INTERNATIONAL TELECOMMUNICATION UNION

Transcription

1 INTERNATIONAL TELECOMMUNICATION UNION CCITT G.651 THE INTERNATIONAL TELEGRAPH AND TELEPHONE CONSULTATIVE COMMITTEE (11/1988) SERIES G: TRANSMISSION SYSTEMS AND MEDIA, DIGITAL SYSTEMS AND NETWORKS Testng equpments Transmsson meda characterstcs Optcal fbre cables Characterstcs of a 50/15 µm multmode graded ndex optcal fbre cable Reedton of CCITT Recommendaton G.651 publshed n the Blue Book, Fasccle III.3 (1988)

2 NOTES 1 CCITT Recommendaton G.651 was publshed n Fasccle III.3 of the Blue Book. Ths fle s an extract from the Blue Book. Whle the presentaton and layout of the text mght be slghtly dfferent from the Blue Book verson, the contents of the fle are dentcal to the Blue Book verson and copyrght condtons reman unchanged (see below). In ths Recommendaton, the expresson Admnstraton s used for concseness to ndcate both a telecommuncaton admnstraton and a recognzed operatng agency. ITU 1988, 007 All rghts reserved. No part of ths publcaton may be reproduced, by any means whatsoever, wthout the pror wrtten permsson of ITU.

3 Recommendaton G.651 The CCITT, consderng that CHARACTERISTICS OF A 50/15 µm MULTIMODE GRADED INDEX OPTICAL FIBRE CABLE (Malaga-Torremolnos, 1984; amended at Melbourne, 1988) (a) graded ndex multmode optcal fbre cables wll be used wdely n future telecommuncaton networks; (b) the foreseen potental applcatons may requre multmode fbres dfferng n: nature of materal, geometrcal characterstcs, operatng wavelength regon(s), transmsson and optcal characterstcs, mechancal and envronmental aspects, (c) Recommendatons on dfferent knds of multmode fbres can be prepared when practcal use studes have suffcently progressed, recommends a graded ndex, multmode fbre, whch may be used n the regon of 850 nm or n the regon of 1300 nm or alternatvely may be used n both wavelength regons smultaneously. Ths fbre can be used for analogue and for dgtal transmsson. Its geometrcal, optcal, and transmsson characterstcs are descrbed below. The meanng of the terms used n ths Recommendaton s gven n Annex A and the gudelnes to be followed n the measurements to verfy the varous characterstcs are ndcated n Annex B. Annexes A and B may become separate Recommendatons as addtonal multmode fbre Recommendatons are agreed upon. 1 Fbre characterstcs The fbre characterstcs dealt wth n 1 are those whch ensure the nterconnecton of fbres wth acceptable low losses. Only the ntrnsc fbre characterstcs (not dependng on the cable manufacture) are recommended n 1. They wll apply equally to ndvdual fbres, fbres ncorporated nto a cable wound on a drum, and fbres n nstalled cables. 1.1 Geometrcal characterstcs of the fbre Core dameter The recommended nomnal value of the core dameter s 50 µm. The core dameter devaton should not exceed the lmts of ± 6% (±3 µm) Claddng dameter The recommended nomnal value of the claddng dameter s 15 µm. The claddng dameter devaton should not exceed the lmts of ±.4% (±3 µm) Concentrcty error The recommended concentrcty error should be less than 6% Non-crcularty Core non-crcularty Fasccle III.3 Rec. G.651 1

4 The recommended core non-crcularty should be less than 6% Claddng non-crcularty The recommended claddng non-crcularty should be less than %. 1. Optcal propertes of the fbre 1..1 Refractve ndex profle For fbres dealt wth n ths Recommendaton, the refractve ndex profles are expected to be near parabolc. 1.. Numercal aperture The optmum value of the numercal aperture (NA) wll depend on the partcular applcaton for whch the fbre s to be used and n partcular on the source couplng effcency requred, the ncreased attenuaton due to mcrobendng effects whch can be tolerated, and the overall baseband response requred. Values commonly employed n practce le wthn the range Whatever actual value s employed should not dffer from the chosen nomnal value by more than Materal propertes of the fbre Fbre materals The substances of whch the fbres are made should be ndcated. Note Care may be needed n fuson splcng fbres of dfferent substances. Provsonal results ndcate that adequate splce loss and strength can be acheved when splcng dfferent hgh-slca fbres Protectve materals The physcal and chemcal propertes of the materal used for the fbre prmary coatng, and the best way of removng t (f necessary), should be ndcated. In the case of a sngle-jacketed fbre smlar ndcatons shall be gven. Factory length specfcatons Snce the geometrcal and optcal characterstcs of fbres are barely affected by the cablng process, wll gve recommendatons manly relevant to transmsson characterstcs of cable factory lengths. Transmsson characterstcs depend greatly on the wavelength used to convey the nformaton. Envronmental and test condtons are paramount and are descrbed n the gudelnes for test methods. The transmsson characterstcs of fbres wll have a statstcal probablty dstrbuton whch wll be a functon of the desgn and manufacturng processes. The specfcaton of lmts for the transmsson characterstcs must therefore take ths dstrbuton nto account. For nstance for certan applcatons a partcular lmt may not embrace 100% of the producton and ndeed may only represent a very small fracton of the total producton. Economc consderatons wll play a large part n the decdng of specfcaton lmts for partcular applcatons..1 Attenuaton coeffcent Optcal fbre cables covered by ths Recommendaton generally have attenuaton coeffcents n the 850 nm regon below 4 db/km and n the 1300 nm regon below db/km. Note 1 The lowest values of the attenuaton coeffcent depend on the fabrcaton process, fbre composton and fbre and cable desgn; values n the range of -.5 db/km n the 850 nm regon and db/km n the 1300 nm regon have been acheved. Note In certan cases, fbres could be used n both wavelength regons.. Baseband response The baseband response ncludes both modal dstorton and chromatc dsperson effects. For certan applcatons the effects of chromatc dsperson are neglgble and can be gnored. Fasccle III.3 Rec. G.651

5 The baseband response s presented n the frequency doman. Admnstratons wshng to use the tme doman wll stll be able to do so by means of mathematcal operatons. For ths purpose the ampltude and phase response should be avalable. By conventon, the baseband response s lnearly referred to 1 km...1 Modal dstorton bandwdth: ampltude response The modal bandwdth ampltude response s specfed n the form of 3 db optcal ( 6 db electrcal) ponts of the bandwdth of the total ampltude/frequency curve corrected for chromatc dsperson. A more complete curve of the total bandwdth response should also be gven. Optcal fbre cables covered by ths Recommendaton generally have normalzed modal dstorton bandwdths greater than 00 MHz. km n the 850 nm regon and n the 1300 nm regon, but not necessarly smultaneously. Note 1 The upper values of the normalzed modal dstorton bandwdth depend on the fabrcaton process, fbre composton and fbre and cable desgn; values greater than 1000 MHz. km n the 850 nm regon and 000 MHz. km n the 1300 nm regon have been acheved. Note In certan cases, fbres could be used n both wavelength regons... Modal dstorton bandwdth: phase response No recommended value s gven as phase response nformaton s only requred n specal cases...3 Chromatc dsperson When requred, the manufacturer of the optcal fbres should ndcate the chromatc dsperson coeffcent values of the fbre type n the operatng wavelength regon(s). The test method s contaned n Annex B, secton V, to Recommendaton G.65. Note 1 For multmode fbres the domnant chromatc dsperson mechansm s materal dsperson. Note Typcal values of the chromatc dsperson coeffcent for hgh grade slca optcal fbres are the followng: Wavelength (nm) Chromatc dsperson coeffcent [ps/(nm.km)] Elementary cable sectons An elementary cable secton as defned n Recommendaton G.601 (term 1008) usually ncludes a number of splced factory lengths. The requrements for factory lengths are gven n of ths Recommendaton. The transmsson parameters for elementary cable sectons must take nto account not only the performance of the ndvdual cable lengths but also, amongst other factors, such thngs as splces, connectors (f applcable) and mode couplng effects whch can affect bandwdth and attenuaton. In addton the transmsson characterstcs of the factory length fbres as well as such tems as splces and connectors etc., wll all have a certan probablty dstrbuton whch often needs to be taken nto account f the most economc desgns are to be obtaned. The followng sub-paragraphs n ths secton should be read wth ths statstcal nature of the varous parameters n mnd. 3.1 Attenuaton The attenuaton A of an elementary cable secton s gven by m A = α n L n + a s x + a c y n = 1 Fasccle III.3 Rec. G.651 3

6 where α n = attenuaton coeffcent of nth fbre n elementary cable secton, L n = length of nth fbre, m = total number of concatenated fbres n elementary cable secton, a s = mean splce loss, x = number of splces n elementary cable secton, a c = mean loss of lne connectors, y = number of lne connectors n elementary cable secton f provded. Note 1 The losses a s and a c of splces and lne connectors are generally defned n equlbrum mode dstrbuton condtons. In operatng condtons apprecable dfferences may occur. Note The above expresson does not nclude the loss of equpment connectors. Note 3 In the overall desgn of a system, allowance must be made for a sutable cable margn for future modfcatons of cable confguratons (addtonal splces, extra cable lengths, ageng effects, temperature varatons, etc.). Note 4 The mean loss s taken for the loss of splces and connectors. The attenuaton budget used n desgnng an actual system should account for the statstcal varatons n these parameters. 3. Baseband response (overall 3 db optcal bandwth) The baseband response s gven n the frequency doman and ncludes the effects of both modal dstorton and chromatc dsperson and can be represented by the expresson: B T = [ B modal + B chromatc ] 1 where B T = overall bandwdth (ncludng modal dstorton and chromatc dsperson), B modal = modal dstorton bandwdth, Bchromatc = chromatc dsperson bandwdth (see Note 3). Note 1 Both the fbre modal dstorton baseband response and the source spectrum are assumed to be Gaussan. Note For certan applcatons the effect of chromatc dsperson s neglgble, n whch case chromatc dsperson can be gnored. Note 3 Bchromatc, the chromatc bandwdth, s nversely proportonal to the secton length and, f the source spectrum s assumed to be Gaussan, can be expressed as: Bchromatc (MHz) = ( λ D(λ) 10 6 L/0.44) 1 where λ = FWHM source lne wdth (nm), D(λ) = chromatc dsperson coeffcent [ps/(nm km)], L = secton length (km) Modal dstorton bandwdth The modal dstorton bandwdth values for ndvdual cable lengths n an elementary cable secton are obtaned from the relevant fbre specfcaton. However, the overall modal dstorton bandwdth of the elementary cable secton may not be a lnear addton of the ndvdual responses due to mode couplng and other effects at splces and, sometmes, along the length of the fbre. The modal dstorton bandwdth for an elementary cable secton s therefore gven by: 4 Fasccle III.3 Rec. G.651

7 Bmodal total = 1 x γ γ modal B 1 n where B modaltotal = overall modal dstorton bandwdth of an elementary cable secton, B modaln = modal dstorton bandwdth of nth fbre n elementary cable secton, x = total number of concatenated fbres n elementary cable secton, γ = modal dstorton bandwdth concatenaton factor. Note The value of γ, the modal dstorton bandwdth concatenaton factor, s typcally n the range 0.5 to 1.0 dependng on the effects of mode couplng at splces, alpha profle compensaton, wavelength of maxmum bandwdth etc. Values below ths range can also be obtaned n certan crcumstances. For a gven fbre, the approprate value of γ whch should be employed can be emprcally derved, and can usually be obtaned from the fbre/cable manufacturer. ANNEX A (to Recommendaton G.651) Meanng of the terms used n the Recommendaton A.1 alternatve test method (ATM) A test method n whch a gven characterstc of a specfed class of optcal fbres or optcal fbre cables s measured n a manner consstent wth the defnton of ths characterstc and gves results whch are reproducble and relatable to the reference test method and to practcal use. A. attenuaton coeffcent In an optcal fbre t s the attenuaton per unt length. Note The attenuaton s the rate of decrease of average optcal power wth respect to dstance along the fbre and s defned by the equaton: where P(z) = P(0) 10 (αz/10) P(z) = power at dstance z along the fbre, P(0) = power at z = 0, α = attenuaton coeffcent n db/km f z s n km. From ths equaton the attenuaton coeffcent s α = Ths assumes that α s ndependent of z. 10 log 10 [ P ( z) / P(0) ] z A.3 bandwdth (of an optcal fbre) That value numercally equal to the lowest frequency at whch the magntude of the baseband transfer functon of an optcal fbre decreases to a specfed fracton, generally to 3 db optcal ( 6 db eletrcal), of the zero frequency value. Note The bandwdth s lmted by several mechansms: manly modal dstorton and chromatc dsperson n multmode fbres. Fasccle III.3 Rec. G.651 5

8 A.4 chromatc dsperson The spreadng of a lght pulse n an optcal fbre caused by the dfferent group veloctes of the dfferent wavelengths composng the source spectrum. Note The chromatc dsperson may be due to one or more of the followng: materal dsperson, wavegude dsperson, profle dsperson. Polarzaton dsperson does not gve apprecable effects n crcularly-symmetrc fbres. A.5 chromatc dsperson coeffcent The chromatc dsperson per unt source spectrum wdth and unt length of fbre. It s usually expressed n ps/(nm km). A.6 claddng That delectrc materal of an optcal fbre surroundng the core. A.7 claddng mode strpper A devce that encourages the converson of claddng modes to radaton modes. A.8 core The central regon of an optcal fbre through whch most of the optcal power s transmtted. A.9 core area For a cross secton of an optcal fbre the area wthn whch the refractve ndex everywhere (excludng any ndex dp) exceeds that of the nnermost homogeneous claddng by a gven fracton of the dfference between the maxmum of the refractve ndex of the core and the refractve ndex of the nnermost homogeneous claddng. Note The core area s the smallest cross-sectonal area of a fbre excludng any ndex dp, whch s contaned wthn the locus of ponts where the refractve ndex n 3 s gven by n 3 = n + k (n 1 n ) (see Fgure A-1/G.651) where: n 1 = maxmum refractve ndex of the core, n = refractve ndex of the nnermost homogenous claddng, k = a constant. Note Unless otherwse specfed, a k value of 0.05 s assumed. 6 Fasccle III.3 Rec. G.651

9 FIGURE A-1/G.651 Some refractve ndex profles A.10 core (claddng) centre For a cross-secton of an optcal fbre t s the centre of that crcle whch best fts the outer lmt of the core area (claddng). Note 1 These centres may not be the same. Note The method of best fttng has to be specfed. A.11 core (claddng) dameter The dameter of the crcle defnng the core (claddng) centre. A.1 core (claddng) dameter devaton The dfference between the actual and the nomnal values of the core (claddng) dameter. A.13 core/claddng concentrcty error The dstance between the core centre and the claddlng centre dvded by the core dameter. Fasccle III.3 Rec. G.651 7

10 A.14 core (claddng) tolerance feld For a cross-secton of an optcal fbre t s the regon between the crcle crcumscrbng the core (claddng) area and the largest crcle, concentrc wth the frst one, that fts nto the core (claddng) area. Both crcles shall have the same centre as the core (claddng). A.15 four concentrc crcles near feld template A template comprsng four concentrc crcles appled to a near feld radaton pattern from a fbre. Note The template s normally used as a global check of the acceptablty of the varous geometrcal parameters of the fbre n one smple process. A.16 four concentrc crcles refractve ndex template A template comprsng four concentrc crcles appled to a complete refractve ndex profle of the fbre. Note The template s normally used as a global check of the acceptablty of the varous geometrcal parameters of the fbre n one smple process. A.17 maxmum theoretcal numercal aperture A theoretcal value of numercal aperture calculated usng the values of refractve ndex of the core and claddng gven by: where n 1 = maxmum refractve ndex of the core, NA t max = (n 1 n ) 1 n = refractve ndex of the nnermost homogeneous claddng. Note The relatonshp between NA ( A.1) and NA t max s gven n Secton I of Annex B, B... A.18 mode flter A devce desgned to accept or reject a certan mode or modes. A.19 mode scrambler; mode mxer A devce for nducng transfer of power between modes n an optcal fbre, effectvely scramblng the modes. Note Frequently used to provde a mode dstrbuton that s ndependent of source characterstcs. A.0 non-crcularty of core (claddng) The dfference between the dameters of the two crcles defned by the core (claddng) tolerance feld dvded by the core (claddng) dameter. A.1 numercal aperture The numercal aperture NA s the sne of the vertex half-angle of the largest cone of rays that can enter or leave the core of an optcal fbre, multpled by the refractve ndex of the medum n whch the vertex of the cone s located. A. reference surface The cylndrcal surface of an optcal fbre to whch reference s made for jontng purposes. Note The reference surface s typcally the claddng or prmary coatng surface. In rare crcumstances t could be the core surface. 8 Fasccle III.3 Rec. G.651

11 A.3 reference test method (RTM) A test method n whch a gven characterstc of a specfed class of optcal fbres or optcal fbre cables s measured strcly accordng to the defnton of ths characterstc and whch gves results whch are accurate, reproducble and relatable to practcal use. A.4 (refractve) ndex profle The dstrbuton of the refractve ndex along a dameter of an optcal fbre. ANNEX B (to Recommendaton G.651) Test methods Both reference and alternatve test methods are usually gven n ths Annex for each parameter and t s the ntenton that both the RTM and the ATM may be sutable for normal product acceptance purposes. However, when usng an ATM, should any dscrepancy arse, t s recommended that the RTM be employed as the technque for provdng the defntve measurement results. Secton I Reference test method and alternatve test method for geometrcal and optcal parameters measurements B.1 Introducton B.1.1 General It s assumed that the geometrcal and optcal parameters, whch are the subject of ths Recommendaton, would be measured only n the factory or n the laboratores of certan Admnstratons wshng to verfy these parameters for system desgn or other purposes. Hence, t s antcpated that the measurements wll be conducted ether on sample fbre lengths or on samples extracted from cable factory lengths. The core dameter and non-crcularty are defned usng the refractve ndex profle as a bass. The remanng parameters can be derved from the refractve ndex profle. Hence, t follows that all the geometrcal and optcal parameters that are the subject of ths Recommendaton, and ther tolerances as approprate, can be obtaned by one sngle basc test. B The four crcle tolerance feld A smple means of verfyng the geometrcal parameters of the fbre s the four crcle tolerance feld method. Ths does not consttute an addtonal requrement on the fbre geometrcal characterstcs, but s an alternatve global check of these characterstcs. If any nconsstency appears between ths method and the check of the ndvdual characterstcs, the latter wll consttute the reference. The four crcle tolerance feld method s based on the template shown n Fgure B-1/G.651 where the two concentrc crcles concernng the core (whose dameter s D co ) have dameters respectvely of D co 4 µm and D co + 4 µm and the two concentrc crcles concernng the claddng (whose dameter s D CL ) have dameters respectvely of D CL 5 µm and D CL + 5 µm. Ths method can be appled to data obtaned ether by the Reference Test Method (four concentrc crcle refractve ndex template) or by the Alternatve Test Method (four concentrc crcle near feld template). Fasccle III.3 Rec. G.651 9

12 FIGURE B-1/G.651 B.1.1. Intrnsc qualty factor The maxmum theoretcal NA, core dameter, concentrcty error, and core non-crcularty devate smultaneously n ways that can ether compound or compensate one another. To properly account for these effects, a theoretcal splce loss can be calculated, usng the values of these geometrcal and optcal parameters measured by exstng test methods. Ether a Gaussan or steady-state dstrbuton of power vs. angle may be assumed. The ntrnsc qualty factor (IQF) can be calculated as the mean of the theoretcal splce losses n the two drectons when the test fbre s splced to a nomnal fbre wth zero msalgnment of the reference surfaces. A value of IQF of 0.7 db s compatble wth the ndvdual tolerances recommended n secton 1 of Recommendaton G.651. If any nconsstency appears between the IQF method and the check of the ndvdual characterstcs, the latter wll consttute the reference. B.1. Geometrcal characterstcs The core dameter and the claddng dameter of the fbre under test, as well as the core and claddng centres, can be determned from an adequate number of ponts sutably dstrbuted on the core/claddng and on the claddng boundares respectvely. If a raster scan s adopted, a hgher number of ponts should be selected, n order to guarantee a suffcently regular dstrbuton. The concentrcty error can be evaluated from the dstance between the core and claddng centres. Core and claddng non-crculartes can be determned from the tolerance feld. B. The reference test methods for geometrcal parameters and the alternatve test method for numercal aperture: the refracted near-feld technque B..1 General The refracted near-feld measurement s straghtforward, accurate and gves drectly the refractve ndex varaton across the entre fbre (core and claddng). The measurement s capable of good resoluton and can be calbrated to gve absolute values of refractve ndex. A schematc dagram of the measurement method s shown n Fgure B-a/G.651. The technque nvolves scannng a focussed spot of lght across the end of the fbre. The launch optcs are arranged to overfll the NA of the fbre. The fbre end s mmersed n a lqud of slghtly hgher ndex than the claddng. Part of the lght s guded down the fbre and the rest appears as a hollow cone outsde the fbre. A dsc s placed on the axs of the core to ensure that 10 Fasccle III.3 Rec. G.651

13 only refracted lght reaches the detector. The detector output s amplfed and dsplayed as the y-axs of an x-y recorder; the x-axs drve s derved from montorng the poston of the focussed spot of lght on the end of the fbre. A typcal ndex profle of a multmode graded ndex fbre s shown n Fgure B-b/G.651. The optcal resoluton and hence the ablty to resolve detal n the profle depends on the sze of the focussed spot of lght. Ths depends both on the numercal aperture of the focussng lens and on the sze of the dsc. However, the poston of sharp features can be resolved to much better accuracy than ths, dependent on stop sze for stepped motor systems, or poston montorng accuracy of analogue drves. B.. Numercal aperture and refractve ndex dfference The maxmum theoretcal numercal aperture s defned as: NA t max = n 1 n The ndex dfference s defned as: n = n 1 n The relatve ndex dfference s defned as: = (n 1 n )/n 1 where n 1 = maxmum refractve ndex of the fbre core, n = refractve ndex of the nnermost claddng. The values of n 1 and n can be determned usng the refracted near-feld technque, hence NA t max, n and. The maxmum theoretcal numercal aperture NA t max, determned n ths way can be hgher (typcally by about 5% to 7%) than the numercal aperture NA determned by the RTM. B..3 B..3.1 Test apparatus A schematc dagram of the test apparatus s shown n Fgure B-3/G.651. Source A stable laser gvng a few mllwatts of power n the TEM oo mode s requred. A HeNe laser, whch has a wavelength of 633 nm, may be used, but a correcton factor must be appled to the results for extrapolaton at dfferent wavelengths. It shall be noted that measurement at 633 nm may not gve complete nformaton at longer wavelengths, n partcular non-unform fbre dopng can affect the correcton. A quarter-wave plate s ntroduced to change the beam from lnear to crcular polarzaton because the reflectvty of lght at an ar-glass nterface s strongly angle- and polarzaton-dependent. B..3. A pnhole placed at the focus of lens 1 acts as a spatal flter. Launch condtons The launch optcs, whch are arranged to overfll the NA of the fbre, brng a beam of lght to a focus on the flat end of the fbre. The optcal axs of the beam of lght should be wthn 1 of the axs of the fbre. The resoluton of the equpment s determned by the sze of the focussed spot, whch should be as small as possble n order to maxmze the resoluton, e.g. less than 1.5 µm. The equpment enables the focussed spot to be scanned across the fbre dameter. B..3.3 B..3.4 Lqud cell The lqud n the lqud cell should have a refractve ndex slghtly hgher than that of the fbre claddng. Sensng The refracted lght s collected and brought to the detector n any convenent manner provded that all the refracted lght s collected. By calculaton the requred sze of dsc and ts poston along the central axs can be determned. B..4 Preparaton of fbre under test A length of fbre of about 1 metre s requred. Prmary fbre coatng shall be removed from the secton of fbre mmersed n the lqud cell. Fasccle III.3 Rec. G

14 The fbre ends shall be clean, smooth and perpendcular to the fbre axs. B..5 Procedure Refer to the schematc dagram of the test apparatus (Fgure B-3/G.651). B..5.1 Fbre profle plot The launch end of the fbre to be measured s mmersed n a lqud cell whose refractve ndex s slghtly hgher than that of the fbre claddng. The fbre s back llumnated by lght from a tungsten lamp. Lenses and 3 produce a focussed mage of the fbre. The poston of lens 3 s adjusted to centre and focus the fbre mage, the laser beam s smultaneously centred and focussed on the fbre. The dsc s centred on the output cone. For multmode fbre the dsc s postoned on the optc axs to just block the leaky modes. Refracted modes passng the dsc are collected and focussed onto a photodode. The focussed laser spot s traversed across the fbre end and a plot of fbre refractve ndex varaton s drectly obtaned. B..5. Equpment calbraton The equpment s calbrated wth the fbre removed from the lqud cell. Durng the measurement the angle of the cone of lght vares accordng to the refractve ndex seen at the entry pont to the fbre (hence the change of power passng the dsc). Wth the fbre removed and the lqud ndex and cell thckness known, ths change n angle can be smulated by translatng the dsc along the optc axs. By movng the dsc to a number of predetermned postons one can scale the profle n terms of relatve ndex. Absolute ndex,.e. n 1 and n can only be found f the claddng or lqud ndex s known accurately at the measurement wavelength and temperature. B..6 Presentaton of results The followng detals shall be presented: a) Test set-up arrangement, wavelength correcton procedure and ndcaton of the scannng technque used; b) Fbre dentfcaton; c) Dependng on specfcaton requrements: ) profles through core claddng centres calbrated for the operatng wavelength, ) profles along the core major and mnor axes calbrated for the operatng wavelength, ) profles along the claddng major and mnor axes calbrated for the operatng wavelength, v) raster scan across the entre fbre, f adopted, v) core dameter 1) v) claddng dameter 1) v) core/claddng concentrcty error, v) core non-crcularty, x) claddng non-crcularty, x) maxmum theoretcal numercal aperture: NA t max, x) ndex dfference: n, x) relatve ndex dfference:. d) Indcaton of accuracy and repeatablty; e) Temperature of the sample and envronmental condtons (f necessary). 1) See Appendx I. 1 Fasccle III.3 Rec. G.651

15 FIGURE B-a/G.651 Refracted near-feld technque Schematc dagram FIGURE B-b/G.651 Typcal ndex profle of a graded ndex fbre obtaned by the refracted near-feld technque Fasccle III.3 Rec. G

16 FIGURE B-3/G.651 Typcal arrangement of the refracted near-feld test set-up B.3 Alternatve test method for geometrcal parameters: the near-feld technque B.3.1 General The near-feld technque can be used for the measurement of geometrcal characterstcs and of the refractve ndex profle of multmode optcal fbres. Such measurements are performed n a manner consstent wth the defnton and the results are reproducble and relatable to the reference test method and to practcal use. The measurement s based on the scannng of a magnfed mage of the output and of the fbre under test over a cross-secton where the detector s placed. When measurng the geometrcal characterstcs of the fbre, the four concentrc crcle near-feld template can be appled to an enlarged mage of the fbre detected wth objectve evaluaton methods, sutable to obtan a hgh degree of accuracy and reproducblty. In partcular the core dameter shall be measured takng nto account the same k factor agreed for the reference test method. B.3. Test apparatus A schematc dagram of the test apparatus s shown n Fgure B-4/G.651. B.3..1 Lght source The lght source shall be ncoherent, adjustable n ntensty and stable n poston, ntensty and wavelength over a tme perod suffcently long to complete the measurement procedure. The full wdth half maxmum (FWHM) spectral lnewdth, shall be recorded. A second lght source can be used, f necessary, for llumnatng the claddng. B.3.. Launchng condtons The launch optcs, whch wll be arranged to overfll the fbre, wll brng a beam of lght to a focus on the flat nput end of the fbre. 14 Fasccle III.3 Rec. G.651

17 For 50/15 µm graded-ndex fbres the overfll launchng condtons are obtaned wth a lght cone whose FWHM ntensty measured from the near-feld be greater than 70 µm and whose FWHM n the numercal aperture (NA) measured from the far-feld be greater than an NA of 0.3. B.3..3 Claddng mode strpper A sutable claddng mode strpper shall be used to remove the optcal power propagatng n the claddng, and to ensure that all the leaky modes are strpped away from the fbre. When measurng the geometrcal characterstcs of the claddng only, the claddng mode strpper shall not be present. B.3..4 Specmen The specmen shall be a short length of the optcal fbre to be measured. Prmary fbre coatng shall be removed from the secton of the fbre nserted n the mode strpper. The fbre ends shall be clean, smooth and perpendcular to the fbre axs. Note Ths measurement can be done on very short peces of fbre (e.g. a few centmetres). In ths case the launchng condtons shall be adjusted to obtan a unform ntensty n the claddng below 15% of the maxmum lght ntensty n the core. B.3..5 Magnfyng optcs The magnfyng optcs shall consst n an optcal system (e.g. a mcroscope objectve) whch magnfes the specmen output near-feld, focussng t onto the plane of the scannng detector. The numercal aperture and hence the resolvng power of the optcs shall be compatble wth the measurng accuracy requred, and not lower than 0.3. The magnfcaton shall be selected to be compatble wth the desred spatal resoluton, and shall be recorded. B.3..6 Detector A sutable detector shall be employed whch provdes the pont-to-pont ntensty of the magnfed near-feld pattern. For example, any of the followng technques can be used: a) scannng photodetector wth pnhole aperture; b) scannng mrror wth fxed pnhole aperture and photodetector; c) scannng vdcon, charge coupled devces or other pattern/ntensty recognton devces. The detector shall be lnear n behavour (or shall be lnearzed) over the range of ntenstes encountered. The senstve area of the detector shall be small wth respect to the enlarged mage of the output end of the fbre and shall be recorded. B.3..7 Amplfer An amplfer shall be employed n order to ncrease the sgnal level. The bandwdth of the amplfer shall be chosen accordngly to the type of scannng used. When scannng the output end of the fbre wth mechancal or optcal systems, t s customary to modulate the optcal source. If such a procedure s adopted, the amplfer should be lnked to the source modulaton frequency. The detectng system should be substantally lnear n senstvty. B.3..8 Data storage The measured near-feld ntensty dstrbuton can be recorded and presented n a sutable form, accordng to the scannng technque and to the specfcaton requrements. B.3.3 B Procedure Equpment calbraton The magnfcaton of the optcal system shall be measured by scannng the length of a specmen whose dmensons are already known wth sutable accuracy. Ths magnfcaton shall be recorded. B.3.3. Measurement The launch end of the fbre shall be algned to the launch beam, and the output end of the fbre shall be algned to the optcal axs of the magnfyng optcs. The focussed mage of the output end of the fbre shall be scanned by the detector, accordng to the specfcaton requrements. The focussng shall be performed wth maxmum accuracy, n order to reduce dmensonal errors due to the scannng of a msfocussed mage. B.3.4 Presentaton of the results The followng detals shall be presented: a) Test set-up arrangement, wth ndcaton of the scannng technque used. Fasccle III.3 Rec. G

18 b) Launchng characterstcs (dmenson and NA of the launchng cone). c) Wavelength and FWHM spectral lnewdth of the source(s). d) Fbre dentfcaton and length. e) Type of claddng mode strpper (f applcable). f) Magnfcaton of the apparatus. g) Type and dmensons of the scannng detector. h) Temperature of the sample and envronmental condtons (f necessary). ) Indcaton of the accuracy and repeatablty. j) Dependng upon the specfcaton requrements: ) profles through core and claddng centres; ) profles along the core major and mnor axes; ) profles along the claddng major and mnor axes; v) raster scan across the entre end face of the fbre, f adopted; v) resultng dmensonal parameters, lke: core and claddng dameters ), non-crculartes of the core and of the claddng, core/claddng concentrcty error, etc. FIGURE B-4/G.651 Typcal arrangement of the near-feld test set-up B.4 Reference test method for the numercal aperture: far-feld dstrbuton B.4.1 Object Ths measurng method s appled to graded ndex fbres n order to determne the numercal aperture by measurng the far feld lght dstrbuton. B.4. Specmen preparaton From the fbre to be measured at one end, a sample of approxmately m length s taken. The sample shall be straght enough to avod bendng losses. The ends of the sample should be substantally clean, flat and perpendcular to the fbre axs. B.4.3 B Apparatus Source The lght source shall be ncoherent, adjustable n ntensty and stable n poston, ntensty and wavelength over a tme perod suffcently long to complete the measurement procedure. B.4.3. Detector The detector must have a lnear characterstc n the requred measurng range. (The output current of the detector must be lnear to the receved lght power.) B Launchng condtons ) See Appendx I 16 Fasccle III.3 Rec. G.651

19 B B B.4.4 B See B.3... Claddng mode strpper See B Dsplay For example, XY-recorder, screen. Procedure Prncple of measurement (Fgure B-5/G.651) The radant ntensty (lght power per sold angle element) s determned as a functon of the polar angle of one plane of the fbre axs (radaton pattern). The dstance d between the end of the sample and the detector must be large compared to the core dameter of the optcal fbre. Possble solutons are: sample fxed, large-area detector fxed; sample fxed, small-area detector lnear dsplaceable; sample lnear dsplaceable, small-area detector fxed; sample fxed, small-area detector angular dsplaceable; sample and rotatable, small-area detector fxed. FIGURE B-5/G.651 Prncple of measurement for far feld dstrbuton B.4.4. B B.4.5 Preparaton The sample s fxed n the sample holder and the lght s launched n accordance wth B Measurement The radant ntensty s determned as a functon of the polar angle n one plane of the fbre axs. Results Fbres covered by ths Recommendaton have a near parabolc refractve ndex profle. Therefore, for the launchng condtons recommended n B (unform mode dstrbuton) the far-feld radant ntensty curve can be approxmated n the regon above 10% of the maxmum ntensty by the followng parabola: P(ϕ) = P(0) [1 (sn ϕ/na) ] The angle ϕ s then determned by the pont of ntersecton of ths parabola wth the abscssa. In general, t s suffcent to determne the angle ϕ by the 5% value of the maxmum radant ntensty out of the full radant ntensty curve. B.4.6 The numercal aperture s Presentaton of results NA = sn ϕ. Fasccle III.3 Rec. G

20 18 Fasccle III.3 Rec. G.651 The followng detals shall be presented: a) Test set-up arrangement, wth ndcaton of the scannng technque used; b) Launchng characterstcs (dmenson and NA of the launchng cone); c) Wavelength and FWHM spectral wdth of the source; d) Type of claddng mode strpper (f used); e) Scannng condtons; f) Fbre dentfcaton and length; g) Temperature of the sample and envronmental condtons, f necessary; h) Indcaton of the accuracy and repeatablty; ) Resultng numercal aperture. APPENDIX I (to Secton I) A possble way to obtan the postons of core and claddng centres as well as the dameters s gven n ths Appendx. I.1 Core centre and dameter The core centre and dameter are determned from an adequate number of scans across a fbre secton wth the approprate k value. Two ponts on the core/claddng nterface are obtaned at each scan. The ponts should be unformly dstrbuted on the permeter of the core, at least approxmately. Let x, y be the Cartesan coordnates of the -th pont of the nterface, a c, b c be the Cartesan coordnates of the core centre, R c be the core radus, z, m be the ntermedate varables z = a c + b c R c and m = x + y. The unknown parameters a c, b c and R c are obtaned by fndng the crcle whch best represents, n the sense of the least squares, the expermental nterface between the core and the claddng. The calculaton algorthm conssts of mnmzng, wth respect to parameters a c, b c and z, the quantty: M = [ ] ) ( ) ( + c c c R b y a x = ( ) + + c c z y b x a y x Cancellaton of the three partal dervatves of M n relaton to a c, b c and z gves a system of three lnear equatons, n the followng matrx form: = c c m y m x m z b a N y x y y y x x y x x The sums are performed from = 1 to = N,

21 N beng the total number of measured ponts Dgtal nverson of ths system gves the values of a c, b c and z, from whch the value of R c s deduced. The core centre s the pont of coordnates a c and b c and ts dameter s the quantty D c = R c. I. Claddng centre and dameter The same calculaton process and defntons as for the core apply to fndng: the coordnates of the claddng centre a g and b g ; the claddng radus R g. The claddng centre s the pont of coordnates a g and b g and ts dameter s the quantty D g = R g. Secton II Reference test method and alternatve test methods for attenuaton measurements B.1 Introducton B.1.1 Objectves The attenuaton tests are ntended to provde a means whereby a certan attenuaton value may be assgned to a fbre length such that ndvdual attenuaton values may be added together to determne the total attenuaton of a concatenated length. B.1. Defnton The attenuaton A (λ) at wavelength λ between two cross-sectons 1 and separated by dstance L of a fbre s defned as: A (λ) = 10 log P 1 (λ) P (λ) (db) where P 1 (λ) s the optcal power traversng the cross-secton 1 and P (λ) s the optcal power traversng the cross-secton at the wavelength λ. For a unform fbre under equlbrum condton, t s possble to calculate the attenuaton per unt length, or the attenuaton coeffcent. whch s ndependent of the chosen length of the fbre. α (λ) = A (λ) L db unt length Note Attenuaton values specfed for factory lengths should be measured at room temperature (.e., a sngle value n the range +10 o C to +35 o C). B.1.3 Descrpton Three methods have been suggested for attenuaton measurements. B The cut-back technque s a drect applcaton of the defnton, n whch the power levels P 1 and P are measured at two ponts of the fbre wthout change of nput condtons. P s the power emergng from the end of the gude and P 1 s the power emergng from a pont near the nput after cuttng the fbre. B.1.3. The nserton loss technque s n prncple smlar to the cut-back technque, but P 1 s the power emergng from the output of the launchng system. The measured attenuaton s the sum of the attenuaton of the nserted length of fbre and the attenuaton caused by the connecton between launchng system and the fbre under test. It s necessary to correct the result for connecton losses. B The backscatterng technque s an ndrect way of measurng the attenuaton by the measurements of the backscattered powers traversng two cross-sectons of the fbre. B.1.4 Feld of applcaton The cut-back technque s generally recognzed as yeldng accurate results. In many stuatons ts destructve nature s a dsadvantage. The nserton loss technque avods cuttng a part of the fbre at the expense of accuracy. Fasccle III.3 Rec. G

22 The backscatterng technque s a sngle ended, non-destructve method, but s lmted n range and sometmes n accuracy. Consderng the advantages and dsadvantages of the three methods, the cut-back technque has been chosen as the reference test method. B. The reference test method: the cut-back technque B..1 B..1.1 Launchng condtons Defnton of launchng condtons The launchng condtons are of paramount mportance n meetng the stated objectves. Launchng condtons should be such as to approxmate equlbrum mode dstrbuton (EMD) whch s understood to exst when the power dstrbuton of feld patterns at the output of the fbre s substantally ndependent of the length of the fbre. B..1. Launchng technques A generc set-up to acheve the launchngs of the EMD s shown n Fgure B-6/G.651. FIGURE B-6/G.651 Generc launchng condtons B Claddng mode strpper A claddng mode strpper encourages the converson of claddng modes to radaton modes; as a result, claddng modes are strpped from the fbre. B..1.. Mode flter The mode flter s a devce used to select, reject or attenuate a certan number of modes, and should assure the establshment of a mode dstrbuton close to the EMD. B Mode scrambler The mode scrambler s a devce used for nducng transfer of power between modes n an optcal fbre, and should provde a mode dstrbuton ndependent of source characterstcs. Note Sutable optcal arrangements can be used whch produce a dstrbuton close to the EMD drectly on the nput end of the fbre under test. In ths case a unque devce s needed for the mplementaton of the three functons of Fgure B-6/G.651. B..1.3 Example For a 50/15 µm low loss homogeneous graded ndex fbre wth a NA of 0. operatng at 850 nm wavelength, an approxmaton of the equlbrum mode dstrbuton can be acheved after the cut-back length of the fbre to be measured, when the followng characterstcs are observed: a) the full wdth half maxmum ntensty value of the lght spot, measured from the near feld s 6 µm; b) the full wdth half maxmum value of the numercal aperture measured from the far feld s 0.11 µm. Both near-feld and far-feld patterns are assumed to be approxmately Gaussan. To obtan ths equlbrum of the modal dstrbuton the arrangement shown n Fgure B-6/G.651 can be used. 0 Fasccle III.3 Rec. G.651

23 The launch beam s ncdent on the launch end of the fbre n the form of a spot, centrally located on the fbre core wth the near feld FWHM ntensty not less than 70 µm and the far feld FWHM Numercal Aperture not less than 0.3 across the central 70 µm of the cone. (For fbre wth an NA < 0.5.) The axs of the launch beam s concdent wth the axs of the fbre. The mode scrambler should comprse a sutable fbre arrangement (e.g., a step-graded-step sequence or a bendng sequence). The mode flter takes the form of a mandrel around whch the fbre under test s wound, wth low tenson and wthn a 0 mm length of the mandrel. The dameter of the mandrel may dffer from fbre to fbre and values n the range 18- mm, wth 5 turns of fbre, are common. The exact dameter of the mandrel s determned by the fbre/cable suppler such that the near feld and far feld patterns from two metres of fbre followng the mode flter and claddng mode strpper, are the same as those obtaned from a stable mode dstrbuton length of fbre or jonted fbres (typcally greater than 5 km). The claddng mode strpper often conssts of a materal havng a refractve ndex equal to or greater than that of the fbre claddng. B.. Apparatus and procedure B...1 Types of measurement Measurements may be made at one or more spot wavelengths, alternatvely a spectral response may be requred over a range of wavelengths. Dagrams of sutable test equpments are shown as examples n Fgures B-7/G.651 and B-8/G.651. Arrangement of test equpment to make spot loss measurement Fasccle III.3 Rec. G.651 1

24 FIGURE B-8/G.651 The cutback technque B... Optcal source A sutable radaton source shall be used, such as a lamp, laser or lght emttng dode (LED). The choce of source depends upon the type of measurement. The source must be stable n poston, ntensty and wavelength over a tme perod suffcently long to complete the measurement procedure. The FWHM spectral lne-wdth shall be specfed such that t s narrow compared wth any features of the fbre spectral attenuaton. The fbre shall be algned to the launch cone, or connected coaxally to a launch fbre. B...3 Optcal detector A large area detector shall be used so that all of the radaton n the output cone(s) s ntercepted. The spectral response should be compatble wth the spectral characterstcs of the source. The detecton must be unform and the detecton must have lnear characterstcs. It s customary to modulate the lght source n order to mprove the sgnal/nose rato at the recever. If such a procedure s adopted, the detector should be lnked to a sgnal processng system synchronous wth the source modulaton frequency. The detectng system should be substantally lnear n senstvty. B...4 Launchng set-up See B..1. B...5 Procedure 1) The fbre under test s set n the measurement set-up. The output power P s recorded. ) Keepng the launchng condtons fxed, the fbre s cut to the cut-back length (for example, m from the launchng pont). The output power P 1 from the cut-back length of the fbre s recorded. 3) The attenuaton of the fbre, between the ponts where P 1 and P have been measured, can be calculated from the defnton usng P 1 and P. Fasccle III.3 Rec. G.651

25 B..3 Presentaton of results The followng detals shall be presented: a) Measurement type, and characterstcs. b) Launchng technque. c) Test set-up arrangement. d) Temperature of the sample and envronmental condtons (f necessary). e) Fbre dentfcaton. f) Length of sample and the cut-back length. g) Attenuaton measured (for the sample) at the selected wavelength. h) Attenuaton quoted n db. In some cases t s possble to convert t nto attenuaton coeffcent n db/km. ) For spectral loss measurements the results should be presented as a plot of attenuaton aganst wavelength. B.3 Frst alternatve test method: the nserton loss technque B.3.1 Launchng condtons The requred launchng condtons are smlar as those descrbed under B..1. B.3. B.3..1 Apparatus and procedure Types of measurements Measurements may be done at one or more spot wavelengths, alternatvely a spectral response may be requred over a range of wavelengths. A dagram of a sutable test set-up s shown as an example n Fgure B-9/G.651 (a calbraton, b measurement). B.3.. Optcal source See B... B.3..3 Optcal detector See B...3. B.3..4 Launchng set-up See B..1. B.3..5 Couplng devce The nserton loss technque requres the use of a very precse fbre to fbre couplng devce to mnmze the couplng losses and to ensure relable results. Ths couplng devce can be a mechancal adjustment vsually nspected or a connector wth a core-to-core postonng. Fasccle III.3 Rec. G.651 3

26 B.3..6 B.3.3 Procedure 1) The measurement set-up s ntally calbrated n order to obtan an nput reference level P 1. ) The fbre under test s set n the measurement set-up and the couplng adjusted to gve a maxmum level on the optcal detector. The output power P s recorded. 3) An attenuaton s calculated accordng to 1.. Ths attenuaton s the sum of the attenuaton of the nserted length of fbre and the attenuaton caused by the connecton between the couplng devce and the fbre under test. Presentaton of results The followng detals shall be presented: a) Measurement type and characterstcs. b) Launchng technque. c) Test set-up arrangement. d) Temperature of the sample and envronmental condtons (f necessary). e) Fbre dentfcaton. f) Length of sample. g) Attenuaton measured (for the sample) at the selected wavelength. h) Connector loss wth ts tolerance. ) Attenuaton quoted n db. In some cases t s possble to convert t nto an attenuaton coeffcent n db/km. j) For spectral loss measurements the results should be presented as a plot of attenuaton versus wavelength. 4 Fasccle III.3 Rec. G.651

27 B.4 Second alternatve test method: the backscatterng technque Note Ths test method descrbes a procedure to measure the attenuaton of a homogeneous sample of optcal fbre cable. The technque can be appled to check the optcal contnuty, physcal defects, splces, backscattered lght of optcal fbre cables and the length of the fbre. B.4.1 Launchng condtons For the attenuaton measurement, the technques descrbed under.1 can be appled. For the other controls, the launchng condtons may be dependent on the characterstcs to be tested. In all cases, n order to reduce the Fresnel reflectons on the nput of the fbre, varous devces could be used such as polarzers or ndex matchng materals. Inserton losses should be mnmzed. B.4. B.4..1 Apparatus and procedure General consderatons The sgnal level of the backscattered optcal sgnal wll normally be small and close to the nose level. In order to mprove the sgnal-to-nose rato and the dynamc measurng range t s therefore customary to use a hgh power lght source n connecton wth sgnal processng of the detected sgnal. Further, accurate spatal resoluton may requre adjustment of the pulse wdth n order to obtan a compromse between resoluton and pulse energy. Specal care should be taken to mnmze the Fresnel reflectons. An example of an apparatus s shown n Fgure B-10/G.651. B.4.. Optcal source A stable hgh power optcal source of an approprate wavelength should be used, such as a semconductor laser. The wavelength of the source should be regstered. The pulse wdth and repetton rate should be consstent wth the desred resoluton and the length of the fbre. Optcal non-lnear effects should be elmnated n the part of the fbre under test. B.4..3 Optcal detecton A detector shall be used so that the maxmum possble backscattered power should be ntercepted. The detector response shall be compatble wth the levels and wavelengths of the detected sgnal. For attenuaton measurements the detector response shall be substantally lnear. A sgnal processng s requred to mprove the sgnal-to-nose rato, and t s desrable to have a logarthmc response n the detecton system. A sutable amplfer shall follow the optcal detector, so that the sgnal level becomes adequate for the sgnal processng. The bandwdth of the amplfer shall be chosen as a trade off between tme resoluton and nose reducton. B.4..4 B.4..5 Launchng set-up See B..1 and B.4.1. Procedure 1) The fbre under test s algned to the couplng devce. ) Backscattered power s analyzed by a sgnal processor and recorded n logarthmc scale. Fgure B-10/G.651 shows such a typcal curve. 3) If the recorded curve has an approxmately constant slope (zone b of Fgure B-10b/G.651), the attenuaton between two ponts A and B of the curve correspondng to two cross sectons of the fbre s A(λ) A B = 1 (V V ) db A B where V A and V B are the correspondng power levels gven n the logarthmc scale. 4) If so requred b-drectonal measurements can be made, together wth numercal computaton to mprove the qualty of the result and possbly to allow the separaton of attenuaton from backscatterng factor. B.4.3 Presentaton of results The followng detals shall be presented: a) Measurement types and characterstcs. b) Launchng technques. c) Test set-up arrangement. Fasccle III.3 Rec. G.651 5