Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
1~7~;89L
D-23.525
1 This invention relates to determining -the integrity of splices
2 in optical fibers, and more particularly to a method for measuring splice
3 loss that may be practiced solely at the location of the splice.
4 Method and apparatus for measuring the insertion loss of a
splice between one ends of a pair of optical fibers are described in the
6 article "Hot Splices of Optical ~aveguide Fibers" by Y. Kohanzadeh,
7 Applied Optics, March I976/Vol. 15, No. 3, pages 793-795. In accordance
8 with the reference, splice loss is defined as
9 10 log (l-Ps/Po) tl)
where P is a measure of light scattered from the splice and P0 is a
11 measure of light transmitted in the input fiber and incident on the
12 splice. Such a splice loss measurement technique has particular
13 advantage in field applications since it can~be practiced solely at the
14 location of~the splice. More specifically, it does not require access to
the other ends of the fibers, which may be spaced many kil~eters from
16 the splice. Applicants have recently discovered that in many~instances
17 such an indication of splice loss, which only utilizes a measurement of
18 light scattered from the splice itself, is not an accurate indication of
19 the true splice loss and the integrity of the splice. An ob~ect of this
invention is the provision of an improved method of measuring splic~ loss
:
21 that may be practiced solely at the location of the splice.
22 ~ ;Sammary of Invention
23 In accordance with this invention, an indication of the loss 10
.
24 log (1-PR/PO) cauaed by a splice between;ends of input and output
optical fibers~is obtained with a measurement PR of radiant power
26 radiated away ~rom the output fiber in at least a significant portion of
27 the length thereof extending downstream of the splice and exhibiting
';
:
~L~7~6~
V-23,525
substantial leaky mode radiation as a result of the splice, for a radiant
power PO in the input fiber and incident on the splice. l~he indication
PR is preferabl~ obtained by integrating radiant power scattered from
the splice and radiant p~er radiated from a significant portion of the
adjacent leaky mode section of output fiber.
Description of Drawings
This invention will be ~ore fully understood from the following
detailed description of preferred embodiments thereof, together with the
drawing in which parts are not drawn to scale. In the drawing,
~ IG. 1 is a schematic representation of apparatus useful in
practicing the ~.ethod of this invention, a plan view of an integrating
cylinder 25 of split block construction being sh~n here;
FIG. 2 is an enlarged side view of a portion of the integrating
cylinder prior to clamping the parts thereof together, and taken along
line 2--2 in FIG. l; and
FIG. 3 is curves which are semi-logarithmic plots of PR/Po
measured along the output fiber for a number of different pairs of
optical fibers that are spliced together, a curve and an associated
splice (not shown) being identified by the sa~e reference character.
Description of Preferred Embodiments
Referring na~ to FIG. 1, apparatus for practicing this inven~ion
comprises an input fiber 11 that is connected to a light source 21, an
output fiber 12 that is connected to a load 22, an integrating cylinder
25 of split block construction, and radiometer means 30. An optical
fiber bundle 28 connects the radiometer to an output port 27 of the
integrating cylinder. It is desirable to be able to produce an accurate
indication and/or measure of the insertion loss of a splice 15 between
the one ends llA and 12A of the fibers.
The integrating cylinder 25 is an enclosure that converts input
light into diffuse light in the interior thereof. Integrating enclosures
are manufactured by Labsphere of New London, New Ha~pshire. The
integrating cylinder 25 is split into halves 25A and 25B by a cutting
7~68~
D-23,52~
plane that is ort~logonal to the plane of the paper, along the line A--A
that includes the axes of the fibers. Flanges 32A and 32B extend around
the circu~ference of the open edges of the top 25A and bottom 25~ of the
cylinder, with a hinge 33 being attached to the rear of the flanges to
facilitate opening and closing the c~linder. Alignment holes and pins
(not shown) may be located along the flanges for providing precision
alignment of the interior surfaces of the cylinder parts. The two halves
of the integrating cylinder are secured together with clamps 35 on the
flanges to form a light tight enclosure.
The top 25A of the cylinder contains an input port 26 comprising
a socket receiving a fiber connector that is adapted for releasably
holding the one end llA of the input fiber prior -to forming the splice,
as is described more fully hereinafter. rrhe bottom of the cylinder
contains the output port 27 comprising a socket that is dimensioned for
receiving a ferrule that is attached to one end of the fiber bundle 28.
rrhe sockets of the input and output ports are oriented so that a light
ray emanating into the cylinder from a fiber in the input port will not
be directly incident on the end of a bundle fiber inside the cylinder.
Each of the cylinder parts includes axially-aligned spaced-apart
semicircular openings that extend through the wall thereof and which are
coaxial with the axis A--A of the cylinder. The semicircular openings,
e.g. openings 41A and 41B, mate for forming a circular opening that is
larger than the diameter of the fibers when the cylinder parts are
clamped together. Semicylindrical channels, such as the channels 43A and
43B in FIG. 2, a~e also forrned in the flanges coaxial with the cylinder
axis A--A and thus associated ones of the circular openings. Associated
channels also rnate when the cylinder parts are clarnped together. The
semicircular channels 43A and 43B, for example, are filled with inserts
45A and 45B of a resilient dielectric material such as a flexible
polyester polyurethane foam that extends above the tops c~ the associated
flanges 32A and 32B when the cylinder is open, see ~IG. 2. Shallcw
troughs 47 are cut into the facing surfaces of the foam inserts to
--3--
~3~7~68~
D-23,525
facilitate locating cormected optical fibers 11 and 12 along ~le cylinder
axis A--A when the cylinder is open. In operation, ~he cylinder is
closed for firmly sandwiching the fibers be~ween the foam inserts which
securely hold them in place without inducing compression modes therein.
The sides of the inserts that face into the cylinder are preferably
coated with a light reflective paint, as is the interior of the sphere.
In an optical com~lunication system, the input fiber 11 and
output fiber 12 are preferably of the same diameter and tJpe. m e one
end o~ the output fiber may be spaced many l~ilometers away from a load 22
which ~ay be an optical receiver. In such a system, the input fiber ~ay
be a pigtail on a light source 21 such as a laser diode. Alternatively,
the one end of the input fiber may be spaced a considerable distance from
an optical transmitter including such a light source. If the input fiber
is not associated with a light source in an existing system, the source
21 may be any suitable device that can be connected to the other end of
the input fiber. m e light source 21 is preferably similar, however, to
what will ultimately be used in a communication link including the
spliced fibers 11 and 12.
It has previously been suggested that the insertion loss for a
splice be determined according to the relationship in equation (1) where
Ps is the radiant power of light scattered primarily from only the
splice when light of a radiant power PO in the input fiber is incident
on the splice. m e measurement PO may be obtained prior to making the
splice by preparing the one end of the input fiber and inserting it into
the input port 26 of the integrating cylinder 25. m e light source 21 is
then energized for transmitting light along the input fiber and into the
closed integrating cylinder which converts it to diffuse light. Bundle
fibers 28 couple diffuse light to radiometer means which provides an
indication or measure PO of the radiant power of reference light
emitted from the free end llA of an input fiber, After the one ends of
fibers are joined together in a splice 15 using conventional techniques
such a fusion welding, epoxy splicing and flame fusion~ the spliced
--4--
D-23,525
fibers are set in grooves of the channel inserts with the splice 15
centered in the cylinder. The cylinder preferably has a very short
length such as two inches when it is used for measuring only light
scattered from t~le splice. The cylinder is then closed and the source 21
energized. This causes light of the intensity PO to be transmitted in
the input fiber and be indicent on the splice. I,ight that is scattered
by the splice is converted to diffuse light in the cylinder ~/hich is
coupled on bundle fibers to the radiometer means for producing the
indication Ps of the intensity of scattered light.
It has been discovered that the output fiber of a pair of
optical fibers connected in a poor splice supports leaky mode radiation
of a significant-measurable level over a length of approximately one foot
downstream of, i.e. moving away from, a splice. That is, in addition to
light in the input fiber being scattered by the splice, scme of the input
light is converted into leaky modes in the core and cladding of the
output fiber. This leaky mode radiation escapes through the
circu~ference of the output fiber and is radiated generally transversely
away from the fiber. It has been determined empirically that when fibers
are joined in a splice of poor quality, light scattered at the splice
itself rnay be only a small part of the total light lost as a result of
radiation from the output fiber and the splice. Stated dirferently, the
a~lount of leaky mode light radiated out o~ the fiber downstream of the
splice r~ay be greater than that scattered from the i~mediate vicinity of
the splice~
In accordance with one aspect of this invention, the insertion
loss of the splice 15 is determined from the relationship
lO log (l-PR/Po) (2)
where P~ is a rneasure of both the amount of light scattered from the
splice and the amount of leaky mode light radiated from a significant
length of output fiber adjacent the splice that exhibits substantial
lealcy mode ~adiation, for light of a radiant po~er PO in the input
fiber and incident on the splice. 1~1e rneasure PO rnay be obtained in
--5--
~7~8~ D-23,525
the rnanner described above. Prior to joining one ends of the fibers in a
splice sucll as with a fusion we]d, the protective plastic coating is
removed frorn the output fiber over a length of approxi~natel~ 1.5 feet
from the one end thereof. A length of the exposed fiber 12 that is
adjacent the completed splice is then set in the grooves of the channel
inserts with the splice located just inside the cylinder. qhe cylinder
preferably extends over a substantial portion of the length, e.g. 12
inches, of output fiber exhibiting leal~y mode radiation, and is securely
closed prior to energizing the light source for illuminating the splice
with light of a radiant power PO which generates leaky mode radiation
in the output fiber. The diameter of the cylinder is preferably srnall,
e.g., one inch. Light that is scattered from the splice and is radiated
from the circumference c~` the output fiber adJacent the splice is
converted to dif~use light and integrated or summed by the operation of
the cylinder. Diffuse light in the cylinder is coupled over bundle
fibers 2~ to the radiometer which provides the indication PR of total
light loss caused by the splice. The decibel value of splice loss is
obtained from equation (2).
This rnethod of indicating splice loss is preferred since it
takes into account most of the light scattered from the splice and/or
radiated away from the output fiber. The sarne relative results are
obtained with this technique in rating the quality of a r~Imber of splices
in optical fibers when the enclosure integrates light ernitted from the
splice and 9-inch and 12-inch lengths of do~nstream fiber 12. Good
results are also obtained with this technique when light scattered from
the splice and only 6 inches of downstream fiber is integrated by the
enclosure, the difference being that the order of rating some of the
fibers is shifted slightly.
In a method er,~odying another modified form of this invention,
the integrating enclosure extends over a limiked portion of the
downstream length of output fiber exhbiting substantial leaky mode
radiation, but not over the splice itself. This length of output fiber
--6--
~7~6~3~
D~23,525
may, by way of example, extend from 3 inches to 12 inches d~nstrearn Or
the splice. This alternate rnethod provides a good indication of splice
loss. It does not give the exact value thereof, however, since there is
some lost light (as a result o~ the splice) that is not 'caken into
consideration. Thi~s alternate method rates the quality of a number of
splices in optical fibers in a slightly different order from that
obtained with the preferred methods in whic~l the integrating enclosure
extends over the splice and the adjacent length of output ~iber.
In apparatus that was built and operated for practicing the
m.ethods of this invention, the integrating enclosure 25 was an
integrating sphere having an inner diameter of 3 inches. Reference to
equation (2) reveals that the insertion loss Q~ the splice is a function
of the power ratio PR/Po and that the highest quality splices have
the lowest power ratios. J~ order to provide a graphic representation of
light scattered from a splice and light radiated from lengths of leaky
mode output fiber 12, splices were formed in a number of fibers with no
special care being taken to obtain precision alignment of the ends of the
~ibers prior to making a splice. The emitted power PR scattered from a
splice and radiated from the immediately adjacent 3-inch length of output
fiber 1~ was measured by locating the associated fiber along the axis of
the sphere with the splice Just inside the sphere and illuminated by
light of an intensity PO in the input fiber. It will be recognized
that the measurement PR here is really an integral, since the sphere
integrates the light emitted from the splice and the length of output
fiber located in the sphere. The sphere was sequentially advanced along
the output fiber in 3-inch increments to obtain measurements of the
radiated power PR scattered by successive 3-inch lengths of output
fiber exhibiting leaky mode radiation as a result o~ the splice. FIG. 3
is curves of this power ratio as a function of do~nstream spacing along
the output fiber 12 for a number of di~ferent optical fibers that are
spliced together. The data is plotted on a semi-logarithmic scale which
compressés the plot in the vertical direction for convenience of
~7
1~71684 D-23,525
illustration. In describing this data, the curve and associated splice
(not shown) are designated by the same reference numeral. In these
curves, points at s=O and 3 inches represent values of PR/Po that are
obtained by integrating lig~lt emitted from the splice and the adjacenc
approximately 3 inches of output fiber 12, and from the length of output
fiber spaced 3 to 6 inches away from the splice, respectively.
Similarly, the plots at s=6 and 9 inches represent values o~ PR/Po
associated with light radiated from lengths of output fiher spaced 6 to 9
inches and 9 to 12 inches, respectively, from the splice. Leaky mode
radiation at points along the output fiber is of a significant level when
this power level is readily measurable and is of a ma~nitude having a
real effect on the overall-integrated value of PR/Po. By way of
example~ it is no longer of a significant value when it decreases by an
order of magnitude from the highest measured value thereof in the
downstream leng~h of output fiber. Consideration of the curves in FIG. 3
reveals that leaky mode radiation produced by all of the splices falls to
a relatively low level at a distance of approximately one foot away from
the splice. The lealcy mode radiation at points on all of the output
fibers in FIG. 3 falls to a level that is at least an order of magnitude
less than a previous high value thereof within a 15-inch do~nstream
length of output fiber. A value for PR/Po corresponding to the
integral of light emitted or lost fron the splice and the adjacent l-foot
length of output fiber ls obtained by summing the values of the power
ratio for s=O, 3, and 9 inches in FIG. 3.
The best rating of the splices represented by curves in FIG. 3
is obtalned in accordance with the preferred embodiment of this invention
in which the power ratio PR/Po is determined by integrating or
summing individual values thereof at~the splice ~and in the three adjacent
3-inch lengths of output fiber (i.e.~ a 12-inch length of output fiber)
that exhibit substantial leaky mode radiation. The rating of the
splices, in decending order of quality (i.e. from best to worst), is 51,
52, 53, 54, 55, 56, and 57. Consideration of these cur~es reveals that
-8-
~ ~ 7 ~ 6 ~ ~ 1}23,525leaky mode radiation in lengths of output fiber 12 downstrearn of the
splice rray be greater than light ernitted from the imrnediate vicinit~ of
the splice for both th~ best splice 51 and the worst splice 57
represented here. It is also clear from these curves that the total
leaky mode radiation from the length of output fiber dcwnstream of the
splice is in most instances considerably greater than that scattered by
the splice. This is sho~1n analytically by sumrning the power ratios
PR/Po at s=3~, 6 and 9 for curve 51, for exarnple, to obtain a power
ratio of 0.033 and comparing it with the corresponding power ratio of
o.oo6 associated with the splice 51 itself at s=O. This is not the case,
however, for the curve 55. It has been determined empirically that these
splices are rated in the same order when this power ratio is deterrnined
from light emitted from the splice and the adjacent 9-inch length of
output fiber exhbiting significant leaky mode radiation. The splices
were also rated in the same order by sumning the power ratios for the
splice and adjacent seg~nents of output fiber until the last measured
power ratio was down by factors of 10 and 5 frcn the previous highest
measurernent thereof. The order in which the splices are rated was only
slightly shifted when the resultant power ratio was determined frc~ the
splice and a 6-inch length of downstream fiber~ and from a length of
output fiber extending from 3 to 12 inches downstream of the splice but
excluding light scattered from the splice itself. The quality of the
splices is rated in a slightly different order when the power ratio
PR/Po is obtained from only lengths of output fiber spaced 3 to 6
inches, 6 to 9 inches, and 3 to 9 inches a~iay from the splice. Thus, it
is desirable to utilize a substantial and significant length of output
fiber supporting substantial leaky mode radiation in obtaining a value of
PR/Po for use in equation (2) and providing a rr.easure of splice loss.
A significant length of downstream fiber exhibiting significant lealcy
rnode radiation is also one for which the integral of light radiated from
the adjacent segment of output fiber is normally of a value not having a
measurable effect on the integral of light radiated from that one length
_g_
~7168~ D-23,525
of output fiber.
Although this invention is described in relation to preferred
embodiments thereof, other ~Jariations and modifications will occur to
those slcilled in the art. By way of example, the novel method is
applicable to radiant energy and light in other than the visible
electromagnetic spectrum. Thus, the words light and radiant energy as
used here mean both visible light and invisible radiant ener~y in the
high and low ends of the frequency spectrum including both ultraviolet
and infrared radiation. Also, the integrating enclosur2 may be of any
convenient shape other than cylindrical and spherical, although it
preferably has a regular shape. Further, the enclosure does not have to
be split into parts of the same size. Additionally, an optical detector
may be located directly in the output port 27 on the enclosure for
producing an electrical signal which is applied to an associated meter
for producing an indication of the intensity o~ light in the integrating
enclosure. Additionally, a single fiber may be located in the output
port for coupling diffuse light to the radiometer means. And if the
intensity of diffuse light coupled to the radiometer means is
insufficient to obtain clear and definite readings, the light source may
be pulsed or the output thereof mechanically chopped at a fixed
repetition frequency for producing a varying electrical current in the
radiometer means for increasing the sensitivity of the detector in a
manner which is well known in the art. The scope of this invention is
therefore defined by the appended claims, rather than the aforementioned
detailed descriptions of preferred embodiments thereof.
--10--
