medium

Typical values of n lie between 1.45 and 1.55.

Light entering with different angles does not follow the same path. Light
entering the center of the fiber core at a very low angle will take a relatively
direct path through the center of the fiber. Light injected at a high angle of
incidence or near the outer edge of the fiber core will take a less direct,
longer path through the fiber and therefore travel more slowly down the
length of the fiber. Each path resulting from a given angle of incidence and
entry point can give rise to a mode. As they travel along the fiber, all the
modes are attenuated.

1-2 Guide to Fiber Optic Measurements

Otdr.bk : OTDR_C1.doc Page 3 Mercredi, Mars 4, 1998 3:31 PM

Types of fibers

Attenuation

The attenuation in a fiber is caused by different factors:

• light absorption. Absorption may be defined as the conversion of light
energy to heat, and is related to the resonances in the fiber material.
There are intrinsic absorptions (due to fiber material and molecular reso-
nance) and extrinsic absorptions (due to impurities such as OH- ions at
around 1240 nm and 1390 nm). In modern fibers, extrinsic factors are
almost negligible.
• Rayleigh scattering. Scattering, primarily Rayleigh scattering, also contrib-
utes to attenuation. Scattering causes the light energy to be dispersed in
all directions, with some of the light escaping the fiber core. A small por-
tion of this light energy is returned down the core and is termed «backs-
cattering».

Note Forward light scattering (Raman Scattering) and

backward scattering (Brillouin scattering) are two additional
scattering phenomena that can be seen in optical materials
under high-power conditions.

Backscattered light
Scattered light
Incident light

Backscattering effect

• bending losses which are caused by light escaping the core due to imper-
fections at the core/clad boundary (microbending), or the angle of inci-
dence of the light energy at the core/cladding boundary exceeding the
Numerical Aperture (internal angle of acceptance) of the fiber due to
bending of the fiber (macrobending).
Singlemode fibers (for example) may be bent to a radius of 10 cm with
no significant losses, however after the minimum bend radius is
exceeded, losses increase exponentially with increasing radius. Mini-
mum bend radius is dependent on fiber design and light wavelength.

Guide to Fiber Optic Measurements 1-3

Otdr.bk : OTDR_C1.doc Page 4 Mercredi, Mars 4, 1998 3:31 PM

Fiber Principles

For a fiber optic span, passive components and connection losses have to be
added to obtain the total signal attenuation.

Input Output
Heterogeneous
Structures
Impurities

Optical
Fiber

Injection Junction
Absorption Diffusion Bending Coupling
Loss Loss
Loss Loss Loss Loss
Macro
or
micro
bending

Loss mechanisms

The attenuation, for a given wavelength, is defined as the ratio between the
input power and the output power of the fiber being measured. It is gener-
ally expressed in decibels (dB).
This attenuation depends on the fiber and on the wavelength. For example,
Rayleigh scattering is inversely proportional to the fourth power of the
wavelength. If we look at the absorption spectrum of a fiber against the
wavelength of the laser, we can notice some characteristics.
The following graph illustrates the relationship between the wavelength of
the injected light and the total fiber attenuation resulting from the contribu-
tion of all the loss mechanisms:

1-4 Guide to Fiber Optic Measurements

Otdr.bk : OTDR_C1.doc Page 5 Mercredi, Mars 4, 1998 3:31 PM

Types of fibers

Attenuation (dB) OH-absorption peak

Scattering

Infrared absorption loss

Wavelength (nm)
850 1300 1550
Attenuation versus wavelength

The main telecommunication transmission wavelengths correspond to the

points on the graph where the attenuation is a minimum. These wave-
lengths are known as the “telecom” windows and are typically as follows:
• first window from 820 to 880 nm
• second window from 1285 to 1330 nm
• third window from 1525 to 1575 nm
Another factor affecting the signal during transmission is dispersion. This
reduces the effective bandwidth available for transmission.
Two main types of dispersion are defined.
• Modal dispersion: when a very short pulse is injected into the fiber
within the numerical aperture, all of the energy does not reach the end of
the fiber at the same time. Different modes of oscillation carry energy
down the fiber down different paths and thus travel further. As an exam-
ple, a 50 µm core multimode fiber may have several hundred modes.
This pulse spreading by virtue of different light path lengths is called
modal dispersion or more simply modal dispersion.
• Chromatic dispersion: the pulse sent down the fiber is usually com-
posed of a small spectrum of wavelengths. This means they go through
the fiber at different speeds. Because propagation speed is dependent on
the refractive index and therefore the wavelength, this effect is known as
chromatic dispersion. It explains why it is important to use test equip-

Guide to Fiber Optic Measurements 1-5

Otdr.bk : OTDR_C1.doc Page 6 Mercredi, Mars 4, 1998 3:31 PM

Fiber Principles

ment which are at the same small spectrum of wavelengths as the wave-
length of operation.
Chromatic dispersion is expressed in picosecond per nanometer per
kilometer: ps / (nm x km). This coefficient, at a given wavelength,
represents the difference after one kilometer between the propagation
time of two wavelengths which differ by a given number of nanometers.
Chromatic dispersion is the dominant dispersion mechanism in
singlemode fibers. In singlemode fibers there is a minimum or zero
(chromatic) dispersion wavelength determined by fiber design and
manufacture, and this wavelength is generally chosen to be near the
operating wavelength of the system. Historically (in standard
singlemode fiber), this was near 1310 nm, but for newer systems, so-
called dispersion shifted fibers are used with the zero dispersion
wavelength moved closer to 1550 nm to take advantage of the lower
fiber attenuation at that wavelength. In some systems, for example,
Dense WDM (Wavelength Division Multiplexing) applications, a slight
positive chromatic dispersion is desirable and fiber designs are available
to accommodate this.
This fiber is ideal for submarine cables because of the increased repeater
spacing and reduced cost. The maximum repeater spacing for high bit
rate transmission is found by measuring the ratio between the maximum
chromatic dispersion tolerated by the system (in ps/nm) and the fiber in
ps / (nm x km). The attenuation of the fiber must also be taken into
account.

Bandwidth limitation

1-6 Guide to Fiber Optic Measurements

Otdr.bk : OTDR_C1.doc Page 7 Mercredi, Mars 4, 1998 3:31 PM

Types of fibers

The two major classes of fibers are those that exhibit modal dispersion (mul-
timode) and those that do not (singlemode) :
• Multimode fibers have much larger core (> 50 µm) than singlemode
fibers permitting many modes of light to travel through the core.
• The core of a single mode fiber is generally 10 µm or less and will allow
only one mode of light (at 1310 or 1550 nm) to propagate, greatly reduc-
ing total dispersion.

1.1.1 Multimode fiber

Multimode fiber, due to its large core, enables different paths (multi-modes)
to transmit the light along the link. This is the reason why this fiber is quite
sensitive to the modal dispersion.
The primary advantages of multimode fiber are it’s ease of coupling to light
sources and to other fibers, reducing the cost of light sources (transmitters),
connectorization and splicing. However, it’s relative higher attenuation and/
or low bandwidth limit it to short distance and low speed applications.

Core
Diameter: from 50 µm to 100 µm

Cladding refractive index < core refractive index

Coating Cladding
Diameter: 250 µm Diameter: 125 µm and 140 µm

Multimode fiber

A. Step index multimode fibers

Step-index fiber guides light rays through total reflection on the boundary
between core and cladding. The refractive index is uniform in the core.
Step-index fibers have minimum core diameter of 52.5 µm and 62.5 µm,

Guide to Fiber Optic Measurements 1-7

Otdr.bk : OTDR_C1.doc Page 8 Mercredi, Mars 4, 1998 3:31 PM

Fiber Principles

cladding diameter of 100/140 µm and numerical aperture between 0.2 and

0.5.
Due to modal dispersion, the drawback to this design is its very low band-
width, expressed as bandwidth-length product in MHz x km. This fiber’s
bandwidth of approximately 20 MHz x km indicates that it is suitable for
carrying a 20 MHz signal only a distance of 1 km, or a 10 MHz signal a dis-
tance of 2 km, or a 40 MHz signal a distance of 0.5 km, etc.
Step-index fibers have been implemented in plastic; their application field
is mostly in short distance links which can accommodate high attenuations.

B. Graded-index multimode fibers

Graded-index (GI) fibers are obtained by giving to the core a non-uniform
refractive index, decreasing gradually from the central axis to the cladding.
This index variation of the core forces the rays to progress in the fiber in a
sinusoidal manner.
The highest order modes will have a longer travel, but outside of the central
axis, in areas of low index, their speeds will increase and the speed differ-
ence between the highest and lower order modes will be smaller than for
step-index fibers.
Typical attenuations are : 3 dB/km at 850 nm
1 dB/km at 1300 nm.
The numerical aperture of graded-index fibers is typically about 0.2.
The bandwidth-length product for Graded index fibers is approximately:
160 MHz x km at 850 nm
500 MHz x km at 1300 nm.
Typical values of the group index :
1.49 for 62.5 µm at 850 nm
1.475 for 50 µm at 850 nm
1.465 at 1300 nm.

1.1.2 Singlemode fiber

The advantage of singlemode fiber is its higher performance with respect to
bandwidth and attenuation. The reduced core diameter limits the light to
propagation of only one mode, eliminating modal dispersion completely.
With proper components, a singlemode fiber system can carry signals in
excess of 10 GHz for over 100 km. The system carrying capacity may be fur-

1-8 Guide to Fiber Optic Measurements

Otdr.bk : OTDR_C1.doc Page 9 Mercredi, Mars 4, 1998 3:31 PM

Types of fibers

ther increased by injecting multiple signals of slightly differing wavelengths

(Wavelength Division Multiplexing) into one fiber.
The small core size generally requires more expensive light sources and
alignment systems to achieve efficient coupling and splicing and connector-
ization is also somewhat complicated. Nonetheless, for high performance
system or systems over a few kilometers, singlemode fibers remain the best
solution.
The typical dimensions of single mode fibers range from 5 to 12 µm for the
core and 125 µm for the cladding. A typical core-cladding angle is 8.5
degrees.
The group index is typically 1.465 for the singlemode fiber.

Core
Diameter: 5 to 10 µm

Cladding refractive index < core refractive index

Cladding
Diameter: 125 µm

Coating
Diameter: 250 µm

Singlemode fiber

The small core diameter decreases the number of propagation modes. In a

single mode fiber, only one ray propagates down the core at a time.

Mode field diameter

The mode field diameter (MFD) of a single mode fiber can be expressed as
the section of the fiber where the majority of the light energy passes.
The MFD is larger than the physical core diameter i.e. an 8µm physical core
could yield a 9.5 µm MFD. This also shows that some of the light energy
also transits through the cladding.

Guide to Fiber Optic Measurements 1-9

Otdr.bk : OTDR_C1.doc Page 10 Mercredi, Mars 4, 1998 3:31 PM

Fiber Principles

1.1.3 Fiber standards and recommendations

There are many international and national standards governing optical cable
characteristics of which only some are cited below.

International standards
For just the international standards, there are 2 main groups :
• The IEC has several standards of which we find:
• IEC 60793-1 and -2 Optical fibers (containing several sections)
• IEC 60794-1, -2, and -3 Optical fiber cables
• The ITU-T (formerly the CCITT) has more standards such as:
• G650 Definition and test methods for the relevant parameters of sin-
gle-mode fibers,
• G651 Characteristics of 50/125 µm multimode graded index optical
fiber
• G652 Characteristics of singlemode optical fiber cable
• G653 Characteristics of singlemode dispersion shifted optical fiber
cable
• G654 Characteristics of 1550 nm loss minimized singlemode optical
fiber cable

National standards
• The CEN is preparing the following recommendations for Europe: EN
186000 (Optical fibre connectors), EN 187000 (Optical fibres), and the
EN 188000 (Optical fibre cables);
• The ETSI provides additional recommendations for Europe;
• The EIA/TIA provides additional recommendations for the USA (FOTP).
Many other standards organizations exist in other countries.

Test equipment standards

• IEC 61350: Power meter calibration
• IEC 61746: OTDR calibration

1-10 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c2.fm Page 1 Mercredi, Mars 4, 1998 3:31 PM

Chapter
Optical Testing
2
2
2.1 Families of optical fiber tests
When analyzing a fiber optic cable over its product life, a series of mea-
surements have to be performed:
• mechanical tests,
• geometrical tests,
• optical tests
• transmission tests.

The three first measurements are only performed once, as there is

minor variation of these parameters during the fiber's life.
Several measurements are made on optical fibers or cables in order to
characterize them before their use for transmission. Many of these mea-
surements are described in the FOTP (Fiber Optic Test Procedure)
propositions of the EIA (Electronic Industries Association) and are
defined by the ITU-T G650 recommendations or the EN 188 000 docu-
ment.

Different kinds of test

Mechanical Geometrical Optical Transmission
Traction Concentricity Index Profile Bandwidth
Torsion Cylindricity Numerical aper- Optical Power
ture
Bending Core diameter Spot size Optical Loss
Temperature Cladding Reflectometry
diameter

Guide to Fiber Optic Measurements 2-1

Otdr.bk : otdr_c2.fm Page 2 Mercredi, Mars 4, 1998 3:31 PM

Optical Testing

2.2 Transmission tests

The main measurements implemented on optical fibers and optical fiber
systems in order to qualify their use for information transmission purposes
are:
• End-to-End Optical Link Loss
• Rate of attenuation per unit length
• Attenuation contribution to splices, connectors, couplers (events)
• Length of fiber or distance to an event
• Linearity of fiber loss per unit length (Attenuation discontinuities)
• Reflectance or Optical Return Loss
Other measurements such as bandwidth or polarization mode dispersion
may also be done, but they are less important, except for some specific
applications.
Whereas some measurements may require access to both ends of the fiber,
others require only one end. Measurement techniques which require access
to one end are particularly interesting for field applications since it will
reduce the time spent travelling from one end of the fiber cable system to
the other.
If we focus on field testing on optical cables, we can see that there are three
main tasks - Installation, Maintenance and Restoration - where testing is
required.

2-2 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c2.fm Page 3 Mercredi, Mars 4, 1998 3:31 PM

Transmission tests

2.2.1 Field tests

Below is a non-exhaustive list of the various tests that can be performed dur-
ing each task (Installation, Maintenance, Restoration). The exact nature of a
testing program will depend on the system design, system criticality and
contractual relationship between the cable and components suppliers, sys-
tem owner, system installer and system user.
Installation testing is performed to ensure that fiber cables received from
the manufacturer are conform to specifications (length, attenuation, etc.)
and have not been damaged in transit, and that they are not damaged during
cable placement. Tests also determine the quality of cables splices and cable
terminations (attenuation, location, reflectance) and that the completed
cable subsystem is suitable for the intended transmission system (end-to-
end loss, system optical return loss) and provide complete documentation of
the cable link for maintenance purposes.
Maintenance testing involves periodic evaluation of the cable system to
ensure that no degradation of the cable, splices or connections has occurred
(cable attenuation, attenuation and reflection of splices and terminations).
In some systems, maintenance tests may be performed every few months
and compared to historical test results to provide early warning of degrada-
tion. In very high capacity or critical systems, automated testing devices
may be employed to test the integrity of the system every few minutes to
give immediate warning of degradation or an outage.
During cable restoration, testing is first performed to identify the cause of
the outage (transmitter, receiver, cable, connector) and to locate the fault in
the cable if the outage is caused by the cable. Testing is then used to assess
the quality of the repaired system (permanent splices), similar to the testing
performed at the conclusion of cable installation.

Guide to Fiber Optic Measurements 2-3

Otdr.bk : otdr_c2.fm Page 4 Mercredi, Mars 4, 1998 3:31 PM

Optical Testing

Pre-installation test on a drum

When installing a fiber network, network topology and equipment specifi-

cations have to be taken into consideration. One of the major parameters to
measure is optical loss budget or end-to-end optical link loss. When calculating
the budget of a fiber link, the following must be considered: the source, the
detector and the optical transmission line. The transmission link includes
the source-to-fiber coupling loss, the fiber attenuation loss, and the loss of
all components along the line (connectors, splices, passive components,
etc.).

Optical loss budget

An optical loss budget lies within maximum and minimum values:
• the maximum value is defined as the ratio of the minimum optical power
launched by the transmitter to the minimum which may be received by
the receiver whist still maintaining communication;
• the minimum value is defined as the ratio of the maximum optical power
launched by the transmitter to the maximum which may be received by
the receiver whist still maintaining communication.

2-4 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c2.fm Page 5 Mercredi, Mars 4, 1998 3:31 PM

Transmission tests

A typical example of a multimode system is described below.

• Transmitter output power (typical) for multimode fiber
(GI) = -12 dBm ±2 dB
• Optical Receiver sensitivity ≤ -27 dBm
• Optical Receiver Dynamic Range ≥ 18 dB
The transmitter specification provides the maximum (-10 dBm) and mini-
mum (-14 dBm) power levels that will occur.
The receiver sensitivity gives us the minimum power level that will be
detected.
The receiver dynamic range provides the maximum power level that can be
detected (-27 dBm + 18 dBm = -9 dBm).
In this example, the maximum optical loss budget is 13 dB :
• Minimum optical power of the transmitter (-14 dBm)
• Minimum receiver sensitivity (-27 dBm)

Example of a typical budget loss

Tx Optical Rx
network

Launched Received
optical optical
power (L) power (R)

L max (dB)
Minimum Optical loss
R max (dB) budget (Bmin)

L min (dB)

Maximum Optical loss

budget (Bmax)

R min (dB)
Optical Budget
B max = Lmin - Rmin
B min = Lmax - Rmax

Optical loss budget

Guide to Fiber Optic Measurements 2-5

Otdr.bk : otdr_c2.fm Page 6 Mercredi, Mars 4, 1998 3:31 PM

Optical Testing

Optical loss budgets should take into account the cable and equipment mar-
gins, which covers allowances for the effect of time and environmental fac-
tors (launched power, receiver sensitivity, connector or splice degrada-
tion...). In order to calculate this budget, typical values of attenuations of the
different fiber components are given, for example:
• 0.2 dB/km for singlemode fiber loss at 1550 nm;
• 0.35 dB/km for singlemode fiber loss at 1310 nm;
• 1 dB/km for multimode fiber loss at 1300 nm;
• 3 dB/km for multimode fiber loss at 850 nm;
• 0.05 dB for a fusion splice
• 0.1 dB for a mechanical splice;
• 0.2 - 0.5 dB for a connector pair;
• 3.5 dB for a 1 to 2 splitter (3 dB splitting loss plus 0.5 dB excess loss).
Once this analysis is performed, the cable installation can be made.

Example of a typical budget loss

NETWORK SHORT HAUL MEDIUM HAUL LONG HAUL

Distance (km) 30 80 200

Fiber loss (dB/km) at 1550 nm 0.25 0.22 0.19

Total Fiber loss (dB/km) 7.5 17.6 38

N° of splices 15 40 25

Average splice loss 0.1 0.1 0.05

Total splice loss 1.5 4 1.25

N° of connectors 2 2 2

Average connector loss 0.5 0.5 0.5

Total connector loss 1 1 1

TOTAL LOSS 10 22.6 40.25

2-6 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c2.fm Page 7 Mercredi, Mars 4, 1998 3:31 PM

Different families of optical testers

2.3 Different families of optical testers

2.3.1 Sources, Power meters and Attenuators

The most accurate way to measure overall attenuation in a fiber is to inject a
known level of light in one end and measure the level when it comes out the
other end. Light sources and power meters are the main instruments recom-
mended by the ITU-T (G651) and the IEC 61350, to measure insertion
loss.
This method required access to both ends of the fiber which is not always
possible.

Light source, power meter and talk set

Guide to Fiber Optic Measurements 2-7

Otdr.bk : otdr_c2.fm Page 8 Mercredi, Mars 4, 1998 3:31 PM

Optical Testing

Light sources
A light source is a device used as a continuous and stable source (CW) for
attenuation measurements.
It includes a source - either an LED or a laser - that is stabilized through
some type of Automatic gain Control:
• LED’s are mainly used for multimode fibers. Lasers are used for single-
mode applications.
• The light output of either an LED or laser source may also have the
option to be modulated (or "chopped") at a given frequency. The power
meter can be set up to detect this frequency. This improves ambient
light rejection. A 2 kHz modulated light source can be used with certain
types of detectors to "tone" the fiber for fiber identification or confirma-
tion of continuity.

Power meter
The power meter is the standard tester in a typical fiber optic craftsman’s
toolkit. It is an invaluable tool during installation and restoration.
The power meter’s main function is to display the incident power on the
photodiode. Features found on more sophisticated power meters may
include temperature stabilization, ability to calibrate to different wave-
lengths, ability to display power relative to "reference" input, ability to intro-
duce attenuation, or high power option.

The requirements for a power meter vary depending on the application.

Power meters must have enough power to measure the output of the trans-
mitter being used (to verify operation) but be sensitive enough to measure
the received power at the far (receive) end of the link. Long haul telephony
systems and cable TV systems use transmitters with outputs as high as
+16 dBm and amplifiers with outputs as high as +24 dBm. Receive powers
can be as low as -36 dBm in systems that use an optical pre-amplifier. In
local area networks, transmit powers are much lower, as are received power.
the difference between the maximum input and the minimum sensitivity of
the power meter is termed the Dynamic Range.
While the dynamic range for a given meter has some limits, the useful
power ranges can be extended beyond that by the of well characterized
attenuators in front of the power meter input; this does limit the low end

2-8 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c2.fm Page 9 Mercredi, Mars 4, 1998 3:31 PM

Different families of optical testers

sensitivity. this high power mode can be an internal or external attenuator :

if internal, it may be fixed or switched.
Typical Dynamic Ranges requirements for power meters are:
• +13 dBm to -70 dB for telephony applications1,
• +24 dB to -50 dB for CATV applications1,
• -20 dB to -60 dB for LAN applications.

Insertion loss and cut back measurements

• The cut back technique is the most accurate measurement, but is also
destructive, and cannot be applied in the field. This is the reason why it
is not used during installation and maintenance. Testing with the cut-
back method requires first measuring attenuation of the length of fiber
under test, then cutting back a part of the length from the source end,
and measuring attenuation of this part as a reference, and then substrac-
ting the two values: the result gives the attenuation of the cut fiber.
• The insertion loss technique is a non destructive method to measure the
attenuation across a fiber, a passive component or an optical link. With
the substitution method, the output from a source and a reference fiber
is measured directly, then a measurement is realized with the fiber to be
measured added to the system. The difference between the two results
gives the attenuation of the fiber.
The purpose of the first or "reference" measurement is to cancel out as
far as possible the losses caused by the various patch cables.

1. Most power meters meet this requirements through two modes of operation, a standard mode (-3
to -70 dBm) and a "high power" mode (+23 to - 50 dBm).

Guide to Fiber Optic Measurements 2-9

Otdr.bk : otdr_c2.fm Page 10 Mercredi, Mars 4, 1998 3:31 PM

Optical Testing

Calibrated
light source Power meter

Measurement P1
Reference pigtail

Calibrated
light source Power meter

Fiber under test

Reference
fiber Measurement P2

Total attenuation of the fiber :

AdB = P1dBm - P2dBm

Insertion loss method (2 steps) to measure the attenuation across a fiber

Significant variations may occur in attenuation measurements if precautions

are not taken with the injection conditions.
Transmitted and received optical power are only measured with an optical
power meter. For transmitted power, the power meter is connected directly
to the optical transmitter’s output.
In the case of received power, the optical transmitter is connected to the
fiber system and then the power level is read with the power meter from the
fiber cable at the point where the optical receiver should be.
Power meter / light source combinations (also defined as loss test sets) mea-
sure cable continuity and cable attenuation.
Link losses are sometimes measured in each direction and averaged to
improve confidence in the measurements.

2-10 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c2.fm Page 11 Mercredi, Mars 4, 1998 3:31 PM

Different families of optical testers

2.3.2 Mini-OTDR
Using the same basic technology as the OTDR (see page 2-12), a new class
of instruments became available in the beginning of the 90’s. Known as
"mini-OTDRs", these fiber test instruments are typically battery-powered,
lightweight, and small enough to be carried in one hand.
The simplest and earliest designs were capable of fault location as a mini-
mum and some rudimentary analysis (attenuation, rate of attenuation, dis-
tance and reflectance) of fiber systems. Modern designs mimic the capabili-
ties of mainframe OTDRs including sophisticated analysis (automatic event
detection, table of events, optical return loss, trace overlay) of fiber links,
data storage capabilities, additional functionality (light source, power meter,
talk set, visual fault locator) and even the modularity formerly found only in
mainframe OTDRs.
A mini-OTDR has become the popular choice for pre-installation and resto-
ration tests where ease-of-use and mobility are important.

Mini-OTDR

Guide to Fiber Optic Measurements 2-11

Otdr.bk : otdr_c2.fm Page 12 Mercredi, Mars 4, 1998 3:31 PM

Optical Testing

2.3.3 Mainframe or full-featured OTDR

OTDRs are the main test equipment used to analyze fiber optics.
Most mainframe OTDRs are modular in design and contain a mainframe
and different plug-in modules which can be implemented to suit the appli-
cation.
The OTDR mainframe contains the controller, display, operator controls,
and optional equipment (such as printer/plotter, external interfaces,
modem, disk drive, etc.). The optical module consists of the laser source
and optical detector and can be changed to allow testing at various wave-
length and fiber type combinations.
Mainframe OTDRs are being rapidly replaced by mini-OTDRs but remain
the choice for laboratory and benchop applications where data acquisition
functions are desired.

Mainframe OTDR

2-12 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c2.fm Page 13 Mercredi, Mars 4, 1998 3:31 PM

Different families of optical testers

2.3.4 Monitoring systems

Test equipment can be integrated into an automated system and connected
to a Network Manager.
Remote systems usually consist of an access point switch, several remote
test units that sit at various central offices, and a centrally located controller.
With traditional field test equipment, it can typically take about six hours
from the failure until the repair is made. The centralized control of a remote
system allows carriers to manage their networks with fewer people. They
also can avoid sending crafts people into the field unless there is an actual
need for service.
In a case of a failure, the system can report the exact location of the prob-
lem, so crafts people and technicians can quickly and easily find the trouble
spot in the field.

CC CC
CC CC
CC
CC
I nt re f ace

L
e d
A
la rm 1 A
la rm 2 M
i cro A
la rm s
K ey
b oard
draw er

Fi ber
Mux

O TDR
Ac u
q it i on
uni t

CC
Co mpu t er

Network CC
CC
CC
CC CC
Operation CC
CC CC
Center CC

CC CC
CC CC
CC
CC
I nt re f ace

A
la rm 1 A
la rm 2 M
i cro A
la rm s
K ey
b oard
draw er

Fi ber
Mux

O
TDR
Unit

O TDR

Ac u
q it i on
uni t

PSTN CC
Co mpu t er

CC
CC CC CC
CC CC CC CC
CC
CC CC
CC CC CC
CC
Interface
L
e d
A la rm 1 A la rm 2 M ic ro A la rm s

Keyb oard
drawer

Fiber
Mux

OT
DRU
n
i t

OTDR
Acqu ition
unit

RFTS Com puter

CC CC
CC CC
CC CC
CC CC

CC
I nt re f ace

L
e d
A
la rm 1 A
la rm 2 M
i cro A
la rm s
K ey
b oard
draw er

Fi ber
Mux

CC CC
O
TDR
Unit

O TDR
Ac u
q it i on
uni t

CC
Co mpu t er

CC
CC
CC CC
CC
CC
CC : Cable Center CC
CC
CC CC CC
CC CC

Remote fiber test system

Guide to Fiber Optic Measurements 2-13

Otdr.bk : otdr_c2.fm Page 14 Mercredi, Mars 4, 1998 3:31 PM

Optical Testing

Most network operators initially will use remote systems to look for and sec-
tionalize catastrophic failure of a link. In this case, the monitoring system is
connected to only one or two fibers in a multifiber link, assuming that in the
event of a catastrophic break all the strands will be cut.

Optical
Switch
Output Fiber not in use
Dark
for fiber
transmission

Fiber under test

NTE cable under test NTE

Cable under test

Out-of-service

Remote monitoring can also be accomplished simultaneously with live traf-

fic being transmitted through the use of Wavelength Division Multiplexing
(WDM) and test equipment operating at wavelengths differing from those
of the transmission system.

2-14 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c2.fm Page 15 Mercredi, Mars 4, 1998 3:31 PM

Different families of optical testers

Optical
Switch
Output

WDM or
lambda test Blocking Filter
Wavelength
Division
Multiplexer
Fiber under test
lambda test

WDM WDM
NTE cable under test NTE

Cable under test

In-service

Guide to Fiber Optic Measurements 2-15

Otdr.bk : otdr_c2.fm Page 16 Mercredi, Mars 4, 1998 3:31 PM

Optical Testing

2.3.5 Other general test equipment

Talk sets
Talk sets transmit voice over installed fiber cable, allowing technicians splic-
ing or testing the fiber to communicate, even when they are in the field.
Both singlemode and multimode talk sets exists.
They can be used to replace mobile or land-based telecommunications
methods which may not be cost-effective or which may not operate at the
distances common to fiber optic links.

OTS talk set

Visual Fault Locators

Visual Fault Locators are red light lasers which visually locate faults, up to
around 5 kilometers.
By sending visual light, the operator can easily see breaks and important
bends in the fiber, as the light escapes out. This function makes them use-
ful for continuity testing of patch cords, jumpers, or short sections of fiber.

2-16 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c2.fm Page 17 Mercredi, Mars 4, 1998 3:31 PM

Different families of optical testers

They can also be used in conjunction with:

• splicing machines to identify fibers to be jointed.
• OTDR to analyze failures which occur within the dead zone.
The most popular fault finders are made with a HeNe source.
Visual Fault Locators can use 635 nm, 650 nm or 670 nm lasers or LEDs,
according to the application:
• 670 nm VFL provides long distance fault location and correct light inten-
sity
• 635 nm VFL provides excellent visibility by shorter fault location.

Fiber Identifiers
Fiber Identifiers are test sets which can detect a modulated signal on a fiber
(usually 2 kHz "tone").

Clip-on testers
These devices are used in conjunction with a suitable light source to enable
power measurements without disconnecting or damaging the fiber. The
clip-on tester is performing measurement by putting a controlled bend in
the fiber and measuring the level of light which escapes out of the fiber.
The measurement can be performed non intrusively (low bend) or intru-
sively (tight bend).

Guide to Fiber Optic Measurements 2-17

Otdr.bk : otdr_c2.fm Page 18 Mercredi, Mars 4, 1998 3:31 PM

Optical Testing

2-18 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c3.doc Page 1 Mercredi, Mars 4, 1998 3:31 PM

Chapter
Principles of an OTDR
3
3
n OTDR (Optical Time Domain Reflectometer) is a fiber optic

A tester characterizing fibers and optical networks. The aim of this

instrument is to detect, locate and measure events at any loca-
tion in the fiber link.
One of the main benefits of the OTDR is that it can fully test a fiber
from only one end, as it operates as a one dimensional radar system. The
OTDR is similar to an accurate radar as its resolution can be between
6 cm and 40 meters.
The OTDR technique produces geographic information with regard to
localized loss and reflective events thereby providing a pictorial and
permanent record which may be used as performance baseline.

3.1 Fiber Phenomena

The OTDR’s ability to characterize a fiber is based on detecting small
signals returned back to the OTDR in response to injection of a large
signal, much like a "radar". In this regard, the OTDR depends on two
types of optical phenomena: Rayleigh Backscattering and Fresnel
Reflections.
The major difference between these two phenomena is as follows:
• Rayleigh scattering is intrinsic to the fiber material itself and is
present along the entire length of the fiber. If Rayleigh scattering is
uniform along the length of the fiber, then discontinuities in the
Rayleigh backscatter can be used to identify anomalies in transmis-
sion along the fiber length.
• On the other hand, Fresnel reflections are "point" events and occur

Guide to Fiber Optic Measurements 3-1

Otdr.bk : otdr_c3.doc Page 2 Mercredi, Mars 4, 1998 3:31 PM

Principles of an OTDR

only where the fiber comes in contact with air or another media such as
at a mechanical connection/splice or joint.

3.1.1 Rayleigh scattering

When a pulse of light is sent down a fiber, some of the photons of light are
scattered in random directions from microscopic particles. This effect,
referred to as Rayleigh scattering, provides amplitude and temporal informa-
tion along the length of cable.
Some of the light is scattered back in the opposite direction of the pulse and
is called the backscattered signal.

Backscattered light
1/1000 of scattered
light

Scattered light
5%/km at 1550 nm

Incident light

The scattering loss is the main mechanism for fibers operating in the three
telecom windows (850 / 1310 / 1550 nm). Typically, a singlemode fiber trans-
mitting light at 1550 nm with a scattering coefficient (αs) of 0.20 dB/km, will
lose 5 % of the transmitted power over a 1 km section of fiber.
The backscattering factor (S) describes the ratio between backscattered
power and the scattered power. S is typically proportional to the square of
the numerical aperture.
Depending on the fiber scattering coefficient (αs) and the fiber backscatte-
ring factor (S), the backscatter coefficient (K) is the ratio of the backscatte-
red power to the energy launched into the fiber.
The logarithmic value of the backscatter coefficient, normalized to a 1 ns
pulse duration, is given by:
Kns (dB) = 10 log K(s-1) - 90 dB
When Kns = - 80 dB, this means that for a 1 ns pulse duration, the backscat-
ter power is - 80 dB below the incident pulse peak power.

3-2 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c3.doc Page 3 Mercredi, Mars 4, 1998 3:31 PM

Fiber Phenomena

Note that -80 dB at 1 ns is equivalent to -50 dB at 1 µs, i.e. :

Kµs (dB) = Kns(dB) + 30 dB
The Rayleigh scattering effect is like shining a flashlight in a fog at night:
the light beam gets diffused -- or scattered -- by the particles of moisture. A
thick fog will scatter more of the light because there are more particles to
obstruct it.
The Backscattering depends on the launched power Po (Watt), the pul-
sewidth used ∆t (seconds), the backscattering coefficient K(s-1), the distance
d (meters) and the fiber attenuation (α) in dB/km:
Backscattering = Po . ∆t . K . 10 -α.d/5

OTDR parameters

A higher density of dopants in a fiber will also create more scattering and
thus higher levels of attenuation per kilometer. An OTDR can measure the
levels of backscattering very accurately, and uses it to measure small varia-
tions in the characteristics of fiber at any point along its length.
While Rayleigh scattering is quite uniform down the length of any given
fiber, the magnitude of Rayleigh scattering varies significantly at different
wavelength as shown in the following diagram and with different manufac-
turer’s fiber.

Attenuation (dB) OH-absorption peak

Scattering

Infrared absorption loss

Wavelength (nm)
850 1300 1550

Attenuation versus wavelength

Guide to Fiber Optic Measurements 3-3

Otdr.bk : otdr_c3.doc Page 4 Mercredi, Mars 4, 1998 3:31 PM

Principles of an OTDR

3.1.2 Fresnel reflection

Fresnel reflection is due to the light reflecting off a boundary of two optical
transmissive materials, each having different index of refraction. This
boundary can occur either at a joint (connector or mechanical splice), at an
non-terminated fiber end, or at a break.
The magnitude of the Fresnel reflection is dependent upon the incident
power and the relative difference between the two indices of refraction.
The amount of light reflected depends upon the boundary surface smooth-
ness and the index difference.

Fiber Pi Reflection is:

Pr (n1 - n2)2
n1 n2 R= =
Pi (n1 + n2)2
Pr
Pr : reflected power
From fiber to air R= 4% (-14 dB) Pi : injected power
n1, n2 : index of refraction

Reflected light from a boundary between a fiber and air has a theoretical
value of -14 dB. This value can be over 4000 times more powerful than the
level of the backscatter. This means that the OTDR detector must be able
to process signals which can vary in power enormously. Connectors using gel
can reduce the Fresnel reflection. The gel acts as an index matching mate-
rial minimizing the glass/air index difference.

3-4 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c3.doc Page 5 Mercredi, Mars 4, 1998 3:31 PM

OTDR block diagram

3.2 OTDR block diagram

Laser diode

Pulse Coupler
Generator Fiber

Photodiode

Time Base Amplifier

Control
Unit
Sampling
& ADC

Averaging
Processing

Display
unit

OTDR block diagram

The OTDR injects light energy into the fiber through a laser diode and
pulse generator. The returning light energy is separated from the injected
signal using a coupler and fed to the photodiode. The optical signal is con-
verted to an electrical value, amplified, sampled and then displayed on a
screen.

Guide to Fiber Optic Measurements 3-5

Otdr.bk : otdr_c3.doc Page 6 Mercredi, Mars 4, 1998 3:31 PM

Principles of an OTDR

3.2.1 Laser diodes

Laser diodes are selected according to the wavelength of the test.
The current wavelengths for OTDR are 850 nm, 1300 nm for multimode,
and 1310 nm, 1550 nm for singlemode.
1625 nm laser diodes are sometimes also used, particularly in remote moni-
toring systems which are carrying live traffic. The purpose of using 1625 nm
is to avoid interference with traffic at 1310 and 1550 nm.

3.2.2 Pulse generator with laser diode

A pulse generator controls a laser diode which sends powerful light pulses
(from 10 mW to 1 Watt) into the fiber. These pulses can have a width in the
order of 2 ns up to 20 µs and a recurrence of some kHz.
The duration of the pulse (pulse width) can be selected by the operator for
different measuring conditions. The repetition rate of the pulses is limited
to the rate at which the pulse return is completed, before another pulse is
launched. The light goes through the coupler/splitter and into the fiber
under test.
The OTDR measures the time difference between the outgoing pulse and
the incoming backscattered pulses hence the word "time domain". The
power level of the backscattered signal and the reflected signal is sampled
over time. Each measured sample is called an "acquisition point" and these
points can be plotted on an amplitude scale with respect to time relative to
timing of the launch pulse. It then converts this time domain information
into distance based on the user entered index of refraction of the fiber. The
index of refraction entered by the user is inversely proportional to the veloc-
ity of propagation of light in the fiber. The OTDR uses this data to convert
time to distance on the OTDR display and divide this value by two to take
the round trip (or two way) into account. If the user entered refractive index
is incorrect or inaccurate, the resulting distances displayed by the OTDR
can be in error.

3-6 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c3.doc Page 7 Mercredi, Mars 4, 1998 3:31 PM

OTDR block diagram

P (Injection)

PP(Reflection)
(Reflection)

Propagation or group delay in fiber :

V (Gp delay) = c/n ~ 3.108 / 1.5 = 2.108 m/s

c = speed of light in vacuum (the real value of c is 2.99792458 m/s)

n = refractive index.

OTDR time to distance conversion (round trip):

L (distance) = V(Gp delay). t/2 = c.t. / 2.n ~ 108 x time (seconds)
E.g. for a 10 ns pulsewidth: L = 108 x 10 ns = 1 m

3.2.3 Photodiode
OTDR photodiodes are especially designed to measure the extremely low
levels of backscattered light, at 0.0001% of what is sent by the laser diode.
As previously stated, the diodes must also be able to detect the relatively
high power of reflected pulses of light. This causes some problems when
analyzing the results of an OTDR (see "Dead Zone" on page 3-11).
The bandwidth, sensitivity, linearity and dynamic range of the photodiode
and its amplification circuitry are carefully selected and designed to be com-
patible with the pulsewidths used and the levels backscattered from the
fiber.

3.2.4 Time base and control unit

The control unit is the brain of the OTDR. It takes all the acquisition
points, performs the averaging, plots them as a log. function of time and
then displays the resulting trace on the OTDR screen.
The time base controls the pulsewidth, the spacing between subsequent
pulses and the signal sampling. Multiple passes are used to improve the sig-
nal to noise ratio of the resulting trace. Since noise is random, by acquiring
many data points at a given distance and averaging them, the noise will tend

Guide to Fiber Optic Measurements 3-7

Otdr.bk : otdr_c3.doc Page 8 Mercredi, Mars 4, 1998 3:31 PM

Principles of an OTDR

to average out toward zero and the remaining data will more accurately rep-
resent the backscatter or reflection level at that point. An OTDR may
acquire up to 32,000 data points and fire thousands of pulses, so the OTDR
processor must be very powerful to deliver fast performance to the user.
The display shows a vertical scale in dB and an horizontal scale in km (or
feet), and plots numerous acquisition points which represent the backscat-
ter "signature" of the fibers under test.

Connector Fusion Connector Fiber Mechanical Fiber

pair Splice pair bend splice end

OTDR

Attenuation
(dB)

Distance (km)

Typical OTDR trace

3.3 OTDR specifications

3.3.1 Dynamic range
The dynamic range is one of the most important characteristics of an
OTDR, since it determines the maximum observable length of a fiber and
therefore the OTDR suitability for analyzing any particular network. The
higher the dynamic range, the higher the signal to noise ratio and the better
the trace will be, with a better event detection. This dynamic range is rela-
tively difficult to determine since there is no standard computation method
used by all the manufacturers.

3-8 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c3.doc Page 9 Mercredi, Mars 4, 1998 3:31 PM

OTDR specifications

Definitions of the dynamic range

One method of determining dynamic range (approved and endorsed by the
IEC 61746) is to take the difference between the extrapolated point of the
backscatter trace at the near end of the fiber (taken at the interception
between the extrapolated trace and the power axis) and the upper level of
the noise floor at or after the fiber end.
• The upper level of the noise is defined as the upper limit of a range
which contains at least 98% of all noise data points.
• The level is expressed in decibels (dB).
• This measurement is performed with a 3 minute period for the averag-
ing.
• This value of the dynamic range was also recommended by Bellcore.
Other definitions of the dynamic ranges are given by different manufactur-
ers, which makes the values comparison very difficult:

dB
Dynamic IEC (98%)

Dynamic range (RMS)

N = 0.1 dB
~6.6 dB
Peak noise level
1.56 dB
SNR=1

km
Dynamic range

• RMS. The RMS (Root Mean Square) also termed SNR=1 dynamic
range is the difference between the extrapolated point of the backscatter
trace at the near end of the fiber (taken at the intersection between the
extrapolated trace and the power axis) and the RMS noise level. You can
compare this value to the IEC 61746 definition by substracting 1.56 dB
from the RMS dynamic range if the noise is gaussian.
• N=0.1 dB. This dynamic range definition gives an idea of the limit to

Guide to Fiber Optic Measurements 3-9

Otdr.bk : otdr_c3.doc Page 10 Mercredi, Mars 4, 1998 3:31 PM

Principles of an OTDR

which the OTDR can measure when the noise level is 0.1 dB on the
trace. The difference between N=0.1 and SNR=1 RMS definition is
approximately 6.6 dB. This means that an OTDR which has a dynamic
range of 28 dB (SNR=1) can measure a fiber event of 0.1 dB up to 21.5
dB.
• End detection: The dynamic range end detection is the one way differ-
ence between the top of a 4% Fresnel reflection at the start of the fiber
and the RMS noise level. This value is approximately 12 dB higher than
the IEC value.
• Bellcore measurement range: The Bellcore measurement range is
defined as the maximum attenuation that can be placed between the
OTDR and an event for which the instrument will still be able to mea-
sure the event within acceptable accuracy limits. The event can be
reflective or non-reflective, or a fiber break. For example, an event can
be a 0.5 dB reflective splice (> 40 dB).
• 4% Fresnel: This is more an echometric parameter than a reflectome-
tric parameter. It represents the ability of the instrument to perceive the
peak of a Fresnel reflection for which the base cannot be perceived. It is
defined as the maximum guaranteed range over which the far end of the
fiber is detected, sometimes with a minimum of 0.3 dB higher than the
highest peak in the noise level;
• Peak level plus 0.3 dB: the dynamic range is the difference between
the front-end backscattered trace and 0.3 dB more than the peak noise
level.

The value of the dynamic range, for each definition can also be given
according to different conditions:
• typical value: this represents the average or mean value of the dynamic
range of the OTDRs which come out of production. An increase of
around 2 dB is usually shown in comparison with the specified value.
• specified value: this is the minimum dynamic range specified by the
manufacturer for its OTDR.
• over a temperature range or at room temperature. At low and high
temperature, the dynamic range decreases usually by 1 dB.

3-10 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c3.doc Page 11 Mercredi, Mars 4, 1998 3:31 PM

OTDR specifications

3.3.2 Dead Zone

Dead zone

OTDR dead zone example

Why do we have dead zone ?

The OTDR is designed to detect the backscattering level all along the fiber
link. It measures backscattered signals which are much smaller than the sig-
nal sent to the fiber. The component which receive those values is the pho-
todiode. It is designed to receive a given level range. When there is a strong
reflection, then the power received by the photodiode can be more than
4000 times higher than the backscattered power and can saturate the photo-
diode. The photodiode requires time to recover from the saturated condi-
tion; during this time, it will not detect the backscatter signal accurately.
The length of fiber which is not fully characterized during the recovery
period is termed the dead zone.
This effect is similar to the one when you are driving a car at night, and that
another car’s headlights dazzle your vision momentarily.

Attenuation dead zone

The attenuation dead zone (defined in IEC 61746) for a reflective or atten-
uating event is the region after the event where the displayed trace deviates
from the undisturbed backscatter trace by more than a given vertical value
∆F (usually 0.5 dB or 0.1 dB). Bellcore specifies a reflectance of - 30 dB, a

Guide to Fiber Optic Measurements 3-11

Otdr.bk : otdr_c3.doc Page 12 Mercredi, Mars 4, 1998 3:31 PM

Principles of an OTDR

loss of 0.1 dB and gives different locations. In general, the higher the
reflected power sent back to the OTDR, the longer the dead zone.
The attenuation dead zone depends on the pulsewidth, the reflectance, the
loss, the displayed power level and the location.
The attenuation dead zone usually indicates the minimum distance after an
event where the backscatter trace can be measured.

ADZ
Attenuation
dead zone

∆F = 0.5 dB or 0.1 dB

Attenuation Dead Zone measurement

At short pulse widths, the recovery time of the photodiode is the primary
determinant of the attenuation dead zone and can be 5 to 6 times larger than
the pulse width itself. At long pulsewidths, the pulsewidth itself is the dom-
inant factor, and the attenuation deadzone is, in effect, equal to the pul-
sewidth itself. The dead zone specified in the literature is generally
measured at the shortest pulsewidth.
Bellcore specifies objectives for two attenuation dead zone, the "front end"
dead zone and the "network" dead zone. Historically, the connection
between the OTDR was highly reflective; this an other factors often caused
the dead zone seen at the front end of the OTDR, to be much longer than
the dead zone resulting from a reflection in the network. Currently, the
OTDR connection has been engineered to have very low reflectance and
there is little difference between the front end dead zone and network dead
zone.

3-12 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c3.doc Page 13 Mercredi, Mars 4, 1998 3:31 PM

OTDR specifications

If the front end attenuation dead zone of the OTDR in use is large, the
effect can be minimized using a launch cable (see "Using launch cables" on
page 4-26).

Event dead zone

Event dead zone is the minimum distance on the trace, where two separate
events can still be distinguished. The distance to each event can be mea-
sured, but the separate loss of each events cannot be measured.
This parameter usually gives an indication of the minimum distance in
order to distinguish between reflective events which occur in close proxim-
ity.
• For a reflective event, the event dead zone definition is the distance
between the two opposite points which are 1.5 dB (or FWHM) down
from the unsaturated peak.

EDZ
Event
dead zone

1.5 dB
±0.1 dB

≤1 dB

± 0.1 dB
Event dead zone

Reflective event Non-reflective event

EDZ Definition EDZ example

Event Dead Zone measurement

• For an non-reflective event, the event dead zone can be described as the
distance between the points where the beginning and ending levels at a
splice or a given value (≤ 1 dB) are within ±0.1 dB of their initial and final
values (this is not the definition).
Event dead zones can also be reduced using smaller pulsewidths.

Guide to Fiber Optic Measurements 3-13

Otdr.bk : otdr_c3.doc Page 14 Mercredi, Mars 4, 1998 3:31 PM

Principles of an OTDR

Front end event dead zone effects can also be minimized from a fiber under
test using a launch cable (see "Using launch cables" on page 4-26).

3.3.3 Resolution
There are four main resolution parameters: display (cursor), loss (level),
sampling (distance) and distance.

Display resolution
The display resolutions are defined as follows:
• The readout resolution is the minimum resolution of the displayed value
(e.g. an attenuation of 0.031 dB will have a resolution of 0.001 dB).
• The cursor resolution is the minimum distance or attenuation between
two displayed points, where a line has been drawn. A typical value can
be 6 cm or 0.01 dB

Loss resolution
The loss resolution is governed by the resolution of the acquisition circuit.
For two near power levels, it specifies the minimum loss difference that can
be measured. This value is generally around 0.01 dB.

Sampling resolution
The sampling (or data point) resolution is the minimum distance between
two acquisition points.
This data point resolution can go down to centimeters depending on pul-
sewidth and range.
In general, the more datapoints that an OTDR can acquire and process, the
better the sampling resolution. The number of datapoints an OTDR can
acquire is therefore an important performance parameter.
a typical value for a high resolution OTDR would be 1 cm sampling resolu-
tion.

Distance resolution
Distance resolution is very similar to sampling resolution.

3-14 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c3.doc Page 15 Mercredi, Mars 4, 1998 3:31 PM

OTDR specifications

The ability of the OTDR to locate an event is affected by the sampling res-
olution. If it only samples acquisition points every 1 meter, then it can only
locate a fiber end within ± 1 meter. The distance resolution is then like the
sampling resolution, a function of the pulse width and the range. This spec-
ification must not be confused with distance accuracy which is discussed
later.

3.3.4 Accuracy
The accuracy of a measurement is the capacity of the measurement to be
compared with a reference value.

Linearity (Attenuation accuracy)

The linearity of the acquisition circuit determines how close an optical level
corresponds to an electrical level, across the whole range.
Most OTDRs have an attenuation accuracy of 0.05 dB/dB. Some OTDRs
can go down to 0.02 dB/dB.
If an OTDR is non linear then with long fibers, the section loss values will
change significantly.

Distance accuracy
The distance measurement accuracy depends on the following parameters:
• Group index : Whereas index of refraction refers to a single ray in a fiber,
group index refers to the propagation velocity of all the light pulses in
the fiber. The accuracy of the OTDR distance measurements depends
on the accuracy of the group index.
• Time base error. This is due to the inaccuracy of the quartz, which can
vary from 10-4 to 10-5. In order to have an idea of the distance error, one
has to multiply this uncertainty by the measured distance.
• Distance error at the origin.
A typical value for the MTS 5100 mini-OTDR is :
± 5 x 10-5 x distance ± 1m ± sampling resolution ± group index uncertainties

Guide to Fiber Optic Measurements 3-15

Otdr.bk : otdr_c3.doc Page 16 Mercredi, Mars 4, 1998 3:31 PM

Principles of an OTDR

3.3.5 Wavelength
OTDRs measure according to a wavelength. The major wavelengths are
850 nm, 1300 nm for multimode, and 1310 nm and 1550 nm for singlemode.
A fourth wavelength is now appearing for monitoring live systems: 1625 nm.
This occurs if the two singlemode wavelengths are used for transmission.
The wavelength is usually specified with a central wavelength and a given
spectral width. The standard spectral width is ±30 nm, but that can be ±10
nm. Some OTDRs display the laser wavelengths used for the measurement.
The attenuation of optical fiber varies with the wavelength, and any mea-
surement should be corrected to the transmission wavelength or to the cen-
tral wavelength (850, 1310 or 1550 nm). Correction is most relevant in the
first window at 850 nm.

3-16 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 1 Mercredi, Mars 4, 1998 3:31 PM

Chapter

4 Using an OTDR 4
T he OTDR is very versatile and has many applications. Firstly, it’s
important to select an OTDR that has the proper specifications
(see chapter 3) for the task at hand. With recent breakthroughs in mod-
ularity, some OTDRs, like the MTS 5100, can be configured
flexibly to perform testing on almost any kind of fiber optic network,
singlemode or multimode, short or long haul.
We can broadly define the use of the OTDR as a two step process :

❏ Acquisition step where the unit acquires data and displays the
results either numerically or graphically;

❏ Measurement step where the operator analyzes the data and makes
a decision based on the results to either store, print, or go the next
fiber acquisition.

4.1 Acquisition
Most modern OTDRs now automatically select the optimal acquisition
parameters for a particular fiber by sending out test pulses in a process
known as auto-conf iguration. Using the Auto-configuration feature, the
user would select the wavelength (or wavelengths) to test, the acquisi-
tion (or averaging) time, and the fiber parameters (e.g. refractive index
if not already entered).
There are about three major approaches to configuration of the OTDR:
• A user might simply let the OTDR autoconfigure and accept the

Guide to Fiber Optic Measurements 4-1

Otdr.bk : otdr_c4.doc Page 2 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

acquisition parameters selected by the OTDR.

• A more experienced user might allow the unit to autoconfigure, analyze
the results briefly and change one or more acquisition parameters to
optimize the configuration for the purposes of his test.
• The experienced user may choose not to use the autoconfiguration fea-
ture altogether and enter acquisition parameters based on his experience
and knowledge of the link under test.
Typically, when testing multifiber cables, once appropriate acquisition
parameters are selected, they are "locked in" and the same parameters are
used for every fiber in the cable (this speeds the acquisition process and pro-
vides for consistency in the data which is helpful when analyzing or compar-
ing fibers).
Below, various acquisition parameters and their effect on the resulting trace
are discussed.

4.1.1 Injection level

Degrading the quality of the OTDR front panel connector through non-
cleanliness will result in poor measurements.
The injection level is defined as the power level which OTDR injects into
the fiber under test. The higher this level, the higher the dynamic range. If
the injection level is low, traces will be noisy and measurement accuracy will
be degraded. Poor launch conditions resulting in low injection levels are the
primary reason for reductions in precision.
The presence of dirt on connector faces and damaged or low quality pigtails
or patchcords are the primary cause of low injection levels. It is important
that all physical connection points are free of dust and dirt in an optical sys-
tem. With core diameters of less than 10 µm in singlemode systems, the
presence of even a 4 µm speck of dirt or dust (approximately the size of the
particulate matter in cigarette smoke) can severely degrade injections lev-
els.
Cleaning kits are available for optical systems from basic tools including iso-
propynol cleaning solution, joseph paper, compressed-air spays, and ready-
to-use impregnated wipes, to more advanced methods with cassette clean-
ers.
Mating of dirty connectors to the OTDR connector, may scratch the OTDR
connector, permanently degrading launch conditions.

4-2 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 3 Mercredi, Mars 4, 1998 3:31 PM

Acquisition

Some OTDRs, like the MTS 5100, will display the measured injection level
during real time acquisition or just prior to averaging. The result is dis-
played on a relative scale on a bar graph rating the injection level from
"good" to "bad".
To determine the relative quality of the injection level, the OTDR "looks"
out a short distance and observes the backscatter returned from the launch
pulse and compares this to an expected value. It is sometimes possible for
the injection level to show "bad" when it is in fact acceptable. This will hap-
pen if there is an attenuator in the system, near the OTDR or if there is a
splitter near the OTDR; in this case, the backscatter level will be lower than
"expected" by the injection level meter. Even though the injection level
increases as pulsewidth increases, the scale displayed is calibrated sepa-
rately for each pulsewidth so the scale is meaningful at any pulsewidth and
increasing pulsewidth will not change a bad injection level to a good one.

4.1.2 OTDR wavelength

The behavior of an optical system is directly related to the wavelength of
transmission. Not only optical fiber will exhibit different loss characteristics
at different wavelengths, but splice loss values will also differ at different
wavelengths.
In general, the fiber should be tested with the same wavelength as that used
for transmission. This means 850 nm and/or 1300 nm for multimode sys-
tems, and 1310 nm and/or 1550 nm for singlemode systems.

If testing is only to be performed at one wavelength, the following

parameters need to be considered:

1. For a given Dynamic range, 1550 nm will see longer distances down the
same fiber than 1310 nm due to the lower attenuation in the fiber:
• 0.35 dB/km at 1310 nm means that approximately 1 dB of signal is
lost every 3 km.
• 0.2 dB/km at 1550 nm means that approximately 1 dB of signal is lost
every 5 km.

2. Single mode fiber has larger mode field diameter (see MFD page 1-9) at
1550 nm than 1310 nm. Larger mode fields are less sensitive to lateral
offset during splicing, but more sensitive to losses incurred by bending
during either installation or in the cabling process.

Guide to Fiber Optic Measurements 4-3

Otdr.bk : otdr_c4.doc Page 4 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

• 1550 nm is more sensitive to bends in the fiber than 1310 nm. This is
shown diagrammatically below. This can also be termed as macroben-
ding.
• 1310 nm will generally measure splice and connector losses higher
than 1510 nm. These results come from a Corning study of over 250
splices where the 1310 nm values were shown to be typically higher
by 0.02dB over the 1550nm values for dispersion-shifted fiber.

Loss (dB)

1310 1550 1580 1620

0.013 0.042 0.094 0.048

0.5

0
1300 1400 1500 1600 1700 λ (nm)

Sensitivity to bending radius = 37,5 mm

Loss (dB)

1310 1550 1580 1620

0.0051 0.123 0.489 2.253

1300 1400 1500 1600 1700 λ (nm)

Sensitivity to bending radius = 30 mm

4.1.3 Pulse width

The OTDR pulsewidth duration controls the amount of light that will be
injected into the fiber. The longer the pulsewidth means the more the light
energy injected. The more light injected means the more light backscatte-
red or reflected back from the fiber to the OTDR.
Long pulsewidths are used to see long distances down a cable. Long pul-
sewidths will also produce longer zones in the OTDR trace waveform where

4-4 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 5 Mercredi, Mars 4, 1998 3:31 PM

Acquisition

measurements are impossible. We call this the dead zone of an OTDR (see
page 3-11).
Short pulsewidths inject lower “levels” of light but reduce this dead zone.

10ns

10µs
30ns

100ns

3 µs

1µs

Different pulsewidths

The pulse width duration is usually given in ns but can also be estimated in
c×T
meters according to the following formula: D = ------------ .
2n
where c represents the speed of light in vacuum (3 x 108 m/s), T the pulse
duration in ns, and n the refractive index.
As an example, a 100 ns pulse could be interpreted as a "10 m" pulse.

Time or Pulse width 5 ns 10 ns 100 ns 1µs 10 µs 20 µs

Distance or fiber length 0.5 m 1m 10 m 100 m 1 km 2 km

Guide to Fiber Optic Measurements 4-5

Otdr.bk : otdr_c4.doc Page 6 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

4.1.4 Range
The range on an OTDR is the maximum distance that the OTDR will
acquire data samples. The longer this parameter, the more distance the
OTDR will shoot pulses down the fiber.
This parameter is generally set at twice the distance of the end of the fiber.
If this parameter is incorrectly set, the trace waveform could contain some
measurement artifacts (see "Ghosts" on page 4-19).

4.1.5 Averaging
The OTDR detector works with extremely low optical power levels (as low
as 100 photons per meter of fiber). Averaging is the process by which each
acquisition point is sampled repeatedly and the results averaged to improve
the signal-to-noise ratio.
By selecting the time of acquisition or the number of averages, the user con-
trols this process within an OTDR.
The longer the time or the higher the number of average, the more signal
the trace waveform will display, in random noise conditions.
The relationship between the acquisition time (number of averages) and
the amount of improvement of the signal-to-noise ratio is expressed by the
equation below:
5 log 10 N
N being the ratio of the two averages.
Note that the noise distribution is considered random for this formula.
As an example, an acquisition with 3 minutes averages will improve by 1.2
dB the dynamic range compared to an acquisition with 1 minute. Averaging
will improve the signal to noise ratio by increasing the number of acquisi-
tions, but the time taken to average the trace is increased. However, accord-
ing to the equation, beyond a certain time, there is no advantage to be
gained as only the signal remains.
In theory, four times more averaging equals + 1.5 dB gain in dynamic range.

4-6 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 7 Mercredi, Mars 4, 1998 3:31 PM

Acquisition

13,5

13

12,5
Helios @ 5ns PW
Dyn. IEC

12
Theoretical
11,5

11

10,5

10
100

140

160

180
120
40

60
20

80

Averaging Time (s)

Dynamic range versus averaging

Guide to Fiber Optic Measurements 4-7

Otdr.bk : otdr_c4.doc Page 8 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

4.1.6 Smoothing
Smoothing is a technique whereby the signal-to-noise ratio is improved by
digitally filtering the acquisition points.
To improve accuracy at lower light levels an OTDR can use filters and aver-
aging techniques to combine the measurements from many pulses.

Two identical fibers - top trace with a smoothing filter

A smoothing function can be performed on the acquisition points. This is

performed by using specific coefficients. A given true point value is modi-
fied to another value which combines previous and subsequent acquisitions
with relevant coefficients.

4.1.7 Fiber parameters

Other parameters related to the fiber can affect the OTDR results as fol-
lows:
• Refractive Index n: this index is directly related to distance measure-

4-8 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 9 Mercredi, Mars 4, 1998 3:31 PM

Acquisition

ments. When comparing distance results from two acquisitions, always

be sure that the appropriate index is being used. It should be noted that
using the refractive index reported by the fiber manufacturer will cause
the OTDR to report fiber length accurately. However, often, particularly
during fault location, the user wishes to determine the cable length.
Fiber length and cable length are not identical and differ due to the
overlength of the fiber in the buffer tube and the geometry (helixing) of
the buffer tubes in the cable. The ratio between fiber length and cable
length varies depending on cable fiber count and cable design, and even
cable manufacturer. While it is possible to have this value (typically
termed the "helix factor") reported by the manufacturer, the precision of
the value still allows for large uncertainty in fault location.
It is often recommended to measure a known length of similarly con-
structed cable and determine an "effective refractive index" that will
cause the OTDR to report cable length instead of fiber length. See "Get-
ting the most out of your OTDR" on page 4-26 for more information on
this.
• Backscatter coefficient K: the backscatter coefficient K tells the
OTDR the relative backscatter level of a given fiber. This coefficient is
entered at the factory and generally the user will not change this param-
eter. Changing it will affect the reported value of reflectance and optical
return loss. While the assumption is made that the backscattered coeffi-
cient for the entire span is consistent, it is possible that there will be very
slight variations from one fiber span to the other. This variation can
cause measurement anomalies such as splices with negative loss values
(or gainers). See section Measurement artifacts and anomalies on page 4-
19 for measurement techniques that minimize the impact of these.
Typical backscatter coefficients at 1 ns are:
- for standard single mode fiber: - 79 dB at1310 nm
- 81 dB at 1550 nm
- for standard multimode fiber: - 70 dB at 850 nm
- 75 dB at 1300 nm

Guide to Fiber Optic Measurements 4-9

Otdr.bk : otdr_c4.doc Page 10 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

4.2 Measurement
Most modern OTDRs will perform fully automatic measurements with very
little user intervention.
In general, there are two types of events: reflective and non reflective.
• Reflective events where a discontinuity in the fiber causes an abrupt
change in the refractive index are either caused by breaks, connectors
junctions, mechanical splices or the undeterminated end of fiber. Con-
nector loss can be around 0.5dB and mechanical splices can range from
0.1dB up to 0.2dB
• Non reflective events occur where there are no discontinuities in the
fiber and generally are produced by fusion splices or bending losses.
Typical values would be from 0.02dB up to 0.1dB depending on the
splicing equipment and operator.

The following measurements can be performed by an OTDR.

For each event: distance location
loss
reflectance
For each section of fiber: section length
section loss in dB
section loss rate in dB/km
ORL (Optical Return Loss) of the section
For the complete terminated system:
link length
link loss in dB
ORL of the link
The OTDR allows the user, at his discretion to perform measurements on
the fiber span in at least three different ways. The user can also use a combi-
nation of these methods:

1. full automatic function: in this case, the OTDR will detect and
measure automatically all the events, sections and fiber end, using an
internal detection algorithm.

4-10 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 11 Mercredi, Mars 4, 1998 3:31 PM

Measurement

Fully automatic trace and table of events (Table mode)

2. semi automatic function: when this is selected, the OTDR will

measure and report an event at each location (distance) where a marker
has been placed. These markers can be placed automatically or
manually. This function is of high interest during span acceptance (after
splicing), where the user desires to completely characterize all events
along the span to establish baseline data. Automatic detection will not
detect and report a non-reflective event with a zero loss, and therefore, a
marker is placed at that location so that the semi-automatic analysis will
report the zero loss. Further analysis of the trace using a PC software
package such as WinTrace ® to perform bi-directional analysis of the
span, then using semi-automatic measurement at fixed marker locations,
will ensure consistency in the number of events from fiber to fiber and
from measurements in the opposing direction.

Guide to Fiber Optic Measurements 4-11

Otdr.bk : otdr_c4.doc Page 12 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

Measurement with markers

3. manual measurement function: For even more detailed analysis or

special conditions, the operator can completely control the measurement
function manually. This means that the operator will place 2 or more
cursors to control the way the OTDR measures the event or value.
Depending on the parameter being measured, the operator may need to
position up to 5 cursors to perform a manual measurement. While this is
the slowest and most cumbersome method of measurement, it is
important to have this capability available for those fiber spans whose
design or construction are very unusual and difficult for automated
algorithms to analyze accurately.

4-12 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 13 Mercredi, Mars 4, 1998 3:31 PM

Measurement

Manual ORL measurement

Guide to Fiber Optic Measurements 4-13

Otdr.bk : otdr_c4.doc Page 14 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

4.2.1 Slope or fiber section loss

The slope of section of fiber, given in dB/km, can be measured either using
a 2-point method (described on page 4-14) or by using a least-squares
approximation (LSA).
The least-squares approximation method tries to determine the measure-
ment line that has the closest fit to the set of acquisition points. It is the
most precise means to measure fiber loss but requires a continuous section
of fiber, a minimum number of OTDR acquisition points, and a relatively
clean backscatter signal free of noise.

Least square approximation : fitting a straight line

The section loss can be reported either in dB or in dB/km. Typical section

losses range between 0.15 to 0.25dB/km for 1550nm systems, 0.25 to
0.35 dB/km for 1310 nm singlemode, 0.5 to 1.5 dB/km for 1300 nm multi-
mode, and 2 to 3.5 dB/km for 850 nm systems.

4.2.2 Event loss

Using manual measurements, there are two ways to measure an event loss:

2-point method
The operator must position a first cursor on the linear level before the
event, and a second cursor on the linear backscatter level after the event.
The event loss is then the difference between these 2 cursor measurements.
This method can be used for a reflective or a non-reflective event. How-
ever, the precision of this method depends on the user’s ability to place the
cursors at the correct positions and can be compromised if the trace has a
large amount of residual noise.
If the trace is very noisy or ‘spiky’, then the user should try to place the cur-
sor on a data point on the trace that is not the top of a spike or bottom of a

4-14 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 15 Mercredi, Mars 4, 1998 3:31 PM

Measurement

trough: this is a sort of visual “averaging” of the trace. If the user is using the
two point method to measure a ‘point’ event (like a splice as opposed to a
length of fiber), then the user should be aware that the result will also
include the effects of any fiber losses between the cursors, because the dis-
tance between the cursors is non-zero.

2-point measurement

5-point method
The purpose of the 5 point measurement method of ‘point’ events is to
reduce the effects of noise on the fiber spans before and after the event by
performing a least squares analysis on the fiber spans, and to minimize the
additional fiber loss that is reported as event loss due to the non-zero dis-
tance between the cursors. In order to do this, the software uses the position
of the 5 cursors to extrapolate the fiber data before and after the event and
take a zero distance measurement of the loss at the event location.
This method is used to measure the loss of both a non-reflective and reflec-
tive events.
To accomplish this, first the operator must make a slope measurement
before and after the event on the linear backscattered level of the trace. The

Guide to Fiber Optic Measurements 4-15

Otdr.bk : otdr_c4.doc Page 16 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

5th point of measurement is placed just before the event where the backs-
catter trace suddenly deviates and the loss measurement is then made at
this event location. This method is more precise than the 2-point as the
OTDR is comparing the difference between 2 linear backscatter levels.

5-point method

5-point measurement

4-16 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 17 Mercredi, Mars 4, 1998 3:31 PM

Measurement

4.2.3 Reflectance and Optical Return Loss

The reflectance of an event represents the ratio of the reflected power to
the incident power at that discrete location in a fiber span. It is expressed in
decibels (-dB). The small negative value indicates a larger reflection than a
large negative value. That is, a reflectance of -33 dB is larger than a reflec-
tance of -60 dB. The larger reflectance will show up as a higher peak on the
trace waveform.

Reflectance measurement

The amount of reflection at a connector, break or mechanical splice

depends on the difference in the refractive index in the fiber and the mate-
rial at the fiber interface (another fiber, air, index matching gel), and geome-
try of the break or connector (flat, angled, crushed, each of which will per-
mit a different amount of reflection to be captured by in the fiber core).
Most mechanical splices use an index-matching gel or fluid to reduce the
amount of change. Smaller changes in the refractive index produce smaller
reflections. Some OTDRs can measure the amount of reflecting light auto-
matically by placing one cursor just in front of the reflection, an another cur-

Guide to Fiber Optic Measurements 4-17

Otdr.bk : otdr_c4.doc Page 18 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

sor at the top of the reflection and by pressing the appropriate button on the
control panel.
The Optical Return Loss (ORL) represents the total optical power return-
ing to the source from the complete fiber span. This includes the backscat-
tered light from the fiber itself, as well as the reflected light from all the
joints and terminations.
ORL = -10 log (Pr/Pi) in dB
with: Pr = reflected power
Pi = incident power
A high level of ORL will degrade the performance of some transmission
links. Analog transmission systems and very high speed digital transmission
systems can be sensitive to ORL. If a system is sensitive to ORL, this is
usually listed in the specifications for the link provided by the manufac-
turer. The MTS 5100 can report a value for total link ORL, by selecting
“ORL = Yes” in the setup menu. The manual ORL measurement is pro-
vided to isolate the portion of the link contributing the majority of the ORL.

ORL of a link

4-18 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 19 Mercredi, Mars 4, 1998 3:31 PM

Measurement artifacts and anomalies

4.3 Measurement artifacts and anomalies

From time to time, unexpected results and events can be seen on the backs-
cattered trace.

4.3.1 Ghosts
False Fresnel reflections on the trace waveform can be observed from time
to time. They can be a result of either:
• strong reflective event on the fiber, causing a large amount of reflected
light to be sent back to the OTDR
• or incorrect range setting during acquisition

OTDR

Ghost
OTDR

Ghosts principle

In both cases, the ghost can be identified as no loss is incurred at the signal
passes through this event. In the first case, the distance that the ghost occurs
along the trace is a multiple of the distance of that strong reflective event
from the OTDR.

Guide to Fiber Optic Measurements 4-19

Otdr.bk : otdr_c4.doc Page 20 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

Example of ghost in the noise

In order to reduce the reflection, you can use index matching gel at the
reflection, or reduce the injected power by selecting a shorter pulse width,
or reducing the power (some OTDRs provide this option) or adding attenu-
ation in the fiber before the reflection.
If the event causing the ghost is situated at the end of the fiber, a few short
turns around a suitable tool (pen, pencil, mandrel etc.) will sufficiently
attenuate the amount of light being reflected back to the source and elimi-
nate the ghost. This is known as a mandrel wrap.

Caution: be sure to select a mandrel of the appropriate diameter for the type
of cable, jacketed fiber, or coated fiber used, so as not cause
permanent damage to the span! It is never recommended to bend
a fiber or cable to introduce attenuation without the use of a
suitable mandrel to prevent excess bending.
Ghosts can also be introduced on the OTDR trace waveform if we incor-
rectly set the distance range.

4-20 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 21 Mercredi, Mars 4, 1998 3:31 PM

Measurement artifacts and anomalies

20 km
OTDR laser distance range
Fiber length 22 km
First pulse Second pulse etc.
OTDR laser pulses

The OTDR’s first pulse is OTDR first pulse

completed at 20 km and the waveform
second pulse is launched
into the fiber.

As the fiber is longer than

the distance range, the
OTDR’s first pulse is still
present on the fiber while OTDR second pulse
the second pulse data is waveform
being acquired. The first
continues 2 km further
down the fiber until it hits
and reflects off the end. 2 km
The first pulse data overlaps with second and consequent pulses and intro-
duces a ghost at 2 km. This distance corresponds to the fiber length minus
the OTDR laser distance range.

4.3.2 Splice "Gain"

It must be remembered that an OTDR measures splice loss indirectly
depending on information obtained from backscattering to calculate splice
loss. It is assumed that the backscatter capture coefficient of the fibers in
the span are identical. If this is not the case, then measurements can be
inaccurate. One common example of this is apparent splice ‘gains’ or ‘gain-
ers’. The inaccuracy is quite small, but with today’s fusion splicing equip-
ment and experienced operators making very low loss splices, it is possible
for the effect to make the splice appear to be a gain.

Guide to Fiber Optic Measurements 4-21

Otdr.bk : otdr_c4.doc Page 22 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

Gain theory
If fibers of different mode-field diameters (core size etc.) are joined, the
resulting OTDR trace waveform can show a higher backscattering level.
This is due to the increased level of backscattered signal reflected back to
the OTDR in the downstream fiber.

A , Ka B, Kb
OTDR

Ka = Kb = Backscatter coefficient S: Splice attenuation

Normal splice

This phenomenon can occur when jointing different types of fiber in multi-
mode or 2 fibers with different backscattering coefficients.

A , Ka B, Kb
OTDR

S1

Ka < Kb Kb-Ba=∆k S1= S+∆K

Positive splice from A to B

B, Kb A , Ka
OTDR

S2

Ka < Kb Kb-Ba=∆k S2 = S-∆K

Negative splice from B to A

S1 + S2
The sum gives the bidirectional or average splice loss value : S = -------------------
2

4-22 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 23 Mercredi, Mars 4, 1998 3:31 PM

Measurement artifacts and anomalies

Bidirectional Analysis
We all know that there is no such thing as a passive amplifier, and that we
can’t get a “gain” in optical power from a fusion splice, but the OTDR will
sometimes report a gain caused by differences in backscatter coefficient.
Note that while these backscatter differences will not always cause a gain to
be reported, they can cause erroneous splice loss readings even if the read-
ing is still a loss.
Bidirectional analysis is a technique used to minimize the effect of backscat-
ter coefficient differences along a span causing these erroneous splice read-
ings. It is used where very accurate baseline data on a span is desired or dur-
ing acceptance testing, where accurate measurement of splicing, often
performed by subcontractors, is desired.
The concept of bidirectional analysis is as follows: If there is a backscatter
coefficient mismatch between two spliced fibers, the sense (algebraically) of
that difference will change depending on the direction of measurement.
That is, if measured in one direction, the difference will appear as a gain, if
measured in the opposing direction, it will appear as a loss. This difference
will combine with the actual splice loss during measurement. However, if
the splice loss reading taken in the two directions is averaged, then the
backscatter effect will subtract out, yielding the actual splice loss.
While the concept is presented here in detail and the manual calculations
presented, in actuality, this analysis is usually performed using programs
such as WinTrace ® which will automatically perform this analysis on much
more complex spans than that shown here.

Example of Bidirectional analysis on a hypothetical span

Fusion Fusion
Connector splice splice Connector
West A B East

Fiber 1 Fiber 2 Fiber 3

Span Architecture

The hypothetical span comprises three fiber sections, fusion spliced

between Connector West and Connector East.

Guide to Fiber Optic Measurements 4-23

Otdr.bk : otdr_c4.doc Page 24 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

The relative backscatter profile of the fibers is shown. In this model, we are
temporarily ignoring the loss in the fiber to show, that if the backscatter
coefficient was sampled at many points along the span, the coefficient
would be higher in the second or middle section.

Fiber 1 Fiber 2 Fiber 3

Backscatter Profile of span

In this case, let’s say that the effect of the backscatter mismatch appears to
the OTDR to be about 0.05 dB. Remember, and this is very important, that
the effect will appear as a gain if going into fiber 2, but as a loss if exiting
fiber 2.

0.05 dB 0.05 dB

Apparent loss/gain at junction due to backscatter coefficient difference

This span has been fusion spliced and the actual fusion splice loss happens
to be -0.03 dB at SPLICE A between fiber 1 and fiber 2, and -0.07 dB at
SPLICE B between fibers 2 and 3. For this example, we will consistently
use the minus sign to represent a loss and no sign to represent a gain.

Connector Connector
West -0.03 dB -0.07 dB East

Actual Fusion Splice Loss

4-24 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 25 Mercredi, Mars 4, 1998 3:31 PM

Measurement artifacts and anomalies

What the OTDR sees……

Connector Connector
OTDR West East

+0.02 dB
(West)
(East)
A
-0.12 dB

Measurement one (W -> E)

When measuring from West to East, and we are showing the fiber loss now,
SPLICE A appears to be a “gain” of 0.02 dB (the actual -0.03 dB plus the
apparent 0.05 dB gain due to backscatter). SPLICE B appears to be a -0.12
loss (the actual -0.07 loss plus the apparent -0.05 loss due to backscatter).

Connector Connector
West East
OTDR

-0.02 dB
(East)
B (West)
-0.08 dB

Measurement two (E -> W)

Guide to Fiber Optic Measurements 4-25

Otdr.bk : otdr_c4.doc Page 26 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

When measuring from East to West, remember that SPLICE B is now on

the left of the OTDR screen and SPLICE A is on the right, then:
• SPLICE A appears to be a loss of 0.08 dB (the actual -0.03 dB plus the
apparent -0.05 dB loss due to backscatter).
• SPLICE B appears to be a -0.02 loss (the actual -0.07 loss plus the appar-
ent 0.05 “gain” due to backscatter).
After taking the two measurements, we can now make a simple chart show-
ing the loss/"gain" of Splices A and B taken in each direction. We can sum
the two readings and then divide by two to take the average. Note that the
result now accurately represents the actual splice losses of the two events.

OTDR

W→E E→W Sum Average

Actual
loss

Splice A +0.02 -0.08 -0.06 -0.03 -0.03

Splice B -0.12 -0.02 -0.14 -0.07 -0.07

Bidirectional analysis

4.4 Getting the most out of your OTDR

4.4.1 Using launch cables

The use of a launch cables in an OTDR measurement enable a number of
effective tasks:
• correct measurement of the insertion loss of the system end connectors
• moves the dead zone caused by the OTDR front panel connector out-
side of the system under test trace waveform
• improves modal equilibrium characteristics in multimode systems so
that measurement are more precise
• allows the user to control the OTDR injection level into the system
under test.

4-26 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 27 Mercredi, Mars 4, 1998 3:31 PM

Getting the most out of your OTDR

The typical length of a launch cable will depend on the system being tested
but generally is between 500 and 1000 m for a multimode test, and 1000 m
for a singlemode test.The fiber used in the launch cable should match the
fiber being tested (core size etc.) and the cable connectors should be of high
quality.

Trace without launch cable

Note If a helper is available at the far end of the span under test, or
if both ends of the span are accessible, some operators use a
“receive cable” (a sufficiently long span of fiber mated to the
far end of the span) to measure the loss of the far end
connector as well

Guide to Fiber Optic Measurements 4-27

Otdr.bk : otdr_c4.doc Page 28 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

System end
connectors

Trace with launch cable and receive cable

4.4.2 Verifying continuity to the fiber end

Sometimes a multifiber cable is installed and you wish to verify that the
cable is continuous between the two exposed ends. You can make an
OTDR measurement on the cable in each direction and that will confirm
that it is continuous. You can also make an OTDR measurement in one
direction and observe the length of the cable as represented on the trace,
however, the length of each fiber in the cable will often vary by a few meters
due to slightly different buffer tube overlength or helix geometry within the
cable. It is difficult, if not impossible, to distinguish a fiber with a much
lower overlength, from a fiber that is broken inside the cable, 1 meter from
the far end.
A simpler way to verify continuity, without having to do a complete OTDR
test from both ends can be accomplished as follows. In this case, you need
access to both ends of the cable, or a helper and communication to the
helper.
Simply hook the OTDR to one of the fibers in the cable, say Fiber 1. Turn
the OTDR on Real Time mode and observe the end of the trace. If the

4-28 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 29 Mercredi, Mars 4, 1998 3:31 PM

Getting the most out of your OTDR

length looks approximately right then do the following (if the length is
grossly short, you know it’s broken).
If you can’t see an end spike (reflective event at the unterminated glass/air
interface at the end of the cable), then have the helper cleave the fiber end
squarely with a hand cleaver. The end spike or end reflection should
become apparent, if it does not, then the helper is not holding the end of
Fiber 1, Fiber 1 is broken somewhere inside the cable near the end. If at
first you do see a large end spike, have the helper dip the end of the fiber in
index matching gel, or alcohol, or wrap the fiber around a small mandrel
near the end. Doing any of these will attenuate the end spike. If it does not,
then the fiber is broken somewhere else near the end of the cable.

4.4.3 Fault location

The OTDR can be an invaluable tool for fault location. Accurate fault loca-
tion depends on careful measurement technique with the OTDR and on
complete and accurate system (cable) documentation. While entire courses
are often taught on the subject of fault location, following the few recom-
mendations below may make the process more accurate and efficient should
the need ever arise.
Cable breaks can be partial or complete (catastrophic). The most common
cause of cable breaks is «dig-ups» (over 40 % of all breaks are dig ups). In
the case of a dig up, fault location does not need to be extremely precise as
the damage can usually be easily located once one is in the vicinity. Other
types of breaks including ballistic (from hunting weapons) or rodent damage
are difficult to find and accurate location with an OTDR can save a great
deal of time and money.
When a cable is damaged the resulting break may be highly reflective or
non reflective. It is generally much easier to determine an accurate distance
to a reflective event. Therefore, it is sometimes helpful to measure several
broken fibers until a reflective break is found. If the break is non-reflective,
it is usually best to let the OTDR software determine the distance to events
using automated analysis. Placing a marker or cursor visually can be inaccu-
rate.
The operator may wish to calibrate the OTDR to display distance in Cable
or Sheath distance by using an «effective refractive index».This is impor-
tant, while the OTDR can accurately determine distances to 5 meters in
10,000, the helix factor of the cable will contribute up to 600 meters of inac-
curacy over a 10,000 meter span. An alternate method of determining actual

Guide to Fiber Optic Measurements 4-29

Otdr.bk : otdr_c4.doc Page 30 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

distance from optical distance is to measure the break from both endpoints
and determining the position of the break relative to the total span length.
This ratio of the optical distance to the break to the total optical length of
the span will be the same as the ratio of the sheath distance to the break to
the total sheath length.
It is important to remember any locations that cable slack is stored. If the
OTDR reads 1800 meters to the break but there are 200 meters of slack
stored at an intermediate handhole, manhole or pole, then the distance to
the break will be similarly shorter.
It is important to remember sag in aerial plant Sheath distance will differ
somewhat from pole distance. After the location of the break is determined,
it should be correlated to a cable sequential marking. Then, when excavat-
ing the cable or examining the aerial plant with binoculars, the correct sec-
tion of cable can be quickly confirmed.
It is always best to measure the distance to the break from the last event
whose physical location is known on the OTDR signature using the cursors.
In this manner, the shortest possible measurement is made on the OTDR
reducing the OTDR contribution to measurement inaccuracy.
During initial cable documentation, take advantage of some of the OTDR
features that permit the addition of notes to events or files. Geographic or
GPS data can be entered here that will be very useful during fault location.
Again, there is absolutely no substitute for complete, detailed, accurate
cable documentation records during fault location.

4.4.4 Effective refractive index

The user is reminded that the OTDR determines the distance to the event
based on time. The refractive index serves as a correlation factor between
time and distance allowing the OTDR to display distance.
If the user knows the refractive index provided by the fiber manufacturer,
he can enter this value on the OTDR thus improving the accuracy of the
optical distance displayed.
In most cable designs, the length of the fiber is greater than the length of
the cable. This can be caused by fiber overlength in the buffer tubes (in
“ loose ” buffer designs) and/or “ helixing ” of the buffer tubes or ribbons
inside the cable. The cable length or physical distance can therefore vary
significantly from the fiber length or optical distance.

4-30 Guide to Fiber Optic Measurements

Otdr.bk : otdr_c4.doc Page 31 Mercredi, Mars 4, 1998 3:31 PM

Getting the most out of your OTDR

In some cases, notably fault location, the users wishes the OTDR to display
cable or physical distance instead of optical distance. This can be accom-
plished by entering a different value of refractive index, sometimes termed
the “effective refractive index” that is adjusted for fiber overlength.
There are two ways to determine effective refractive index :

1. Using cable records or knowing the cable or physical distance (Leff)

between two known events on the OTDR trace, the user must obtain
from the OTDR the following data :
Optical distance between 2 known events (Lopt)
Refractive index used by the instrument (RIopt)
The effective refractive index (RIeff) can then be calculated using the
formula : RIeff = (Lopt* RIopt) / Leff

2. On some OTDR’s like the MTS 5100, the RIeff can be calculated
automatically by delimiting the two known events with two cursors and
changing the refractive index until the OTDR reports cable or physical
distance instead of optical distance.

Guide to Fiber Optic Measurements 4-31

Otdr.bk : otdr_c4.doc Page 32 Mercredi, Mars 4, 1998 3:31 PM

Using an OTDR

4-32 Guide to Fiber Optic Measurements

Otdr.bk : otdr_a1.doc Page 1 Mercredi, Mars 4, 1998 3:31 PM

Chapter
Glossary
5
5
ased on IEC 50 chapter 731, EIA-440-B and other documents.

B
Absorption: in an optical fiber, loss of optical power resulting from
conversion of power into heat.

Adaptor: female part of a connector in which one or two connector

plugs are inserted and aligned.

APD (Avalanche Photodiode): photodiode which operates in the

avalanche mode, providing internal gain that is advantageous in
reception.

Architecture: The protocol that defines computer communication

networks. With respect to optical fiber cabling, this term refers to the
layout of the cabling in star or ring configuration, for example.

Armored cable: A fiber cable that includes a layer of corrugated steel

to prevent rodent ingress. Primarily for direct buried applications,
occasionally used in aerial applications where squirrels are a severe
problem
Attenuation dead zone: for a reflective or attenuating event it is the
region after the event where the displayed trace deviates from the
undisturbed backscatter trace by more than a given vertical value ∆F
(usually 0.5 dB or 0.1 dB). Bellcore specifies a reflectance of - 30 dB, a
loss of 0.1 dB and gives different locations. In general, the higher the
reflected power sent back to the OTDR, the longer the dead zone.
The attenuation dead zone depends on the pulsewidth, the reflectance,
the loss, the displayed power level and the location. It usually indicates

Guide to Fiber Optic Measurements A-1

Otdr.bk : otdr_a1.doc Page 2 Mercredi, Mars 4, 1998 3:31 PM

Glossary

the minimum distance after an event where the backscatter trace can be
measured.

ADZ
Attenuation
dead zone

∆F = 0.5 dB or 0.1 dB

Attenuation Dead Zone measurement

ATM (Asynchronous Transfer Mode): A network standard that specifies

fixed length cells to transmit data, voice and video information. ATM is
scalable in that it can operate at different transmission speeds such as 51,
100, 155, 622 Mb/s and beyond.

Attenuation: in optical fibers, loss of average optical power due to

absorption, scattering and other radiation losses. It is generally expressed in
dB without a negative sign.

A-2 Guide to Fiber Optic Measurements

Otdr.bk : otdr_a1.doc Page 3 Mercredi, Mars 4, 1998 3:31 PM

(µm)

Attenuation coefficient: The rate of optical power loss with respect to

distance along the fiber, usually measured in decibels per kilometer (dB/
km) at a specific wavelength. The lower the number, the better the fiber’s
attenuation. Attenuation is specified at 850 and 1300 nm for multimode
fiber. and 1310 and 1550 nm for singlemode fiber, over a temperature range
of -60°C to +85°C.

Backscattering: portion of scattered light which returns in a direction

generally reverse to the direction of propagation.

Bandwidth: difference, expressed in Hertz (Hz), between the highest and

the lowest frequencies passing through the fiber.
Note : This term is often used to specify the bandwidth (MHz x km) of a
multimode fiber.

Guide to Fiber Optic Measurements A-3

Otdr.bk : otdr_a1.doc Page 4 Mercredi, Mars 4, 1998 3:31 PM

Glossary

Attenuation
20
Twisted
screened pair
TG22U Coaxial
RG217U
Coaxial
15 RG220U

Graded Index
λ = 1.3 µm
10
Graded Index
λ = 0.85 µm
Monomode
5
λ = 1.3 µm

0 Frequency
0.1 1 10 100MHz 1 10 100 GHz

Attenuation versus frequency

Bend Radius (minimum): The radius a fiber can bend before increased
loss or mechanical damage occurs.

Broadband: A signal technique that involves modulating the signal on a

carrier before transmission. This allows multiple information signals to be
transmitted simultaneously on different carrier frequencies.

Buffer tube: A thermoplastic tube which is a component of fiber optic

cables serving to segregate the fibers into “groups” and to mechanically
decouple mechanical forces on the cable from the fibers by permitting the
fibers to “float” in the tube. The tubes can be filled (in outdoor cables) or
unfilled (in indoor cables).

Building backbone cable: A cable that connects the building distributor

to a floor distributor. Building backbone cables may also connect floor
distributors in the same building.

Building Distributor: A distributor in which the building backbone

cable(s) terminate(s) and at which connections to the campus backbone
cable(s) may be made.

A-4 Guide to Fiber Optic Measurements

Otdr.bk : otdr_a1.doc Page 5 Mercredi, Mars 4, 1998 3:31 PM

Cable: A structure carrying multiple fibers, usually more than 4 (less than 4
fibers in a structure is usually referred to as “CORD”) and providing
mechanical and environmental protection, tensile strength and fire
resistance.
Two different design concepts exist :
Loose tube cable: This design allows primary coated optical fiber or bundles of
primary coated optical fibers to lie loosely inside a polymer tube or "former"
thus taking advantage of the minimum strain configuration within the tube
or former whist protecting them from abrasion and other external forces.
The tube or former may be filled with compounds to prevent ingress and
propagation of moisture which may affect the optical fibers.
These cables are designed to withstand the mechanical stresses involved
when cables are pulled through extensive duct systems and are particularly
suitable for external use.

Thermoplastic
Outer jacket tube
polyethylene
Bloisture
Central FRP blocking gel
strength member

Flooded core Multiple 250

micron fibers

Aramid strength
element
Loose tube cable

Tight buf fered tube cable: This design features secondary coated (buffered)
optical fibers within a flexible and durable construction. The cables are of
generally low fiber count with aramid strength element protection layers
and a polymer outer sheath. This design is particularly suited to internal
applications.

Guide to Fiber Optic Measurements A-5

Otdr.bk : otdr_a1.doc Page 6 Mercredi, Mars 4, 1998 3:31 PM

Glossary

Thermoplastic
jacket
Outer jacket
polyethylene
Aramid
strength
Central
element
member

900 micron
Tight Buffered
fibers
Overall
polyester tape
barrier

Tight buffered tube cable

Cabling system:

Campus Building Floor Telecommunication Terminal

distributor backbone distributor Outlet Equipment

Campus Building Horizontal Work area

backbone cabling backbone cabling cabling cabling
subsystem subsystem subsystem

Generic cabling system

Campus backbone cable: A cable that connects the campus distributor to

the building distributor(s). campus backbone may also connect building
distributors directly.

Campus Distributor: The distributor from which the campus backbone

cabling emanates.

A-6 Guide to Fiber Optic Measurements

Otdr.bk : otdr_a1.doc Page 7 Mercredi, Mars 4, 1998 3:31 PM

Chromatic dispersion: A type of dispersion that causes broadening of

input pulses along the length of the fiber. Chromatic dispersion is due to the
different wavelengths of light traveling at different speeds through the fiber.
It is at a minimum value at the fiber zero dispersion wavelength.

Cladding: The glass layer surrounding the core of an optical fiber. The
lower index of refraction of the cladding as compared to the core causes the
light within the core to be totally internally reflected and remain in the core.

Coating : An acrylate polymer material put on a fiber during the draw

process to protect it from the environment and rough handling.

Connector: A junction which allows an optical fiber or cable to be

repeatedly connected or disconnected to a device such as a source or a
detector.

Coupling ratio/loss (Cr, Cl): ratio /loss of optical power from one output
port to the total output power, expressed as a percent.

Core: The central region of an optical fiber through which light is

transmitted.

CPE: Customer Premises Equipment.

Cutoff wavelength: In singlemode fiber, the shortest wavelength at which

a single mode can be transmitted. Beyond this wavelength, several modes
transmit simultaneously, and the fiber becomes multimode.

CW: Abbreviation for continuous wave.

Dense WDM (Wavelength division multiplexing) : Technique used to

multiplex several signals on the same fiber within a narrow wavelength
band.

Dead Zone: Distance sections in trace, which are associated with every
reflective event and represent the distance between the beginning of the
event and the point where a consecutive event can be detected.

Guide to Fiber Optic Measurements A-7

Otdr.bk : otdr_a1.doc Page 8 Mercredi, Mars 4, 1998 3:31 PM

Glossary

DFB: Abbreviation for distributed feedback laser. This laser has a Bragg
reflection grating in the active region in order to suppress multiple
longitudinal modes and enhance a single-longitudinal mode.

Dispersion: The cause of bandwidth limitation in a fiber. The spreading

(or broadening) of a light pulse as it spreads along a fiber. Major types are :
• modal dispersion cause by differential optical path lengths in a
multimode fiber,
• chromatic dispersion caused by a differential delay of various wavelengths
of light passing through a fiber.

Distributor: The term used for the functions of a collection of components

(e.g. : patch panels, patch cords) used to connect cables.

DTE: Data Terminal Equipment, generally.

Dynamic range:
IEC Dynamic Range (introduced by Bellcore)
The difference between the extrapolated point of the backscatter trace at
the near end of the fiber (taken at the interception between the extrapo-
lated trace and the power axis) and the upper level of the noise floor at or
after the fiber end. The upper level of the noise is defined as the upper
limit of a range which contains at least 98% of all noise data points. The
dynamic range is expressed in decibels (dB). This measurement is per-
formed for 180 seconds usually with largest pulsewidth of the OTDR.
RMS Dynamic Range
The difference between the extrapolated point of the backscatter trace at
the near end of the fiber (taken at the intersection between the extrapolated
trace and the power axis) and the RMS noise level.

A-8 Guide to Fiber Optic Measurements

Otdr.bk : otdr_a1.doc Page 9 Mercredi, Mars 4, 1998 3:31 PM

dB

Dynamic IEC (98%)

Dynamic range (RMS)

N = 0.1 dB
~6.6 dB
Peak noise level
1.56 dB
SNR=1

km
RMS dynamic range

EDFA : Abbreviation for Erbium Doped Fiber Amplifier. Device which

amplifies an optical signal without employing O/E and E/O conversions.

Electromagnetic Spectrum: It is a term used to describe the entire range

of light radiation, from gamma rays to radio.
Electromagnetic spectrum

Type of Radiation Frequency Range Wavelength Range

20 24
Gamma-rays 10 - 10 <10-12 m
X-rays 1017 - 1020 1 nm - 1 pm
Ultraviolet 1015 - 1017 400 nm - 1 nm
Visible 4.1014 - 7.5x1014 750 nm - 400 nm
Near infrared 12
10 - 4.10 14
300 µm - 750 nm
Infrared 1011 - 1012 300 µm
Microwaves 8
10 - 10 12
3m - 300 µm
Radio waves 100 - 108 >1 mm

Ethernet: A network protocol specified for operation to 10 Mbit/s.

Standards are being developed for 100 Mbit/s and beyond.
Event dead zone: minimum distance on the trace, where two separate
events can still be distinguished. The distance to each event can be mea-
sured, but the separate loss of each events cannot be measured. This param-

Guide to Fiber Optic Measurements A-9

Otdr.bk : otdr_a1.doc Page 10 Mercredi, Mars 4, 1998 3:31 PM

Glossary

eter usually gives an indication of the minimum distance in order to

distinguish between reflective events which occur in close proximity.
• For a reflective event, the event dead zone definition is the distance
between the two opposite points which are 1.5 dB (or FWHM) down
from the peak. The reflectance of the event shall be specified: as an
example Bellcore gives a reflectance of -30 dB.
• For an non-reflective event, the event dead zone definition is the dis-
tance between the points where the beginning and ending levels at a
splice or a given value (≤ 1 dB) are within ±0.1 dB of their initial and final
values. Usually this dead zone is a fixed value and depends only on the
pulsewidth and the fiber. This definition is not often used.

EDZ
Event
dead zone

1.5 dB

≤1 dB

Event dead zone

Reflective event Non reflective event

Event Dead Zone measurement

Ferrule: A mechanical fixture, generally a rigid tube, used to confine and

align the polished or cleaved end of the fiber in a connector. Generally
associated with fiber-optic connectors.

Fiber Distributed Data Interface (FDDI): A standard for 100 Mbit/s fiber-
optic local area network.

Fiber optic span: A series of one or more terminated optical fiber elements
which may contain complex passive components.

A-10 Guide to Fiber Optic Measurements

Otdr.bk : otdr_a1.doc Page 11 Mercredi, Mars 4, 1998 3:31 PM

Floor distributor: The distributor used to make connections between the

horizontal cabling, other cabling subsystems and active equipment.

FTTB: abbreviation for Fiber-To-The-Building

FTTC / FTTK: abbreviation for Fiber-To-The-Curb / Kerb

FTTH: abbreviation for Fiber-To-The-Home

FTTO: abbreviation for Fiber-To-The-Office

Fusion Splice : A permanent joint accomplished by the application of

localized heat sufficient to fuse or melt the ends of the optical fiber
together, forming a continuous single fiber.

Fusion splice by electrical arcing

Graded-index fiber: Fiber design in which the refractive index of the core
is lower toward the outside of the core and increases toward the center with
the peak at the centerline. This multimode fiber design reduces the time
difference between the arrival of different modes, minimizing modal
dispersion and maximizing bandwidth.

Group index: The factor by which the speed of light in vacuum has to be
divided to yield the propagation velocity of light pulses in the fiber.

Hub : Houses the network software and directs communications within the
network.

Index of Refraction: see refractive index

Guide to Fiber Optic Measurements A-11

Otdr.bk : otdr_a1.doc Page 12 Mercredi, Mars 4, 1998 3:31 PM

Glossary

Insertion loss: The increase in the total optical attenuation caused by the
insertion of an optical component in the transmission path.

Joint: an assembly designed to connect 2 or more optical fibers.

Jumper :A cable unit or cable element without connectors used to make a

connection on a cross-connect.

LAN (Local Area Network): A geographically limited communications

network intended for the local transport of data, video, and voice. It’s a high
speed transmission (Mbit/s) which facilitates information transfer.

Laser (Light Amplificated by Stimulated Emission of Radiation): A

device that produces monochromatic, coherent light through stimulated
emission.

0.5
Laser : 0.6 nm

839 840 841 Wavelengths (nm)

Spectral bandwidth of a laser

Launch fiber: A length of fiber used to create an equilibrium modal

distribution in multimode, and to measure the first connector of the
network in both multimode and singlemode systems.

Light Emitting Diode (LED): A semiconductor device used to transmit

light into a fiber in response to an electrical signal. It typically has a broad
spectral width. Its spectral width typically is 50 to 60 nm.

A-12 Guide to Fiber Optic Measurements

Otdr.bk : otdr_a1.doc Page 13 Mercredi, Mars 4, 1998 3:31 PM

LED : 65 nm

Laser : 0,6 nm

700 800 1000 Wavelength (nm)

Comparison of Laser and Led

Mechanical splice: A fiber splice accomplished by fixtures or materials

rather than thermal fusion.

Mechanical splice

Micro bend: Small distortion of a fiber caused by external factors such as

cabling.

Mode field diameter (MFD): A parameter which expresses for a single

mode fiber the section where the majority of the light energy passes. It can
be expressed as the diameter of optical energy in the fiber. Because the
MFD is greater than the core diameter, MFD effectively replaces core
diameter in practise.

Multimode fiber: An optical fiber in which light travels in multiple modes.

Guide to Fiber Optic Measurements A-13

Otdr.bk : otdr_a1.doc Page 14 Mercredi, Mars 4, 1998 3:31 PM

Glossary

Cladding
Size: 125 µm
Index of refraction: 1.46

Core
Coating Size: 62.5 µm or 50 µm
Size: 900 µm Index of refraction: 1.48
Multimode fiber

Multiplex: Combining two or more signals into a single bit stream that can
be individually recovered.

Node: A point of flexibility and/or interconnection within the fiber optic

cabling system.

Numerical aperture : The number that expresses light gathering capacity

of a fiber related to the acceptance angle.
The sine of 5% optical power angle (corresponding to -13 dB) is used to
measure the Numerical Aperture.

-13 dB (5% of max power density)

α0
0 dB Max. power density
Fiber under test

NA = 2 α0 -13 dB

Optical Loss budget: The amount of signal loss that can be tolerated in a
system before errors occur.

ORL (Optical Return Loss): The ratio (expressed in dB) of the reflected
power to the incident power from a fiber optic system or link
ORL = -10 log (Pr/Pi) or ORL = 10 log (Pi/Pr)

A-14 Guide to Fiber Optic Measurements

Otdr.bk : otdr_a1.doc Page 15 Mercredi, Mars 4, 1998 3:31 PM

OTDR: Abbreviation for Optical Time Domain Reflectometer. An

instrument used to characterize a fiber optic link. Useful in estimating fiber
link attenuation, attenuation coefficient, discrete reflections, splice/
connector loss, and point defects, all as a function of fiber distance.

Patchcord: A cable assembly, permanently assembled at both ends with

connector components (principally for cross-connection within a patching
facility).

Pigtail: A short length of optical fiber permanently attached to a connector

and intended to facilitate jointing between that connector and another
optical fiber or component.

Point to point: A connection established between two specific locations or

devices such as a hub and a workstation or between two buildings.

Reflectance: The ratio of reflected power to incident power of an event or

connector R = 10 log(Pr/Pi).

Refractive index: A property of light transmitting materials defined as the

ratio of the velocity of light in vacuum (c) to its velocity in a given
transmission medium (v).
n = c/v
n = Refractive Index
c = 2.99792458 . 108 m/s
e.g.: n (air)= 1.0003 ; n (water) = 1.33 ; n (Glass)= 1.5

Repeater: A device used to regenerate an optical signal to allow an increase

in the system length.

Scattering: A property that causes light to deflect out of the core area of the
fiber, thereby contributing to attenuation.

Singlemode fiber: An optical wave guide (or fiber) in which the signal
travels in one mode.

SONET (Synchronous Optical NETwork): It is a transport interface that

enables the public network to carry various kind of services. SONET is the

Guide to Fiber Optic Measurements A-15

Otdr.bk : otdr_a1.doc Page 16 Mercredi, Mars 4, 1998 3:31 PM

Glossary

North American optical fiber standard that supports transmission rates that
start at 51.84 Mb/s and reach to 2.488 Gb/s.

Telecommunications closet: An enclosed space for housing

telecommunications equipment, cable terminations, and cross-connect
cabling. The telecommunications closet is a recognized cross-connect
between the backbone and horizontal cabling subsystems.

Telecommunications outlet: A fixed connecting device where the

horizontal cable terminates. The telecommunications outlet provides the
interface to work area cabling.

Transmitter: An electronic package used to convert a signal carrying

electronic information to a corresponding optical signal for transmission by
fiber. The transmitter can be a light emitting diode (LED), laser diode, or
vertical cavity surface emitting laser (VCSEL).

SNR (Signal to Noise Ratio): The ratio of the received optical signal power
divided by the RMS noise floor for the detector.

Splice : A permanent junction between optical fibers

Splitter: A passive device which devises optical power among several

output fibers from a common input.

Step-index fiber: A fiber whose index of refraction (n) changes sharply at

the interfaces of its core and cladding.

Visual Fault Locator: The visual fault locator is a visual light source used
to locate breaks or point of excess loss in fiber cable. The common
wavelengths are 635 nm, 650 nm and 670 nm.

A-16 Guide to Fiber Optic Measurements

Otdr.bk : otdr_a1.doc Page 17 Mercredi, Mars 4, 1998 3:31 PM

WDMs: Abbreviation for Wavelength Division Multiplexers. Passive fiber

optic components which combine optical channels on different
wavelengths.

WAN (Wide Area Network): A network used for the transport of information
over many miles.

Guide to Fiber Optic Measurements A-17

Otdr.bk : otdr_a1.doc Page 18 Mercredi, Mars 4, 1998 3:31 PM

Glossary

A-18 Guide to Fiber Optic Measurements

Otdr.bk : Notes.fm Page 1 Mercredi, Mars 4, 1998 3:31 PM

1 Notes

Guide to Fiber Optic Measurements N-1

Otdr.bk : Notes.fm Page 2 Mercredi, Mars 4, 1998 3:31 PM

Notes

N-2 Guide to Fiber Optic Measurements

Otdr.bk : Notes.fm Page 3 Mercredi, Mars 4, 1998 3:31 PM

Guide to Fiber Optic Measurements N-3

Otdr.bk : Notes.fm Page 4 Mercredi, Mars 4, 1998 3:31 PM

Notes

N-4 Guide to Fiber Optic Measurements

Otdr.bk : Notes.fm Page 5 Mercredi, Mars 4, 1998 3:31 PM

Guide to Fiber Optic Measurements N-5

Otdr.bk : Notes.fm Page 6 Mercredi, Mars 4, 1998 3:31 PM

Notes

N-6 Guide to Fiber Optic Measurements

Otdr.bk : OtdrIX.fm Page 1 Mercredi, Mars 4, 1998 3:31 PM

1Index

A front end dead zone 3-12

Accuracy using launch cable 4-26
distance accuracy 3-15 why a dead zone? 3-11
level accuracy 3-15 Dispersion
Acquisition 4-1 chromatic dispersion 1-5
Attenuation 1-3, 1-4, 1-8 dispersion shifted 1-6
Attenuator 2-7 modal dispersion 1-5
Auto-configuration 4-1 Distance error 3-15
Averaging 4-6 Dynamic range 2-8, 3-8

B E
Backscatter coefficient 4-9 Electronic Industries Association
Backscatter profile 4-24 2-1
Backscattering 3-3 Event loss measurement
Backscattering factor 3-2 2-point method 4-14
Bending losses 1-3 5-point method 4-15
Bidirectional Analysis 4-23 F
C Fault location 4-29
Cladding 1-1, 1-9 Fiber Identifier 2-17
Clip-on tester 2-17 Fiber Optic Test Procedure 2-1
Continuity check 4-28 Fresnel reflection 3-4
Core 1-1, 1-9 G
Cut back measurement 2-9
Gain 4-21
D Ghosts 4-19
Dead zone Graded-index multimode 1-8
attenuation dead zone 3-11 Group delay 3-7
event dead zone 3-13 Group index 1-8, 1-9

Guide to Fiber Optic Measurements 1-1

Otdr.bk : OtdrIX.fm Page 2 Mercredi, Mars 4, 1998 3:31 PM

Index

I Optical Return Loss (ORL) 4-13, 4-

IEC 1-10 18
Injection level 4-2 OTDR
Insertion loss 2-9, 4-26 block diagram 3-5
ITU-T 1-10, 2-1 definition 3-1
description 2-12
L measurement 4-10
Laser diode 3-6 specifications 3-8
Launch cable 4-26 use 4-1
Least-squares approx. (LSA) 4-14
Light absorption 1-3 P
Photodiode 3-7
M Power meter 2-8
Measurement 4-10 Pulse generator 3-6
artifacts 4-19 Pulse width 4-4
event loss 4-14
full automatic function 4-10 R
ghosts 4-19 Range 4-6
manual measurement function Rayleigh scattering 1-3, 3-1
4-12 Reflectance 4-17
reflectance 4-17 Refractive index 1-2, 4-8
section loss 4-14 Resolution
semi automatic function 4-11 display resolution 3-14
slope 4-14 distance resolution 3-14
Mini-OTDR 2-11 loss resolution 3-14
Modal equilibrium 4-26 sampling resolution 3-14
Mode 1-2 S
multimode 1-7
Section los 4-14
singlemode 1-8
Section loss 4-14
Mode field diameter (MFD) 1-9, 4-
Singlemode fiber 1-8
3
Slope 4-14
Monitoring system 2-13
Smoothing 4-8
Multimode fiber 1-7
Source 2-8
N Standards 1-10
Numerical aperture 1-1, 1-8 Step index multimode 1-7

O T
Optical loss budget 2-4 Talk set 2-16

1-2 Guide to Fiber Optic Measurements

Otdr.bk : OtdrIX.fm Page 3 Mercredi, Mars 4, 1998 3:31 PM

Testing V
installation testing 2-3 Velocity 1-2
maintenance testing 2-3 Visual Fault Locator 2-16
Time base
description 3-7 W
Time base error 3-15 Wavelength 3-16, 4-3
Transmission tests 2-2

Guide to Fiber Optic Measurements 1-3

Otdr.bk : OtdrIX.fm Page 4 Mercredi, Mars 4, 1998 3:31 PM

Index

1-4 Guide to Fiber Optic Measurements