AU688725B2 - Liquid crystal optical waveguide display system - Google Patents

Description

OPI DATE 07/06/93 APPMN. 10 24039/92 II l JIJHiI AIJP DATE 05/08/93 PCT NUMBER PCT/US92/06418 Illl! l 111111l11111l 11111111111111 AU9224039 INTERNATIONAL APPLICATION PUBLISHED UNDER THE PATENT COOPERATION TREATY (PCT) (3r) International Patent Classification 5 G02B 6/26, G09G 3/08 H01J 5/16 Al (11) International Publication Number: (43) International Publication Date: WO 93/09454 13 May 1993 (13.05.93) (21) International Application Number: PCT/US92/06418 (22) International Filing Date: 3 August 1992 (03.0.92) Priority data: P-1PGT S9459 9810 9--30 Otbr 99 -ft9)-9 WO( (34) Countries for which the regional or international application wasfiled: US et al.

(71X72) Applicant and Inventor: ROCKWELL, Marshall, III [US/US]; 303 Grenola Street, Pacific Palisades, CA 90272 (US).

(74) Agents: HIGGINS, Willis, E. et al,; Flehr, Hohbach, Test, Albritton Herbert, 4 Embarcadero Center, Suite 3400, San Francisco, CA 94111-4187 (US).

(81) Designated States: AU, BR, CA, JP, KR, US, European patent (AT, BE, CH, DE, DK, ES, FR, GB, GR, IE, IT, LU, MC, NL, SE).

Published With international search report.

*3) 1 '5 i c nf (54)Title: LIQUID CRYSTAL OPTICAL WAVEGUIDE DISPLAY SYSTEM (57) Abstract -96 A controllable waveguide display based on a cladding Tap Brightness supercladding (1 and 6) and/or core utilizing liquid crystals is Brightness Compensation described. An electric field applied through a fiber causes the liquid D ata

U

n it crystal layer to become aligned. Changes in the refractive index of 98 97 the liquid crystal layer causes light to switch out of the fiber. In one T embodiment light is coupled into a supercladding running Memo Tap Sensor alongside the core and reflected out of the fiber at a reflector pit onReceiver cut in the fiber. Parallel arrays of fibers are used to cover a substrate and make large viewing screens. A tapered supercladding 89 helps improve the contrast ratio of screens using fiber taps. A thin 87 S cladding and closely spaced dark cladding also help improve the screen contrast ratio. Color techniques based on a three core 3 fiber that shares a single supercladding is introduced. An illumi- 9 3 nation method is also taught which breaks white light into coiored f components (73) with dielectrc filters (103) to efficiently utilize white light (102). An electronic feedback system (96) is introduced which provides screen brightness and color uniformity under varying temperature and environmental conditions.

I I WO 93/09454 PCT/US92/(' 418 1 2 3 4 6 7 8 9 11 12 13 14 16 17 18 19 21 22 23 24 26 27 28 29 31 32 33 34 36 37 38 LIQUID CRYSTAL OPTICAL WAVEGUIDE DISPLAY SYSTEM BACKGROUND OF THE INVENTION 1. Field of the Invention The invention relates to large area, thin-panel, video screens. More particularly, this patent teaches improvements in the field of optical waveguide displays.

Specifically, techniques for tapping light out of optical fibers with a liquid crystal switching element are described.

2. Prior Art It has been established that displays can be made using optical waveguides. Methods employing the electro-optic, acousto-optic and thermo-optic effect have previously been described. However, further improvements are required in WO 93/09454 PCT/US92/06418 2 1 the switching devices and system architecture to improve 2 the resolution, image quality, power efficiency and price 3 of this technology.

4 Examples of related U.S. patents include: 4,737,014; 6 3,871,747; 4,786,128; 5,009,483; and 5,106,181.

7 8 Important prior art is contained in the following 9 publications: Chen S. Tsai "Guided Wave Acousto- Optics", Springer-Verlag, Berlin, 1990; R.G. Hunsperger 11 "Integrated Optics: Theory and Technolog", Springer- 12 Verlag, Berlin, 1991; K. lizuka "Engineering Optics", 13 Springer-Verlag, Berlin 1987; Bernd R. Hahn and Joachim 14 H. Wendorff, "Piezo and pyroelectricity in polymer blends of poly(vinylidene flouride)/poly(methylmethacrylate)" 16 Polymer, 1985 Vol. 26, October Pages 1611-1618; Bernd R.

17 Hahn and Joachim H. Wendorff, "Compensation method for 18 zero birefringence in oriented polymers", Polymer, 1985, 19 Vol. 26, October Pages 1619-1622; Bernd Hahn and Joachim Wendorff, "Dielectric Relaxation of the Crystal-Amorphous 21 Interphase in Poly(vinnylidene fluoride) and Its Blends 22 with Poly(methylmethacrylate), Macromolecules 1985, 18, 23 Pages 718-721; M. Gottlieb G. B. Brandt, "Temperature 24 Sensing in Optical Fibers Using Cladding Jacket Loss Effects", Nov. 15, 1981, vol. 20, No. 22, Applied Optics; 26 Karl F. Schoch and Howard E. Saunders, "Conductive 27 Polymers", IEEE Spectrum, June 1992, Vol 29, Number 6, 28 Pages 52-56.

29 31 SUMMARY OF THE INVENTION 32 33 1. Objects of the Invention 34 Accordingly, it is an object o invention to show how 36 a liquid crystal c it ormed inside a flexible waveguide 37 struc -e--an be used to switch light.

The present invention provides an optical waveguide display system comprising: at least one optical waveguide with electro-optic switchable elements placed along a length of said at least one optical waveguide; at least one photosensor coupled to receive light from said at least one optical waveguide; and a control circuit coupled between said at least one photosensor and said electro-optic switchable elements to control said electro-optic switchable elements based on light sensed by said at least one photosensor.

So

S*

(N:\LIBTTO1356:hrw WO 93/09454 PCT/US92/06418 1 DRAWING DESCRIPTIONS 2 3 FIG. 1 is an end view of a single core electro-optic 4 fiber.

6 FIG. 2 is an end view of a three core electro-optic fiber.

7 8 FIG. 3 is an end view of a liquid crystal fiber with 9 ridges in the supercladding main chamber and intermediate cladding layers.

11 12 FIG. 4 is an end view of a fiber with a supercladding that 13 reflects light into the supercladding main chamber from 14 the side.

16 FIG. 5 is a side view schematic of an electro-optic fiber 17 which shows how light is tapped out using a wave 18 description of light.

19 FIG. 6 is a side view schematic of an electro-optic fiber 21 which shows how light is tapped out using a ray-optics 22 description of light.

23 24 FIG. 7 is a side schematic view which shows how light is scattered in a Raman-Nath diffraction device.

26 27 FIG. 8 is a side schematic which shows how light is 28 scattered in a Bragg diffraction device.

29 FIG. 9 is a plot which shows how the index of refraction 31 in a liquid crystal material can be adjusted by applying a 32 large, changing, electric fields.

33 34 FIG. 10 is a schematic view of the control electronics used in a waveguide display system.

36 37 FIG. 11 is a schematic view of a feed-back system used to 38 correct waveguide tap irregularities.

WO 93/09454 PCT/US92/06418 1 2 FIG. 12 is a fiber-end view of showing how light is output 3 to the viewer through a forward scattering diffusing 4 layer.

6 FIG. 13 is a fiber end view of showing how a Fresnel lens 7 is used to compensate for lateral fiber spacing 8 irregularities to create a straight pixel line along the 9 direction of light flow in the fiber.

11 FIG. 14 is a fiber end view showing how a reflective 12 element can reduce pixel column irregularities.

13 14 FIG. 15 is a schematic view of an efficient color display which works .by separating broad-spectrum light into 16 separate colors.

17 18 FIG. 16 is a schematic view of a rotating filter disk 19 system used to make color images.

21 FIG. 17 shows an end view of a finished waveguide ribbon.

22 23 FIG. 18 is a backside view of a finished waveguide ribbon.

24 FIG. 19 is a front view of a waveguide ribbon that shows 26 how staggered reflector pits on separate fibers enables 27 long interaction length taps to be used and still maintain 28 high screen resolutions.

29 DRAWING REFERENCE NUMBERS 31 32 1 Supercladding main chamber 33 2 Dark cladding 34 3 Cladding 4 Core 36 5 Electro-optic material, or 37 Liquid crystal chamber, or 38 Liquid crystal material WO 93/09454 PCT/US92/06418 1 2 3 4 6 7 8 9 11 12 13 14 16 17 18 19 21 22 23 24 26 27 28 29 31 32 33 34 36 37 38 6 7 8 9 10 11 12 14 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39a 39b 40a 40b 41a 41b 42a Tapered supercladding Control electrode Conductive separation barriers Reference electrode Reference electrode Intermediate cladding Ridged intermediate cladding or core Irregularly shaped supercladding main chamber Electric field Light collecting in supercladding main chamber Core and cladding have same refractive index Light escaping from the core Control electrode charged with a voltage Control electrode with no voltage Light guided in the core Large amount of light in core Intermediate amount of light in core Small amount of light in core Light beam Grating spacing Interaction length Width of interaction medium 2nd order beam 1st order beam Oth order beam 1st order beam 2nd order beam Reflector pit cut in supercladding main chamber Grating elements Oth order beam 1st order beam Highest voltage Increasing refractive index High constant bias voltage High constant refractive index Lowest voltage Decreasing refractive index Low constant voltage WO 93/09454 PCT/US92/06418 1 2 3 4 6 7 8 9 11 12 13 14 16 17 18 19 21 22 23 24 26 27 28 29 31 32 33 34 36 37 38 42b 43a 43b 44a 44b 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 66 67 68 69 70 71 73 74 75 76 77 78 79 Low constant refractive index Highest voltage Increasing refractive index Intermediate voltage Intermediate refractive index Fiber reference electrode connector Light detector Viewing window Selected tap electrode row Tap electrodes Selected intensity modulator Intensity modulator control electronics Optical waveguide or fiber Light Tap control electronics Light scattered to the viewer Ridged scattering material Light barrier Fiber or optical waveguide Light emitted from fiber Protected clear layer mounted on fiber front Fresnel lens element Reflective surface Forward scattering diffusing layer Reflective element Transparent support medium Ribbon waveguide element Microlense array Focusing lens 3 colors of light enter at 3 different angles Green light Red light Mirror Light detector Mirror Mirror Collimating lens Collimating lens WO 93/09454 PCT/US92/06418 1 80 "Cold" dielectric mirror 2 81 "Hot" dielectric mirror 3 82 Light source collimating mirror 4 83 Light source 84 3 element coloring filter 6 85 Electric motor to spit filtering element 7 86 Synchronization signal 8 87 Signal applied to control electrode 88 9 88 1st control electrode or fiber tap 89 Signal applied to control electrode 11 90 2nd control electrode or fiber tap 12 91 Photodetector 13 92 Photodetector at fiber tap 14 93 Light in supercladding main chamber 95 Sensor receiver, i.e. A/D converter 16 96 Brightness compensation unit 17 97 Tap controller 18 98 Memory 19 99 Light tapped out of first tap 100 Light in 2nd tap's supercladding main chamber 21 101 Beam shaping optics section 22 102 Collimated beam of filtered visible light 23 103 Red reflecting only dielectric mirror 24 104 Blue-green light 105 Blue reflecting only dielectric mirror 26 106 Green light 27 107 Support film 28 108 Upper Fresnel lens element 29 109 Fiber 110 Vertical ribbon control line 31 111 Light traveling through ribbon core 32 112 Ribbon feed-through 33 114 Waveguide ribbon back view 34 115 Ground support layer or film 116 Metal ground layer 36 37 38 WO 93/09454 PCT/US92/06418 1 DETAILED DESCRIPTION 2 3 FIBER TYPES 4 FIG.'s 1-4 show different types of fiber waveguide end 6 profiles. In FIG. 1 light is carried through the core 4.

7 Light is confined by a low index cladding 3 surrounding 8 three sides of the core 4. The fourth side contacts a 9 thin, low index intermediate cladding 11 on the top side of the core 4. Intermediate cladding 11, the use of which 11 is optional, is made from a low index material closely 12 matched to the core 4 index of refraction. The 13 intermediate cladding 11 is thin enough to enable the 14 evanescent field of core 4 guided light to extend into the electro-optic layer 5. It is the electro-optic layer 16 that switches light out from the core 4 and into the 17 tapered supercladding 6.

18 19 The addition of intermediate clading layer 11 enables the construction of a fiber which can guide long distances.

21 Since the intermediate cladding layer 11 is made out of a 22 high quality optical material similar to that used in the 23 cladding 3, the losses caused by the electro-optic layer 24 are minimized. Light is mostly confined within traditional cladding materials in regions 3 and 11, with 26 only a fraction extending into the electro-optic region 27 The cladding 3 is typically very low index with respect to 28 the core 4.

29 The use of the intermediate cladding layer 11 means less 31 of the light will extend into the electro-optic layer 32 Therefore, longer tap lengths will be required to get 33 light out of the core. Long tap lengths translate into 34 decreased screen resolution. Therefore, fiber guiding properties can be traded off against tap switching 36 efficiency and tap length by the addition of the 37 intermediate cladding layer 11.

38 WO 93/09454 PCT/US92/06418 j0 1 There is a relationship between the index of refraction of 2 intermediate cladding 11 and how thin the intermediate 3 cladding 11 must be to enable the evanescent field of core 4 4 guided light to extend into the electro-optic layer If the refractive index of the intermediate cladding 11 is 6 much lower than the core 4 refractive index, the 7 intermediate cladding 11 must be very thin to enable 8 evanescent field of core guided light to extend into the 9 electro-optic layer 5. If the refractive index is slightly less than the core's 4 refractive index, the 11 intermediate cladding 11 can be much thicker, and still 12 have the evanescent field of core 4 guided light extend 13 into the electro-optic layer 5. Multiple intermediate 14 cladding layers 11 (Not Shown) may be used with 'ifferent indexes of refraction to achieve complicated light 16 distribution patterns within the core 4/electro-optic 17 layer 5 boundary if desired.

18 19 Many electro-optic materials used in layer 5 significantly scatter and absorb light. In fact, the very property that 21 makes electro-optic materials attractive for use as a 22 switching element, such as easily changing their index of 23 refraction, tends to increase scattering losses due to 24 index of refraction non-uniformity's in the electro-optic material 5. The intermediate cladding layer 11 helps 26 minimize losses due to irregularities in the electro-optic 27 layer 28 29 A different method of making the fiber guide long distances (Not Shown) is to make the low index cladding 31 material 3 wrap around a majority of the core. In this 32 configuration layer 11 would be replaced by the cladding 3 33 and a small channel, narrower than the width of the core 4 34 would contain the electro-optic material 5. This channel contacts the core 4 directly. For example, the narrow 36 channel may contain a liquid crystal material. By making 37 the electro-optic channel very narrow, light can still be WO 93/09454 PCT/US92/06418 1 tapped out, but with a greater tap interaction length, 2 than if a wider region is used.

3 4 In general, the tap length can be reduced by increasing the core's 4 contact with the electro-optic material 6 Thus, electro-optic materials on more than one side of the 7 core 4 will reduce the tap interaction length. For 8 example, liquid crystal materials placed on either side of 9 the cnre will shorten the tap length. Also, a very wide, narrow thickness, core 4 will have a short tap length.

11 U.S. Patent 5,009,483 shows examples of a wide, narrow 12 thickness core, with electro-optic materials on either 13 side. This configuration can be used with liquid crystals 14 to form a very short length tap.

16 Switching layer 5 is preferably made of liquid crystals.

17 It may be necessary to make the liquid crystal switching 18 layer 5 wider than the fiber core 4 so that the liquid 19 crystals aligned along the vertical wall (the y-axis as one looks at Fig. 1) of the liquid crystal layer 5 do not 21 cause core light to leak by forming a high index region 22 near the walls of the liquid crystal layer 5. This wider 23 electro-optic layer 5 separates surface aligned liquid 24 crystals on the vertical walls from the core 4. Thus, the liquid crystals closest to the intermediate cladding 11 or 26 core 4 are uniform. Uniformly aligned liquid crystals 27 translate into a uniform electro-optic switching region 28 This helps make the fiber guiding and switching properties 29 more predictable.

31 An electric field extends between a control conductor 7, 32 on the outside of the fiber, and a ground or biasing 33 reference conductor 9 or 10. In the simplest case the 34 biasing conductor 9 and control electrode 7 form a field all the way through the fiber. The field between the 36 electrodes 7 and 9 causes the liquid crystals in liquid 37 crystal cavity 5 to become aligned. This alignment 38 reduces the index of refraction along the orientation of WO 93/09454 PCT/US92/06418 Ca 1 light flow in the liquid crystal layer 5. Thus, when a 2 field is applied between 7 and 9 the fiber guides light 3 well. However, when the field is removed, the liquid 4 crystals in layer 5 becomes less aligned and increase their refraction. This increase in refractive index 6 causes guided light to exit the fiber core 4 into the 7 tapered supercladding 6.

8 9 Since the electric field extends completely through the fiber when electrodes 7 and 9 are used, the drive voltage 11 to switch the fiber tap is high. A conductive material 12 10 imbedded in the fiber enables lower drive voltages to 13 achieve the same field densities. By reducing the distance 14 between the biasing electrode 10 and the control electrode 7 lower voltages can be used to switch the fiber tap.

16 Biasing electrode 10 runs the length of the fiber.

17 18 Conductive plastics made with loaded metal or graphite 19 particles may be used to make biasing electrode 10. New polymers which are inherently conductive may also be used.

21 Such conductive polymers are discussed by Karl F. Schoch 22 and Howard E. Saunders, in an article entitled "Conductivw 23 Polymers" published in the June 1992 IEEE Spectrur, 24 Either way, the conductive element should be a thermoplastic, or glassy compound so it can be put in the fiber 26 preform and stretched in the fiber drawing process. If 27 glass is used to make the fiber, a number of inorganic and 28 organic dopants can be to the glass to make it conductive.

29 A different approach to decreasing the separation of the 31 biasing 9 and 10 and control 7 electrodes is to groove the 32 fiber (Not Shown) deeply either on the sides or on the 33 top. The groove can be metalized to bring the biasing 34 electrode close to the electro-optic cladding layer This general idea of getting the electrodes closer 36 together to reduce the drive voltages is shown in the 37 fiber shown in FIG. 4.

38 WO 93/09454 PCT/US92/06418 1 It may be necessary to coat the inside of the liquid 2 crystal cavity 5 with a material which can not be 3 dissolved by the medium contained in the cavity. Possible 4 mediums contained in cavity 5 include gases, liquids and solids. For example, liquid crystal materials can cause 6 the walls of the electro-optic layer 5 in a plastic fiber 7 to dissolve. Consequently, it may be necessary to line or 8 coat the liquid crystal chamber walls 5 in plastic fibers 9 to prevent liquid crystals from dissolving the plastic chamber walls. Possible liners in a PMMA fiber may 11 include PVDF or PVDF/PMMA blends. Teflon and Teflon 12 derivatives may also be used to protect the walls of the 13 liquid crystal cavity 14 The fiber, when it is formed in the fiber drawing process, 16 will naturally cause the liquid crystals to become aligned 17 along the length of the fiber elongation. This alignment 18 is caused by the longitudinal orientation of molecules in 19 the liquid crystal chamber 5 walls along the direction of light flow. Liquid crystal surface alignment along the 21 axis of light flow may be used advantageously to increase 22 the switching speed of the fiber tap. This effect is 23 discussed in the next section on surface mode liquid 24 crystal devices.

26 Since the supercladding 6 has a higher index of refraction 27 than the core 4 and electro-optic cladding 5, light 28 switched out from the core 4 will be angled in a direction 29 normal to the direction of light flow through the core 4.

Therefore, once light enters into the tapered 31 supercladding 6 from the core 4 it is angled in direction 32 toward the main supercladding chamber 1. Refracting 33 light up and away from the core 4 is needed for the 34 subsequent trapping of light in the supercladding main chamber 1.

36 37 The supercladding 1 and 6 must have a high refractive 38 index relative to the core 4 and claddings 5 and 11 in

I

WO 93/09454 PCT/US92/06418 1 order to trap light in the main supercladding chamber 1.

2 Assuming a PMMA core 4, and a fluorinated acrylic cladding 3 3, a high index supercladding 6 can be made from 4 polystyrene or polycarbonate. If glass is used to make the fibers, Schott Glass Co. has a large range of glasses 6 that are suitable to achieve the desired refractive index 7 differences between the core 4, electro-optic cladding 11, 8 and supercladding 6.

9 Tapering the supercladding 6 to a constricted point before 11 the supercladding main chamber 1 helps accomplish several 12 important functions. First, it helps contain light 13 switched out of the core 4 in the main chamber 1. In the 14 same way a small hole in a large light-tight box causes light entering the hole to become trappeA, light that 16 enters the main chamber 1 will have a low probability of 17 re-exiting through the small opening. Making the main 18 supercladding 1 have a higher index of refraction than the 19 tapered supercladding 6 section also helps increase the amount of light that stays in the supercladding main 21 chamber 1.

22 23 Increasing the relative fraction of light carried in main 24 chamber 1 vs. the tapered section 6 is critical to increasing the screen contrast ratio. The screen contrast 26 ratio is partly determined by the amount of light that 27 "bleeds" from one reflector pit 35 to the next along the 28 length of a fiber. There will always be some cross-talk 29 between consecutive reflector pits 35 along the length of a single fiber. In other words, all the light in the 31 supercladding main chamber 1 will not be entirely 32 reflected out at a single supercladding reflector pit 33 34 Light will travel between consecutive reflector pits 3f because the pit is not cut far enough into the 36 supercladding 1 and 6 to keep all the light in the tapered 37 part of the supercladding 6 from continuing through the 38 fiber and being reflected at the next reflector pit WO 93/09454 PCTr/US92/0641 8 1 However, the ratio of light in the tapered supercladding 2 region 6 vs. light in the main chamber 1 can be adjusted 3 by making the tapered opening into the supercladding main 4 chamber 1 very narrow. A majority of the light will then be trapped and carried in the main chamber 1. Collecting 6 the light in the supercladding main chamber 1, in turn, 7 decreases the amount of light that makes it through the 8 tapered supercladding 6 to the next reflector pit 9 As light bounces against the walls in the taper 6 the 11 angle of its lateral incidence, (left-right as one stares 12 into the fiber end of FIG. increases. For example, if 13 the light in a 30 uM core 4 fiber has an angle of 14 propagation of 5 degrees, and it enters the main chamber through an opening 5 microns wide, the angle of 16 propagation will have increased to 30 degrees at the 17 constriction just before the supercladding main chamber 1.

18 The angle of incidence caused by the constriction should 19 not exceed the critical guiding angle of the supercladding 6 and 1/cladding 3 interface. It helps to have a very 21 well collimated beam of light in the core 4 from the light 22 source. Well guided core light can be 23 24 The tapered region 6 also serves to physically space the core 4 apart from the main chamber 1. Separation is 26 needed because the reflector pits 35 formed in the fiber, 27 usually via a cutting process, must not damage the core 4 28 and liquid crystal layer 5. Spacing the core apart from 29 the supercladding main chamber 1 with the tapered supercladding region 6 minimizes potential damage to the 31 sensitive core 4 and electro-optic cladding 5 during the 32 pit formation process. The tapered region 6 also allows a 33 large core 4 to be used in a fiber, and still trap an 34 acceptable amount of light in the supercladding main chamber 1. A large core 4 is easier to couple light into 36 from the light source 83.

37 WO 93/09454 PCT/US92/06418 1 The supercladding taper 6 between the core 4 and 2 supercladding main chamber 1 can be made in many different 3 shapes. It is shown in FIG.'s 1-4 only as linear or 4 straight line. However, it may also have a parabolic, exponential or even shape. In fact, an exponential 6 shaped taper would allow the biasing element 10 to be 7 placed closer to the control electrode 7 to redcce the 8 waveguide tap drive voltages.

9 The supercladding main chamber 1 and taper 6 may include a 11 pigment or dye to selectively filter broad-spectrum light 12 carried in the core 4. In this way white light coupled 13 out of the core 4 will be colored as it travels through 14 the supercladding 1 and/or 6 enabling colored images to be displayed.

16 17 The low index cladding layer 3 is made thin to help reduce 18 the manufacturing cost of the fiber. Presently, low index 19 cladding materials suitable for use in plastic fibers are expensive. Currently, they are made out of fluorinated 21 acrylics and cost around $200 U.S./pound. Thin cladding 3 22 layers reduce the amount of needed fluoro-polymers and 23 thus significantly lowers the fiber manufacturing cost.

24 Keeping the electro-optic layer 5 thin also helps reduce 26 the fiber cost. Liquid crystal material presently cost 27 around $5 U.S/cc in large quantities. Thus, confining the 28 liquid crystal material just in the thin, narrow, electro- 29 optic switching region 5 reduces the amount needed of this expensive material.

31 32 A thin cladding 3 also allows the dark cladding 2 to be 33 placed closer to the supercladding 6 and core 4. The 34 close placement of the dark cladding 2 helps prevent stray light from traveling long distances down the fiber 36 cladding 3 before being absorbed. Placing the dark 37 cladding close to the supercladding 1 and 6 and core 4 is 38 very important to improve the contrast ratio of the WO 93/094544 PCr/US92/06418 1 display. The dark cladding 2 prevents stray light from 2 traveling through the cladding 3, and core 4 and 3 supercladding 1 and 6 in non-guided modes, between 4 adjacent reflector pits 35 along a fiber.

6 The dark cladding 2 contains a pigment or dye which 7 absorbs visible light. Ideally, the dark cladding 2 should 8 have an index of refraction that matches the cladding 3.

9 Thus, stray light in the cladding 3 will enter the dark cladding 2 with little or no reflection or refraction at 11 the cladding 3/dark cladding 2 interface and be absorbed 12 by the pigment. The outside of the dark cladding 2 may 13 also be metalized to further minimize cross-talk between 14 separate fibers.

16 FIG. 2 shows how a single fiber can be used to make a 17 color display. Separate colors are carried in each of the 1U three cores 4. Different colors of light switched out of 19 the cores 4 will make it through the tapered supercladdings 6 to the main chamber 1, where they mix and 21 are reflected at a reflector pit 35. Integrating three 22 separately switchable cores 4 with a single supercladding 23 main chamber 1 into a single fiber provides space savings.

24 It also makes it easier to couple the 3 different colors of light into the fiber cores since the core to core 26 spacing can be precisely controlled during the fiber 27 manufacturing process.

28 29 The close placement of the cores 4 in FIG 2 may cause cross-talk between control electrodes 7 which determine 31 the amount of light taped out of each core 4. This cross- 32 talk can be minimized by integrating a conductive material 33 8 in the fiber between the electro-optic claddings 5 and 34 cores 4 of each fiber. Layer 8 shields the electro-optic layers 5 from electrical cross-talk from the control 36 electrodes 7. Conductive separation layer 8 may be biased 37 at different voltages with respect to the control 38 conductors 7. This electrical isolation can also be made WO 93/0)9454 PCT/US92/06418 1 by making grooves that extend from the outside into the 2 fiber. (Not Shown) 3 4 FIG. 3 shows grooves formed both in the intermediate cladding region 12 and the main supercladding 14. These 6 grooves serve two completely separate functions. The 7 intermediate cladding grooves 14 help align the liquid 8 crystals along the surface of the intermediate cladding 9 12. Liquid crystal surface alignment helps electro-optic layer 5 switch faster. These grooves can also be formed 11 directly on the surface of the core 4 to accomplish the 12 same liquid crystal alignment function if the intermediate 13 cladding 11 is not used.

14 The irregular surface in the inside of the main chamber 14 16 helps scatter light so it is less likely to exit the main 17 chamber 1 into the tapered region 6. Making the main 18 chamber 1 in different shapes also can be used to increase 19 the appearance of uniformity of the screen pixels. Since consecutive pixels are typically staggered on separate 21 fibers, as shown in FIG.'s 13-14, they will have a 22 staircase, or other irregular appearance. By using 23 lenses, and different shaped supercladding main chambers 1 24 for each fiber, this non-uniformity can be minimized.

26 FIG. 4 shows a method whereby light is launched into the 27 supercladding main chamber 1 from the side. This approach 28 enables the close placement of the control 7 and biasing 9 29 conductors. This fiber geometry could be further improved by making two, instead of one, tapered supercladding 6 31 channels on either side of the main chamber 1. Or in a 32 different configuration, a series of the shaped 33 tapered supercladdings 6 stacked on top of each other 34 could allow three colors to be used for color operation.

FIG. 4 also demonstrates how a different liquid crystal 36 orientation in the electro-optic cladding 5 can be used to 37 guide light in the fiber core.

38 WO 93/09454 PCr/US92/06418 1 LIQUID CRYSTAL LAYER 2 3 Liquid crystals undergo a large refractive index 4 anisotropy, or change in birefringence, in an electric field. Three different types of crystals: Nematic, 6 Cholestreric and Smectic are of interest in this design.

7 In each, the molecules are elongated and it is the 8 relative orientation of these molecules which determine 9 the index of refraction of the liquid crystal material.

The axial orientation of the elongated molecules is 11 described using a mathematical vector called the 12 "director".

13 14 This description will focus on fiber taps employing Nematic and Cholesteric liquid crystals. In a Nematic 16 liquid crystal, the orientation of all the molecules are 17 parallel, but the molecules are randomly arranged 18 perpendicular to the directors. A Cholesteric liquid 19 crystal uses molecules with directors arranged in a helical fashion across the material. A complete turn of 21 the helix will extend one pitch length along the helix 22 axis.

23 24 Nematic liquid crystals are used in the preferred embodiment. They can be used in two important ways. The 26 first utilizes surface alignment properties. When the 27 surface material in cavity 5 is matched with the 28 appropriate liquid crystal material, the liquid crystals 29 will tend to lay down flat on the surface along the axis the fiber was stretched in. These surface aligned liquid 31 crystals will have a higher index of refraction along the 32 direction of fiber light propagation than those aligned in 33 the electric field perpendicular to the fiber.

34 It takes a stronger electric field to cause surface 36 aligned liquid crystals to become aligned in an electric 37 field than liquid crystals distantly separated from the 38 electro-optic cavity 5 surface. Furthermore, when the WO 93/09454 PCT/US92/06418 1 electric field is removed, the liquid crystals near the 2 surface of the electro-optic cavity 5 realign faster than 3 those further away. Consequently, surface alignment 4 effects can be utilized to make very high speed fiber switching devices. Surface alignment effects are 6 discussed in Optical Shields Technology report describing 7 their surface mode device.

8 9 Chemical treatment of the electro-optic cladding layer walls after the drawing process may be necessary to induce 11 surface alignment of the liquid crystals. The ridged 12 surface of the intermediate cladding 12 or core 4 may also 13 be used to increase the alignment. The proper choice of 14 materials for the core and intermediate cladding 12 or 11 can also increase the degree of surface orientation of the 16 liquid crystal layer.

17 18 Liquid crystals aligned along the electro-optic cladding 19 walls will have a higher index of refraction than those aligned with the electric field. Thus, the liquid crystal 21 layer aligned near the surface will have a high index and 22 act as part of the core 4 to carry light. Liquid crystals 23 further from the surface will be aligned with the electric 24 field and will have a low index of refraction and act as a cladding to confine light in the core. The liquid crystal 26 transition region between the cladding and core will have 27 a varying index of refraction depending on the applied 28 electric field and the alignment properties of the liquid 29 crystal layer. However, when a field is applied normal to the direction of light propagation, the liquid 31 crystal/cladding region will typically have a graded index 32 profile ranging from high near the core to low in the 33 center of the electro-optic switching cavity 34 Anti-alignment agents may be used to prevent the liquid 36 crystal molecules from aligning on the surface walls. In 37 this case the entire liquid crystal cavity 5 will 38 uniformly switch when an electric field is applied. In WO 93/09454 PCT/US92/06418 1 this case the liquid crystals near the walls will align in 2 the electric field as easily as liquid crystals in the 3 center of the cavity. This approach can be used to make 4 high speed evanescent switching devices with a predictable, uniform, response. A non-surface aligned 6 liquid crystal cavity 5, in conjunction with an evanescent 7 coupling design, is considered to be an important high 8 speed switching innovation.

9 Ferroelectric polymers could also be used in the electro- 11 optic cladding layer 5. Ferroelectric polymers tend to be 12 self aligning. They also have very fast switching speeds.

13 The disadvantage with current ferroelectric material is 14 their irregular orientation, which increases scattering.

However, future ferroelectric polymers look promising and 16 they may also be used in the electro-optic switching layer 17 18 19 TYPES OF OPTICAL WAVEGUIDE TAPS 21 Depending cn the dimensions of the core 4, electro-optic 22 cladding 5 and supercladding 6, and the specific type of 23 liquid crystal material utilized in the fiber, different 24 classes of fiber taps can be constructed.

26 FIG. 5 shows what happens if the core 4 is made very 27 narrow. Only the zero order mode 22 of light can travel 28 through core 4 when the core 4 is made very narrow. A 29 single mode fiber is useful because it increases the interaction of the core guide light 4 with the electro- 31 optic layer 5. Increased interaction with the electro- 32 optic layer 5 decreases the tap interaction length and 33 improves screen resolutions. A fiber with a narrow single 34 mode core 4 will be called a "single mode" device and a wider multi-mode core a "multi-mode" device. Both single 36 mode, and multi-mode devices can be used in this display.

37 WO 93/09454 PCT/US92/06418 1 If the distance between the core 4 and supercladding 6 is 2 made very close, in other words, if the electro-optic 3 cladding 5 is made very thin, the evanescent field of the 4 core 4 guided light will extend 23, 24 and 25 into the supercladding 6. Evanescent coupling effects can be used 6 to enable small electro-optic layer 5 refractive index 7 changes to cause large changes in the amount of light 8 switched into the tapered supercladding 6.

9 An evanescent fiber tap increases the fiber switching 11 speed since the refractive index of the liquid crystal 12 layer 5 does not need to change very much to move the 13 evanescent field of the core guided light 22 into the 14 supercladding 23, 24 and 25. Utilizing the full range of liquid crystal alignment in the electro-optic layer 5 is 16 consequently not necessary to switch light out of an 17 evanescent waveguide tap.

18 19 Evanescent couplers can be used with single and multi-mode fibers. A multi-mode evanescent tap will require longer 21 tap distances to get light out of the core. Multi-mode 22 cores may be used to increase the light guiding efficiency 23 of the fibers. This is because light in the core 4 spends 24 more time away from the electro-optic switching layer traveling in the center of the core 4 and cladding 3. A 26 single mode fiber used in an evanescent configuration will 27 be called a "single mode evanescent device." A multi-mode 28 fiber used in an evanescent configuration will be called a 29 "multi-mode evanescent device." 31 Evanescent couplers have different tap efficiencies 32 depending on the wavelength of light traveling through 33 them. In other words, different light wavelengths are 34 more easily tapped out of a given coupler than others.

his color sensitivity can be reduced by having a large 36 change in the index of refraction of the electro-optic 37 layer 5. Since liquid crystals are capable of large 38 refractive index changes, the color sensitivity of a WO 93/09454 PCT/US92/06418 1 liquid crystal evanescent fiber taps is typically not an 2 issue. However, solid non-linear materials presently are 3 capable of much smaller refractive index changes than 4 liquid crystals. (On the order of An<10^-4 vs.An<10^ -2 for liquid crystals.) Therefore, solid electro-optic 6 switching layer 5 taps may require the use of multiple 7 evanescent couplers, each specifically designed to work in 8 a specific range of wavelengths. In this case a fiber 9 structure similar to that shown in FIG. 2 may be used.

11 Evanescent couplers are very sensitive to the index of 12 refraction of the electro-optic layer. Consequently, 13 temperature and biasing conditions will greatly effect 14 their operation. It is likely a photo-sensor will be needed at the end of each fiber to provide feedi-back 16 information to compensate for varying tap efficiencies 17 caused by temperature fluctuations and manufacturing 18 variances.

19 In a multi-mode evanescent device higher order modes will 21 couple more easily into the supercladding 6 than lower 22 order modes. Consequently, if a multi-mode fiber is used, 23 it may be advantages to couple only high order modes into 24 the fiber from the light source.

26 If a great change in the liquid crystal refractive index 27 is used, as shown in FIG. 6, the entire liquid crystal 28 layer 5 can be made to increase its index of refraction to 29 the same, or higher, level as the core 4. A waveguide tap based on large refractive index changes does not need to 31 utilized evanescent coupling to get light 19 out of the 32 core 4. Therefore, the supercladding 6 does not need to 33 be closely spaced next to the fiber core 4 to facilitate 34 coupling. As a result, by using the entire switchable refractive index range in the liquid crystal layer 5, the 36 tight tolerances needed in an evanescent tap can be 37 relaxed.

38 WO 93/09454 PCT/US92/06418 1 In a large refractive index change device light tapped out 2 of the core 4 travels through the electro-optic switching 3 layer 5 until it exits into the supercladding 6 region.

4 Light will ultimately refract out of the electro-optic cladding layer 5 if the supercladding 6 has a higher index 6 of refraction. A tap which carries light long distances 7 through the electro-optic region 5 will switch more slowly 8 since it will take longer to fully align the liquid 9 crystals to tap light into the supercladding 6.

Furthermore, the tap interaction length will be longer 11 because light will tend to travel through the electro- 12 optic switching layer 5 greater distances before exiting 13 into the tapered supercladding 6.

14 A tap that employs a large refractive index change can 16 also be used without a supercladding 6. In this 17 configuration light passes through the liquid crystal 18 layer 5 and enters directly into a diffusing layer (Not 19 Shown). This fiber configuration is particularly useful if the core is made very wide and narrow. This simple 21 design may find use in low cost/simple displays. This 22 type of waveguide electro-optic waveguide tap is disclosed 23 in FIG. 8 of U.S. Patent 5,106,181.

24 It is possible to make the entire supercladding 6 and/or 1 26 out of a liquid crystal material. This simplifies the 27 fabrication of the waveguide fiber by requiring fewer 28 materials and layers; since the intermediate cladding 11 29 and electro-optic cladding 5 can be combined into one large supercladding 6 and 1. Simplicity of construction 31 is made possible since the superclading 6 and 1 and 32 electro-optic layer 5 can be combined into a single large 33 cavity filled with liquid crystal material. A variant of 34 this simplified approach is to make the core out of the liquid crystal material and make the cladding 3 and 36 supercladding 6 solid.

37 WO 93/09454 PCT/US92/06418 1 A different type of waveguide modulator (Not Shown) 2 utilizes Raman-Nath and Bragg diffraction. If a plurality 3 of closely spaced electrode fingers are oriented 4 lengthwise along the direction of light flow, a diffraction grating can be formed in the electro-optic 6 material 5. If the electro-optic layer 5 is wide, (in the 7 x axis when looking at FIG. 1) and narrow (in the y-axis 8 when looking at FIG. light will be diffracted toward 9 the sides (left and right using FIG. 1 as a reference) of the fiber core. Supercladdings at the sides of the core 11 5 (Not Shown) can collect this light and combine it in a 12 single supercladding main chamber 1 so it can be reflected 13 at a pit. This type of tap has the advantage that a wide, 14 narrow, core 4 is easier to couple light into from the light source. An acous r-optic integrated optic device 16 similar to this is described by Manhar Shah in Applied 17 Physics Letter, Vol. 23, No. 10, November 15, 1973.

18 19 FIG.'s 7 8 show schematically how Raman-Nath and Bragg devices work. If Raman-Nath diffraction is used, as shown 21 in FIG. 7, a series of short length 28 gratings 27 22 oriented along the axis of light flow 26 couple light into 23 ,higher order modes 30, 31, 32, 33 and 34 in a tap region.

24 If Bragg diffraction is used, as shown in FIG. 8, a continuos set of gratings at the Bragg angle 65, off-axis 26 to the direction of light flow 28, will be needed to meet 27 the Bragg conditions. Bragg and Raman-Nath type 28 modulators can be made to operate in an electro-optic 29 core, or supercladding. Also, these sorts of modulators could be made with either liquid crystal and/or solid 31 electro-optic materials. An electrode may be needed over 32 the multiple finger electrodes to provide a continuously 33 aligned medium if liquid crystals are used in order to 34 allow the fiber to guide light.

36 All of the modulators described above can use either 37 surface aligned liquid crystal materials, or non-aligned WO 93/09454 PCT/US92/,6418 1 materials. The basic breakdown of modulator types is 2 thus: 3 4 Single mode: non-evanescent 6 surface aligned LC 7 non-suface aligned LC 8 evanescent 9 surface aligned LC non-surface aligned LC 11 12 Multimode 13 non-evanescent 14 surface aligned LC non-suface aligned LC 16 evanescent 17 surface aligned LC 18 non-surface aligned LC 19 The distinctions between these different classes of 21 modulators depends on the dimensions, arrangement, and 22 choice of materials employed in a particular fiber. The 23 exact classification between a specific modulator type 24 will often be difficult to clearly identify. For example, even with anti-alignment agents, there will still be some 26 surface alignment of the liquid crystals in the electro- 27 optic cavity 5. Consequently, modulators designed to 28 operate specifically using one approach may have 29 operational characteristics similar to fiber taps using other approaches.

31 32 MATERIALS 33 34 This section describes commercially available materials that are considered important for making optical waveguide 36 displays.

37 WO 93/09454 PCT/US92/06418 1 Liquid crystal materials made by Merck are likely to be 2 used in the electro-optic cladding layer 5. The Merck 3 Nematic liquid crystal product line has an index of 4 refraction ranging from between 1.478 1.5. Thus, Merck Nematic liquid crystals may be used with a wide range of 6 different plastics and glasses that have different indexes 7 of refraction.

8 9 One approach is to make optical fibers out of glass or silica. Glass, which is often used in fiber optic 11 applications, is the most promising inorganic compound.

12 Schott Glass Co. sells glasses with a variety of different 13 refractive indexes that are readily combined with liquid 14 crystals to make a glass-fiber display. Glass fibers can be pulled with excellent uniformity and with hollow cores 16 running through their centers. These hollow glass fibers 17 can be filled with liquid crystal material after drawing 18 to make an electro-optic cladding layer 5. Collimated 19 Holes, in Campbell California presently manufactures hollow core glass fibers that could be adapted to large 21 screen displays.

22 23 A new class of low temperature glasses being researched by 24 Corning Glass works may also be used. These glasses are discussed in the July 1992 Scientific American. Low 26 temperature glass can be doped with organic dyes to 27 achieve non-linear effects. Furthermore, certain in- 28 organic materials, lithium niobate, with large non- 29 linear effects can also be mixed or embedded in a glass matrix to achieve the electro-optic properties needed for 31 a display.

32 33 The most promising materials, however, are polymers. Of 34 these, Poly (Methylmethacrylate) or PMMA, Poly (Ethylmethacrylate) or EA, Poly (Vinyladeneflouride) or 36 PVDF, Polycarbonate, and Polystyrene are the most 37 attractive. These plastics are inexpensive, thermo- WO 93/09454 PCT/US92/06418 1 processible, commonly available, and have excellent 2 optical transparency.

3 4 For example, PMMA made by Rohm Hauss under the brand names V 825, VS 100, VM 100, and VLD is preferably 6 used to make many different fiber structures.

7 Specifically, the 825 is the highest purity PMMA and is 8 most promising for use in the core 4. The VS-100 is a co- 9 polymer mixture with EA and tends to have a lower index of refraction compared with V-825. Therefore, the VS-100 is 11 useful as a intermediary cladding 11. The VS-100 has also 12 been used as the cladding 3 since it is also a co-polymer 13 system with EA.

14 PVDF made by Ato-Chem of North America and sold under the 16 brand name of Kynar is also an important plastic. Kynar 17 in the 710-780 family of products may be used in the dark 18 cladding 2, and as a possible liner for the electro-optic 19 chamber 5. Pure PVDF is also an electro-optic material.

Aligned and poled PVDF sheet are sold by Ato-Chem under 21 the brand name Kynar Piezo film. A poled PVDF-based 22 material may be used as a solid electro-opZic cladding ?3 layer 5. Presently, the optical properties of pure PVDF 24 are poor since it is semi-crystalline, and the amorphous and crystalline regions have different refractive indexes 26 and tend to scatter light. This scattering, however, is 27 reduced in PVDF/PMMA blends, and therefore blends show 28 great promise for use in the electro-optic switching 29 layer.

31 PVDF and PMMA blends may be used to form a low-cost, low- 32 index cladding material. These two plastics form an 33 amorphous co-polymer system with a low index of refraction 34 when mixed in ratios of 40% or more of PMMA. This copolymer system may be used in the cladding 3 or, more 36 likely, in the dark cladding 2. The optical clarity of 37 this mixture depends significantly on the temperature and 38 extruder mixing conditions. A well mixed PVDF/PMMA WO 93/09454 PCT/US92/06418 1 composite has excellent optical properties. In the dark 2 cladding 2 a pigment or dye will be added to the PVDF/PMMA 3 co-polymer system to allow it to absorb stray incident 4 light. Avecor, in southern California, sells pigments and dyes which may be mixed to darken the plastic. PMMA/PVDF 6 mixes are considered to be very important in lowering the 7 cost of fibers.

8 9 A translucent, poorly mixed combination of PVDF and PMMA can be used as a scattering material. Polystyrene and 11 PMMA is similarly translucent and can also be used as a 12 scattering material. Rohm Hauss also makes a plastic 13 that tends to selectively forward scatter light. Forward 14 scattering films are very important for use in the diffusing layer in the output section.

16 17 Low index fluoropolymers, specifically those made of 18 fluorinated acrylic, may be used in the cladding 3.

19 Polymer Technology, Hoechst Celanese and Mitsubishi Rayon all make fluorinated acrylic which can be used as a low- 21 index cladding 3 layer. A different type of low-index 22 plastic that may be used in the cladding 3 is sold under 23 the brand name Cytop, and is made by Asahi Chemicals of 24 Japan.

26 Solid electro-optic materials, in particular polymers, may 27 also be used in region 5. Polymers that are both cross- 28 linked and amorphous can be employed. A manufacturing 29 technique utilizing cross-linked polymers, which are poled immediately after the fiber is drawn, is discussed in a 31 pending patent that focuses on manufacturing techniques.

32 33 A high index plastic is needed in the supercladding 6 and 34 1. Polycarbonate and Polystyrene are both high index plastics which may be utilized.

36 37 The mounting substrate and output optic layers may be made 38 out of flexible materials to enable the screen to be WO 93/09454 PCT/US92/064 18 1 rolled up. Since these layers may need to be thick to 2 focus and scatter light properly, they must be made very 3 pliable. Rubber-based, silicon-based and/or polyurethane 4 compounds are important materials for use in screen substrate, focusing, and scattering components.

6 7 DRIVING INDIVIDUAL WAVEGUIDE TAPS 8 9 The polarity between the controlling electrode 7 and the biasing electrode 10 or 9 may need to be reversed 11 periodically to prevent long lasting liquid crystal 12 alignment in the liquid crystal cavity 5. Extended 13 periods of alignment in a non-changing electric field can 14 cause the liquid crystals to become semi-permanently aligned. The result is that even when the electric field 16 is removed, liquid crystals do not change their refractive 17 index. Such alignment problems causes the guiding and 18 switching properties of the fiber to change over time.

19 Periodically reversing the polarity of the biasing electrodes 10 or 9 and control electrode 7 helps prevent 21 long-lasting liquid crystal alignment.

22 23 Certain electro-optic materials may have a higher index of 24 refraction when an electric field is applied than they do without the electric field. The drive signal approaches 26 discussed in this section should include these sorts of 27 materials. Therefore the terms guiding and non-guiding, 28 when an electric field is present, should be used 29 interchangeably. It will be assumed for simplicity, however, that the index of refraction decreases, as it 31 does in most liquid crystal configurations, when an 32 electric field is applied.

33 34 Fiber optic liquid crystal taps can be electrically driven in at least three different ways. First, it is possible 36 to operate the liquid crystal fiber tap in a simple on/off 37 voltage mode. When the voltage is on, or high, the 38 crystals will be aligned, and the fiber will guide. When WO 93/09454 PCT/US92/06418 1 the voltage is off, the liquid crystal layer 5 will 2 increase its refractive index and the core 4 will lose 3 light into the supercladding 6.

4 The second major approach biases the tap with a DC 6 component in the off state to facilitate the use of 7 surface aligned liquid crystals. Surface aligned liquid 8 crystals can be used with biasing to increase the tap 9 speed. For example, when "on" an electric field is applied just above the threshold needed to align surface 11 liquid crystals and make the fiber guide. When the 12 voltage is dropped from this high level, the surface 13 liquid crystal layer will quickly fall back in alignment 14 with the liquid crystal cavity 5 wall. This realignment is accompanied by an increase in the surface liquid 16 crystal refractive index and causes light to couple out of 17 the fiber.

18 19 In a tap based on surface alignment effects, the bias voltage level used in the off, or non-guiding state, and 21 the high voltage level used in the on, or guiding state, 22 is typically higher than that used in a non-surface mode 23 device. This is because the voltage needed to align the 24 liquid crystals near the surface walls is higher than the voltage needed to align liquid crystals in the middle of 26 the electro-optic cavity 5. The exact voltage required 27 depends on the alignment properties of the liquid crystals 28 themselves and cavity 5 wall surface properties.

29 The third major biasing and drive approach is useful in 31 both surface mode devices, and non-aligned evanescent and 32 non-evanescent devices. It is based on the fact that the 33 range of refractive indexes a liquid crystal can achieve 34 in an electric field is very large. By not using the entire liquid crystal refractive index range, and using 36 large voltage swings, the time needed to achieve a given 37 refractive index can be reduced.

38 WO 93/09454 PCT/US92/06418 14 1 FIG. 9 shows how, this approach works. A large voltage 2 swing 39a is used t:o switch the liquid crystal layer from 3 one semi-aligned state 39b to another 40b. The large 4 voltage change accelerates the rate of liquid crystal alignment 39a, or un-alignment 41a, to quickly achieve the 6 desired index of refraction. Once the desired index of 7 refraction is achieved, the new refractive index is fixed 8 by adjusting the voltage level 40a and 42a to a point 9 needed to maintain alignment.

11 This driving method can be used to increase the switching 12 speeds of evanescent couplers. Evanescent couplers need 13 only a small refractive index change 40b, 42b and 44b to 14 achieve large changes in their light coupling efficiency.

Therefore, they can be used with the large voltage swing 16 technique to achieve high switching speeds. It is 17 believed this large voltage swing switching technique is 18 quite different than any liquid crystal drive technique 19 mentioned in the prior art.

21 Grey scale operation can be realized by operating the 22 liquid crystal layer at a level between full and non- 23 alignment. Intermediate levels of alignment will cause 24 only a fraction of the light to be taped out of the fiber core. Coupling intermediate amounts of light out at 26 intensity modulators 50 enables gray-scale images to be 27 formed.

28 29 The amount of light coupled out of a fiber tap is not a linear function of the applied voltage. In other words, 31 half the voltage will not always switch out twice the 32 amount of light from a fiber tap. Particularly in 33 evanescent taps, or taps which employ surface effects, the 34 switching characteristics of the tap will depend in a very complicated way on the applied voltage.

36 37 Circuitry may be needed to translate a desired tap 38 brightness level into a voltage that can be used to drive WO 93/09454 PCT/US92/06418 1 the tap control electrode 7. This circuitry may need to 2 have a very high precision. For example, if the function 3 of light intensity for a given voltage is anything other 4 than a first order linear function, voltage control with a precision higher than the desired output gray scale 6 resolution will be needed to achieve a desired gray scale 7 resolution. In other words, if a standard digital to 8 analog converter is used to drive the taps to make gray 9 scale values, a resolution of more than 8 bits may be needed to achieve 8 bits worth of brightness at the fiber 11 tap.

12 13 An entirely different approach to achieving grey scales is 14 to conrtrol the time the tap is turned on. In this case, the voltage is switched "LOW", to cause the tap to switch 16 light out, only for the amount of time needed to get a 17 desired brightness. The total brightness of light at a 18 fiber tap is thus determined by integrating the amount of 19 light tapped out during it's "HIGH"/"LOW"/"HIGH" voltage interval. This approach is promising because simple 21 digital electronics can be used to achieve grey scale 22 operation. Using time duration to achieve grey scale is 23 particularly effective when used with high speed taps.

24 If the tap switches slowly, the amount of light coupled 26 out of a fiber tap is, like the analog approach, not a 27 linear function of the time the tap is actively turned.

28 Therefore, circuitry may be needed to translate a desired 29 tap brightness level into a time interval that can be used to drive the tap. It is possible a Digital Signal 31 Processor (DSP) may be used to translate the desired 32 brightness to a corresponding intensity modulator time 33 interval.

34 The capacitance of fiber taps, using both liquid crystal 36 and solid electro-optic materials, is very small. Since 37 the light is guided in a narrow channel, very small area 38 control electrode plate 7 is needed to aligned the small WO 93/09454 PCT/US92/06418 1 amount of material in the liquid crystal layer and cause a 2 significant change in the guiding properties of the fiber.

3 Low capacitance fiber taps can be driven with low-power, 4 low-cost, FET's, MOSFET's and bipolar circuits.

6 CONTROL ELECTRONICS 7 8 Before high quality screen images can be formed, the 9 output brightness of the screen needs to calibrated.

There are two sources of brightness irregularity in a 11 waveguide screen. The first is caused by manufacturing 12 variations and will cause a fixed pattern of brightness 13 non-uniformity. The second is caused by temperature and 14 changing mechanical conditions in the environment where the screen is operated.

16 17 The first source of brightness irregularity can be 18 corrected by making a map of the screen brightness. A 19 calibrated video camera or photo sensor aimed at the screen is used to measure and record the relative 21 intensities of image pixels under controlled conditions.

22 This brightness map is saved in a memory device and 23 combined with image data in the display controller at the 24 time of image generation to make brightness corrected images.

26 27 The second source of brightness irregularity is caused by 28 temperature variations. Fiber taps are very sensitive to 29 the index of refraction of the electro-optic layer.

Liquid crystals change their refractive index dramatically 31 with temperature. Thus temperature conditions will cause 32 the switching properties of fiber taps to change.

33 34 Temperature irregularities can be minimized by introducing a feed-back element into the system as shown in FIG. 11.

36 This feed-back system measures the amount of light making 37 it to the fiber ends with a photodetector 91 or 95. By 38 systematically going down the length of each fiber and WO 93/09454 PCT/US92/06418 1 measuring the efficiency of each tap, for example 88 or 2 89, for a given applied electric field, 87 or 89, the tap 3 efficiency can be measured. In this way the screen 4 brightness can be dynamically compensated to provide a uniform brightness under different temperature and 6 environmental conditions.

7 8 A single large photosensor 46 can be used to measure the 9 output of multiple fibers in the feed-back loop as shown in FIG. 10. Alternatively, many small photosensors 91 or 11 95 may also be used to measure the output of individual, 12 or small groups, of fibers in parallel. Photo sensors can 13 be made from photodiodes. It is also possible to use 14 pyrodetectors. For example, poled PVDF film may be laminated to the front of a waveguide ribbon to make an 16 integrated pyroelectric sensor for detecting light 17 intensity.

18 19 A control electronics system translates the input video signal into a set of control voltages needed to produce a 21 screen image. The control electronics system typically 22 works by digitizing an analog video input signal, or, 23 buffering a digital signal. Screen information is stored 24 in a dual port video RAM. This memory is accessed by a screen controller circuit that maps video-RAM contents 26 into screen brightness variations.

27 28 The screen control electronics typically sequence through 29 all the taps 49 in the screen. A selected row of taps 48 dumps all the light out of the fiber when the voltage is 31 lowered. The intensity modulators 58 produce the desired 32 brightness at the selected tap row by reducing the amount 33 of light 50 traveling through the waveguide fibers 52.

34 As previously mentioned there are two methods of achieving 36 grey scales. One uses analog values to control the 37 intensity modulators 50. The other uses the time of 38 intensity modulator 50 activation to create grey values.

WO 93/094854 PCr/US92/06418 1 2 If the analog approach is used, at least one D/A converter 3 will be needed to convert digital image information into 4 analog voltages that are used to charge capacitors in the intensity modulator 51 drive controller. Once the gray 6 scale information for all of the intensity modulators is 7 locked in the capacitors, the capacitor charges are 8 transferred, in parallel, to the gate of the driver 9 transistors. The drive transistors connected directly to the intensity modulator electrodes 7 control the amount of 11 light in the intensity modulators.

12 13 If intensity modulation is accomplished with length of 14 duration of a tap switching element, then a means is needed to determine the tap activation interval. Both the 16 analog drive approach and intensity modulation approach 17 may use a DSP to convert a desired pixel brightness into 18 the associated analog value or tap time duration length.

19 DSP adjusted drive voltages are converted by a D/A converted and used by the intensity modulator controller 21 to make an image.

22 23 Compensation for longitudinal ribbon alignment errors 24 encountered during the manufacturing process may be necessary to make a uniform screen image. A digital 26 buffer with ribbon offset information can be used to 27 adjust for ribbon alignment non-uniformities.

28 29 LIGHT SOURCE INPUT OPTICS 31 Most of the power is consumed in an emissive display 32 generating light. Light generation and its efficient 33 subsequent use is critical to lowering the amount of power 34 used in a screen. Power consumption is especially important in large area screens; where significant 36 illumination is needed across large surfaces.

37 WO 93/09454 PC-/US2/06i8 1 In a waveguide display it is possible to direct upward of 2 80% of the light generated from the light source to the 3 viewer. Dielectric filters enable white light to be 4 broken into red, green, and blue components and guided across the screen in separate fibers. Since different 6 colors of light are carried in separate waveguides, nearly 7 all the illumination from a light source can be utilized.

8 9 This is in contrast with traditional back-lit active matrix color liquid crystal displays. These displays 11 transmit only 5-8% of the original light from the light 12 source to the viewer. This low efficiency causes their 13 power consumption to be very high. The new approach of 14 using optical waveguides to guide separate colors provides significant improvements over existing display 16 methodologies.

17 18 Turning to the drawings, FIG. 15 shows the light source 19 83. Ideally, the light source should have a high electrical/light conversion efficiency. It should also 21 produce a well collimated beam. Beam collimation, or lack 22 of divergence, is important to increase the efficiency of 23 light coupled into the fiber cores. A collimated beam is 24 more readily focused into small fiber cores with a microlens array. Collimated light also enables smaller 26 refractive index changes in the electro-optic layer 5 to 27 switch light out of the fiber core 4, Small refractive 28 index changes translate into increased tap switching 29 speeds since small refractive index changes are more quickly achieved in a liquid crystal material. Therefore, 31 beam collimation directly effects the screen brightness, 32 refresh rate and resolution.

33 34 The most promising light sources are presently quartz halogen incandescent lamps and xenon arc lamps. Quartz 36 halogen lamps are commonly available and relatively 37 inexpensive. However, they have poor electrical/light 38 conversion efficiency and produce a divergent beam when WO 93/09454 PCT/US92/06418 1 compared with xenon arc lamps. Xenon arc lamps produce an 2 extremely intense, small, arc which is readily collimated 3 into a beam.

4 Xenon lamps, with integrated reflectors 82, are the 6 preferred method of illumination for a waveguide display.

7 An integrated reflector arc lamp is very compact, since 8 the reflecting mirror is built into the lamp housing. ILC 9 Corporation in Sunnyvale produces a Xenon lamp named Cermax which is ideally suited for waveguide display 11 applications.

12 13 FIG. 15 shows how light from the lamp is processed before 14 being focused into the fiber cores. Lamp light first passes through-band-pass dielectric "hot" 81 and "cold" 16 filtering mirrors. These mirrors block infrared and 17 ultra-violet portions of the light spectrum. Only visible 18 light makes it to the fiber ends 67. Efficient light 19 filtering is essential to reduce the heating and chemical bond degradation effects on sensitive plastic fiber caused 21 by IR and UV. It may be necessary to utilize multiple 22 "hot" 81 and "cold" 80 stage filters in series to reduce 23 UV and IR to acceptable level. Other filter types which 24 may be used include glass absorption filters, and water filled IR filters.

26 27 Filtered, visible light is passed t'rough collimating and 28 beam shaping optics 79 and 78. This collimated optical 29 section 101 increases the uniformity and collimation of the light beam. Many different collimating techniques may 31 be used in section 101. For example, an aperture could be 32 used at the focal point of lens light 79 to increase the 33 light collimation, at the expense of brightness. Or a 34 spatial filter could be placed in the light path to increase the beam uniformity. Also, beam shaping optics 36 may be used to produce a square beam 102, to more 37 efficiently utilize the lamp light. The specific ordering WO 93/09454 PCT/US92/06418 1 and use of these light path elements would be clear to 2 anyone who has designed projector illumination systems.

3 4 The filtered, collimated, visible white light 102 is passed through a series of dielectric band-pass mirrors 6 103 and 105. Low-pass filter 103 reflects the red 7 portion of the light spectrum 73 and passes through the 8 yellow and blue 104 light to filter 105. High-pass filter 9 105 reflects the blue portion of the light spectrum and passes through the green light 106. The remaining green 11 light 71 is reflected by mirror 74.

12 13 Using a small, collimated, beam 102 at filtering mirrors 14 helps reduce the size of the dielectric filters. It also allows the dielectric filters to be placed closer toge ier 16 and hence decreases the size of the light source optical 17 system. Light reflected from the dielectric filters 73 is 18 expanded by lenses 71 and 69.

19 Each color thus hits the micro-lens array 68 at a 21 different angle 70. A single microlense element is used 22 to focus the three separate primary colors into three 23 separate fiber cores. Each color hits a shared micro- 24 lense element at a different angle. Color fibers, like those shown in FIG. 2, or a closely spaced set of single 26 core fibers like those shown in FIG. 1, receive the colors 27 and in their fiber cores 4. Each fiber core 4 carries 28 one color of light across the screen.

29 A micro-Lens arrays manufactured by Corning under the 31 brand nane SMILE may be used to focus light into the fiber 32 end. However, SMILE lenses, since they are round, will 33 only couple approximately 70% of the light into the fiber 34 cores. Square micro-lense arrays produced by diffusion techniques can couple nearly all of the light into the 36 fiber core since they do not loose light in the corners 37 like round lenses. A planar graded lense can also be used WO 93/09454 PCT/US92/0(141X Ito 1 in conjunction with a micro-lense array to focus light 2 into the fiber ends.

3 4 A different method of making a color display is shown in FIG. 16. white light is focused on to a rotating disk 84 6 with thres diferent color reflective dielectric filter 7 segments. As the disk turns each segment sequentially 8 reflects a different color into the ribbon 67 fiber cores.

9 By sequentially 75 painting buffered colored screen images one after another, a full color image is forred. This 11 technique has the advantage that it utilizes extremely 12 simple optics. Furthermore, color operation can be 13 achieved with 1/3 the intensity modulator drive electronic 14 connections as the three color approach in FIG. Therefore, while FIG. 16 is not as light efficient as the 16 illumination approach taught in FIG. 15, the 17 simplification in optics and drive electronics provides 18 useful manufacturing cost advantages.

19 Anti-reflective coatings are used in all optical elements 21 in the light path. Furthermore, a fan may be needed to 22 cool optical elenmts. .Li %S fan should have a filter to 23 prevent dust rand other small particles from being 24 depositing on the optical components.

26 OUTPUT OPTICS 27 28 Light 59 tapped out of a fiber reflector pit 35 typically 29 will need to pass through an output optics section before it is seen by a viewer 55. The output optics section re- 31 directs, scatters and shapes light reflected from the 32 reflector pits 59 before it is seen by the viewer.

33 34 Separate, parallel, fibers 58 will typically be needed to make a continuous light column along the direction of 36 light flow. Separate fibers are necessary because the tap 37 length is too long to enable consecutive pixels to be 38 formed on a single fiber and still maintain high screen WO 93/09454 PCI'/US92/06418 1 resolutions. Consequently, consecutive pixels along the 2 direction of light flow will be achieved by staggering 3 long interaction length taps on separate fibers.

4 Unfortunately, if a single vertical column is illuminated 6 with multiple fibers, there will be a staircase, or other 7 irregular pattern, as a result of the horizontal shift 8 inherent in carrying light in separate fibers. An 9 irregular column pixel pattern along the direction of light flow can be eliminated by adjusting the light path 11 taken by the light emitted 59 at a reflector pit 35 before 12 it seen by a viewer.

13 14 This adjustment can be caused by using a diffusing medium 56, a mirror 64, or lenses 61 in between the reflector pit 16 35 and the side of the diffusing screen the viewer sees 17 56. FIG.'s 12-14 show how a scattering layer 56, mirror 18 structure 64 or lenses 61 can be used to eliminate pixel 19 non-uniformity along the length of the fibers.

21 FIG. 12 shows a simple forward scattering diffusing screen 22 approach. Fibers 58 are mounted to a mounting substrate 23 60. Diffusing layers 66 and 56 positioned to direct light 24 59 from the fibers 58 to the viewer 55. Light exiting the fibers 59 first travels through optical medium 66. The 26 optical layer 66 may be used to slightly absorb the 27 intensity of the light 59 to keep ambient and back- 28 scattered light from reflecting around within region 66.

29 The film 60 or fiber 58 surface is coated with a black layer to prevent light reflected within cavity 66 from 31 traveling long distances and causing large pixel sizes 32 through diffusion in the cavity 66. Light exiting the 33 fiber 59 hits the diffusing screen 56 and is scattered to 34 the viewer 55. The ridges on the outside of layer 56 prevent surface reflections on the screen side observed by 36 the viewer. Barriers 57 keep scattered light in a well- 37 defined vertical column.

38 WO 93/09454 PCT/US92/06418 1 FIG. 13 shows a method whereby Fresnel lenses 61 are used 2 to eliminate the horizontal multi-fiber pixel shift 3 problem. Fresnel lens array 61 directs light 59 from each 4 fiber to a specific area of the forward scattering diffusing screen 63 and 56. 3M makes linear Fresnel lens 6 arrays under the name "Scotchlens" which may be used as a 7 lens array 61. A barrier 57 like that shown in FIG. 12 8 could also be used with Fresnel lens approach to further 9 define light into vertical columns.

11 FIG. 14 shows a method whereby mirrored reflectors are 12 used to compensate multi-fiber horizontal shift problems.

13 Light 59 exiting reflector pits bounces off a mirror 14 surface 62 on reflecting structure 64 and hits the forward scattering diffusing screens 63 and/or 56. Scattered 16 light 55 travels to the viewer and can be seen. The top 17 of the mirrored reflecting structure 64 will typically be 18 made of a dark material to increase the screen contrast 19 ratio seen by the viewer.

21 A ridged diffusing layer 56 is commonly used in existing 22 rear projection large screen displays. However, it is 23 presently made out of a plastic, which is easily scratched 24 or torn. It is desirable to make the ridged layer 56 out of a less scratch prone materials such as glass or coated 26 plastic to prevent damage. A flat glass panel may also be 27 placed between the ridged diffusing screen 56 and viewer 28 to prevent damage.

29 WAVEGUIDE RIBBON 31 32 Multiple ribbons 114, each in turn with multiple fibers 58 33 mounted on it, cover a substrate to make a screen. A 34 fiber ribbon is typically 25-100mm wide and has square fibers ranging in size from .125-.5 mm in diameter mounted 36 on it. Metal electrodes are deposited on the back 110 of 37 the ribbon to control the fiber taps 7 and intensity 38 modulators electrodes 7.

WO 93/09454 PCr/US92/06418 1 2 In the prior art, the intensity modulator electrodes 3 run along the bottom of the screen. The tap electrodes 49 4 are distributed along the remaining length of the ribbon.

The control electronics is shown being placed over the 6 intensity modulators on the bottom of the ribbon.

7 8 Significant manual labor is needed to connect the taps 49 9 across separate ribbons in the approaches taught in the prior art. A new technique which eliminates the manual 11 connection of tap electrodes is introduced here which is 12 based on adding layers 115 and 107 to the back of the 13 waveguide ribbon. These layers have vertical electrical 14 conductors 110 and feed-through holes 112 that enable signals to control the intensity modulator 7 and tap 16 conductors 7 from control electronics positioned at the 17 bottom of the display.

18 19 A ground plane 116 may be positioned between the vertical ribbon control lines 110 and the electrode layer 115 and 7 21 to prevent the electric field of the vertical control 22 electrodes 110 from inadvertently switching light out of 23 the fiber taps. Feed-throughs 112 that go through the 24 ground layer 116 connect the vertical control lines 110 to the tap control electrodes 7.

26 27 As a result of this vertical conductor 110 innovation a 28 printed circuit board containing the control electronics 29 may be placed over the control lines 110 at the bottom or edge of the display and control the entire screen image.

31 Also, the waveguidc, ribbon 114 can be mounted on the 32 mounting substrate with the reflector pits face up 35 from 33 the substrate, or, face down. Depending on the screen 34 configuration front and back ribbon orientations with respect to the substrate may be used.

36 37 38 WO 93/09454 PCT/US92/06418 1 COMPLETE DISPLAY 2 3 The screen controller printed circuit board (Not Shown) 4 can be connected to the vertical control electrodes 110 with a zebra connector (Not Shown). In this connection 6 approach the bottom of the PC board compresses a zebra 7 connector between conductors on the bottom of the PC board 8 and vertical control conductors 110 on the back of the 9 ribbon 114. The PC board is pushed against the zebra connector by plates which are tightened against blocks 11 attached to the back of the waveguide ribbon.

12 13 It is also possible to mount an integrated circuit 14 directly on the back of the waveguide ribbon 114 (Not Shown). In this approach a control chip is bonded 16 directly on the back of the ribbon. A protective case is 17 placed over the flip mounted control chip to protect it 18 during ribbon handling. Connections in the protective 19 case enable electrical and drive signal connections to made to control the ribbon.

21 22 Self-contained ribbons using mounted control chips may be 23 made with a clear adhesive layer deposited on the front 24 surface 108 of the ribbon (Not Shown). This adhesive layer allows self-contained ribbons to be directly applied 26 to large windows virtually anywhere. This will enable 27 large screens to be made anywhere there is a window where 28 a ribbon can be affixed to.

29 Separate screens can be connected together by joining them 31 along the edges. A material with a refractive index 32 matched to the substrate can be placed in the seems 33 between separate screens.

34 The control electronic box can be curved to make the 36 display thinner or flatter. The box can be placed above, 37 below, behind or on the sides of a screen. The screen can WO 93/09454 PCT/US92/06418 1 be cooled by convection heating in the space between the 2 wall and the back screen which will cause chimney affect.

3 4 RAMIFICATIONS AND CONCLUSIONS 6 Economical wall-size, or even theater-size, screens are 7 possible based on the technology described in this patent.

8 Consequently, it is important to anticipate the 9 environmental effects large moving images will have on individuals living and working in proximity to large wall- 11 size displays.

12 13 Dizziness, or even nausea, may be experienced by many 14 people who stand near a large screen. The motion caused by large moving images can be very disorienting.

16 Furthermore, the content of images displayed on a large 17 screen will have a far greater psychological impact than 18 the same images shown on small screens.

19 Taken together, these motion and psychological effects 21 could have negative consequences for people in the 22 vicinity of the screen. A means must be provided whereby 23 any person near a screen can easily regain control over 24 their environment. Specifically, a method is needed to conveniently turn a big screen off.

26 27 Impact sensors should be integrated into the screen to 28 allow a person to hit or kick the screen and have the 29 picture(s) being displayed change. Once the screen is hit images should shrink, disappear, or be replaced by less 31 disturbing images. Also, a micro-phone providing speech 32 input could be used in conjunction with a computer to 33 enable voice input to change the screen state. A person 34 could speak or scream the words "Go Away!" to make the screen images disappear.

36 37 These control features are felt to be important to 38 minimize the negative consequences of this technology. It WO 93/09454 PCT/US92/0641 !8 1 is sincerely hoped that features which enable people to 2 turn these screens off are integrated into each and every 3 screen by manufacture who produce them.

4 While this application contains many specifics, the reader 6 should not construe these as limitations on the scope of 7 the invention. Rather, they are merely exemplifications 8 of the preferred embodiments. Those skilled in the art 9 will envision many other possible variations that are apparent give the ideas presented here. Accordingly, the 11 reader should determine the scope of the invention by the 12 appended claims and their legal equivalents, and not 13 solely by the examples which have been given.

14 16 17 18 19 21 22 23 24 26 27 28 29 31 32 33 34 36 37 38

Claims (8)

1. An optical waveguide display system comprising: at least one optical waveguide with electro-optic switchable elements placed along a length of said at least one optical waveguide; at least one photosensor coupled to receive light from said at least one optical waveguide; and a control circuit coupled between said at least one photosensor and said electro-optic switchable elements to control said electro-optic switchable elements based on light sensed by said at least one photosensor.

2. The optical waveguide display system of claim 1 in which said electro- optic switchable elements are liquid crystal switchable elements.

3. The optical waveguide display system of claim 1 in which said at least one photosensor comprises multiple, separately addressable photosensors.

4. The optical waveguide display system of claim 3 in which said multiple separately addressable photosensors are coupled to either individual or groups of separate waveguide elements in a one-to-one correspondence.

The optical waveguide display system of claim 1 in which said control Scircuit compensates for refractive index irregularities caused by temperature induced refractive index changes. 20

6. The optical waveguide display system of claim 1 in which said control circuit adjusts for screen brightness irregularities to achieve a uniform and calibrated screen brightness profile.

7. The optical waveguide display system of claim 1 in which said at least one photosensor is placed at or near the end of said waveguides.

8. An optical waveguide display system, substantially as hereinbefore described with reference to the accompanying drawings. Dated 8 January, 1998 Marshall A Rockwell, Iil Patent Attorneys for the Applicant/Nominated Person SPRUSON FERGUSON