AU592066B2 - Coupler for optical fibres - Google Patents

Description

AUSTRALIA

U(

Patents Act COMPLETE SPECIFICATION

(ORIGINAL)

Class It. Class Application Number: C7 s Lodged: Complete Specification Lodged: Accepted: Published: Priority *Related Art: Nam. s ofApi0n) APPLICAkNT'S REF.: ICI CASE NO. Z/QM.33160/AU IMPERIAL CHEMICAL INDUSTRIES PLC :Address(es) of Applicant(s): Imperial Chemical House Millbank, London SWI 3JF England vkctual Inventor(s):

S

000560 S S Eric Reginaid Cox Philipe Jacques Charles Le liaert A00 ddress for Service is: PHILLIPS, ORMONDE AND FITZPATRICK Patent and Trade Mark Attorneys 367 Collins Street Melbourne, Australia, 3000 Complete Specification for the invention entitled: COUPLER FOR OPTICAL FIBRES The following statement is a full description of this invention, including the best method of performing it known to applicant(s): P19/3/84 QM33160iv COUPLE~R FOR OPTICAL FIBRES This invention relates to the field of fibre optics and in particular to non-invasive coupling of optical fibres to enable light to be tiransferred from one or more fibres to one or more other fibres.

Optical fibres consist essentially of a transparent core surrounded by a cladding of lower refractive index than the core to irovide cotsistent propagation of light along the fibre by total internal refelection. Around c-d h fibre or bundle of fibres are usually One or more protective sheaths. In various optical fibre comrstunication applications it is desirable to transfer a predetermined'portion of light travelling along one fibre (the l"m~in fibre") to one s or more other fibres ("tap fibres"), or likewise to feed light from the'tap fibres into the main fibre.

Various invasive couplers are knowhi, but these require substantial skills, eg. to remove cladding and fuse individual cores together in a carefully controlled manner, or to be inefficient, e~g. through lose of light were fibres are broken and connected to a preformed cou,ling device, According to the p~a~ inig, n; oif fr-iy- *s optically coupling a main fibre to at least one tap *fiore having a light receptive portion comprise meanks to hold an intermediate por#-ioO of t main *fibre and to conform it to a predoteri ed path which 11 ~executes a Laend haying a radius o curvature see 0 sufficiently small to cause e of any light f lowing through the main fibre leak from the cladding aroiand the outer e of the bend, and means to hold the tap f re with its light, receptive portion positione o receive at least a portion of the light leak' from the cladding around the outer edge of the 1N1 -lbt Li

I

I

I

I

U

If '1 Li -4 According to the present invention there is provided S S S S S 5 S St 5555 S 5 1 S S S is its t 1,551 it St S S C IS C S '~C I S I S S 5 Cf 55 4 1 St S it S 5 a device for optically coupling a main optical fibre comprising a core surrounded by cladding of lower refractive index, and having a protective layer surrounding the cladding, to at least one tap fibre having a light receptive portion in the form of an exposed end face, the device comprising means to hold an intermediate portion of the main fibre with cladding and protective layer intact and to conform it to a predetermined path which executes a bend having a radius of curvature sufficiently smalil to cause some of any light flowing through the main fibre to leak from the cladding around the outer surface of the bend; the holding means comprising means to hold the end portion of the tap fibre with its end portion adjacent the outer surface of the bend to receive at least a portionof the light leaking from the cladding around the outer surface of the bend, in which the means for holding the main fibre comprises a first assembly having a mating surface positioned with respect to said predetermined path such that the cladding on the outer surface of the bend in the main fibre lies flush with the mating surface, and the means for 20 holding the tap fibre comprises a second assembly lockable against the mating surface in a predetermined position wherein the light receptive portion of the tap fibre is positioned. to receive at least a portion of the light leaking from the cladding around the outer surface of the bend in the main fibre, and including a stripping blade fixed to the second assembly which blade can t. slid along the mating surface to strip a portion of the protective layer from the outer surface of the bend which is proud of the mating surface to expose the cladding as the second assembly is moved to its lockable position.

WNENEENN

-2- 0 ti 15 t ti c ft t I t I2 r r 20 Ici We find that by bending the supply fibre sufficiently to obtain a substantiai leakage of light (determinable by measuring the accompanying decrease in the amount of light transmitted);by clamping or otherwise securing the two fibres accurately in predetermined positions, very reproducible proportions of light in the supply fibre can be' bled into the tap fibre. Moreover, our preferred devices according to this invention can achieve such reproducibility with very high efficiencies, and, as the technique is non-invasive, the main fibre remains continuous throughout its passage therethrough with no more preparation required than removal of a small area of any protective layers it may have (but not of cladding). This can be achieved by factory floor mechanics without specialist fibre optic skills, and can be achieved without any residual damage to the otpical fibre.

A simply constructed device according to the invention is one for use with at least one tap fibre having an end portion in which the light receptive portion is in the form of an exposed end face, the means for holding the tap fibre comprising means to hold the end portion thereof with its exposed end face adjacent to the outer surface of the bend in the main fibre thereby to enable it to receive portions of light leaking from the bend when light is passed through the main fibre.

A preferred three port coupler is one in which the means for holding the main fibre comprises a first assembly having a mating surface and the means for holding the tap fibre comprises a second assembly lockable in a predetermined position against the ii I r C Cr 1 C l~t

C

1111CC I: C

I-

~-~---_ILIIIIII. I

-I

-3t (C Ct C C rC C CI Ctt: mating surface, the configuration of. the first assembly being such as to hold the main fibre with its cladding on the outer surface of the bend lying flush with the mating surface, whereby removal of any protective layers proud of the mating surface will expose said cladding. The first assembly could be formed of a single block drilled to accommodate the main fibre, but we prefer it to be 4plittable to provide opposing cleavage faces at least one having a configuration for receiving and holding the main fibre in a predetermined position, e.g. by having a channel or series of pegs into or around which the fibre may be located, thereby to facilitate absembly, the two parts interlocking or being hinged or otherwise held 15 together to provide a secure assembly. For convenience we prefer to provide the second assembly with a stripping blade which can be' slid along the mating surface to strip the protective layers from the bend in the main fibre, and hence avoid any need for the operator to have his own separate stripper. A particularly preferred device is one in which the second assembly is locked in the predetermined position by sliding it along the mating surface, the stripping blade being located ahead of the tap fibre 25 in the direction of sliding whereby stripping of any protective layers proud of the mating surface is effected automatically when locking the two assemblies together.

This three port device is preferably 30 proportioned such that when light travelling through the main fibre approaches the bend:along a first axis, the end portion of the tap fibre ip held along a second axis approximately parallel to the first axis cC C.

C Ci

(((CCC

C

r .r I Cr;; r d tr 1~

KT

~1 -4- C is1 ft t t i: 2 t; 20 but displaced from the first axis in the same direction as the bend in the main fibre.

This displacement is preferably sma~Il, i.e. less than the diameter of the supply fibre. The tap fibre will then rnormally only receive light from the main fibre when that light has travelled along the main fibre towards the bend along said first aiis and not from the other direction. When the axes are displaced in this manner, we prefer the exposed e., 1 d face to be set at an acute angle to the main axis of the tap fibre it terminates, preferably within the range 15 to 350.

These variable parameters, i.e. end face angle, axis displacement and interaxis angle, are all interrelated in their effects on the efficiency of the device, and the selection of specific values for each will affect the optimum values for the other two.

Besides the above three port coupler, another particularly useful device for coupling tap fibres through exposed end faces, is a directional four port coupler comprising means to hold two tap fibres with their exposed end faces adjacent t6 the outer surface of the bend in the main fibre but positioned substantially symmetrically with respect to said bend thereby to enable each to receive light travelling along the main fibre in the opposite direction to that received by the other fibre. Light injected through the two tap fibres would likewise be caused to travel along the ianin fibre in opposite directions, and light injected by one would travel along the main fibre in the same direction as light which can be accessed by the other. Such a device, thus provides a very convenient way of coupling a communications V CC C C

C

C C

CCC

C C 4

I

IIr i

I

ii 1

V

t 15 r tr rt IT fC tI C 20 station non-invasively into a ring main or data bus communications link, having a detector connected by one tap fibre and a source (with itj attendant modulator) connected to the other.

The amount of light which is caused to leak through the cladding of the main fibre depends on the radius of curvature of the bend, th6 smaller the radius of curvature the greater is the amount of light leaked. Hence, for example, to divide the light into two equal portions (although smaller through losses), requires a bend having a smaller radius of curvature than when only one tenth of the supply is required to be tapped. However, the actual curyature required will depend on the size and construction of the fibre, but will in any case be substantially smaller than the normal minimum radius usually quoted in respect of commercial fibres. 'For example, for dividing the light into two equal parts, the radius of curvature may typically be about 2 mm, and it is desirable to use such devices with optical fibres which do not degrade through stress cracking under such circumstances. In general, therefpre, we prefer to use these devices with polymer fibres clad polymethylmethacrylate fibres) rather than glass fibres, although some glass fibres seem to withstand such bending.

The angle through which the main fibre is bent is a factor in the overall efficiency, in that the larger the angle the wider is the 'spread of leaked light, and the smaller is the portion of light that can be caught by the fibre. However, at lower angles, especially below about 20*, very little light will be c.

C. C

'I

C C C( P~

CC

C C C

C

Lf L. I i leaked. Thus although there are no precise limits to the range T of bend angles that will leak light (given a suffici-'ently small radius of bend curvature) a practical range is about 20 to 900, with our preferred range being 35 to 500.

It should be noted here that where an apparatus comprises several such devices with a plurality of tap fibres extracting light from a common main fibre, these may be arranged so close together that their respective bends in the main fibre are lilt"merged into a single extended bend. Under these conditions the bend may extend to angles even greater than 90' as the light which could not be caught by a single tapping fibre becomes available for the adjacent tapping fibres arrayed around the extended bend, and thereby maintains good efficieny.

Efficiency can also be improved by providing index matching gel between the bend and the light receptive portion and heat treating the supply fibre at the bend to anneal the fibre may also improve efficiency.

It is not essential to use a preformed light-receiving portion, as this can be formed in situ in non-invasive manner by bending an intermediate portion of the tap fibre in essentially the same manner as the main fibre. This can be t r achieved in a device in which the means for holding the tap fibre is desinged to cause an intermediate portion of the tap fibre to conform to a predetermined path which executes a bend having a radius of curvature sufficiently small to cause some of any light flowing through the fibre to leak from the cladding around the outer edge of the bend, and to hold S. adjacent to each other the outer edges of the bends in the two fibres whereby light leaked from one fibre can be caught by the other. As may be realised, in such a configuration there need be no asymmetry to identify one fibre as the main fibre and the other as the tap.

The invention is illustrated by reference to specific embodiments shown in the drawings, in which I Figure 1 shows in elevation a body portion ofA-o~.bzimntfrom which a cover plate has been removed to show internal details, Figure 2 shows a main fibre and a tap fibre in positions corresponding to those in which they are held in use, but spaced rl S Q-6- 4, Ii

LI

ij apart slightly for clarity, Figure 3 is a longitudinal section through an assembled coupler with a tap plug poised for insertion, Figure 4 shows part of the coupler of Figure 3 holding a main fibre, Figures 5-8 show various arrangements of fibres as they can be held in appropriate three and four port couplers according to the invention, Figures 5a-8a and 5b are diagrammatic representations showing how the coupler forming the configuration of Figures 5-8 can be used in practice, Figure 9 shows how a plurality of the couplers of Figure 6 can be used in a unidirectional communications ring main, and Figure 10 shows how they can be used in a bidirectional S data bus.

IL

The body portion 1 shown in Figure 1 is fabricated from a U block of polymethyl methacrylate, in the surface of which are milled three interconnecting C C c, -7-

L

7Z -8ti

I

I,

1, if 1,

I

k 10 15 t C 4' 20 IC t1 channels, 2.2 mm wide and deep. Of these a first channel 2 has an axis which is parallel to but displaced by a very small amount by 0.2 to mm for this size of fibre) from thal of' a second channel 3. The third channel 4 is formed at an angle of 50' to the other two, and the inside edge of the obtuse angle formed between the first and third channels has an accurately formed 1 ir. radius of curvature, provided 1y inserting a 2 mm diameter rod after milling of the channels. The cover plate (not shown) is simply a flat plate of polymethylmethacrylate which is bolted to the body portion through holes 6, 7 thereby closing off the three channels to form tubes.

The optical fibres 11, 12 of Figure 2 are 1 mm diameter polymethyl methacrylate fibres coated with fluoropolymer cladding, and around this is a protective sheath, giving a total diameter of about 2.2 mm. In use the main fibre 11 has ,a portion of protective sheath removed to expose the fibre 13, complete with its cladding. It is then fitted into the first and third channels thereby forming a tight bend 14 around the rod 5 with the exposed portion of cladding forming tho outer side of the bend. The tap fibre 12 has an exposed end portion 15 slanted at an angle of 300, and is pressed into the second channel with the slanting end face facing upwards in the illustrated orientation, and located adjacent to the outer side of the bend in the main fibre, this being formed by the exposed cladding 13. All the space within the junction of the channels is then filled with index matching gel, this beiqg an option to improve efficiency. The cover plate is then bolted onto the front of the body portion to hold the fibres in place.

I C C tC CC C IC .C CC I I.

Jr

II

1 L F~~rr~ -9- Although this embodiment has'been described as carried out in a particular order, this order need not be adhered to, and where a sensor or other optical device is to be provided with light from a ring main fibre via a tap fibre, it could weil be marketed with a device according to the invention already secured to the end of the tap fibre, ready to,be clamped on an optical fibre ring main at the point of use. This f provides a substantial practical advantage over other coupling devices for such an application in that because there is no need to break into or otherwise degrade the main fibre, the point of use can easily be moved along supply fibre as required, without any loss of efficiency.

Although this embodiment has been described in i *terms of removal of light from a ring main, it can also be used as a feeding coupler, for supplying light to the ring main via the tap fibre. The following two Examples are provided to illustrate how this kind of embodiment can be used in practice and to give results we have obtained with two such devices.

EXAMPLE 1 A coupling device for tapping out half of the light in a main fibre, was assembled as described in respect of the above specific embodiment, with the main fibre being bent through 50° around a 2mm diameter rod, giving a radius of curvature at the centre of the core of about 2 mm, and using a gel of refractive index 1.46. When 1 mW 9 f visible light was fed into the supply fibre where it extended from the first channel, and about 400 pW of light emerged from the other end of the main fibre and also from the tapping fibre. This gave an excess loss of less than 1.5 dB.

i is EXAMPLE 2 A further device was assembled which was I) substantially the same as that used in the preceeding example except that the main fibre was bent through L400 around a curvature-defining rodiof '10 mm diameter.

5 This c;ave a radius of curvature at he centre of the core of about 6 mm. This greater radius of curvature j caused less light to leak, and the ratio of light emerging from the tap fibre to that continuing along the main fibre was only about 1 10, that emerging from the main fibre being about 750 FW and that from i the tap fibre about 75 pW. The excess loss was thus less than IdB.

Figures 3 and 4 show an injection moulded three port coupler comprising a first assembly 21 for holding a main fibre 22 and a second assembly 23 for holding a tap fibre (not shown). The first assembly has a mating surface 24 with two longitudinal under cut ridges and the second assembly is correspondingly P shaped with grooves,(not shown) to receive the ridges and enable one to slide along the other while holding the second assembly against the mating surface of the first. At one end is an end stop 25 with an integral resilient latch 26, such that when the second assembly ;is slid along the first until it abuts the end stop

(CC

S 25 25, the latch on the latter engages a latching recess C C E 27 on the former, and thereby locks the two assemblies ~together.

C t tThe first assembly is split longitudinally with an integral hinge 28 along one edge. In one part is a channel 29 to receive the optical'fibre 22, and the other part locks against it to hold the fibre securely in place. The channel has two straight portions I, meeting at an angle at the mating surface. The fibre is shown only in Figure 4, and this has been provided 35 to show how the fibre (complete withits protective

I

.1 [-ft t' -11sheath) is held by the channel with'a tight bend 31, the protective sheath 32 around the bend standing proud of the mating surface. In the second assembly there is also formed an integral stripping blade 33 (shown in isolation in Figure 4) which'slides along the mating surface during assembly of the coupler and strips the protective sheath from the fibre around the bend, so as to expose the cladding. Ahead of the blade is a recess 34 to receive the stripped sheath material.

The second assembly has a passage 35 drilled through it at an angle, parallel to one of the arms of the channel in the first assembly, the two being very nearly coaxial, as shown in Figure'3 and in the prototype shown in Figure 1. The passage is drilled to receive a plug 36 of varying diameter (this being a third plastics moulding), which in turn is drilled to S"receive a tap fibre (not shown) with its sheath I' removed from its end portion to fit the plug.

|i 20 Extending on both sides of the plug are latching S\.plates 37 which, when the plug is inserted into the assembly, lie on either side of the latter, locking the plug into position and preventing it rotating.

The narrow end of the plug has a slanting end face 38, and the tap fibre is shaped to lie flush with this.

SWhen the plug is located in the assembly, it holds its A end face and hence also the end of the tap fibre against the exposed cladding of the main fibre.

Figure 5 shows a main fibre, 51,52 and a single tap fibre 53, held as described above in connection S" with Figures 1-4, but showing variable parameters St Cwhich can be preselected, according to the S' configuration of the coupler, to provide the desired j characteristics, e.g. a 1/1 splitter or a 10/1 splitter, the figures representing the relative I. i. i I i 4 proportions of li~ In the drawing the matching gel, and Ob bend ar Rc radius ac end fac et viewini Such three splitter or as a c as shown in FigurE parameters are se~ j results are as fo~ As a spliti ?xcess losg As a combi' 20 (crosstallk When used resu:Lts are CC 25 ti C Excess losi iJ tt:: Figure 6 s fibres 63,64 held of which there ap CC 30 conventional coup is eihown diagramm, -12iht Output, ignorin g ir.2Lhereft losses.

shaded portion represents index the variable parameters are as igl1e of curvature ~e angle angle port couplers can be used as a ~ombiner, and this can be represented ~s 5a and 5b. When th6 above variable Lected to give a 50/5O split, typical L lows: Ler A to B 38% A to C ft38% Total 76%' iis typically 1.2 dB.

-xer, inputting 100% at 5 and at C B to A C to A )B to C and C to B: less than as a 10% tap as shown in 5b, typical A to B 73% "Ato C 7% Total so0.

B is typically hows a main fibre 61,62 and two tap in a directionai four port couopler, pears, to be no 'Krown' equivalent among lers for glass fibres. This coupler atically in Figure 6a, typical 4 C C:I J J -13results being as follows With 100% input of A With 100% input at D: A to B 62% D to A 0.5% (crosstalk) A to C 16% D to'B -f A to D t 40.5% (crosstalk) D toC 1 (crosstalk) 78% excess loss 1.1 dB total loss 1 10.0 dB Such directional couplers are particularly suitable for use with a communication ring main (local area network), to inject light into and tap light from the optical ring main. So far we have effectively used up to eight such couplers around a single ring main, for transferring data between various peripherals and the ring main.

Figure 7 shows a main fibre 71,72 and two tap fibres 73,74 formed from the same fibre, held in the configuration of a transmissive star coupler. This configuration is symmetrical, there being no essential [difference between the main fibre and the fibre forming the taps. This coupler is shown diagrammatically in Figure 7a, and has essentially the same configuration and the same effect as known star couplers designed for coupling the ends of four glass fibres, except that this is non-invasive, the two fibres being removable from the coupler without any optical deterioration, the only loss being of a short length of protective sleeve.

.o Typical results are as follows, when 100% of light is X •input at A A to B 67% A to C A to D 40.1% (crosstalk), 72% I, excess loss 1.4 dB

I,

-14- Figure 8 shows, a multiple tap, with a main fibre 81,82, and two tap fibres 83,84 positio ned and aligned so that both will receiv e light that hbis approached the bend from the sme direction, i.e. along;portiol 81 of the main f.ibre and not along portion 82. This is also shown diagrammatically in F'igure Ba, and it can be used as a splitter, e.g. as a demultiplexer, 6r as a combiner, e.g.

as a multiplexer. Typical results are as follows, where each input is 100%.

Splitter (demultiplexer) A to B A to C X to D excess loss 1.0 dB Combiner (multiplexer) B to A: 40% C(D) to At INB to C :10.35% (cropstalk) C(D) to B (crosstalk) B to D :10.35% (crosstalk) C(D) to D(c) 1%(cros-stalk) 40% total 1099 4.0 dR* total loss 7.0 dB.

in Figure a plurality of, couplers 91. of the iJ kind shown in Figure 6 non-invasively access a ring main 92. The ring main in each case is* inserted into thecoupler with its cladding intact, and corresponds to

J.

reference numerals 61 and 62 of Figure 6. The two tap p fibres corresponding to reference numerals 63 and 64 are connected respectively to a light source S (having means to modulate the light according to the data to be S 30 carried) and a detector D. In each case the couplers are orientated such that light injected by the various sources travels 4nticlockwise (in the view shown in Figure 9) around the single fibre ringmairn, this also being the direction which can be picked up for transmission to the detectors.

Figure 10 shows diagramimatically how the invention can be used to feed and access a bidirectional commtunications link in the form of a duplex fibre 101, 102. This requires 'two devices 103, 104 according to the invention, one to access the up fil~re .101 and the other to access the down fibre 102, and they each operate precisely as described above for those accessing the ring main, each having its own source S arid detector D.

lj However, the two detectors anid the two sources use a ii 10 conmmon electrical im~put 105 and a qommon electrical output 106 respectively, and are thus most conveniently b formed into a single unit as indicateA by the hatched line 107.

4' Figures 9 and 10 both illustrate a significant advantage over previously known invasive couplers. in the latter, where the main fibre is chopped, terminated and the terminated ends inserted inito a coupler, any malfunction occurring at any one coupler can affect the whole ring main or data bus, generally leading to a, total shut down of the whole system. Devices of the present invention are non-invasivel, so the complete fibre A continues through the coupler, and should the latter if become loose or otherwise fail, the remainder of the ring main (or data bus) with its accessing stations can continue to function amongst themselv'es. Similarly C C r Cadditional stations can be coupled into the main fibre r without interrupting service to t1jose. already in operation.

C C

Claims (2)

1. 2. A device as claimed in claim 1 wherein locked in the predetermined position by mating surface, the stripping blade being the second assembly is sliding it along the located ahead of the 7'T777~ r i VP 4 17 tap fibre in the direction of sliding whereby stripping of any protective layers proud of the mating surface is effected automatically when locking the two assemblies together.
2. 3. A device as claimed in claim 1 comprising means to hold two tap fibres with their exposed end faces adjacent to the outer surface of the bend in the main fibre but positioned substantially symmetrically with respect to said bend thereby to enable each to receive light travelling along the main fibre in the opposite direction and that received by the other fibre. 0 0 0 °r o 4. A device as claimed in claim 1 wherein the exposed end face S: is set at an angle of 15-350 to the main axis of the tap fibre it S. terminates. o 0 A device according to Claim 1 substantially as herein b4 cuA 1o \o described with reference to any one or more of Figs -4et-1 of the accompanying drawings. 0000 0o a 0 o0 DATED: 28 August, 1989. o PHILLIPS ORMONDE FITZPATRICK SAttorneys for: IMPERIAL CHEMICAL INDUSTRIES PLC. 4" 1 f 4 /1