Issue

Date:
01/09
Issue
Date:
03/10
Supersedes:
Supersedes:--/-05/09

Single-Mode Fiber

Single-Mode Fiber
Splicing and OTDR splice measurements

Application Note

Introduction
The continuous increase of bandwidth used by consumers,

fusion splicing is the commonly used method. Although fusion

government and enterprises causes a rapidly expanding worldwide

splicers have advanced in ease of use and speed, people who are

optical fiber telecommunications network. With the building of Fiber-

responsible for and those who perform fusion splicing do need

To-The Home (FTTH) networks and a general move from long-haul

specific knowledge about fiber, splicing and testing of the results.

to access networks the average installed length of optical fiber

cable is decreasing. The combined effect is that the amount of fiber

This Application Note explains all aspects of fusion splicing on

splices made each year increases even more than the fiber use.

Draka single-mode products, ESMF and BendBright-XS. This

includes the testing of spliced fibers.

Fibers can be connected to each other by fusion splicing,

mechanical splicing and by the use of connectors. Of these three,

1.

Fusion splicing of optical fibers

Fusion splicing is the method of joining two optical fibers end-to-end

alignment, cladding alignment or local injection and detection of

using heat. The goal is to join the two fibers together in such a way

light (LID)

that optical signal passing through the fibers is not attenuated or

6. The two fibers are fused together

reflected back by the splice. The splice and the region surrounding

7. Visual inspection of the splice and splice loss estimation

should be almost as strong as the fiber itself. The source of the

(available on most splicing machine). Redo the splice (step 1 to 7)

necessary heat is usually an electric arc.

in case an error is found

8. Check mechanical strength of the splice (normally done by the

The following steps are necessary to splice optical fiber:

splicing machine)

1. Clean the fiber

9. The bare fiber area around the splice is protected with a splice

2. Stripping the coating off the two fibers that will be spliced

protector

together
3. Cleaning of the stripped fiber

Alternatives to fusion splicing include using optical fiber connectors

4. Each fiber must be cleaved so that its end-face is perfectly flat

or mechanical splices both of which have in general higher insertion

and perpendicular to the axis of the fiber

losses, lower reliability and higher return losses than fusion splicing.

5. Aligning of two end-faces of the fibers. This is normally done by

the splicing machine by means of: fixed V-groove, optical core

2.

Splicing Draka ESMF

Standard Single-Mode (SMF) fiber is the prevailing fiber for

estimates that more than 5 million fiber splices are made using

telecommunication applications carrying digitized voice, internet

Draka ESMF each year.

data, analog video and many other data. Enhanced Single-Mode

Fiber (ESMF) is the standard Draka single-mode fiber and is also

Figure 1 shows the aggregated statistics of splice tests on Draka

the most widely sold type of optical fiber by Draka.

ESMF spliced to itself and to other fibers coming from various fiber
vendors using different manufacturing process (Corning, OFS, and

Draka ESMF is made of highest quality silica glass that enables

Fujikura) at 1550 nm. These splice tests were made with Fitel S177,

easy splicing. The excellent geometry parameters make sure that

Fujikura 50S and Sumitomo T39 splicers. It can be underlined that

splice attenuation is very low and repeatable. This is proven in a

average values and statistic distribution widths are remarkably low

large number of tests as well as proven field performance. Draka

and at the top-of-the-art level.

Draka Communications

Netherlands:

Tel: +31 (0)40 29 58 700

Fax: +31 (0)40 29 58 710

[email protected]

France:

Tel: +33 (0)3 21 79 49 00

Fax: +33 (0)3 21 79 49 33

www.drakafiber.com | www.draka.com

USA:

Toll free: 800-879-9862

Outside US: +1.828.459.9787

Fax: +1.828.459.8267

ESMF splicing results

Axial Deposition) fiber that may looks different to a PCVD fiber. This

35%

Draka - Draka

different appearance of the left and right fiber in the splicer display

Draka - Other

is normal so that there is no reason to reject the splice.

30%

Percentage

25%

In the rare case that 1) a profile/core alignment splice machine

20%

reports an error due to the machines inability to properly recognize

15%

a fiber, and 2) a quick solution is needed, a quick fix can be to

10%

change the splice program from automatic (or SM) to MM. Herewith
5%

splicing machines do not need to recognize the fiber core and

0%
0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

therefore can always splice every fiber. Due to the superb geometry
parameters of Draka fiber, the increase in average splice loss is

Splice attenuation 1550 nm[dB]

very limited, and the mechanical quality of the splice is not

Fig. 1

compromised. Because of the inconvenience not having valid splice

Splice loss distribution of Draka ESMF

estimation values in MM setting this method should be limited to

Especially since the incorporation of the Alcatel optical fiber division

emergency use and the failing splicer should undergo full

in 2004, Draka produces large volumes of this single-mode fiber in

maintenance as soon as possible.

several factories all around the world. In combination with the

APVDTM (Advanced Plasma Vapor Deposition) technique, Draka

A common practice in the field operation relates to a visual

uses two different methods to produce the preforms (where fibers

evaluation of the splice with the optical imaging system of the

are drawn from). Those methods are called FCVD (Furnace

splicer. Presence of defects such as kinks, neck-downs, bubbles

Chemical Vapor Deposition, originally from Alcatel) and PCVD

and vertical white/dark lines often correspond to a poor splice

(Plasma Chemical Vapor Deposition, originally from Draka). Fibers

quality that can impact on splice loss, mechanical behavior or return

from both types fully comply with all relevant international

loss.

specifications (ITU G.652 and IEC 60793-2-50).

When splicing fiber on some splicers, a presence of a light/dark
Modern fully automated splicers will recognize an inserted fiber and

circle (or halo) at the splice joint vicinity is sometimes visible (similar

choose the right splicing program automatically. The splicing

to those shown in the pictures, Figure 3). Depending on operator

process is then shown on the splicer screen. For some machines

experience, such splice could be misinterpreted as a bad one that

that use profile or core alignment, the difference between fibers of

needs to be redone. Such halo-shaped features give no impact on

different manufacturing methods may be visible on the screen of a

splice quality in term of loss, mechanical resistance and return loss.

splicing machine (see example picture, Figure 2).

Fig. 2

Splicer screen showing fibers from different manufacturing

process

This is because of the differences in refractive index and glass

types. An FCVD fiber may looks different from an OVD (Outside

Fig. 3

Pictures of two splices between external (right) and Drakas fiber

(left) for two different splicer models

Vapor Deposition) fiber that looks different from a VAD (Vapor

Draka Communications

Netherlands:

Tel: +31 (0)40 29 58 700

Fax: +31 (0)40 29 58 710

[email protected]

France:

Tel: +33 (0)3 21 79 49 00

Fax: +33 (0)3 21 79 49 33

www.drakafiber.com | www.draka.com

USA:

Toll free: 800-879-9862

Outside US: +1.828.459.9787

Fax: +1.828.459.8267

Due to the presence of dopants located in the cladding, a complex

imaging system of the splicer. This may be seen on the screen of

refractive index profile is produced close to the splice joint (along

the splicer as a typical halo-like pattern. Due to the difference of the

radial and longitudinal directions) during arc discharge and heat-

optical systems and imaging methods, the corresponding visibility

induced diffusion. This visual feature originates from changes in the

can noticeably differ from one splicer to another.

refractive index, during joint formation, that are caught up by the

3.

Splicing Draka BendBright-XS SMF

Draka has been on the forefront in the development of bend-

have updated software for bend- insensitive fibers, including

insensitive fibers. With the 2006 introduction of the BendBright-XS

BendBright-XS, it is possible that the machine may not align

single-mode fiber (SMF), Draka has answered the growing market

correctly the fiber. If this is encountered, this can easily be

demand for bend-insensitive SMF driven by the mass deployment

overcame by simply adopting the machine setting. (see Table I ).

of Fiber-To-The Home (FTTH) access networks. For detailed

Table I:

information on characteristics and application of Draka

Splice machine setting recommendations for

BendBright-XS

BendBright-XS, a datasheet and Application Note are available on

the Draka website (http://www.drakafiber.com).
The bending losses have been improved by two orders of

MODEL

RECOMMENDED PROGRAM
CORRESPONDING ALIGNMENT
METHOD

magnitude thanks to the addition of a trench in the optical cladding.

FSM-11S

Automatic mode

where the refractive index is lower than the refractive index of the

FSM-17S

Automatic mode

optical cladding in the vicinity of the core.

FSM-18S

Automatic mode

FSM-30S

SMF

FSM-40S

MMF

includes splicing BendBright-XS to itself, to other bend-insensitive

FSM 50S

BendBright-XS

fibers, and to standard single-mode fibers.

FSM 60S

BendBright-XS

S122A

Standard SM

S175 (All
version)

BendBright-XS
(US only)
Standard SM
with cladding
alignment*

FUJIKURA

The trench is a perfectly delimited and homogeneous annular zone

Draka and all major splice machine manufacturers have conducted

extensive splicing tests of BendBright-XS and have found all
machines are capable of splicing BendBright-XS effectively. This

FURUKAWA
FITEL

Splice machines use either outside diameter (OD) or core/profile

alignment. OD alignment machines whether single or mass fusion,
are compatible with BendBright-XS. However, some single fiber
core/profile alignment systems with auto recognition software may

S176
S177A

BendBright-XS

Type-25

SM settings

Type-45

SM settings

Type-37

SM
Diameter Alignment*

Type-39

BBxs Diam

Type-65

Standard SM

Type-66

Standard SM

be confused by the new profile design. Because of the close vicinity

of the core and the surrounding trench in BendBright-XS, some
older machines do not have the program to discern the difference

Fig. 4

ERICSSON

CORNING
(SIECOR)

SUMITOMO

between the core and trench (See figure 4).

The trench in BendBright-XS showing up on the fusion splicer

visualization screen

When using an older profile/core alignment machine which does not

M90i
OptiSplice
LID Micro

MMF
(VIDEO mode)
MMF
(VIDEO unequal
pairs)

RSU12

Standard SM

FSU995

Standard SM

Fixed VGroove
Fixed VGroove
Fixed VGroove
Core
alignment
Cladding
alignment
Core
alignment
Core
alignment
Fixed VGroove
Cladding
alignment
Cladding
alignment
Core
alignment
Fixed VGroove
Fixed VGroove
Cladding
alignment
Cladding
alignment
Fixed VGroove
Fixed VGroove
Cladding
alignment
Cladding
alignment
Fixed Vgroove
Core
alignment

ALTERNATIVE
SETTING

MMF
Automatic
mode
Automatic
mode
SM with clad
alignment
SM with clad
alignment
-

Setting should be done by qualified personnel

Draka Communications

Netherlands:

Tel: +31 (0)40 29 58 700

Fax: +31 (0)40 29 58 710

[email protected]

France:

Tel: +33 (0)3 21 79 49 00

Fax: +33 (0)3 21 79 49 33

www.drakafiber.com | www.draka.com

USA:

Toll free: 800-879-9862

Outside US: +1.828.459.9787

Fax: +1.828.459.8267

Table I lists most common splice machines on the market. It is

differ that much when BendBright-XS is spliced to another fiber.

intended to provide guidance and recommendations in case

Two possible splicing cases can be thus distinguished:

alternative settings are required. It should be noted that the splice

- Splicing BendBright-XS to standard single-mode fibers

machine manufacturers have already updated or are in the process

- Splicing BendBright-XS to itself

of updating software to recognize BendBright-XS. Standard SM

settings can be used for outside diameter/cladding alignment

BendBright-XS to SMF

machines, including mass fusion splice machines.

Splicing the trench-assisted BendBright-XS fiber to a standard SMF
Care should be taken when using splicing equipment that uses the

will occur frequently at the edge of an access network or when

LID (Light Injection and Detection) fiber alignment method. Because

splicing fiber pigtails in passive components like power splitters.

the transmission length between the injection and the detection

Although the optical field confining trench represents a very small

points is rather small, the injected power might very well be

part of the total fiber cross-section only, it does influence the

propagated by the inside-trench area of the optical cladding instead

softening temperature of the fiber end slightly. Most fusion splicers

of the fiber core. This effect could lead to fiber misalignment.

have suitable splice programs to perform splicing in a correct

Therefore it is advised to not use the LID alignment system.

manner (See table II). Table III and Figure 5 show the result of
splice test performed by Draka between different commercial

In the case of older profile/core alignment splicers or LID based

available G.652.D fibers and BendBright-XS, performed with

splicers (not reported in Table I) that have no suitable program for

several fusion splicers.

splicing BendBright-XS, a simple approach is to use the Multimode

BBXS splicing results

Fiber fusion program setting. This provides a lower temperature and

35

a higher fusion time (without compromising the mechanical quality

BBXS-SMF

30

BBXS-BBXS

of the splice). Usually this program can be selected from the

25

valid splice estimation values in MM setting this method should be

limited to emergency use and the splicer should be updated with
new software or preferably replaced by a more suitable machine.

Percentage

splicers program library. Because of the inconvenience not having

20
15
10
5

As shown in the table I, for some splicers, best results are obtained
0

by switching the alignment method from core/profile to cladding or

0.00

0.01

0.02

0.03

diameter alignment. In Table I, core alignment machines that can

be switched from core alignment to OD alignment are noted with an

0.04

0.05

0.06

0.07

0.08

0.09

0.10

> 0.10

Splice attenuation 1550 nm [dB]

Fig. 5

Splice losses distribution of BendBright-XS at 1550 nm

* This setting change should be performed by an experienced

technician since it can result in total failure of the splicer when done
Table II: Splice loss statistics for fusion splice tests BendBright-XS
with commercial enhanced single-mode fibers (ITU-T
G.652.D). Measurement performed with bidirectional
OTDR method

improperly.

Do not hesitate to contact the local distributor of the splicing

equipment for up-to-date information and updating procedures.

Splice test results:

As the BendBright-XS fiber design allows a backwards compatibility

with already deployed fibers (standard G.652 single-mode fiber), it
is also important to guarantee compatibility with existing
deployment procedures. As far as fusion splicing operation is
concerned, it is important to ensure that splicing conditions do not

SETTINGS

Fusion
Splicer
Model

Program

Alignment

S177A
S177A
S175
T-39
FSM-50S

SM
MM
SM
SM
Auto

FSM-11S

Auto

FSM-17S

Auto

FSM-40S

MM

Cladding
Cladding
Cladding
Cladding
Core
Fixed
V-groove
Fixed
V-groove
Cladding

Tel: +31 (0)40 29 58 700

1310 nm
Standard
Average
Deviation
dB
dB

1550 nm
Standard
Average
Deviation
dB
dB

0.041
0.040
0.044
0.043
0.039

0.024
0.022
0.023
0.023
0.026

0.032
0.040
0.038
0.041
0.047

0.015
0.020
0.034
0.019
0.023

0.056

0.025

0.052

0.021

0.030

0.019

0.030

0.014

0.063

0.017

0.052

0.014

Draka Communications

Netherlands:

Fax: +31 (0)40 29 58 710

[email protected]

France:

Tel: +33 (0)3 21 79 49 00

Fax: +33 (0)3 21 79 49 33

www.drakafiber.com | www.draka.com

USA:

Toll free: 800-879-9862

Outside US: +1.828.459.9787

Fax: +1.828.459.8267

Table III: Splice loss statistics for fusion splice tests BendBright-XS
with itself. Measurement performed with bidirectional
OTDR method

BendBright-XS to BendBright-XS

Splicing BendBright-XS to itself works like splicing every other

SETTINGS

Fusion
Splicer
Model

Program

Alignment

commercially available splicing machines. In any case, average

S177A
S177A
S175
T-39
FSM-50S

SM
MM
SM
SM
Auto

splicing losses and distribution widths of the splice losses

FSM-11S

Auto

distributions are remarkably low.

FSM-17S

Auto

FSM-40S

MM

Cladding
Cladding
Cladding
Cladding
Core
Fixed
V-groove
Fixed
V-groove
Cladding

standard SMF. Given that spliced fibers have identical chemical

compositions, splicing conditions are usually more relaxed than
splicing with dissimilar fiber. As a result, fusion splicing usually
exhibits better performance than when splicing with heterogeneous
fibers. This is shown in table III and Figure 5 using different

4.

1310 nm
Standard
Deviation
dB
dB

1550 nm
Standard
Deviation
dB
dB

Average

Average

0.026
0.047
0.038
0.045
0.027

0.034
0.023
0.022
0.032
0.016

0.019
0.038
0.036
0.044
0.015

0.035
0.013
0.027
0.027
0.021

0.036

0.027

0.036

0.015

0.036

0.036

0.033

0.039

0.066

0.031

0.053

0.024

OTDR measurements on splices

Optical Time Domain Reflectometers (OTDR, see example picture,

Figure 6) are widely used in the telecommunication industry for

Pulsed
LASER

testing bare and cabled fiber, including final link commissioning.

Fiber under test

Detector

OTDRs can measure the attenuation coefficient of fiber and are

extremely useful to analyze discreet events in a link such as splice

Electronics

points or connector pairs. These instruments are also extremely

Display

useful in locating damaged or distressed cable or broken fibers.

Fig. 7

Rayleigh backscattering basic principles

A representation of the OTDR power P(z) at wavelength

backscattered by a point z along an optical fiber is [1]:

P( z ) = C.

2
MFD 2 ( z )

.Pi . W .10

2
.z
10

[1] where:
Fig. 6

Pi is the input OTDR pulse power

w is the input OTDR pulse width

OTDR is based on raylight scattering. A small fraction of the light is

z is the distance from the origin

spread in all directions when it encounters in-homogeneities of the

is the attenuation coefficient of the fiber (assumed

Example of OTDR for field use

size of its wavelength. A small part of this light is captured by the

constant to simplify the equation)

core of the fiber and propagates backwards (see Figure.7). It is

MFD(z) is the fiber mode field diameter at point z

called backscattered light. It is a local phenomena and any variation

C is a constant, which depends on several parameters

of the backscattered level along the fiber is due to a defect or to an

alteration of the properties at a local point.

such as the fiber material

Equation [1] shows the backscattered power depends on e.g. the
pulse-width, the fiber attenuation coefficient and the MFD. The

When measuring, a light pulse is first injected into the fiber(s) under

optical echo, as given by the above equation, is conventionally

test. Then the OTDR measures the amount of backscattered light

represented on a logarithmic graph: it therefore appears as a

as a function to the time from the initial pulse.

(theoretically) straight line, where the slope represents the

attenuation coefficient of the fiber, .

Draka Communications

Netherlands:

Tel: +31 (0)40 29 58 700

Fax: +31 (0)40 29 58 710

[email protected]

France:

Tel: +33 (0)3 21 79 49 00

Fax: +33 (0)3 21 79 49 33

www.drakafiber.com | www.draka.com

USA:

Toll free: 800-879-9862

Outside US: +1.828.459.9787

Fax: +1.828.459.8267

Combining fibers by means of mechanical or fusion splices may

by the optical transmission systems. The second goal is to ensure a

result in loss of the optical signal because of a number of aspects:

mechanically stable and reliable fiber splice that will last as long as

radial offset of the core centers

angular offset (tilt)

the fibers and the optical fiber cable.

axial offset (mechanical splices)

To achieve the first goal a maximum average splice loss should be

mismatch of MFDs

set. To comply with the system budget, a maximum individual splice

This section focuses on the MFD mismatch because that is in most

loss is not needed, because the systems functionally depends on

cases an important contributor to apparent splice loss displayed

the total link loss, not on an individual splice.

on an OTDR and it can cause confusion when measurement results

are wrongly interpreted (See Section 5).

The true loss due to MFD mismatch is given by equation [2]:

MFD1 .MFD2
Loss due to MFD Mismatch [dB] = 20 Log 2
2
2
MFD1 + MFD2

The induced (true) splice loss due to MFD difference is quite small,
as shown in Figure 8. Typically this loss can be ignored due to its
small order of magnitude. As an example, when MFD1 = 9.2 m and
MFD2 = 9.0 m, the induced loss is only 0.002 dB.
Splice loss from MFD difference
0.010

Fig. 9
Splice loss (dB)

0.008
0.006

Illustration of field installation. In the foreground, a splicing

machine and, in the background, a connection box.

The second goal, mechanical reliability of the splice, is not covered

by attenuation alone. Mechanical reliability is assured by performing

0.004

a splice proof test, which is performed automatically on most

0.002

machines (see example picture, Figure 9) after the actual splicing is

0.000
- 0.5

- 0.4

- 0.3

- 0.2

- 0.1

0.1

0.2

0.3

0.4

0.5

finished. This is to secure a minimum level of mechanical reliability.

Worldwide specifications for splice loss on SM fiber do vary, but

MFD difference (m)

they agree on two main points:

Fig. 8

True splice loss due to MFD difference

Splice loss specification should be based on averages

The measurement of splice loss is done by averaging bi-

The specification of splice losses serves two goals. The first is to

directional OTDR measurements. This is described in

ensure that the optical link loss is within the loss budget as it is set

more detail in section 5

5.

Practical considerations on splice loss determination

During installation, the splice loss is estimated by the optical image

To determine the loss of a splice, in general two possibilities are

processing system of the splicer unit. Based on this estimation, the

available: power metering and OTDR. Since a power meter

splice can be approved or rejected. When commissioning an optical

measures end to end loss, the OTDR is the only tool available to

link, splice losses usually are checked again by OTDR testing from

measure the loss of individual splices. With the OTDR technique,

either one side or from two sides of the fiber link. Note, that for

special care should be taken since OTDRs do not directly measure

testing splices in networks with optical splitters, special procedures

splice loss. There is a special effect referred to as Apparent Gain

exists.

and Apparent Loss, which are related to the OTDR principle of

Draka Communications

Netherlands:

Tel: +31 (0)40 29 58 700

Fax: +31 (0)40 29 58 710

[email protected]

France:

Tel: +33 (0)3 21 79 49 00

Fax: +33 (0)3 21 79 49 33

www.drakafiber.com | www.draka.com

USA:

Toll free: 800-879-9862

Outside US: +1.828.459.9787

Fax: +1.828.459.8267

operation.

The error in the unidirectional measurement (apparent gain/loss), is

estimated by the following equation [3]:

The OTDR backscatter power is very sensitive for the local MFD, as

MFD1

Loss OTDR Error [dB ] = 10 Log

MFD2

it can be seen from equation 1. This backscattered power is

inversely proportional to the MFD of the fiber. Assuming all other
parameters in equation 1 are equal, this means that in a fiber with
smaller MFD, the backscattered power is higher than that in a fiber

This apparent loss or gain does not have an effect on the system

with larger MFD.

loss budget and should not be considered as true splice loss. In the
previous example with MFD1 = 9.2 m and MFD2 = 9.0 m, the error

When fibers with different MFDs are spliced together, the

in the measured unidirectional loss equals 0.1 dB, much larger than

backscatter power level just before and just after the splice will

the true loss due to the MFD mismatch (0.002 dB), see Figure 11.

differ. When measuring from a fiber with a larger MFD into a fiber
with a smaller MFD, the amount of light being backscattered from

The splice effect displayed by the OTDR is a combination of the

the fiber at the launch side of a splice is lower than that from the

actual true splice loss and the apparent gain or loss resulting from

fiber at the other side of the splice. The OTDR may interpret this

the different MFDs. When no true loss is present, the effective gain

difference as an apparent increase in power. If the apparent

and loss shown by both bidirectional OTDR traces are identical in

increase in power is greater than the true loss of the splice itself,

size.

the OTDR trace shows an overall increase in optical power or

Splice loss from MFD difference

apparent gain. Measured in the opposite direction (light is launched

0.200

through the smaller MFD fiber into the larger MFD fiber), the OTDR
trace for the same splice junction will show an exaggerated loss

MFDs.

0.100
Splice loss [dB]

(apparent loss). This directionality is strictly because of the differing

0.000

-0.100

Especially in unidirectional measurements, this particular

True loss

characteristic of OTDR testing may cause misinterpretations in case

of MFD mismatch. Apparent gain and loss should not be

-0.200
-0.5

-0.4

-0.3

Apparent loss/gain

-0.2

-0.1

0.1

0.2

0.3

0.4

0.5

MFD difference [m]

misinterpreted with true optical loss. This commonly known

phenomenon is an artifact of the unidirectional OTDR measurement

Fig. 11

which in general is well understood in the industry. Figure 10 shows

Error in unidirectional OTDR measurements (apparent gain or

loss) in comparison to true splice loss due to MFD difference

an example of unidirectional measurement with apparent gain and

Normally some true loss will be present (e.g. because of offset, tilt,

loss.

etc), because of which the effective gain and loss shown by the
OTDR are asymmetric, as shown in Figure 11. From a
unidirectional measurement this true loss can NOT be determined.

The apparent unidirectional splice loss as shown by the OTDR can

than be written as equation [4]:

OTDRSplice Loss / Unidirectional Meas. [dB ] = True Splice Loss + LossOTDR Error
In bidirectional measurements LossOTDR Error changes sign in case of
dissimilar MFDs. From this, its clear that the true loss of a splice
Fig. 10

Example of and OTDR for field use. Typical OTDR trace with
splices, including gainers

can only be determined by performing a bidirectional OTDR

measurement and averaging both OTDR loss readings (see Figures
12 and 13), as standardized in the Int. Standard IEC 60793-1-40

Draka Communications

Netherlands:

Tel: +31 (0)40 29 58 700

Fax: +31 (0)40 29 58 710

[email protected]

France:

Tel: +33 (0)3 21 79 49 00

Fax: +33 (0)3 21 79 49 33

www.drakafiber.com | www.draka.com

USA:

Toll free: 800-879-9862

Outside US: +1.828.459.9787

Fax: +1.828.459.8267

(Measurement methods and test procedures Attenuation) and the

A second alternative method is to use an optical power-meter.

Industry test procedure EIA/TIA FOTP-61 (Measurement of Fiber or

Taking into account the fiber linear attenuation, it gives the total loss

Cable).

of the link and the splices average loss.

In summary, the maximum apparent splice loss seen by

unidirectional OTDR measurement can be calculated with equation
[5]:

MFDmax
Maximum LossOTDR Error [dB ] = 10 Log
MFDmin
Fig. 12

Two bidirectional OTDR traces showing a splice between fibers

with unequal fiber attenuations (1 > 2) and a true splice loss.
Because of the assumed equal mode fields both traces show the
same true loss.

Table IV: Examples of calculated apparent loss from MFD

differences

Minimum

MFD (m)
Maximum

Apparent
Loss (dB)

8.9
8.8
8.7
9.0
8.6
8.6

9.1
9.2
9.3
9.4
9.4
9.0

0.10
0.19
0.29
0.19
0.39
0.20

Table IV shows the apparent loss for some MFD differences. For
Fig. 13

The same two bidirectional OTDR traces, now assuming

difference in mode (MFD1 > MFD2), causing an equal Loser and
Gainer, which - due to the presence of true loss - are unequally
shown by the OTDR.

Draka ESMF with a MFD1310 specified as 9.0 0.4 m, the worst

case apparent loss/gain is 0.39 dB.

Finding two fibers opposite of each other that are on both limits of
While the official procedure to determine true splice loss is the

the MFD range is of course very unlikely to happen.

average of two bidirectional OTDR measurements, these two-way

OTDR tests are not always possible (e.g. when one side of a

Unidirectional OTDR tests can be useful for making a first analysis

spliced link is not accessible, or when the link is too long). In such a

of a link to select splices that should be redone. Taking into account

case the true splice loss can NOT be determined any more from the

that modern fiber can be spliced with low loss, a relatively high

single OTDR trace and Draka recommends then to use alternative

threshold (e.g. > 0.4 dB) should be set in order not to redo splices

methods, which in general are less accurate.

that will show as a gainer from the other side.

One alternative method is to rely on the estimated splice loss given

As OTDR measurements mostly are performed when all splice

by the splicer in combination with a unidirectional OTDR check on

closures are closed and buried or mounted, redoing splices will

the average attenuation per km of the total link. (In a unidirectional

require time consuming operations like traveling, road-guarding,

OTDR test a gainer splice is expected to be followed by a loser

equipment setting up, etc. Because these additional operations

splice and vice versa).

account for high additional costs, redoing of splices should be

prevented as much as possible by selecting an adequate
measurement strategy.

Draka Communications

Netherlands:

Tel: +31 (0)40 29 58 700

Fax: +31 (0)40 29 58 710

[email protected]

France:

Tel: +33 (0)3 21 79 49 00

Fax: +33 (0)3 21 79 49 33

www.drakafiber.com | www.draka.com

USA:

Toll free: 800-879-9862

Outside US: +1.828.459.9787

Fax: +1.828.459.8267

Conclusion
Draka ESMF fiber (G.652.D) can be spliced to itself and to other

machine are a guarantee. Drakas bend-insensitive SM fiber

SMF complying to ITU G.652 with splice losses that very rarely will

BendBright-XS can be spliced to itself and to other SMF without any

exceed 0.1 dB.

decrease in splice quality.

Same conclusion applies to BendBright-XS, Drakas bend-

Splice loss is measured by averaging bidirectional OTDR

insensitive SMF introduced in 2006. The work carried out with the

measurements.

splicer manufacturers and the tests performed on the most popular

Specification of splice loss should be based on average splice loss.

Acknowledgements
This document was edited in collaboration with:

Fujikura U.K.

Fitel U.K.

Dirksen Opleidingen

Sumitomo U.K.

Simac Electronics

References
[1]

IEC 62316 TR Ed. 2.0: Guidance for the interpretation of OTDR backscattering traces.

[2]

ANSI/TIA/EIA-455-8 (May 2000): Measurement of Splice or Connector Loss and Reflectance Using an OTDR.

BendBright-XS is a trademark of Draka Communications. Other products and company names listed are trademarks or trade names of their
respective companies.

Draka Communications

Netherlands:

Tel: +31 (0)40 29 58 700

Fax: +31 (0)40 29 58 710

[email protected]

France:

Tel: +33 (0)3 21 79 49 00

Fax: +33 (0)3 21 79 49 33

www.drakafiber.com | www.draka.com

USA:

Toll free: 800-879-9862

Outside US: +1.828.459.9787

Fax: +1.828.459.8267

Find out more?

Draka Communications has offices and production facilities all over the world. To get in touch with the Optical Fiber segment and
find out how we can help you with your fiber needs, visit our website at www.drakafiber.com | www.draka.com or contact us at:

Netherlands
Draka Communications Optical Fiber

France
Draka Communications Optical Fiber

USA
Draka Communications Optical Fiber

Zwaanstraat 1
5651 CA Eindhoven

Parc des Industries Artois-Flandres

644, Boulevard Est BILLY BERCLAU
62092 Haisnes Cedex

2512 Penny Road

NC 28610 Claremont

Phone: +31 (0)40 29 58 700

Fax: +31 (0)40 29 58 710

Phone: +33 (0)3 21 79 49 00

Fax: +33 (0)3 21 79 49 33

Toll free: 800-879-9862

Outside US: +1.828.459.9787
Fax: +1.828.459.8267

Draka Communications
[email protected]
www.drakafiber.com | www.draka.com