CA1205899A - Fiber optic energy sensor and optical demodulation system and methods of making same - Google Patents

Abstract

ABSTRACT
A fiber optic energy sensor and optical demodulation system is disclosed wherein the signal energy to be sensed or detected causes an etched single mode fiber to be stretched or compressed which causes the optical path length for electromagnetic radiation traveling in the core of the optic fiber to change. The change in optic path length modulates the electromagnetic radiation traveling in the fiber. The optical demodulation system locates each energy sensor between the members of a pair of length limited Bragg reflectors formed inside of the optical fiber. Each pair of reflectors as arranged, forms a Fabry-Perot type interferometer which only contains resonances for those portions of the electromagnetic spectrum in which the Bragg reflectors operate. Since each energy sensor is located between the reflectors of a pair, then as signal energy is detected the sensor's resulting optical path length change causes the resonances of the Fabry-Perot interferometer, to shift spectrally. The system then partially demodulates this spectral shift using a second Fabry-Perot interferometer, called the analyzer interferometer, whose resonances have a spectral separation relative to that of the interferometer containing the energy sensor so as to cause amplification of the spectral shift. The output of the combination of the energy sensor interferometer and the analyzer interferometer shifts spectrally more than the original spectral shift by an amplification factor given by equations presented in the detailed des-cription of the invention. The system also allows the use of more than one analyzer interferometer, each causing different amplification. The resulting amplifications can be arranged to provide outputs each of which corresponds to a separate digit of the number expressing the original spec-tral shift thus reducing the bandwidth of the electronic detectors and time demodulator.

Description

OS8'Lt~
1936-1~21D

This is a di.visional application of copending Canadian patent application Serial No.37~,300 :Eiled March 31, 1981 and assigned to Chevron Research Company.
Technical Field The invention relates to the field of radiant energy modulation in optical fibers.
Background of Prior AIt The prior art for either phase modulation or frequency modulation of, for example, light in an optical fiber utilized the acousto-optic effect wherein the signal to be imposed onto the light carried in the fiber is used to mechanically or acoustically excite the fiber. This mechanical or acoustical excitation gives rise to a variation in the optical index of the core of the fiber. The result is a variation in the optical path length for the light traveling in the fiber. This light is therefore modulated in phase and frequency by the signal. For glass fibers the change in optical index is quite small for a given energy of mechanical or acoustical excitation. In order to obtain sufficient modulation, this necessitates either high signal energy or long interaction lengths where the interaction length is the length of fiber which must be acoustically excited wherein modulation occurs. The sensitivity of optical fibers to direct acoustical modulation is discussed by J. A. Bucaro in Applied Optics; Vol. 18; No.6; March 15, 1979.

5~3~ '', Tlle invention constructs single mode fiber which has reduccd clad thickness for use in sensors which allow signal energy to stretch single mode fibers to cause phase modulation. The invention also constructs low order mode optical fibers from large diameter optical fibers. The invention accomplishes these two efforts by etching presently available optical fibers.
As discussed in "Acoustic Sensitivity of Single Mode Optical Power Dividers", by S.K. Sheem and J.H. Cole in Optics L,etters, Vol. ~; No. 10;
October, 1979, the prior art etches-single mode fiber in order to decrease its light guiding capabilities withou* any regard to its increased or de-creased acoustic sensitivity or change in mode structure. Such an effect, decreased light guiding capability, is considered detrimental to the inven-tions purpose and the invention specifically provides means to minimize it.
This invention also uses length limited distributed wavelength reflectors to reflect light within a single mode optical fiber. Such reflec-tiOlI is discussed in our U.S. Patent No. 4,268,116, of May 19, 1981, entitled METHOD AND APPARATUS FOR RADIANT ENERGY MODULATION IN OPTICAL FIBERS, and in "Photosensitivity in Optical Fiber Waveguides: Application to Reflection Filter Fabrication" by K.O. Hill et al in Applied Physics letters, 32(10), 15 May 1978.
The invention also uses reflectors which cause reflection within an optical fiber in an arrangement which resemblas a Fabry-Perot interferometer.
Such an arrangement is discussed in "Fiber Optic llydrophone: Improved Strain Configuration and Environmental Noise Protection" by P.G. Cielo in Applied Optics, Vol. 18, No. 17; 1 September 1979. The invention provides a novel detection system which uses this arrangemen* of reflectors as one of its many components, Brief Summary of Invention The invention comprises a novel type o~ fiber optic energy sensor, a method of manufacturing such type, and an optical demodulation system which can be used to convert the output of this type of energy sensor as well as other types of lZ~5 energy sensors into a more easily handled electronic signal.
The invention uses etched single mode fiber for the energy sensorn The energy sensor operates as follows:
The s-ignal energy to be sensed or detected is caused S to stTetch an etched single mode fiber. An etched single mod3 fiber is a single mode glass clad fiber whose cladding thick-ness has been reduced a specific amount so ~s to lower i~s strength. The invention pro~ides that,~Yhen necessary to main-tain the etched single mode fiber's light guiding properties, the portion o the glass clad which was removed is replaced with a plastic material whose optical index is lower ~han that of the s-ingle mode fiberls core material and whose elastic - modulus is lower than that of the glass clad which it replaces.
Such an etched singlls mode fiber is more sensitive to s*retch-ing or compression because it is weaker. For a given amountof ~ignal energy~ a single mode fiber will stretch a greater amount after it has been etched.
The pr;or art teaches that stretching a length of single mode fiber causes ths optical path length for electro-magnetic radiation traveling in its core to change. The artfurther teaches that this change in optical path length in-creases as the amount which ~he single mode iber is stretched increases. The prior art uses this change in optical path length to modulate the elec*ro-magnetic radiatiorl traveling in the core of the fiber. The prior art also teaches that the amount of modulation increases as the optical path length change increases in magnitude. Therefore, a fiber optic energy sensor constructed with etohed single mode fiber and operating by longitudinally stretc:hing or compressing the etched single mode fiber will for a given amount of signal energy produce greater modulation w}lich results in greater sensitivity.
The invent.ion also uses the etching process to produce vptical fibers with low modal dispersion from op~ical fibers having larger diameters.
The invention also provides a manufacturi~g prvcess to construc~ devices using etched optical fiber~ This process causes a form to be constructed of materials which will not affect the etching process. These forms are used to maintain the fiber to be etched in the same configuration as it is to be . .

~ 58~

in the actual device. Various means ~re also detailed which permit the form to be removed if it is to be absent in the actual de~ice.
Finally, the invention provides an optical demodu-lation system which renders energy sensors more useful byactually optically demodulat~s ~he output of the energy sensor, thus substantially reclucing the previously enormous bandwidths required of the electronic demodulation equipment. The optical demodulation system presented also allows for multiplexing several energy sensors on~o ~he same optical ~iber, thus sub-stantially reducing cc,sts for a multi-sensor system such as a hydrophone array.
The optical demodulation system locates each energy se.nsor between the members of a pair.of length limited Bragg reflectors formed inside of the optical iber. Each pair of reflectors as arranged., forms a Fabry-Perot type inter~er-ometer which only ont.ains resonances ~or those portions of the electro-magnetic spectrum in which the Bragg reflectors operate. Since each energy sensor is located between ~he re-flectors of a pair, then as signal energy is.detected thesensor's resulting optical pa~h length change causes the resonances of ~he Fabry-Perot intererometer to shit spectrally.
The system then partia.lly demodulates this spectral shift using a second Fabry-Perot interferometer, called the analyzer inter~erometer, whose resonances have a spectral separation relative to that of the interferometer containing the energy sensor so as to cause amplification of the spectral shiftO
The output o~ the combination of the energy sensor interfer- I
ometer and the analyzer interferometer:shifts spec~rally more than the original spectral shift by an amplification factor given-by equations presented-in the-detailed description of the invention. The system also allows ~he use of more than one analyzer interferometer, each causing different amplifi-cation. The resulting amplifications can be arranged tG pro-vide outputs each of which corresponds to a separate digit o~the number expressing the original s~ectral shift thus re-ducing the bandwidth of the electronic detectors and time demodulator.

~;~OS~9 ~he optical demodulation system finally allows for multiplexing several energy sensors onto the same fiber hy causing each reflector ~air corres~onding to each sensor to have different reflection bands from all other reflector pairs.
~he system uses a wavelength scanning laser, the output of which scan~3 over the resonances of one reflector pair at a time.
This divis:;onal application is particularly concerned with apparatus and method for phase modulating electro-ma~netic radiation travelling in the core of an optical fiber and appa-ratus and method for guiding electro-magnetic radiation.
The latter apparatus comprises an etched optical fiber having low modal dispersion, wherein the optical fiber having low modal dispersion is coated with a material having an optical inde~ lowelr than that of the core of the fiber.
The phase modulating apparatus comprises an etched optical fiber; and means for longitudinally stretching or compressing the etched optical fiber, wherein the etched opti-cal fiber is coated with a material having an optical index lower than that of the etched optical fiber's core and having a modulus of elastlcity lower than that of the material which was removed by etching.
The invention will now be described in greater detail with reference to the accom anying drawings, in ~hich:
FIGURE 1 is a cross sectional view of a single mode optical fiber greatly magnified;
FIGURE 2 is a greatly magnified, cross sectional view of the single molle optical fiber of FIGURE 1, which has been etched in accordance with the invention, FIGURE 3 is a greatly magnified cross sectional view of a large diame~ter fiber of core material;

5~9 FIGURE ~ is a grea-tly magnifled cross sectional view of the large diameter fiber of FIGUP~E 3 after having been e-tched and coated in accordance with -the invention;
FIGURE 5 is an end view of an acoustic ener~y sensor shown in EIGURE 6;
FIGURE 6 is a cross sectional illustration on line ~-6 of the acoustic energy sensor of FIGURE 5;
FIGURE 7 is an enlarged partial illustration of the acoustic energy sensor of FIGUR~S 5 and 6;
FIGURE 8 is an end view of another acoustic energy sensor provided by the invention, FIGU~E 9 is a eross section on line 9-9 of the acoustic ener~y sensor of FIGURE 8;
FIGURE 10 is an illustration of a form and a single mode optical fi~er as arranged by the invention for the purpose of manufacturing fiber optie energy sensors;
FIGURE 11 is a eross seetion of only the form of FIGURE 10 after having been coated with a guard material in accordance with the invention;
FIGURE 12 is an illustration of the form and single mode optical fiber of FIGURE 10 after etchiny and coating in aeeordanee with the invention;

-5a-PIG. 13 is an end view of a collapsible ~orm which could be used in the etching process;
PIG. 1~ is a schematic diagram of an optical de-modulation system provided by the invention;
FIG. 15 is an illustration of a typical transmission of a reflector pair of the pairs 25 in FIG. 14;
FIG. 16 is a graph of a laser output suitable for use in the optical demodulation system of FIG. 14;
PIG. 17 is a schematic diagTam of a multiple analyzer interferometer demodulator which the invention p~ovides as a substitute for the part of the optical demodulation system which is enclosed by dashed lines W in FIG. 14; and PIG. 18 is a schematic diagram o an example time demodulator circuit illustrated in FIGS. 14 and 17.
Detailed Description of the Invention __ The invention comprises a highly sensitive fiber optic, energy sensor and an optical demodulation system which can convert the output of the energy sensor to an electric analog signal. First the energy sensor will be discussed and second the demodula~ion system will be discussed.
The present art of fiber optic energy sensors teaches *hat if a single mode optical iber is compressed radially, or stretched, or compressed longitudinally~ then the optical path length for electro-magnetic radiation tTaveling in the core of the single mode optical fiber changes. The art further teaches that as the amount which the single mode fiber is stretched or compressed increases, then the change in the optical path length also increases. The present art uses this change in optical path length to cause phase modulation of the light traveling in the core. The length of optical fiber in which modulation occurs is called the interaction length.
The invention pro~ides etched single mode fiber for use in fiber optic, energy sensors. Single mode fiber is a fiber constructed so as to allo~y only the lowest order mode to propagate. This lowes~ order mode fo~ some single mode fiber constructions i5 two fold degenera~e. In ~hese cases, the lowest order mode contains ~wo states of propagation which are distinguished by the fact that their polarizations are mutually perpendicular.

7~20s~t~
Etched single mode flber is defined herein as single mode optical fiber ~hose clad thickness has been re-duced by a chemical reaction ~e.g., etching in a bat~ of hy-, drofluoric acid or a Datn of hydrofluoric as,id b,uffered with S ammonium ~luoride), or ion millin~., , FI~. 1 is a magnified cross sectional ~iew of a fiber prior to etching. FIG. 2 is a magnified cross sectional view of the fiber after etching. In FIG. 1, the glass clad generally designated,as 2-l, is shol~n to have a thic~ness designated K.
10' In FIG. 2, the clad 2-2 is shown to have a reduced thickness designated as R. In both FIGS. 1 and 2~ the core designated as 1 1 and 1-2 has a diameter V ~hich remains unchanged by the nature of the etching process, which occurs only at the exposed surface of the fiber.
The utility of such a fiber is e~plained first in terms of sensitivi~y followed by an explanation of the ease ~f manufacture of de~ices employing etched single mode fibers.
For a given amount of signal energy, E, to be detected, the , fiber of length L and total cross sectional area Sl, of FIG. 1', will stretch an amount ~Ll as belo~:

~Ll ~ ~ EQ I

where YO is the modulus of elasticity of the fiber material and for the explanation may be assumed constant and equal to that of fused quartz. Using the same derivation above, but substituti~g the reduced cladding-thickness into EQ l gi~es the amount of stretch ~L29 which the etched fiber will undergo for the same given amount of signal energy E, ~L2 ~ ~ EQ II
where S2 is the cross sectional area o~ the etched fiber Since S~ is greater t:han S2 then from FQ 1 and II, ~L~ ~ ~Ll.
The present art of fiber optic sensing teaches therefore, that for a given amount oi' signal energy the e~ched single mode fiber will have a greater change in optical path length than the normal single mode fiber, resulting in a greater amount of

2~ 3 phase modulation o light traveling in the core.
Further utility can be understood realizing that fibers ha~ing very small overall diameters are difficult to construct using present methods and even if constructed, are difficult to handle. Through the teachings of this invention, devices which might use fibers with reduced clad thickness can be construc~ed with readily available larger diameter fibers.
When such devices are assembled to the point ~here the larger fiber is in place, the fiber can ~hen be etched7 thereby elimi-nating further handling o~ a thin fiber or a fiber with a Te-duced clad. A more detailed explanation of this process follows later.
Further utility can be recognized when the need to construct fibers with small core diameters arises, and the invention allows such fibers ~o be constructed from larger diameter fibers. FIG. 3 shows the cross section of a large diameter optical fiber, 3-3, which is a core material, (e.g., silica glass), of diameter F. The large diameter fiber is etched thus producing the thin fiber whose cross section is shown as 4-4 in PlG. 4, having a reduced diameter G. The in-vention further provides that fiber having a diameter G can then be coated with a material 5-4 having a lower index o~ re- ¦
fraction, such as RTV ~70 silicone rubber produced by ~eneral Elec~ric Corp., than the~ fiber itself thus producing an optical fiber wi*h a small core diameteru Such a small core diameter fiber is useful for having a low number of guided optical modes.
As an example of ~he etching process, the fiber 3-3, FIG. 3, may have a pre-etching diameter in the range o~ ~3JU~
to lOO~m and the etched core 4-4, PIG. 4~ may have a diameter in the range of 50~m 1:o 5~m .
The invention provides the particular hydro-acoustic energy sensor shown in l:IGS. 5 and 6, and in part, in FIG. 7.
FIG. 5 is an end view oi the sensor illustrating its cylindrical shape~ FIG. 6 is a sec1ional view of the sensor. This sensor consists o~ a rigid cyllndrical skeleton9 perhaps made of aluminum, designated 6 in PIG. 6. The outside surface of this cylindrical skeleton has a reduced diameter between planes H
and J. About this cylîndrical skeleton is a membrane of 2~

compliant material ~enerally designated 7, in which are radial turns of sing~le mode optical fiber generally designated as 8. Such a compliant material may be~ ~or example, silicone rubber or PVC. This sleeve is cemented as at 13 and/or clamp-ed as at 14, to the larger diameter ends 13' of the cylindrical s~eleton, thus creating a space 9 between the compliant membrane and the rigid cylindrical skeleton where the diameter of the rigid form is reduced. In the wall of this rigid cylindrical skeleton where the diameter is reduced are equalization holes 10 which extend from the inside wall of the cylindrical skeleton to the space between the compliant membrane and -the rigid skele-ton. On the inside walls of the cylindrical skeleton are pro-trusions designated as 11 in FIG. 6. Also, lnside the cylindri-cal skeleton is stretched a compliant bladder 12 which ser~es as ballast supply and creates a reservoir 16, which communi-cates with space 9 by holes 10. The spaces 16 and 9 are filled with a second Viscou!; compliant material, such as air, He, or a silicone oil. There is also provided end caps 17 which create an additional space -:L6' sho~n in FI~. 6, and which are provided with holes 15, which extend through the thickness of each of the end caps. The hydrophone sho~n in FIGS. 5, 6; and 7 operates as follows:
The hydrophone is immersed in the fluid containing the acoustical waves to be measured. At any particular clepth the invention causes the stat;c pressure in the spaces 9 and 16 to be equalized with the static pressure in the fluid exterior to the hydrophone by allowing some of this ~luid to enter the hydrophone through holes 15 and then to stretch bladder 1~
around the protrusions 11, as illustrated by the dashed lines labeled 12', thus compressing ~he second viscous compliant material in spaces 16 and 9O l~hen the pressure in spaces 16 and 9, ~lus the additional pressure generated in stretching the bladder 12 is equal to the exteT;or pressure, the ~luid stops flowing through holes 15. The holes 15 and/or the equali-zation holes 10 are sufficiently small so as to slo~ the rate o~equalization to time periods much longer than the time periods bet~een the acoustic pressures to be measured.
The acoustic signals to be measured or sensed by the hydrophone consist of alternating changes in the surrounding ~luid pressure. Because these chan~es are not equalized by ~~ -lO- ~ ~O S~ 9 ~ ~
he above bladder mechanism, they instead cause the compliant membrane 7 to expand an(l contract radially, thus longitudinally stretching and compressing the etched single mode fiber 8.
For ~hose applications which require the hydro-acoustic 5 sensor o~ FIGS. 5, 6 an~l 7 to be in motion while it is being used to sense hydro-acoustic signals, the invention provides strength strands, for ecample, fibers 8' in FIGS. 6 and 7 , attached parallel to the axis o~ the rigid cylinder. The fibers 8' are cemented to the outside and/or inside surface o ~he lO `compliant membrane 7 and extend under each clamping ring 14 about the rigid cylinder 6. The clamping surfaces are the portions 13' of the rigiid cylinder to which the compliant mem-brane 7 is attached. Such strength fibers 8' may be of Kevlar, a tire-cord fiber made by DuPont, OT glass. Such strength 15 fibers 8' are placed to increase the longitudinal strength of the compliant membrane ;r. Therefore, if the hydro-acoustic sensor of FIGS. 5, 6 ancl 7 should be accelerated in the direction of the axis of the rigicl cylinder 6 then the resulting deform-ation of the compliant membrane 7 will have been lessened by 20 the strength fibers ~' 3n FIGS. 6 and 7. Further, the strength fibers 8' when placed pclrallel to the axis of the rigid cylin-der 6 will not substant;ially increase the resistance of the compIiant membrane to rcLdial contraction as caused by the acous-tic signals to be sensecl. FuTther, the mass of the compliant 25 membrane may be varied as well as the density of the single mode fiber turns 8 which wil~ ha~e the affect of shifting its hydro-acoustic frequency responseO
The invention also provides the hydro-acoustic sensor of FIGS. 8 and 9. FIG. ~, is an end view of the hydro-acoustic 30 sensor and FIG. 9 is a c:ross sectional view of the sensor of FIG~ 8 wherein 7-9 designates a compliant membrane within which is a helix of single mocle fiber 8-9. The assembly also in~ludes an inner cylinder-202 oi resilient complian~ material such as silicone rubber, which is in contact with the inside wall of 35 the compliant membrane 7-9. Strength strands, for example, fibeTs 201 are located parallel to the axis of the compliant membrane 7-9 and a~e in mechanical contact with the inner cy-linder material 202 so as to increase ~he longitudinal strength of the inner cylinder w;thout significantly modifying its radial compliance. The strength fibers Z01 may be made of Xevlar or glass an(l may also be continued in length beyond the ends of the comp]iant material and then may be used to anchor the sensor ;n place. The inner compliant cylinder 202 may also be continued in length to aid in positioning or anchoring the sensor. The in~ention also provides tha~ the streng~h fibers 20l be mechanically attached to the outside of the compliant membrane 7-9 parallel to the axis of membrane 7-9 to increase it5 longitudinal strength. Increasing the 1~ longitudinal strength of the sensor not only increases sensor durability but also reduces the amount of radial expansion and contraction which results from longitudinal acceleration of the sensor, without reducing i~s ~esponse, i.e., radial expansion - and contraction, to acoustic signals.
The hydro-acoustic sensor of FIGS. 8 and 9 operates as follows: The sensor is immersed in the solution contain-ing the acousticsignals. The periodic alternations o pres-sure present in an acoustic signal cause ~he compliant ~embrane 7-9 to expand and contract. As the membrane 7-9 expands and contracts, it stretches or compresses the etched single mode optical fiber 8-9, thus, as previously set forth~ modulating the electro-magnetilc radiation traveling inside the core of the fiber 8-9. Furth~r, since the inner cylinder is also radially compliant it will ofer less resistance to the ex-~5 pansion and contractill~n of the compliant membrane. The in-ven~ion provides ~h~ ~he sensor of FIGS. 8 and 9 may use etched single mode Eiber for the single mode fiber 8-9.
The etched single mode fiber of the invention has utility in all eneTgy sensors, which use a signal energy to longitudinally stretcll or compress a single mode fiber to cause a change in the fibers optical path length~ Some energy sensors may employ :Low order mode optical fiber when the modal dispersion of such Eibers is low enough to maintain sufficient optical coherence throughout the interaction length. For these energy sensor; the invention supplies the thin fiber of FIG. 4. It is to be noted ~hat any optical fiber can be etched to inCTeaSe its sensi1:ivity to longitudinal stretching and com-pression, such as m~ ;imode step index or graded index fibers.

-12~ S ~

In cases wherein the glass cladding of step index or graded index ~iber is removed by etching so as to impair its electro-magnetic racliat.on guiding abilities, the invention provides that the resul1ing fiber may be coated with a material-as at 2-2' PIG. 2 of the dral~ing, having an optical index lower than the fiber core such as RTV 670 silicone rubber to restore its ability to guide electro-magnetic radiation. Therefore, the invention includes in its scope "etched optical ~iber" as well as "etched single mode optical fiber" and throughout the specification and claims these terms can be interchanged when-ever etched fibeT has low enough modal dispersion to suffi-ciently maintain optical coherence for proper operation of the device and when the purpose of using etched optical fibers is to increase sensitivity to longitudinal stretching OT compres-sion, or when the p1l:rpose is to provide a low order mode-fiber, i.e., one with low modal dispersion, or for both of these pur-poses simultaneously.

Manufacturing -~
The inventioll also provides the following method of manufacture of fiber optic energy sensors which may employ etched single mode fiber. First, a orm is made which will maintain the fiber 1:o be etched in the same configuration or arrangement as it is to be used in the particular sensor being constructed. In the case of the hydrophone of FIGS. 5~ 6 and 7, the fiber is config-lred as a helix. A suitable form for this hydrophone is a cylinder 18 as shown in FI~. 10, around whichL is cut a spiral groove 1~9 in ~Yhich is wound an unetched optical fiber 20'. If it is desired that the form be remo~ed af~er etching, the form material should be a material which can be melted or dissolved into the liquid state at temperatures or with solvents whichLwill not cause dam-age to the fiber or the compliant membrane material. Such a form material is beeswax. ~urther, because some form materials may jeopaTdize even etching of the fiber ~wax may rub onto the fiber in places, thus shielding it from the etchant~, forms of such materials are first lightly coated by dipping or spraying with solution of a guard material 21' in FIG. 11, which upon hardening will not affect the etching process.

- 1 3 ~ )S~

Suitable guard materials are: ~ype 139 Low Andex Plastic Cladding Solution produced by Optelecom, or Kynar, a viny-lidene fluoride manufactLred by Pennwal~ Chemical Co In cases where the fiber to be etched will not have a sufficient s glass clad ~o guide lig~t, the invention provides ~hat the guard material have.a lower optical index of refraction than that of the fibeT core. Type 139 Low ~ndex Plastic Cladding Solution or Kynar have op~ical indices lower than that of silica glass.
; 10 ~hen necessary, the invention also provides that the iber be cemented to the form at peThaps the ends of the etching core as shown at 22' in FIG. 10. The guard materials already mentioned will suffice as ceme~t.
. When it is necessary to protect portions of the fiber from the etchant, these portions may also be coated with the guard materials already mentioned as shown at 23 7 in FIG. 10.
If the fiber optic energy sensor is *o employ etched singled mode fiber, the invention next causes the form 18' with fiber 20' in place as shown in FIG. 1~ to be placed in a bath of either hydrofluo~ic acid or hydrofluoric acid buffer-ed with ammonium bifluoride 7 or any other chemical which can dissolve or remove the glass clad of the ~iber. In general, this etching bath is u.ltrasonically agitated, if necessary~ to facilitate entry of the etchant around all portions of the fiber which must be etched.
After the etching period, (which may be determined empirically),has concluded the form wi*h the IIOW etched fiber in place is removed from the bath, washed in water, dried and then dipped in and Tem~oved from a bath of dissol~ed or molten coating material, or sprayed OT othe~wise coated with a solution o the material which when cured, dried, or cooled beeomes the compliant membrane mat.erial. Ultrasonic agitation of the bath is carried out when necessary to facilitate entry of the coat-ing liquid around all portions of the fiber. The invention also provides that the appl.ication of the coating material may take place in a ~acuum to aid uniformity of coating and elimination of air pocXets.

In cases using a fiber l~hich after etching does no~
have sufficient claddi~ng thickness to cause electro-magnetic radiation to be guided in the core, the invention provides that the coating material have an index of refraction which is lower than the core material. Such coating baths may be either of the guard materials already mentioned or silicone rubber such as General Electric Company's RTV 670. The viscos-ity of the coating bat}l may be varied as a means of control-ling the thickness of the coating which remains on the form upon its removal from 1:he bath. Lo~er coating bath viscosi-ties will provide thimler coatings. The form having been r~-moved from the coating bath is then rotated until the coating has hardened in order to achieve an even coat in the presence of gravity. ~IG. 12 depicts the form and fiber of FIG. lO
after ~he etching and dipping processes have been completed.
The etched fiber is den,oted as 20-E and the compliant membrane material is denoted as 124.
After the cc~ating in FIG. l2 has solidified, holes are drilled through''the coating and guard materials which ex-tend into the form mat:erial. The loation of such holes mustbe chosen so as to allow the form material to be removed by melting or dissolving but without damage to the fiber within the coating. Such a hlole 125 is shown in ~IG. 12. The form may be of a material such as teflon which may be cooled with liquid nitrogen to caulse the ~orm to shrink away from the com-pliant membrane and guard material 9 thus facilitating its re-moval through a much larger opening 1~6 in FIG. 12 formed by cu~ting away the membrane at the plane marked P in PIG. 12.
FuTther, it is contemp~lated the fo~m may be collapsed to facili*ate its removal. For the hydrophone of FIGS. 5, 6 and 7, a suitable form whiic]i could be collapsed is shown in end viewl PIG. 13.
FIG. 13 is an end view of a cylinder 257. A ~e-movable section of th'is cylindeT called a key is shown at 256. The key 256 exten(ls parallel to the axis of the cylinder and for the entire leng1:h of the cylinder. The arrow labeled ZZ illustrates the mo1:ion of the key ~o facilitate its removal.
Upon the key's remova], the cylinder 257 ~-ill collapse radially thus allowing its removcLl from the compliant membrane material after etching and dipping.
If it is desired to produce a fiber optic energy sensor which uses a compliant sleeve o~ envelope t:o contain unetched optical fibers, then the invention also allows the etching ancl washing steps to be eliminated from the above-detailed manu~acturing process.
Optical Demodulation System The optical demodulation sys-tem included in the invention is shown in FIG. 14. Referring to FIG. 14, 24 designates an optical fiber upon which are mounted pairs 25 of length limited distributed Bragg reflectors. Length limited distributed Bragg reflectors are, as used in the invention, devices which cause particular wavelength bands of electro-magnetic radiation travel-ing in the optical fiber to be in part reflected back to the source and in part transmitted onward through the optical fiber and allow light which is spectrally outside of these part:icular wavelength bands to be transmitted on-ward through the optical fiber nearly unaffected. Such reflectors may be con-structed by inducing spatially periodic perturbations of the optical index of the clad surrounding the core of an optical fiber so that the spatial period exists in a direction parallel to the axis of the core and the required length of the spacial period does not exceed the length over which optical coherence is maintained Eor the coptical fiber. Spatially periodic perturba-tiOIls can be inducecl by partially removing the clad from a length of tho fiber and then placing tho Eiber ugainst arl optical grating so that the teeth of the grating aro perl~elldicular to the axis of the core. The ma~nitude of tho reElectivity m.ly be increclse~l or decreasecl by removing more or less of the clad thus placing the optical gratitlg nearer or further from the core as disclose~l in above-lllentionod U.S. Patent No. ~,268,116. Such reflectors may also be constructed using the metllod developed by ~lill et al and described in "Photosensitivity in Optical ~iber Waveguides: Application to Reflection ~ilter labricatior~ pplied Physics Letters; #32~10), 15 May 1978, where it is shown that a reflection wavelength band occurs at:

-16- ~ ~S ~

~ -~ 2nd ~ EQ III

where ACM i.s the ce~ter of the reflection wa~elength b~nd for a particular ~21ue of 'I;
n is the effective optic index o~
- refraction ~or the optical fiber core, d is the sp~tial period of ~h.e per-turbations which create the Bragg reflector;

and M is an inte~er which is ~eater than zero ænd will be called the order o~
- the reflec~ion band.

The width, ~cM, is the full spectral width of a particular reflection band meas~ired at half of the total reflected in-tensity o which the particular Bragg reflector is capable.It is shown in the prior art.to be:
~cM ~ c~2 EQ IV
2nQ
- where ~ is t;he length of the length limited Bragg reflector.
Referring again to FIG. 14,the pairs 25 of reflec-tors are labeled A, B, C, .O.. Both reflectors in each pair are made to partially reflect the same wavelength bands and to have the same transmis;ion spectra by for instance~ adjust;ng d and Q. However, each pair is made so that it reflects partic-ular wavelength bands which are spectrally different than the reflection wavelength ~ands of all other pairs, again by ad-justing d and Q in accordance with EQ II1 and EQ IV, 50 that : there is a wa~elength :interYal h.I. which contains at least one of only these particular wavelength bands for each reflector pair to be used.

~ )s~

Each pair 25 forms a Fabry-Perot type interferometer .
inside of the single mode fiber 24. This Fabry-Perot type interferometer is sensitive only to electro-magnetic radiation which is spectrally within the reflection ~Yavelength bands o~
the distributed Bragg reflectors ~hich form the particular pair.
FI`G. 15 is an illustration of the tTansmission of a particular reflector paiT. Referring to FIG. 15, the ordinate represents the tTansmission of electro-magnetic radiation through the par-ticular reflector pair and the absicca represents the wave-length of electro-magnetic radiation which is traveling inside the fiber 24 and is incident on the reflec~or pair.
Electro-magnetic radiation which is spectrally outside of the reflection wavelength bands of a particular pair is transmitted pTactically unaffected. Such radiation is shown as regions lS a in ~I~. 15.
The maximum amount of electro-magnetic radiation traveling inside the fiber and spectrally within the reflector bands of a particular reflector pair will be transmitted on-ward through the reflector pair when the wavelength .
~ _ 2( , ~ EQ V
where OPL is the opti~al path lQn~th between the reflectors;
and N is a positive integer.
If ~ = 2(0PL) EQ VI

then a minimum ~mount of the elec~ro-magnetic radi~tion will be tra~s~ltted onward throu3h the T~e~lecto~ ~ir.
This results in a spectrally periodic tTansmission as is sho~Yn in ~egion b of FIG. 15.
As is taught in the field of interferomet~y, the spectral width of the transmission peaks which are designated 300 in FIG. 15 may be altered ~Yith respect to the spectral : separation of the tra.nsmission pea~s ~ b~ changing the ~~ -18~ J~ ~ 3 magnitude oE the reflectivity of the length limited Bragg re~lectors which for~l the pair of re~lectors responsible for the transmission peaks. This can be achieved as previously discussed.
The numbe* of peaks 300 in wavelength region b of FIG~ 15 is given by:

.
X ~ Z EQ VII
Q
where Z is the geometric le~gth bet~reen the ref:Lectors as mea~ured along the axis of 1;he single mode ~iber;
~nd Q is 1;he length of the distributed .Bragg re~lectors as measured along the fiber axis .
As the optical path le~ngth between the two reflectors of a pair changes, the transmission peaks shown in FIG. 15 within wavelength region b will shift spectrally within this region b, as is indicated by EQ V.
The invention causes some or all of the length of the optical fiber 24, located between the two reflectors of a pair, to be the interaction length of a fiber optic energy sensor, e.g., the acoustic energy sensor of FIGS. 5, 6 and 7.
As previously explained, such sensors operate by allowing signal energy being detected 1:o longitudinally stretch or compress a length of optical fiber thus changing its optical path length.
Therefore, for a reflector pair B, for example, inside of which is located an interaction length of a fiber optic energy sensor which is detecting a signal the transmission peaks of region b of FIG. 15 o~ this pair B will shift spectrally as caused by ; the signal energy being detected.
Referring again to FIG. I4, the invention uses a wavelenght scanning laser 26 to supply electro-magnetic radia-tion which is injected with suitable focusing lenses 27 into the single mode fiber 24 upon which are located the reflector pairs 25. The output of the laser 26 is scanned or chirped over a particular wavelength range. FIG. 16 contains a graph of a laser output suitable for the invention. The scan range is ~L, and is so labeled in FIG. 16. The scan time interval is ~T and is also labeled as such in FIG. 16. The scan rate is L The invention chooses the scan range of the laser 26 to ~e the wavelength interval W.I. previously men~ioned so that a reflector wavelength band region b of FIG. 15 of each pair 25 in FIG. 14 lies spectrally within the scan range.
.Once again referring to FIG. 14, the assembly includes a beam spli~ter 127 to direct a portion of the laser output beam to a Fabry-Perot interferometer 28 hereafter called "ref-erence Fabry-Perot".interfe~ometex. When the output wave-length ~L of the la.ser is such that (Q-1-2) ~ D EQ VIII

where Q is a positi~o integer and D is the optical path length between the reflectors formi~g the Fabry-Perot inter~erome~er 20, then the reference ~abry-Perot interferometer 28 will transmit some of this radiation to the photodetector 2g of FIG. 14, whic~
will then produce a:n electrical reference signal. Photodetector 29 is a commercial device, e.g. ~TIXL 45~ produced by Texas Instruments Inc., the output of which is an electrical signal : ~ 25 the amplitude of which is a known function of the amplikude of the incident radiation. If the laser is scanning as in FIG. 16, the transmitted output of the reference Pabry-Perot interfero-meter 28 will be a ;eries of temporally separated peaks each one corresponding to a resonance of the reference Fabry-Perot interferometer 28.
The invention arranges the optical path length D of the reference Fabry-Perot interferometer 28 and chooses the spatial periods of the reflectors in each reflector pair 25 so that for the laser scan range ~L a transmission peak of the reference Fabry-Perot interferometer 28 occurs at a wave-length very near the reflector wavelength band of each reflect-or pair 25.
Again referri.ng to FIG. 14, the output end of the -20- ~ Z(~

single mode ~iber 24, i.e.~ the opposite end from that into which the laser beam in injected, is associated with a suitable focusing mechanism 3,2 focused into a Fabry-Perot inter~erometer 130 as shown in FIG. 14. The output o the interferometer 30 is directed onto a photodetector 31.
The prior art of interferometry and the preceding explanatiOn of the spectral transmission of a reflector pair ; 25 shows that if the scanning laser is at some particular time injecting a particular wavelength AL of electro-magnetic radiation into the iber which falls wi~hin the wavelength region b, shown in ~IG. lS, of a particular reflector pair A, then-this electro-magnetic radiation will be transmitted through the partioular reflector pair A~ through the remaining fiber, *hrough all other reflector pairs (since the invention causes all other reflection wavelength bands of all other re~lector pairs, B, C, etc. to be different) 9 through the Fabry-Perot interferometer 30 and onto the photodetector 31 at maximum intensity whenever the injected radiation l~avelength ~L is spectrally centered on a par~icular ~ransmission peak of PIG. lS of the particular reflector pair A and when this particular transmission peak is also spectrally coincident with a transmission peak of the Fabry-Perot interferometer 30~here- -after called the analyzer Fabry-Perot interferometer.
The optical pal:h length TR ~(n)(Z) between the reflectors ~of, e.g., reflector pair B is arranged so that for a par~icular wavelength region bet~reen ~lD and ~2D the reflector pair B will produce SR trnasmission peaks. This will occur if:
.
TR ~ /1 - 1 ~ EQ IX
~A1D ~2D) 30 FOT the same wavelengl~h region between ~lD and ~2D 9 the analyzer ~abry-Perot interferometer 30 will produce SA trans-mission peaks if:

¦_ A _ ¦ EQ X
¦ (~lD . A2D~
where TA is the optical path length between the reflectors o~ the analyzer Fabry-Perot inter~erometer 30.

-21~ S ~ ~-t~ ~

As previously explained, i~ a signal is detec~ed by a fiber optic energy sensor which is located between the length limited Bragg reflectors of, for example, pair B, then the transmission pea};s of region b of FIG. 15 of pair B will undergo a spectral shift, ~SR , within the region b. By adjusting the relative values of SA and SR using EQ IX and X
the invention actually amplifies this spectral shift,a~sR by causing the resulting spectral shift, ~SA of the transmission~
of pair B and the analyzer interferometer 30 combined as in 10 FIG. 14 to be:
a~SA ~z u~sR EQ XI
where U is the ampli~cation and is, for example:
U ~ fSR _ EQ XII
fsR _ SA
for sA = (f)(sR) ~ 1 and SA and SR ~ 2 and f i5 a positive integer To better explain the demodulation system and to demonst~ate some of .the ].ess evident constraints to be con-sidered for its imple:mentation, an example of the system of FIG. 14 with the addition of energy sensors placed within the pairs .25 as previously detailed ~ill be detailed chrono-logically through two laser scan intervals. The laser scan begins at ~l which does not fall within the.reflection wave-length bands of any of the pairs ~5. As the laser outputwa~elength srans in t.ime, it will eventually begin to scan across the transmission peaks of a particular pair A. A~
this time the reference interferometer .28 transmits a pulse _ of laser light to the photodetector 29 ~hich then delivers an electrical pulse to t.he time demodulator 33. This electrical Teference pulse is used in the time demodulator 33 to reset and start an electrical c:Lock. The time demodula*or 33 ~lso counts ~he reference pulses in one scan interval and dependin~ on the number of the pulse, directs the final output of the electric 1 clock to one of the electric outputs corresponding to the parti-cular reflector pair ~-~hose transmission pea~s are being scanned at that time. Such ell-ctronic circuitry can be readily a~quired from present commercia.lly available products.

Returning to the e~ample of the system, the laser ou~put is nol~ beginning to scan over the ~ransmission peaks of pair A. When the laser output wavelength is within the first peak of pair A, at ~2, the laser light travels through pair A and all other pairs and eventually to the analyzer interferometer 30O For the sake of the explanation, assume .that the system of FIG. 14 is designed to provide an ampli-fication U = 100 by ~laking use of EQ XII with SR =~10. Also for simplicity, assume that the interval between ~lD and ~2D
of EQ X and XI for ea.ch reflector pair in ~he example system is spectrally coincid.ent with region b of ~IG. 15 for each reflector pair. Therefore, for U = 100, SR = 10, then SA = 99.
- Assume also that the analyzer in~erferomete~ 130 has a peak which is spectrally coincident with the first peak of pair A. TherefoIe, the laser light is tTansmitted on~o the photodetector 31 which produces an electrical output which when delivered to the time demodulator 33 stops the electric clock the final output of which is an electric signal corresponding to the time on the clock and is delivered to the electrical lead or leads labeled A. As the la.ser continues to scan eventually the wave-length o its output nears the transmission peaks of pair B.
Again the reference i.nter~erometer 28 transmits a pulse of light which causes photodetector 29 to produce a pulse which resets and starts the clock and prepares electrical lead or leads B for the final output of the clock.
As the sigr.Lal being detected by the energy sensor of pair A changes, the t:ransmission peaks of pair A shift spectrally.
Assume that the signa!l has caused the peaks to shift ~9 t~A~
sometime prior to the second laser scan. When the second laser scan begins, the output wavelength is again ~l . Soon after the beginning of the scan, the laser output again nears the first transmission peak of pair A, the output wavelength is approxi-mately ~2 ~ .. This wavelength, however, is not coinci-dent with a pea~ of 1:he analyzer interferometer 30 so no light is transmitted to the photodetector 31 to stop the clock. However, as the laser continues to scan its output will at a later time be ~2 -~ gl9a~ + a~, which is the n~w spectral location of the second transmission peak of pair A. By the previous equa~ions for the amplification 1~ of this example system, ~2 ~ A
is also the spectral location of the analyzer inter~e-rometer 30 transmission peak ~hich is spectrally nex~ to that peak 5 located at ~2 . There~:Eore, transmission occurs through analyzer interferometer 30 and the photodetector 31 produc~s a signal which stops the clock. Even though the transmission peaks of pair A shifte(l only ~ ) , the output of the com-bination cf pair A and the interferometer 30 did not occur un-lO til the laser outpu$ wavelength reached ~2 ~ result-ing in a spectral amp]L:ification of lO0. The r9e~ainder of the second scan interval p:roceeds as described in the first laser scan interval.
The.implementa1::ion of the demodulation system ~equires 15 particular attention to the band~idth of the optical fiber 24.
The bandwidth must be sufficiently high so as to maintain the narrowness of the retu-rning reflector pair transmission peaks.
Single mode fiber wil:L suffice in most cases. Note that the . optical demodulation system can be used whenever it is desired 20 to determine the spectral motion of the fringes of a Fabry-Perot interferometer 1~ith or without the use of optical fibe~
It is also allowed that the laser light may pass through the analyzer Fabry-Perot :interferometer first and then to the Fabry-Perot interferomete~ ~hose spacing one is attempting ~o me~sure.
25 Ho~ever, if optical f:iber is used to carry ~he laser light to the Fabry-Perot inter:Eerometer being measured, and if the laser light is to pass through the analyzer interferometer fïrst, then it is necessary to select optical fiber ~ith low dispersion for.carrying light from the analyzer interferometer to the Fabry-30 Perot inte~ferometer ;ince this ligh~ will have an additionalamplitude temporal dependence as caused by the spectrally periodic ~ansmission of the analyzer interferometer. Further, the finesse of both t].l.e analy~er interferometer and the sensor interferometer must be chosen so that if none of the transmission 35 peaks is exactly coincident spectrally there will still be only su.fficient o~erlap to produce a meaningful combined output.
Finally, the example system will produce ambiguous outputs if .

3~:3 the spectral motion of the reflector pair transmission peaks is allowed to equal or exceed .1~ or is below gg~A~
Finally, the invention provides that the electrical reference signal may be derived from the same signal which s causes the laser to scan. The cTi~eria for a Teference signal is only that it must have a known position in time with respect to that position in time of any particular wavelength of the laser scan. Further, the criteria for a suitable laser scan is: first, the scan interval must occur su~ficiently often in a period of time in order to detect the highest frequency of the oscillation of the temporal position of the output of the combination of a reflector pair and analyzer in~erferometer;
second, the output wavelength of the scanning laser must be a known function of time.
To eliminate the said ambiguity for the excessive motion of the peaks the invention provides the addition o a second analyzer interFeroTneter 30B shown in FIG. 17. FIG. 17 is a schematic of a rlsplacement subsystem -for that subsystem W enclosed by the das;~ed lines in FIG. 14. This additional analyzer 30B is arranged by means of, e.g., EQ XII, to provide a lower magnification when used with the same output of the reflector pairs. From the previous explanation a lower ampli-~ication combination can provide a higher threshold for trans-mission peak motion at which ambiguity first occurs. If EQ XII is used to establish the amplification U then the threshold spectral shift is:
fA~ . EQ XIII
Such an arrangement using two analyzer interferometers ea~h causing different amp:Lifications rould be implemented as follows: the first interferometer 30 could have-as in the previous example of the demodulation system SA = 99 transmission peaks between ~lD and ~2D. The re~lector pair cculd have SR -- lO
-peaks between ~lD and ~2D and the additional interferometer 30B could have SA - 9 peaks between ~lD and ~2D If~ for example, time demodulators 33 and 33B provided analog outputs then for a particular shift ~SR corresponding to pair A

-25- ~ ~6)~ 8~3~D

~he electrical OUtpllt of time demodulator 33B corresponding to pair A would be a voltage e such that el = K~AsR Ul EQ XIV
whe.re K is a constant Ul is the amplification which equals 10 for SR ~ 10 and SA = 9 and A~SR is the spectral shift of ~he peaks of r~gio~ b of FIG. 15 corresponding to ~lect:rical lead A
The output e2 of demodulator 33 would be:
e2~ SR U2 EQ XV
= .LOOKA~SR
where 100 is the amplification, U2, for SR = 10 and SA = 99 Such an arrangement of course can be expanded to include many such analyzer interferometers with different amplifications simp:l~ by adding more beamsplitters such as 127' to di~ide the output of *he reflector pairs among them.
Note that from the previous explanation analyzer interfero meter 30B m-ay begin to produce ambiguity for spectral shifts smaller than .laA. However, analyzer 30 will as described produce meaningful ~utputs for spectral shifts below .l~.
One may therefore contemplate adding analyzer interfsrometers which would cause e:ither lower or higher amplification than those analyzer interferometers already pTesent in a systemO
The time ,lemodulator is an electrical device which performs two unctions: first~ producing an e~ectrical signal which by means o, :foT example, its amplitude, frequency of oscillation or phase of oscillation contains or conveys the time between the receipt of a reference pulse and the receipt of an additional pulse ca:Lled ANAL pulse, which is the pulse from the photodetec~or recei~ing electro-magnetic radiation from the analyæer Fabry-:Perot interferometer, and; second~ directing this electric signa:l to a particular output wire or to a par-ticular group of output ~ires. There are many electrical -26 ~Z~ 3 ~) circuits which can accomplish this one of which is shown achematically in FIG. 18. Re~erring to FIG~ 18, Ul and U2 are voltage comparators, for example, part #LM311 produced by National Semiconductor Corp., U4 and U5 are counters, e.g.
part #74161 produced by Texas Instruments~ U3 is a clock generator, e.g., part #74LS124 produced by Texas Instruments, U6 is a demux, e.g. 9 part ~74155 produced by Texas Instruments and U7, U8, and U9 are latches, e.g., part #74175, also pro-duced by Texas Instruments.
The circuit operates as follows: Ul and U2, voltage comparators, serve to convert the reference pulse and the additional pulse into standard TTL logic voltage levels for use in the demodulato~r. The regularly spaced reference pulses serve to reset counter U4 which is continually counting at a rate approximately 16 times as fas~ as the reference pulse repetition rate, as driven by clock generator U3. The result-ing output of counter U4 is a number which starts at 0 when the reference p~lse ic; received and counts upward until again reset to 0 by another reference pulse, starting its count anew.
Meanwhile, each time aL reference pulse is received, counter U5 is incremented. It is set to automatically return to 0 after counting the appropriate number of channels (in this case, 3).
When an ANAL pulse comes, it is routed to the appropriate latch (U7, U8, or U9) via demux U6. The counter output number is latched into the appropriate channel latch and represents the time bet~een the reference and ANAL pulses. The next ~AL
pulses causes the ~ext channel latch to store the number Tepre-senting the time between those reference and ANAL pul$es, and so on. Each time a new time count is latched~ the trailing edge of the latch pulse notifies the user that new data is available.
Statement of Industrial Application , An improved fîber optic energy sensor and method of manufacturing the sens~r, and an improved optical demodulation system is provided which is particularly sensitive to stretch-ing or compressing by signal energy to be sensed or detected.

Claims (4)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. Means for phase modulating electro-magnetic radiation traveling in the core of an optical fiber comprising: an etched optical fiber; and means for longitudinally stretching or com-pressing the etched optical fiber, wherein the etched optical fiber is coated with a material having an optical index lower than that of the etched optical fiber's core and having a modulus of elasticity lower than that of the material which was removed by etching.

2. Means for guiding electro-magnetic radiation com-prising an etched optical fiber having low modal dispersion, wherein the optical fiber having low modal dispersion is coated with a material having an optical index lower than that of the core of the fiber.

3. A method of phase modulating electro-magnetic radia-tion comprising: directing the electro-magnetic radiation into an etched optical fiber and longitudinally stretching or compressing the etched optical fiber, said method further comprising coating the fiber with a material having an optical index of refraction lower than that of the etched optical fiber's core and having a modulus of elasticity lower than that of the material which was removed by the etching process.

4. The method of guiding electro-magnetic radiation to provide low model dispersion comprising injecting the radiation into an optical fiber produced by etching a fiber of optical fiber core material, wherein the optical fiber produced by etch-ing a fiber of optical fiber core material is coated with a mat-erial having an optical index lower than that of its core.