DE2445150B1 - Imaging system for integrated optics - with magnifying or reducing simply by changing cross-section of layer waveguide - Google Patents

Abstract

The system has a flat, prism-shaped or pyramid-shaped waveguide with reflecting boundary surfaces. The length L along the waveguide axis between an object (A) and an image (A') and at least one value (Weg) representing a typical transverse dimension of the waveguide are linked by an equation wherein the product of the length L and the wavelength is equal to 4 hn1W(eg)2. Where n1 is the refractive index of the waveguide material, whilst h is an integer 1, 2, 3, etc. The waveguide is pref. in the form of a layer whose thickness is taken as the typical transverse dimension.

Description

From the magazine "Applied Optics", 9 (1970), 753, there is an imaging system known, in which dielectric optical waveguides are used for imaging, which have a parabolic refractive index profile in one or both directions transverse to the Have the direction of propagation of the light and are referred to as self-focusing light guides will. There is a significant technical difficulty with this imaging system however, the manufacture of the layers or fibers, e.g. B. by Diffusion, with the required parabolic refractive index profile.

The object of the invention is to provide an imaging system that is suitable for the image does not have conventional components such as lenses above mirrors, only a waveguide is needed that is simple and without the self-focusing Light guides occurring difficulties can be produced.

The invention solves this problem by what is specified in claim 1 Imaging system.

Further developments of the invention are characterized in the subclaims.

Waveguides according to the invention have the advantage that they are spatial can have constant refractive indices and essentially flat walls or interfaces, so that they are technically very easy to manufacture, for example in layered or Strip shape by high vacuum evaporation, by cathode sputtering or by other methods known per se. An optical shift leader that only directs light leads in the dimension perpendicular to the layer, conveys a one-dimensional one optical image of its one end face on the opposite end face, if its dimensions are chosen correctly. An optical stripline that does that Light leads in the two dimensions perpendicular to the strip axis in these two dimensions an optical image of its end faces on top of one another, if it has a rectangular or square cross-section with the required Width or height. But there are also waveguides with other polygonal, z. B. triangular or hexagonal cross-sections are possible.

The images can be upright or inverted, enlarged or reduced, real or virtual, and they have a certain depth of field. The lateral resolution is only limited by the subtleties of the waveguide's mode spectrum. Through corrective actions and apodization, the point mapping function can be further improved. A resolving power from two to three wavelengths is possible. The generally relatively small field of view a single waveguide can be achieved by stacking and bundling individual conductors to be made bigger.

Possible applications for such imaging systems arise for example in image transmission in systems for optical image processing, as a laser resonator in various designs, or as a filter. Of special The advantage is the suitability of the imaging system that can be implemented in the smallest of dimensions for various purposes of integrated optics.

A major advantage of the invention is the possibility of a magnifying or downsizing imaging system simply through continuous expansion or To create tapering of the cross section of the waveguide.

The principle of image creation in the imaging systems described here differs significantly from that in ordinary optical imaging systems: While in conventional imaging systems the so-called optical path length between an image point Pl and the associated object point Pi, P2 end, Pl for all possible Light paths s as exactly the same as possible should be, it is for the imaging systems described here characteristic that the optical path lengths along the various possible, discrete light paths differ by integer multiples of the wavelength A. In this regard, the imaging systems described here belong in a class with the Fresnel lenses. Because of the existence of the mentioned path length differences they are best suited for images using monochromatic light. As the spectral width of the imaging light increases, it deteriorates Image quality (chromatic error). The mapping takes place according to the invention by breaking down the light emanating from the object into the wave functions ("Modes") of the waveguide, through independent propagation of these waves along of the conductor and their subsequent in-phase composition to the picture.

The invention is explained in more detail using preferred exemplary embodiments. The drawing shows F i g. 1 shows a cross section through a dielectric layer conductor, F i g. 2 an arrangement for the experimental verification of the figure, FIG. 3 one dielectric stripline, FIG. 4 a phase corrected waveguide, F. i g. 5 waveguides formed from a stack or bundle of individual conductors, F i g. 6 waveguides with modified end faces, FIG. 7 a representation for Explanation of the creation of a virtual image, FIG. 8 an abbreviated image Waveguide, FIG. 9 shows a representation for determining the maximum possible shortening of a waveguide, FIG. 10 a waveguide with metal layers serving for apodization, 11 and 12 waveguides with a locally variable cross section, FIG. 13 a integrated optical arrangement with a waveguide, FIG. 14 an imaging acoustic waveguide, Fig. 15 shows an arrangement for transmitting sub-millimeter wave radiation with an imaging waveguide, Fig. 16 and 17 arrangements for coupling a Laser with optical layer conductors, Fig. 18 an arrangement for coupling two Layer conductors, FIGS. 19 to 22 different, implemented by a waveguide Resonators, FIG. 23 an amplifier for integrated optics purposes; and F. i g. 24 a filter to separate two different radiations.

First, the simplest imaging systems will be described. F i g. Fig. 1 shows, in cross section, a dielectric optical layer guide, the a one-dimensional optical imaging enables the waveguide consists of an optical material 13 which is transparent at the light wavelength A used with the refractive index m. The right-angled coordinate system of FIG. 1 is with his x-direction oriented parallel to the direction of propagation of the light. The perpendicular to it The direction within the plane of the drawing in FIG. 1 is the z direction and the y direction stands perpendicular to the plane of the drawing. The waveguide is limited in the x-direction through the two surfaces x = 0 and x = L, and in the z-direction through the two Surfaces z = 0 and z = Wz. The surfaces mentioned should be as optically flat as possible, i.e. H., their deviations from an ideal plane should be small compared to the wavelength used A. In the y-direction the dimension of the layer conductor is very large compared to the wavelength, z. E.g. 1000 to 10,000 # or more.

The shift supervisor according to FIG. is on its top and bottom of two other optical materials that are also transparent at wavelength A. 10, 12 of the refractive indices n0 and n2, respectively. Among the noptic materials, should this also includes the free space (vacuum) created by the refractive index n = 1 is marked. So that the light is guided through total reflections in the conductor the refractive indices no and n2 must both be smaller than the refractive index of the conductor material, i.e. no <n1 and m <m. If these conditions for total reflection are not fulfilled, the shift supervisor is »leaky« and the image quality that can be achieved gets worse. However, the imaging property is not completely lost. That the same also applies in the event that materials 10 and 12 are metals. the Dimensions of materials 1Q and 12 in the z direction must be if they are clear are about one wavelength or more. In the case of metallic materials 10 and 12, thinner layers are sufficient, but they must be at least as large as that the interfaces n1 / n0 and m / m are reflective.

The in F i g. 1 shift leader shown causes a one-dimensional Mapping of the area 0 <z <Wz of the input level x = 0 on the corresponding Area 0 <z <Wz of the starting level x = L. This strip-shaped areas 0 <z <Wz thus play the role of the entrance and exit pupils of the Imaging system.

As will be explained later, the image is only sharp if if the dimensions of the shift leader satisfy the relationship L. # = 4 h2n1Weq.z² (1) Here h7 = 1, 2, 3, 4, ... is an integer, the quantity Weq.z is an equivalent one Layer thickness. It is the thickness of an optical layer conductor with ideally metallic reflecting walls, their possible electromagnetic field distributions (modes) match those of the given real shift supervisor in the range 0 <z <Wz. Because of the so-called "goos chicken effect" is the equivalent layer thickness in the case of the totally reflective layer conductor (n0 <n1 and 112 <m) slightly greater than the geometric thickness W of the layer conductor and also hangs very slightly on the polarization of the imaging light. Is the light with the electric Vector polarized perpendicular to the plane of the drawing in FIG. 1 (so-called TE polarization), then Weq, z = + (i / 2) [(n1² - n0²) -½ + (n2i - n22) 112] (2) If the light is against it with the magnetic vector perpendicular to the plane of the drawing in FIG. 1 polarized (so-called TM polarization), then Weq.z = W + (# / 2 # n1²) [n0² (n1² - n0²) ½ + n2² (n1² - n2²) -½]. (3) In cases in which no or m are greater than m, one can still find the equivalent layer thickness from equations (2) and (3) calculate if the imaginary expressions that occur are simply deleted.

For an explanation of the one-dimensional mapping, see also Fig. 1 referenced. The image scale of the shift supervisor shown there is 1: 1. The image drawn from the input plane x = 0 to the output plane x = W is upright, when hz in equation (1) is an even number. The picture is on the other hand for odd Hz vice versa.

The mapping is one-dimensional because the shift supervisor does it Light only in the z-direction through reflections at the interfaces n / no and nl / n2. in the y-direction the side walls of the layer conductor are so far away that the light practically does not reach them.

There is thus no guidance and no mapping in this direction. From an object that is in the input level x = 0 and that from the left is illuminated, therefore only the structure is imaged along the z-direction. Structures of the object along the y-direction become in the one-dimensional mapping not reproduced.

A shift supervisor is used to illustrate equations (1) to (3) considered, for example, from a liquid with the refractive index n1 = 1.500 which is enclosed between two polished plates of equal size fused quartz, no = n2 = 1.457.

The layer thickness is Wz = 50 μm and the wavelength of the imaging Light A = 633 nm. Equations (2) and (3) give the equivalent Layer thickness of this system at Weq.z = 50.70 µm with TE polarization and Wzq'z = 50.66 around at TM polarization. The one required for a mapping according to equation (1) The length of the layer conductor is approximately L = 24.4 mm with TE polarization and L = 24.3 mm with TM polarization. These lengths are the shortest conductor lengths left enable a mapping (hZ = 1). For any integer multiple of these lengths but you also get sharp images.

An experimental proof of the figure is shown in Figure 2 Arrangement possible. In the case of a shift supervisor according to FIG. 1 was in the entrance level x = 0 a gap of e.g. B. 4 µm width (in a chrome layer on a glass substrate) attached, the edges of which were parallel to the y-direction. The gap was from the left illuminated with the red light of a HeNe laser (Aw = 633 nm). The starting level x = L = 24 mm was covered with a microscope cover slip, so here, at the end of the Liquid layer, an area of optical quality was present. The starting level could be observed through a microscope. By making small corrections to the layer thickness W by means of a pressure screw could then in the starting plane a sharp, inverted, real picture of the gap in natural size can be obtained. That it is here was a real image, is shown, among other things, by the use of several, gaps of different widths (3 to 50 μm) at x = 0. In each case the picture had same size as the original. The gap became in the entrance plane in the z-direction moved up and down, the image in the initial plane moved in the opposite direction Direction.

In a further experiment, the layer thickness Wz of the liquid layer, through thin metal sheets defined between the quartz plates was reduced from W = 50 µm to W = 35.7 µm, with the length L = 24 mm unchanged the shift. For the new layer thickness, equation (1) is fulfilled with hz = 2. The image was sharp again, but it was now upright: it was moving up and down in the same direction as the gap itself. In all these observations stood a rotating diffuser in the laser beam between the laser and the gap, around the to avoid disturbing speckles in the images observed under the microscope. There was also a polarizer in front of the column, with which the light was optionally available Could be TE or TM polarized. It turns out that the settings of the Layer thickness for optimal sharpness of the slit image in both directions of polarization were slightly different according to the difference between equations (2) and (3).

The exact fulfillment of equation (1) is for a sharp image essential. For this reason, practically some way of getting a or to change several of the quantities in equation (1) so that the equation (1) is exactly fulfilled.

In the experiment described above with the liquid layer the layer thickness W was adjusted by adjusting the distance between the quartz plates by means of a Pressure screw has been changed slightly. Other options for setting the focus the figure consist in changing the refractive index nl of the layer, e.g. B. by changing the temperature or by applying an electric field. One Another possibility is to vary the wavelength A.

The wave guide principle described above for the shift supervisor and mapping in the z-direction can easily be applied to the y-direction as well will. An optical stripline of generally rectangular shape is then obtained Cross-section which, with the correct choice of dimensions, produces a 2-dimensional image perform. Such a dielectric stripline is shown in FIG. 3 shown. With the thick, rectangular stripline, the waveguide is in the y and z-directions almost independent of each other. Therefore must for a sharp image in addition to equation (1), a corresponding mapping condition for the y-direction must be fulfilled: L) = 4hyfl1W2qy. (4) The equivalent stripe width occurring herein Wzq, y is calculated analogously to equations (2) and (3) from the geometric width Wy of the strip. Instead of no and m, however, there are the refractive indices n3 and Use n4 of the materials on either side of the strip (Fig. 3).

The image again generates a real image of the input plane x = 0 in the initial plane x = L. The image field size is Wyx Wz. The mapping takes place on a scale of 1: 1. It is erect when hy and hZ are both even numbers. Otherwise the image is reversed in one or both coordinate directions. the The simplest form of a two-dimensional imaging waveguide is square Cross section, d. H. Wy = Wz. Then of course hy = hz also applies and the image is either upright (with even hy = hz) or vice versa (with odd hy = hz).

The imaging conditions according to equations (1) to (3) can be derived theoretically as well as information about the achievable resolution and the permissible tolerances of the waveguides be made. For the sake of simplicity only the 1-dimensional imaging, symmetrical (no = n2) shift leader is used here considered in more detail.

It is known that the electromagnetic field in a shift conductor can be represented as a superposition of a number of modes: V (x, z) = E a ", F ", (z) exp (itj ,,, x). (5) In the case of TE polarization, V means the y component of the electric field, with TM polarization V is the y component of the magnetic Field.

The quantities am are amplitude factors, Fm (z) are the normalized, real field distributions of the modes, and Bm are their propagation constants. A Time factor exp (- iwt) should be added to all field sizes.

Let there be a certain field distribution Vo (z) in the input plane x = 0 of the layer conductor, which is due to the fact that an object is brought into this plane and illuminated from the left with a polarized monochromatic light source. This field Vo (z) excites the modes capable of propagation in the waveguide. Their amplitudes are a ,,, = f VO (z) F ", (z) dz, (6) and they propagate independently of one another according to equation (5) along the waveguide. For the plane x = L results It is assumed here that all modes from m = 0 up to a maximum mode number m = M are excited and capable of propagation. From equation (8) one can see that a mapping of the plane x = O into the plane x = L always takes place when exp [iL (ßin, - fl0)] = 1 applies to all excited modes at the same time. (8) Under this condition the fields of the modes interfere at x = L in exactly the same way as in the object plane x = 0. The factor exp (ißoL) in equation (7), which is common to all modes, is meaningless (it gives the temporal phase position of the image relative to the object). Under the condition of equation (8), the figure is erect and natural in size.

Before proving that equation (8) applies to a shift supervisor can be fulfilled, we shall show that the fulfillment of another proposition of conditions exp [iL (p1,1-1J0)] = (-1) "'(9) by the excited modes m = 0 ... M results in an image, namely an inverted one and in natural size. This follows from the symmetry character of the modes of a shift supervisor. In the usual Representations of the TEm and TMm modes of a symmetrical shift leader (no = n2) the fields Fm (z) with an even mode number m have even symmetry with respect to the Median plane of the layer, and the fields with odd mode number m have odd symmetry: F ", (z) = (-1) m Fm (W2 - z). (10) By substituting this The relation in equation (7) then follows that under the condition of equation (9) V (L, z) = V (0, Wz - z) (11) holds, i.e. that is, there is an inverted mapping.

The feasibility of the conditions in equations (8) and (9) for a Layer leader with spatially constant refractive index m is derived from his dispersion equation derived: xmWz = 2 # m + m #. (12) Here m = 0, 1, 2, ... is the mode number, which is identical to the number of parallel nodal planes of the mode function Fm (z), and xm = = (k2 2 ~ / ,, 2 1/2 (13) the transverse component of the propagation vector in a shift leader theory with zigzag plane waves, and (-2 # m) is the reflection phase of these waves at the interface ni / no. Furthermore, k = 2 # / # the constant of propagation of light in a vacuum. With the introduction of a polarization index a (for TE polarization: o = 0, for TM polarization ol = 1) the reflection phase to be expressed as #m = arctan [(n1 / n0) 2 # (ßm² - k²n0²) ½ / xm]. (14) Not in one Phases #m of most modes are too thin layer conductors with many modes near #m # # / 2. Therefore, the right hand side of equation (14) becomes a series develops #m = # / 2 - D10 x + higher terms. (15) For all fashions with not too high The mode number applies (with ßm = kn;) Dlo = k (n0 / n1) 2 # (n1² - n20) 112. (16) After Substituting equation (15) into dispersion equation (12) is obtained if it is neglected of the higher terms of equation (15) = (m + 1) # / II eq.z, (17) where here the equivalent Thickness was introduced as Weq.z = Wz + 2 D10. (18) Finally, it follows from the equations (17) and (13) after some transformations, where still (ß0 + ßm) # 2km is set, the phase position #m of the m-th mode, based on the basic mode (m = 0) to #m = (#m - ß0) L = - # (m² + 2m) [L # / 4 n1 Weq.z] (19) It follows from this relation then directly the validity of the conditions of equations (8) and (9) for the shift supervisor, provided, however, that the square brackets in equation (19) are a whole Number hz is equal. The latter requirement, however, is already in equation (1) asserted imaging condition that is thus proven. The ideal phase positions the modes are Tm = - (m² + 2m) h2 #. (20) To improve the quality of the image by means of a shift supervisor becomes an infinitely thin object at z = zo assumed in the object plane x = 0, and the resulting light distribution of the image calculated at x = L. For this purpose Vo (z) = # (z-zo) is set, where # is the so-called Called delta function.

According to equation (6), the amplitudes am = Fm (zo), and the image results from equation (7) Here, equation (8) has been assumed to be valid, and the phase factor exp (ißoL) has been omitted. For the evaluation, the mode functions of the shift supervisor were used in a form that corresponds to the approximations made in the previous section.

For the sake of simplicity, the difference between the geometric thickness Wz and the equivalent thickness Wzq'z are neglected. (The index "Z" or "eq, z" in W is therefore often left out in the following.) For shift supervisors with many modes, however, the results obtained remain qualitatively correct. With xm from equation (17) the mode functions are explicitly F ", (z) = 21/2 sin xmz. (22) and the evaluation of the sum in equation (21) yields the amplitude distribution sin [(2M + 1) # (z - z0) / 2 W] 1 sin [(2M + 1) # (z + z0) / 2 W] S (z, L) = ½ - ½ (23) sin [# (z - z0) / 2 W] sin [# (z + z0) / 2 W] Except when z or zo are in the immediate vicinity of the interfaces z = 0 and z = W of the layer, you can add the 2nd link here neglect the right side. The first link then shows that the light distribution in the image depends only on the distance (z- zo) from the image center 20, but not on the absolute position 20 of the image. This behavior has also been observed experimentally. The full half-width r of the intensity distribution l V (z, L) 12 is roughly r = WIM. (24) This width also has the meaning of a spatial resolution for the shift supervisor as an imaging system: Two object points only result separately recognizable images if their distance is at least r. The statement of the equation (24) is therefore also that the maximum number of solvable points is equal to the number of Mder zur Figure is the amount of modes of the waveguide.

From the relation r = W / M it follows that the resolution is the better, the more modes M constructive to the mapping.

This number M of contributing modes is due to various factors limited: a) The number M cannot be higher than the number Mo of those capable of spreading at all Modes of the shift supervisor. In order to keep Mo large, the refractive index differences are allowed (ni - no) and (nt - n2) should not be too small, and W> A should be. b) The derivation the imaging conditions happened under several neglects, e.g.

Omission of the higher terms in equation (15).

The impact of these neglects is very little on the Modes with low number (m <Mb), but it can increase rapidly at higher m. The actual phase systems give way for the highest of the modes possible according to (a) ç> m in general even differs considerably from the ideal value given in equation (20) away. These highest modes therefore do not contribute to the improvement of the image quality, but cause an undesirable reduction in brightness contrasts and can even create wrong structures in the picture. For this reason it is advantageous to if M is significantly smaller than Mo, for example M <Mo / 2. This shows again the advantage of a high total number Mo of possible modes. The necessary limitation of M can practically take place in that the aperture of the illumination optics is limited will.

Apart from this, however, it is also possible to exert the influence of the above To reduce neglect through constructive measures.

These neglects are partly opposite Influences on the phases 1pm.

Therefore, in principle, by varying the refractive indices n0 and n2, the thickness Wz and possibly the polarization the phases #m for a possible make a large number of modes nearly equal to the ideal values of equation (20). At this Optimization must of course be the dispersion equation (12) for all modes exactly (numerically) be solved. An even further reduction in phase deviations and a associated improvement in image quality can be achieved by attachment one or more thin, dielectric interlayers between the material ni of the shift supervisor and the adjacent media with refractive indices n0 and n2.

In Fig. 4 is a waveguide with phase correcting thin layers 41 made of dielectric material. c) For a given imaging system the deviations of the phases from the ideal value are smallest when using the smallest possible number hz, i.e. for hz = 1 or 2. In the higher-order images (hz> = 3) the phase errors of the modes are correspondingly larger, and the image quality is correspondingly larger is worse. d) Last but not least, the number M is the contributing factor Modes limited by the tolerances in the mapping equation (1) Sizes L, # 1 W no, n1, n2. These tolerances have the effect that the corresponding to the hz expression in equation (19) is practically not exactly an integer. If the out of it resulting deviation of a phase #m reaches the value # / 2, the relevant hears Fashion to contribute constructively to the creation of the image. Wear all higher fashions then of course nothing more.

This results in a mapping to which M modes still contribute should that the relative deviations # L / L, ## / #, 2 # W / W from the equation (1) sufficient theoretical values must all be small against 1/2 hz (M2 + 2A «).

The refractive index m goes into the imaging condition in two ways Equation (1), namely on the one hand directly, on the other hand indirectly via paths. Since the the latter quantity, which decreases steadily with increasing n1, exists for given materials no and m always have a very specific refractive index ni, for which the imaging condition according to equation (1) is insensitive to small fluctuations of nl. These Excellent refractive index can be determined from the condition # (n1Weq.z) / # n1 = 0.

If one selects nl = m, an image is focused once stay sharp even with small changes in m. Because of the two above opposite influences of nl on equation (1) it is in a corresponding way even possible through a suitable choice of the refractive index m smaller temperature-related To compensate for changes in L, W and ni against each other. The condition for a such temperature-compensated mapping reads ö (fl2We2qjL) (DT = 0. e) Finally it should also be noted that there is an imaging error in asymmetrical (120 ¢ m) layer conductors can arise because the symmetry relation equation (10) is only approximate is applicable. This error only occurs in asymmetrical lines and only in reversing lines Figures (odd hz) on. Like the other errors mentioned, it decreases with increasing Mode number M and can limit the image quality. In symmetrical ladders (no = n2) this error does not exist.

If the sheathing of the shift leader is not symmetrical (no + m), the expression Weq, z = W2 + Dto + D12, where D12 is defined analogously to equation (16). In this case it was the equivalent Layer thickness has already been given in detail in equations (2) and (3).

This is described in detail using the example of the dielectric layer conductor The mapping principle can also be applied to more general types of waveguides.

First of all, it is clear that the material has the refractive index m of the dielectric Waveguide can be solid, liquid or gaseous if it is used at the Wavelength is sufficiently transparent. The shift supervisor (ni) or one or Both of the outer materials (no, m) can also be from free space (air, vacuum) educated will. In any case, however, the interfaces n1 / n0 and n1 / n2 must be in the case of an almost grazing Reflecting incident light. The higher the reflectivity, the brighter becomes the resulting image.

The reflectivity should also be with a slightly steeper incidence of light still be high, so that the higher modes of the waveguide which are important for the resolution can also be transmitted well.

The highest reflectivity generally gives total internal reflection. It exists if the refractive numbers no and m of the external media are smaller than that Refractive index m of the inner medium. As a special example of a totally reflective one Waveguide is a guide for X-rays called here, which is in the interior empty (vacuum, m = 1) and surrounded by some kind of matter on the outside. Their refractive index (no or n2) for X-rays is known to be somewhat smaller than a, independent on the chemical composition. The practical implementation of such a ladder for A = 0.15 nm, e.g. B. consist of two aluminum vapor-coated glass plates, which are 2.5 cm long and include a gap of 1 m.

In this ladder m = 1 and no = n2 x 1.8 10-6, and it can be in it Mo 53 modes spread out.

In the infrared spectral range and especially in the submillimeter wave range and microwaves can also create high reflectivity by using the walls highly conductive metals for the exterior materials can be obtained.

If the incidence is very grazing, the so-called Fresnel's Almost complete reflection. Therefore, one gets with the waveguide described here a picture even if one (or both) of the external media has a higher one The inner medium has the refractive index nl. In this case, however, takes the reflectivity usually decreases rapidly with steeper incidence. In the resulting The modes are then all the more »leaky« the higher their mode number is. This effect can limit the number of contributing modes and hence the resolution. This case n <no and m <n2 is of practical importance in waveguide lasers.

The essential mathematical requirement for the validity of the Imaging conditions according to equations (8) and (9) is that the differences in the Propagation constants, (ßm - ßo), are in the ratio of whole numbers to each other. According to equation (20) 5, these differences correspond to the sequence of numbers for the shift supervisor (m2 + 2m). The possibility of making an optical image with a waveguide, is thus determined by the spectrum ßm of its modes: Every prismatic waveguide with any cross-sectional shape and any distribution of the refractive index over the cross-section can provide an optical illustration if all distances (Bm - ßo) integer multiples of a smallest distance dß in its mode spectrum are. The length 2 / Aß is then one of the 15 possible mapping lengths. If the said The condition is not exactly exactly or not fulfilled for all modes, becomes an illustration only be possible with reduced quality.

A rectangular 20 strip conductor that is thicker in terms of wavelength (Wz>; t; as Al as shown in Fig. 3 is a waveguide in which the above condition can be met for rational relationships between the mode spacings is. It then conveys a two-dimensional image. In the rectangular waveguide the modes depend on two mode numbers, my and mz.

The field distributions are approximated by the form Fmx, m (y, z) = 2 sin (X ,,, y Y) sin (x ,,,; z) (25) 30 with xmy = (my + 1) sT / Weq y (26) = (»1 + 1): qWeq (27) 35 These approximations apply to mode numbers my and felt which are not too large. The spectrum of the propagation constants is 40 ßmy, 3m = (k²n1²-xmy² Also for small my and felt you can find the phase positions of the modes as The condition for rational relationships between these mode spacings is fulfilled when the transverse dimensions Weq, y and Wzq'z of the waveguide are related to one another like the square roots of small integers. The simplest design is then the waveguide with a square cross-section. Its imaging properties have already been discussed above.

Stand in waveguides with cross-sectional shapes other than polygonal the mode spacings, as far as known, not in approximately rational relationships to each other, so that no images in the sense considered here are possible are. The dependence of the mode spectrum on the cross-sectional shape of the conductor opens up however, the possibility of correcting the mode spectrum of a rectangular or square waveguide to affect its imaging properties to improve. The correction is, for example, a very minor one Modification of the rectangular or square Cross-section so that it is slightly barrel-shaped or is distorted in the shape of a pillow. The resulting shifts in the spectrum the ßmymz are of different sizes for the modes with low and high mode numbers. This different behavior can then be used to create undesirably large ones Deviations of the phases m from their ideal values according to equation (20) for as possible many modes to reduce the residual error as small as possible, so that the number The modes that contribute constructively to the illustration are enlarged. The exact shape and the correct amount of optimal cross-sectional distortion depend on the values the refractive indices and can in individual cases by numerical calculation of the mode spectra can be determined in the case of a systematically distorted cross-sectional shape.

Another possibility for correcting the phases Ym is that in the case of the one shown in FIG. 4 shown layer conductor mentioned application of thin dielectric Layers on the wall of the actual Waveguide medium the refractive index M. This possibility also exists for conductors with rectangular or different cross-section.

For some applications, imaging systems are required that are compatible with given resolution r have a field of view that is larger than the size W = M / r, dealing with a single layer conductor or a rectangular waveguide can be achieved practically. In these cases, waveguides can be stacked or bundled to get a larger field of view. Such a stack of shift leaders 51 is shown in FIG. 5 (a) in a perspective view and in FIG. 5 (b) in cross section shown. Equally strong conductor layers with the refractive index m are created by thin intermediate layers 52 of a material with a lower refractive index no separated.

The intermediate layers must have such a thickness that adjacent Conductor layers are "optically isolated" from one another. If the difference in the refractive index is sufficiently high, z. B. (m -) 2o) III> 0.1, a thickness of the intermediate layers of about one is sufficient Wavelength. If the thickness Wz of the waveguide layers and their length L according to Equation (1) are dimensioned, each layer maps its end faces on one another. So that the stripe-shaped images of the individual layers in the initial plane of the stack continuously to form an overall picture, all individual layers have to be be dimensioned for operation with even Hz. With a simple stack according to FIG. 5 (a) is therefore a one-dimensional, upright image of a large field of view possible. Its size is equal to the height of the stack and can therefore in principle be arbitrary to be made big.

A special version of a single shift supervisor or one Stack is given when its width B (see Fig. 5a) is equal to the length L (or a Multiples) is made. A ladder or stack of this size has the remarkable Property that it has its four side surfaces perpendicular to the plane of the layer maps onto the opposite side surfaces.

F i g. 5 (c) shows the bundling of square strip conductors 53. A two-dimensional image with a large field of view can thus be achieved. The individual strip conductors 53 must again be straight in accordance with equations (1) and (4) Numbers hy and be measured, and for their optical "isolation" this applies to them stacked shift supervisor 51 mentioned. The optical behavior of such a bundle is comparable to that of a so-called fiber optic plate. Compared to the latter is, however, with a bundle of the type described here according to FIG. 5 (c) a higher one spatial resolution r possible, with a significantly smaller number of individual fibers and at the same time increased packing density.

In the case of the fiber optic plate, the resolution is determined by the fiber diameter limited to about 6 .mu.m, while with a stripline a resolution r z 1 .mu.m or better attainable. In the present case, however, there is a need low tolerances of the individual conductors and monochromatic lighting.

The terminology "one-dimensional" or "two-dimensional" used here so far The illustration of the layer or strip conductor refers to the illustration of the waveguide end faces on each other, and it corresponds to the fact that the walls of the waveguide are the Restrict the free propagation of light in one or in two dimensions. Well educates a waveguide but not just the one pair of planes x = 0 and x = L on top of each other but all pairs of planes whose distance L satisfies equation (1), such as z. B. the pairs AA '> BBt CC' etc. in Fig. 6 (a). But that means that the volume of the waveguide is mapped. In this sense, therefore, a waveguide enables also a mapping in the 3rd dimension (x-direction).

The above consideration shows that the figure considered so far the end faces on top of each other still exist even if these faces are not perpendicular to the direction of the conductor, but the same angle lx + 90C include with the conductor. Corresponding waveguides are shown on the basis of Fig. 6 explained. From the illustration in FIG. 6 (a) it follows that in the case of a reversing Figure (odd number hz) the end faces must be inclined towards each other like in Fig. 6 (b), while in an upright image, the end faces are parallel must be as in Fig. 6 (c). The size of the angle o; is arbitrary. This angle should but not be too small, otherwise the number of modes contributing to the mapping is restricted. If in a shift supervisor the M-th mode is still for illustration should contribute, then os must be greater than the angle between the x-direction and the direction of the two plane waves that make up fashion can think: aM = arcsin (zM / knl) ~ (M + 1) 42 q Weq. (29) Due to the in F i G. Figure 6 (a) of the 3rd dimension is also an even more extensive modification of the waveguide end faces possible without loss of the imaging properties. the End surfaces can have any shape, regularly or irregularly curved have such. B. is shown in Fig.6 (d). If only the shapes of entry and The exit area is the same (or, in the case of an odd number hz, the other way around) and the If the waveguide has the dimensions prescribed above, it becomes the entrance surface mapped onto the exit surface, and vice versa. These ways, inclined or to map curved end faces on top of one another, of course, exist with rectangular ones or prism-shaped strip conductors in the same way as with the layer conductor.

Referring to FIG. 7 should now be shown that a figure in the 3rd dimension there is not only between two volume elements, both of them lie within the waveguide as in FIG. 6 (a) but also still when one or both volume elements are located entirely outside the waveguide. In Fig. 7 there is an illuminated or self-luminous object A at a distance d in front of the entry surface x = 0 of waveguide 73 between a polarized monochromatic light source 71 and the objective of a microscope 72 serving for observation is arranged. The light emanating from object A creates a certain, complex amplitude distribution based on the correct dimensioning of the waveguide the exit area x = L is mapped. The one emerging from this area x = L Light thus has the same spatial distribution as the light emitted by the Plane x = 0 would run to the right if the waveguide were not there. The light in the space x> L therefore seems to emanate from an object that is on Places x = L - d, namely from the virtual image A 'of the object A. From the the above derivation can also be seen that the areas 0 <z <Wz of the entrance and exit surfaces are the entrance and exit pupils of the imaging system. It can also be seen that this is for a shift leader made considerations in an analogous manner for the rectangular or prism-shaped Strip conductors apply and that the shape of the conductor end faces does not matter either play, provided that the symmetrical or antisymmetrical in the sense of F i g. 6 are. Finally, it should be mentioned that the mapping shown in Figure 7 is reversible. If you project an image into the entry surface of the waveguide at the point x = + d, it is mapped to x = L + d and can be captured there on a screen will.

An imaging layer or strip conductor, with the object A and Real image A's both outside of the waveguide is shown in Fig. 8 Cross section shown. For the sake of simplicity, it is assumed here that the optical Media upstream and downstream of waveguide 83, d. H. in the spaces x <0 and x> Lv, have the same refractive index nl as the waveguide. Let the object A be small (height H «Wz) and is located near the center plane of the waveguide at x = - d at a short distance from the beginning of the waveguide.

It is now assumed first of all that the length of the waveguide is equal to a value L that satisfies equation (1) (shown in dashed lines in FIG. 8). Then the waveguide creates an image A 'of the object at the point x = L - d If M denotes the mode number of the highest excited mode in the waveguide, then the image A 'is designed only from those rays of light which are at angles of at most OtM run against the x-axis, with lxM according to equation (29). For this reason remain in FIG. 8 areas marked with double hatching are dark. Consequently it can't affect the quality of the picture A 'if one in the specified Region Lv ... L omits the reflective walls of the waveguide, d. H. the waveguide shortened to length Lv.

Thus the image A 'comes outside of the length Lv of its walls defined waveguide: The previously virtual image is a real one become.

In the illustration by means of a waveguide shortened in this way means the length L, which goes into the basic equation equation (1), no longer the length of the waveguide, but rather the distance of the image from the object. Similar This distance is like a 1: 1 imaging, conventional optical system L independent of the distance d of the object.

The possible shortening (d + d,) of the waveguide is of course limited. As the shortening increases, the quality of the image A 'decreases because the above ray-optical one Argumentation is only admissible in the vicinity of picture A '. The maximum possible The shortening is determined by the thickness Wzof the conductor and the angle xM, and it also depends on the size of the object to be imaged. This is in F i G. 9 illustrates. According to the considerations in connection with FIG an image will only be sharp if the image point A 'is doubled in the one shown in FIG hatched angular range at the end of the waveguide 93 lies. As the figure must be reversible, it must at the same time be required that the object point A in the marked angular range at the beginning of the waveguide. From Fig. 9 it follows: The more the waveguide is shortened, the closer the points Pl and Pl move at which the image field has the size 0, and the smaller it becomes in focus mappable field of view. According to FIG. 9 maximum possible shortening amounts to each end dmaz = dniz # W2 / 2 (AM ~ n1Wz² (M + 1) i. = L / 4 hz (M + 1) in the example of the layer conductor considered above (m = 1.5; W = 50 Fm; # = 632 nm; L # 24 mm, M = 20) equation (30) provides the value for this d'maz 0.3 mm. Of the maximum reductions According to equation (30), only part of it can be used in practice, because it shrinks once yes for d. dmax or d '. dZmax the image field size to zero, and to the other step already when the object or image point approaches the limits of those marked in FIG Areas with stronger image defects. This follows from a closer examination of the Light distribution at the border of the dark rooms marked in FIG.

The in F i g. Diaphragm 91 shown in FIG. 9 will be explained further below will.

Above it was assumed that the spaces x <0 and x> Lv der 8 are fulfilled by a medium of the same refractive index n as the waveguide. This requirement can be dropped. When the end faces just cut off are as in Fig. 1 and if the object space x <0 with a medium of the refractive index nG is fulfilled and the image space x> Lv with a medium of the refractive index na, see above the refraction of light at the end faces causes a shift of the object and pixels in the x-direction. Which then into the mapping equation equation (1) incoming quantity is L = L, + (nGd + nBd ') / n1 (31) when deriving this Relationship was accepted men that all imaging light rays at small angles (<MM) to the x-axis.

The possibility of mapping with a shortened shift supervisor was experimentally carried out on the above-mentioned, liquid-filled layer conductor (Fig. 2) Observed: For this purpose, the thickness Wz of the conductor layer was determined by reducing the adjustment pressure P slightly enlarged. According to equation (1), this became the image length L is greater than the length Lv of the quartz plates, and the image of the gap should be outside of the plates arise at x> Lv.

In fact, the observation microscope had to be removed from the waveguide remove something to see the slit image in focus again. Conversely, immigrated to Reduction of the plate spacing, i.e. the thickness Wz, the gap pattern in the area x <L, as could also be determined by readjusting the microscope.

In the case of an imaging waveguide of the type described here, it may be useful to provide apodization. In equation (23) was the amplitude distribution S (z, zo) expressed, which prevails in the image that a waveguide of an ideal designs point-shaped (or line-shaped) object.

The associated intensity distribution has the central maximum a series of relatively strong secondary maxima. They come from the fact that S (z, zo) is the Fourier transform is the light distribution in the mode spectrum and that this distribution is abrupt at the mode number M has been cut off. But if so such higher Modes should nevertheless be stimulated appreciably, which equations the imaging conditions (8) or (9) do not meet, then in the point mapping function S (z, zo) numerous other maxima and fine structures occur in the "image" of a point are highly undesirable. These disturbances and the mentioned secondary maxima can be attenuated or be eliminated if the high fashions m> Man of image creation succeed to be less involved or not at all. For this so-called apodization exist especially when imaging with a waveguide (layer or strip conductor) two possibilities: either the propagation of the higher modes on the waveguide attenuated, so that they arrive at the location of the image in a severely weakened manner, or it takes place a spatial filtering of the image.

In a thick waveguide with lossless core material (m) an attenuation of the mode propagation by absorption of the guided light in the surrounding materials or by refraction of light into them Materials (no> m; n2> fix). The absorption constant increases for both damping mechanisms proportional to (m + 1) 2 with the mode number m. Both mechanisms are suitable therefore good for apodization. In practical terms, it just depends on the strength of the losses set so that they only take effect from the desired mode number Mab. Crap z. B. the highest mode up to which the imaging conditions equations (8) and (9) are valid. KM denotes the effective absorption constant of this Mode of the waveguide, the apodization requires kLKM = 1.

If the attenuation is based only on absorption in the substrate material (no), so one finds the necessary absorption constant zo (imaginary part of no) of this material with TE polarization as x0 ~ (n2, n0) 3/2 (M + 1) -2 WZ / 2 h, n, i. (32) In the example of FIG. 2 gives this formula zo = 2.7. 10-3 what an absorption of the solid material (no) of 240 dB / mm. The substrate material must so be a practically opaque black glass.

A very similar effect can also be achieved if one according to FIG. 10, very thin metal layers 101 at the interfaces ni / no and nl / n2 provides. These layers can e.g. B. made of aluminum, gold, nickel, or chromium, their thickness must be 10 to 100 nm. The metal layers can also be combined with dielectric layers, which may be used to correct the mode spectrum these interfaces are applied.

Metal layers can also be used as reflection-increasing layers, namely in so-called leak waveguides, which are characterized by no> m or m> m are. There is no special, reflection-increasing measures in these leakage waveguides namely the attenuation so high that there is an undesirably strong apodization, the results in a reduced resolution r. By a partially reflective, in limited Dimensions of absorbing metal layer (101) of suitable thickness (10 to 100 nm) is the Leak attenuation is reduced and the apodization can be adjusted to the desired level will.

The disturbing modes m> M differ from the lower ones Modes in that they have beam directions with larger angles, olm> olM. Therefore can the light of these higher modes through a simple diaphragm 91 in larger Distance behind the image can be filtered out, as shown in FIG. 9 is indicated.

The size of the aperture 91 must be dimensioned so that it covers the angular range 2lxM can pass. Even more advantageous than the abruptly cutting aperture a mask whose transparency decreases continuously from the center to the edges according to a desired apodization function. If you watch it from the waveguide generated image then through the aperture resp.

Mask, the mentioned disturbing secondary maxima remain invisible, because the corresponding high spatial frequencies are missing.

A waveguide 113, the cross section of which gradually widens (FIG. 11) or tapers in the direction of propagation, can convey an enlarged or reduced image. The linear enlargement or reduction ratio is equal to the ratio of the equivalent conductor thicknesses at the end and the beginning of the conductor. The possibility of such a scale-changing mapping is based on the principle known from microwave technology that in a waveguide with a cross-section that can be changed in the x-direction, a mode that has been excited adapts to all cross-sectional changes, provided that these changes take place very slowly, ie over distances of very many wavelengths. Therefore, at the beginning of the waveguide according to FIG. If the modes are excited with certain amplitudes am, there is still the same distribution of amplitudes am at the end of the waveguide, but the mode functions Fm (z) have changed. According to equations (22) and (17), their change in the layer conductor corresponds to a scale transformation in the ratio of the equivalent layer thicknesses at the beginning and the end of the conductor. However, an enlarged (or reduced) image only arises in the exit surface when all modes are put together again in phase. The conditions according to equations (8) and (9r are again exp (S2m) = 1 and exp (Apm) = (- 1) m, with the relative phase positions lym through given are. With the assumption that in every mode of light adiabatically adapts to the local cross-sectional shape and runs with the propagation constant fim (x) corresponding to the local cross-section, the phase positions for can be calculated. In the same approximation as in equation (19) one finds am = - (m2 + 2 m) n [L / 4 q Weq, z] (34) with an averaged equivalent layer thickness By comparing equation (34) with equation (19) it can be seen that all the results obtained for the layer conductor with uniform thickness also remain valid for the layer conductor with locally slowly changing cross-section, if the equivalent thickness Weqz is divided by the Equation (35) replaced mean thickness. In particular, the fundamental mapping equation equation (1) as well as the considerations about inverted and upright mapping remain valid, and there are also all mentioned possibilities of one-dimensional and two-dimensional mapping with shift supervisor or

Stripline further. The number M of the constructive to the creation of the picture contributing modes is, however, determined by the narrowest cross-section. she is thus smaller in the conductor considered here than in a uniform conductor of the same type medium thickness WeqSz. This fact is of importance for the practically achievable Enlargement or reduction ratios, because the narrowest cross-section must always allow a not too small number MP 1 of modes, thus one at all meaningful mapping can come about.

Another consequence of equation (35) is that slight local Differences in thickness, e.g. B. as a result of unevenness of the interfaces flilfil, does not lead to a loss of the imaging properties as long as only the middle one Thick Weq, z satisfies equation (1). The tolerance A wi calculated above W therefore only refers to the mean layer thickness, but not to local deviations, which can be bigger.

The in F i g. 12 is the wedge-shaped layer conductor 123 shown simplest form of a waveguide with a locally variable cross-section, since it has flat walls or interfaces 124, 125. For this conductor results in equation (35) that the mean thickness Weq, z is equal to the geometric mean of the equivalent Layer thicknesses at the beginning and at the end of the conductor is = = [We, .2 (0) s We! Z (L) 1 '. (36) As an example, let us again consider a shift supervisor similar to the one in FIG. 2 shown considered. When its layer thickness widens in a wedge shape from 25 µm to 100 ltm, the averaged equivalent layer thickness is again Wzq'z 5t: 50 μm. You get then with a conductor length of L = 24 .mu.m and red He-Ne laser light a reversed one Image that is enlarged in relation to the conductor end faces, i.e. 25: 100. At the at the thin end there can be about Mo = 25 modes.

If M = 15 of these contribute constructively to the creation of the image, is the spatial resolution at the thin end r = WIM x 1.7 µm and at the thick end r Y 7 um.

The possibility of using a waveguide limited by flat surfaces Obtaining an enlarged image is of particular interest to the problem the imaging by means of X-rays. Also important is the possibility that one in the case of a layer conductor according to FIG. 12 with liquid or gaseous conductor material with the refractive index m can change the magnification ratio in the simplest way. For this it is only necessary to move one of the interfaces 123, 124 of the conductor relative to the to move others, in such a way that the layer thicknesses at the end faces in change the desired way, but with the geometrically averaged layer thickness must be preserved.

In the mode of operation explained above, it was assumed that the medium with the refractive index nl as a light-guiding film on a substrate with the Refractive index no is applied (Fig. 1). The expansion of the shift leader in the y-direction was assumed to be very large versus wavelength.

F i g. 13 shows in perspective a different possibility, the principle of imaging an object A as an image A 'by means of a light source 131 fed waveguide 133 to use in integrated optics. The substrate 132 has the refractive index ns. Of the The waveguide shown has a refractive index m> ns in the y-direction only a thickness Wz of the order of magnitude of a light wavelength, see above that only a few modes are viable in the y-direction. Here is the For the sake of simplicity, assume that it is only a single fashion. With the right choice of his Dimensions (L, Wz), the waveguide 133 provides a one-dimensional image its one end face at x = 0 on the opposite end face at x = L. The mapping equation equation (1) applies again. Instead of the refractive index Here, however, m is the so-called "effective index" No of the shift supervisor mode used to use. The effective thickness Wzq'z in the z-direction is because of the goos chicken effect again very slightly larger than the geometric width Wz of the waveguide. But it can be used for the waveguide of FIG. 13 not according to equations (2) and (3) must be calculated, but must e.g. B. be determined empirically.

There is no particular restriction for dimension Wz.

The practical implementation of such an imaging waveguide the integrated optics could consist of a quartz substrate with a refractive index ns = 1.458 consist, on which a film of (BaO + SiO2) with the refractive index by cathode sputtering = 1.60 is applied. With a layer thickness Wz = 0.5 µm, the effective Index of TEo mode for He-Ne laser light about No - 1.55. This waveguide yields a mapping with hz = 1 if its dimensions, for example, approximately L = 4 mm and Wz = 20 µm. In the z-direction of this conductor, about Mo = 17 modes are able to exist. If m = 10 of these contribute constructively to the mapping, one obtains in the z-direction 13 shows a spatial resolution of approximately r = 2 μm.

It is evident that with waveguides according to FIG. 13 with appropriate Modification of the dimensions, also enlarging and reducing images possible are. Further, according to the above-mentioned stack, there can be an arrangement more numerous provide the same conductor tracks next to one another in the z-direction on the same substrate, to get a larger imaged field of view.

It is known that most of the geometrical-optical laws, which describe the propagation, reflection, refraction and diffraction of light corresponding formulation also apply to sound waves. Instead of light frequency, The terms sound frequency and sound wavelength occur in light wavelength and speed of light Aw, and the speed of sound v, and instead of the optical refractive index, in the case of sound the reciprocal value 1 / v of the speed of sound v can be used. Also dismantling one Wave field in a sum of orthogonal modes is just as possible with sound as with light. For these reasons, the law described by equation (1) exists for imaging by means of a waveguide also for sound waves in solid, liquid and gaseous media.

An imaging waveguide 143 for sound waves is made according to FIG. 14 from a layer or from a rectangular prism of a material that is acoustically is permeable, i.e. hardly absorbs sound. For sound waves also come Metals as conductor material in question. A waveguide based on total internal reflection, is surrounded on the outside with other acoustic materials that allow higher speeds of sound (w and V2) have as the conductor material (vl). Instead, it is enough if the acoustic impedances of the conductor material and the environment differ greatly, z. B. a solid conductor in a gaseous environment. In Fig. 14 is an imaging acoustic Shift supervisor shown in cross section, who uses a sounder A as an object "Sound image" A 'depicts that used by a consumer or detection device 144 will. If V (x, z) is the local sound amplitude (i.e. the maximum particle displacement) is referred to, then a general sound field can be correspondingly used in the case of light the sound amplitudes V (x, z) in the waveguide according to equation (5) as a superposition represent a number of simultaneously excited modes of the waveguide. The mode function for compression waves and for the two possible polarizations of shear waves differ slightly from each other and from the mode functions of light. The differences lie mainly in the different penetration depths D des Sound into the acoustically thinner materials, which are also TE and TM polarized Distinguish light. With sufficient thickness Wz of the layer (Wz> A) applies to the the lowest modes, however, in each case D g <Wz, and the modes of the waveguide and the imaging conditions are given by equations (12l (13), (15l (17)) through (20) and described by equation (1). For reasons of analogy it then follows for the sound waveguide also the validity of all the considerations made above Resolving power, phase corrections, tolerances, two-dimensional imaging, waveguide stacks and bundles, modification or end faces, imaging with shortened waveguides, Apodization and zooming in and out. Furthermore, one of the Fig. 13 corresponding imaging system with surface acoustic waves is possible.

The simplest example is a two-dimensional imaging acoustic Stripline mentioned. It is cantilevered and made of fused quartz. Its cross-section is square 2 x 2 mm, its length is L = 200 mm. This Ladder creates an inverted mapping (hy = hz = 1) of its end faces onto one another for a sound wavelength of about A = 80 µm, which is when using compression waves corresponds to a sound frequency of about 75 MHz: At 10 constructive for illustration A spatial resolution of r = 0.2 mm is to be expected for the contributing modes.

Below are some more arrangements for particularly expedient Application examples of a waveguide according to the invention described.

The simple transfer of an optical image from a face an imaging waveguide on the opposite end face has already been explained.

If it is an ordinary, two-dimensional image, a rectangular or square conductor cross-section is appropriate. The illustration can be chosen upright or inverted, enlarging or reducing. For a bundle of conductors can be used for a large field of view. As monochromatic Light is required, this type of imaging is especially important for optical systems Image evaluation and processing of interest. The transmission can e.g. B. from the Inside a vacuum vessel to the outside or vice versa, for example in the case of an image recording tube, or from rooms of high pressure, high temperature or high radioactivity or in such spaces into it.

When building systems that use monochromatic radiation, for example in the wavelength range A = 50 CLm ... 5 mm, there is occasionally the problem to transmit such radiation in the laboratory over distances of a few meters without thereby significantly worsening the mode purity of the beam (e.g. TEMoo). Optical transmission systems with lenses are for this purpose due to the lack of good, low-absorption systems Materials not very suitable A transmission in a metallic hollow tubular waveguide, which only has a single mode causes considerable absorption losses. In one heavily oversized waveguides, the losses are low, but it is difficult to to excite a single mode in a targeted and clean manner in such a multimode waveguide. An imaging waveguide can also be used here with advantage. if the radiation to be transmitted has a sufficiently small cross-section, it can pass through the coupling-out hole 152 (FIG. 15) of a sub-millimeter laser 151 directly into the one The end of a waveguide (153) mapping according to equation (1) is irradiated, without taking into account the multimode properties of this waveguide is. At the other end there is a "picture" of the entrance, so that the one at the exit A beam fed to a consumer 154 has the same spatial structure as the Input beam.

An example is the transmission of the radiation from an HCN laser Shaft length A = 0.337 mm.

The radiation will z. B. through a hole 4 mm in diameter coupled out to one of the resonator mirrors of the laser 151. This radiation is supposed to go through an imaging waveguide over a distance of e.g. B. L = 4.75 m transferred will. This can, as shown schematically in cross section in FIG. 15, by means of of a square metallic hollow tube happen. If that one end of the pipe is placed directly in front of the decoupling hole 152 of the laser, then emerges from the other At the end of the hollow tube, a beam of the same diameter (4mm) as the one Laser beam directly behind the decoupling hole, provided that the hollow tube satisfies the mapping conditions of equations (1) and (4). In the present case does that mean z. B. a hollow tube with a square inner cross-section of Wy = Wz = 20mm. The figure is reversed (hy = hz = 1).

The principle of transmission and mapping just described can one use in the integrated optics to the radiation of a GaAs injection laser or amplifier 161 to be coupled into a given light-conducting layer 162 (Fig. 16). A problem with this coupling is that the active zone or Layer 164 determined the active zone of the laser not easily in a plane with the light-guiding layer 162 can be brought. When the laser hits the substrate 165 is put on, instead there remains a height difference of e.g. H = 10 µm, which is caused by the cover and contact layers of the laser.

As shown in Fig. 16, this height difference can be converted by a reversing Figure to be overcome with a shift leader 163. The exit surface 168 of the laser is as an object through this layer conductor 163 of the thickness Wz in the end 169 of the layer 162, the thickness WF of which is, for example, 1 µm. About the manufacture avoiding a sharply cut end face of the conductive layer 162 brings about a cover layer 166 of low refractive index nn with a sharp edge is applied to this, which defines the conductor end of layer 162. At the same time, the refractive index m des The material of the layer conductor 163 is equal to the refractive index nF of the layer 162 is selected so that the layer conductor 1163 optically extends over the length indicated by L. is partly formed by the layer 162, that is to say by both layers together.

For a high-resolution image it is advantageous to use the shift supervisor 163 to make symmetrical, d. H. for its cover plate 167 a material with the same Refractive index m = no like the substrate 165 to be selected. The thickness Wz and length L of the shift leader 163 must be measured with an odd hz according to equation (1).

So that the image of the light exit surface is exactly at the level of the layer 162 is designed, Wz = WF + H must be chosen. For example, if WF = I um and H = 10 µm, then for AS = 8900 A, m = 1.51 and hZ = 1 an image length must be of L = 0.82mm can be used. It can be advantageous for the exact setting of the figure be the layer conductor 163 made of a suitable liquid or plastic material to manufacture. The figure can then be moved up and down and by tilting the Cover plate 167 can be optimally adjusted. Then the cover plate 167 is in fixed in their position, e.g. B. by polymerizing the material of the layer conductor (1).

Another solution to this coupling problem is possible if the Height H of the active laser zone above the base is so great that it is impractical would result in large imaging lengths L according to equation (1). In this case it is it is possible to arrange the GaAs laser 171 laterally next to the substrate 175, as in FIG F i g. 17 is shown in cross section. In this arrangement there must be the difference in height Ht and the layer thickness Wz can be adjusted until an optimal image is achieved is.

It is advantageous that the edge K2 of the substrate 175 is not needs to be perfectly trained, because it is located at a point of Darkness. The edge Kl of the cover plate 177, on the other hand, should be as perfect as possible and be straight, because it is almost in the light beam of the laser. If the Edge Ki is rough, part of the laser light is scattered. The light-guiding layer 172 and the cover layer 176 correspond to those according to FIG. 16.

The arrangement shown in FIGS. 16 and 17 can be modified somewhat Form used to make two layer leaders of the integrated optics so with each other to couple so that the incoming light in one conductor is transmitted into the second will. A well-known possibility of coupling is the use of a thin auxiliary layer conductor, which is applied to the conductor ends to be connected. The three participating conductors gradually taper towards their ends. A better solution the coupling problem for monochromatic light offers the use of an imaging Shift supervisor 183 according to FIG. 18, for connecting two conductors on the same substrate or on separate substrates. The simplest case shown is the connection two identical conductors 182, 182 '(WF = Wfl) on the same or on two of the same Substrates 185, 185 '. A material is then used for the imaging layer conductor 183 the same refractive index selected, m = nF = nF ', so that the ends of the conductors 182 and 182' again be defined by the sharp edges of cover layers 186 and 186 '. The distance L between the cover layers and the layer thickness Wz of the imaging layer conductor 183 are according to equation (1) for an image of the wavelength used with hz = 2 dimensioned. So the figure is upright. In the middle between the end faces E, The conductor creates an inverted intermediate image of the conductor ends. Since this Image is on the side of the cover plate 187, it is on the opposite side the layer of the layer conductor 183 (refractive index nl) dark. So it doesn't It matters whether there is a common continuous substrate there or whether there the joint of the edges K, K 'of two separate substrates lies. It bother too small irregularities in these edges do not affect the figure, provided that their extension in the x-direction does not exceed the value dmax defined in equation (30) significantly exceeds.

The refractive index n2 of the cover plate 187 needs for the one used here upright image not necessarily equal to the refractive index no = no 'of the substrate material to be. In the case of ideal imaging, the coupler only couples the same modes in each case conductors 182 and 182 'together.

Various modifications of the coupling arrangement of FIG. 18 are possible possible: If the layer thicknesses of the two conductors or their refractive indices are different are (WF # WF 'or nFz nF'), n9, a coupling of two identical modes can be can be achieved in that the image is selected to be enlarged or reduced by making the layer thickness Wz wedge-shaped. The enlargement ratio must be equal to the ratio of the transverse propagation constant z ,: z, b 'der Mode y to be coupled can be made in conductors 182 and 182 '. By choosing one different modes (z »: zv ') can also be used for different modes, c and v selectively couple the conductors together.

In the case of different refractive indices (nF Z nF ') the one to be coupled Conductor it is possible to omit the cover layers 186 and 186 'if the refractive index n of the imaging conductor is chosen so small (nl <nF; m <nF ') that this itself acts as a cover layer for the two conductors 182, 182 '. In this case the 18, the length of the imaging condition equation (1), denoted by L2, and the end faces E and E 'of the conductors should be as sharp and rectangular as possible be.

19 shows a laser resonator with an imaging waveguide 193. As shown, a longer piece of waveguide is flat at both ends If mirrors are provided, a cavity resonator is created. The possible waveforms this resonator can initially also be called a longitudinal mode number designated number q of the nodal levels x = const. classify the electric field. In addition to the number q there is a number of transversal modes, the number of which is at Shift supervisor of the so-called transversal mode number felt of the node levels z = const. is equivalent to. The rectangular stripline has two transverse mode numbers my and felt required. The field images of the various transverse modes are identical to the fields of the corresponding waveguide modes with the same mode numbers felt (or the same myund mz).

In a long cavity resonator (L> Wz) owner. in general the different transverse modes belonging to the same number q are different Resonance frequencies. This is undesirable for the construction of lasers. That's why open cavity resonators are advantageously used for lasers in which the higher transverse modes (felt '1) suffer increased diffraction losses and thus can be suppressed. Is now an imaging Waveguide provided, d. H. if its length satisfies equation (1), then all have transversals Modes have the same resonance frequency. This can be seen from the fact that the imaging condition Equation (1) is equivalent to the conditions of equations (8) and (9).

They mean that all waveguide modes in a resonator have the length L are in resonance at the same time.

In a laser whose resonator is dimensioned according to equation (1), but all transversal modes become m = 0.1, 2. . . same longitudinal mode number q not only vibrate with the same frequency, but also with the same phase, because in a laser with frequency-degenerate modes, as is well known, any one is sufficient small non-linearity or other disturbance to phase lock the modes. A resonator with an imaging waveguide therefore appears particularly suitable for Construction of lasers of high spectral purity (»single frequency laser«). Of the Resonator corresponds in a certain way to a Fabry-Perot resonator with concentric spherical mirrors. In contrast to the latter, however, the resonator with imaging occurs Waveguides do not show any increased diffraction losses in the event of degeneration, since all of them lead to Figure contributing modes are guided modes of the waveguide.

The mentioned mode coupling can also be done intentionally and in a well-defined manner Shape can be brought about by using sufficiently small mirrors 201, 202 at the resonator ends according to FIG. 20. The size of these mirrors becomes appropriate selected to be somewhat larger than the resolving power rdes defined in equation (24) Waveguide 203, and its position must be such that the waveguide touches the mirror mutually map each other. One of the mirrors should be partially transparent to decouple the generated laser light.

Because of the small size the mirrors have light reflected from them a relatively large, diffraction-limited aperture. According to the derivation of equation (24) the reflected light propagates in the waveguide in the form of the M lowest Fashions off. Because of the assumed imaging property, almost anything works Light on the other mirror, from which it is reflected back into the resonator will. The convergence of light exactly on the small (r <W) mirror means almost complete mutual cancellation of the M excited modes in the modes not from Areas occupied by the mirrors (W - r) of the waveguide end faces. This is but only possible if there are fixed phase designations between the M modes.

The most important property of the laser resonator according to FIG. 20 is that in a simple manner a large, almost uniformly illuminated volume optically amplifying material so can be used in a laser or amplifier. that a single, spatially coherent, diffraction-limited output beam can be achieved is. The almost uniform illumination is the result of the simultaneous excitation a large number (M) of modes. The only places where fashions completely extinguish each other are the mentioned areas of the end faces next to the mirrors. In all the rest Volume of the waveguide, the intensity distribution of the light is more even, than it would be if only a single fashion was suggested. This prevents the effect the so-called »spatial hole-burning«, which z. B. with the Nd: YAG laser easy to Instabilities give rise to.

A possible modification of the arrangement according to Fig. 20 consists in using a small mirror only on one waveguide end, on the other end but a large mirror as in Fig. 19 which covers the entire end face. In in this case only that small part of the large mirror is effective, which is the The image of the smaller mirror corresponds to The distance L between the mirrors is in all cases to be dimensioned according to equation (1).

Another modification shown in Fig. 21 uses an imaging Waveguide 213, the length Ls of which is only half as great as according to equation (1) for a figure would be necessary (Ls = L12), and the one at one end with a perpendicular to the axis of the waveguide arranged plane mirror Sp 1 is completed.

If the mapping equation (1) is satisfied with an odd hZ, then forms this waveguide reverses its open end on itself. This roughly follows from the properties of the waveguide equivalent to Fig. 21, which is made from the piece the length Lsselbst and the mirror image of this piece designed by the mirror Sp 1 consists. The half-length waveguide is made by setting up the two mirrors Sp 2 and Sp 3 are a resonator at symmetrically located points of the open end. If the size of these mirrors Sp 2 and Sp 3 is sufficiently small, the resonator according to FIG FIG. 21 very similar properties to that according to FIG. 20. Symmetrical in place of the two arranged mirror can of course also be a single, arranged in the axis Mirrors are used.

The small mirror of the examples described above can also go through a slotted or perforated diaphragm B inserted into a waveguide 223 can be replaced is, which at both ends with large flat mirrors Sp 1 resp.

Sp2 is complete, as shown in FIG. 22 is shown.

If the lengths La and Lb of the two waveguide pieces on the right and to the left of the faceplate B are dimensioned so that each side depicts the faceplate in itself, this creates a resonator with the advantages of the one shown in FIG. 20 shown. One the output mirror, e.g. B. the mirror Sp 2, should be used to couple the laser radiation be partially permeable again. If there is another piece of pipe behind it the length Lb connects, the output radiation appears in the image B 'of the diaphragm bundled.

Of great interest to integrated optics amplifiers and lasers are layer or strip conductors that contain rare earth ions. as the most important examples are neodymium-doped yttrium-aluminum-garnet (Nd: YAG) and neodymium pentaphosphate (NdPP) are mentioned. Because of the limited luminance available 'Incoherent pump light sources, it is advantageous to use the amplifying waveguide to pump from the side where it offers the largest area.

This means that a sufficiently large fraction of the pumping light that is seen is absorbed in the waveguide, the layer thickness Wz of the waveguide should be comparable be with the absorption length of the reinforcing material, d. H. about 2 mm at Nd : YAG and about 50 µm in the NdPP. On the other hand, it would be for numerous uses the integrated optics are advantageous if the waveguide only has a single mode led, so would be very thin (about 1 Fm). A possible compromise between these conflicting requirements is to use a thick waveguide, but to operate only one of the many viable fashions in it. this however, it gives rise to very difficult adaptation problems. Here, too, offers the use an imaging waveguide advantages.

The arrangement according to FIG. 23 corresponds in structure and mode of operation largely that shown in FIG. The light to be amplified arrives in one individual mode of the thin layer or strip-shaped conductor 232, which is expediently consists of a material with the same refractive index as the optically reinforcing material, from which the imaging waveguide 233 is made, i. H. nF = nl. As shown this is shaped in such a way that well-defined end surfaces Ki, K2 to the conductor 232 arise, which the waveguide 233 maps into one another. The figure must be upright (hz just). Because of the assumed equality of the refractive indices nl = nF, it is optical effective thickness of the imaging waveguide Wz = W + WF.

The operation of this amplifier is such that that in conductor 232 incoming light at the end face Kl in the thick waveguide 233 passes. In it it spreads out as a mixture of many fashions, is amplified at the same time, and is then fed back into the conductor 232 at the end face K2. Of the Waveguide 233 can be much thicker than conductor 232, so that one improved utilization of the pump light. Is the reinforcing material for example made of NdPP with a refractive index m = 1.60, then at one wavelength A = 1.05 µm and an amplifier length of L = 2 mm, the conductor thickness Wz = 12.5 µm be. This means an order of magnitude better utilization of the pump light, than would be possible in a single mode layer conductor (WF> 0.9 µm). In the Arrangement according to FIG. 23 it is necessary to have reflections on the two end faces of the waveguide 233 to avoid. These areas are therefore irregular as shown designed. Instead, they could also be provided with a suitable absorber.

F i g. 24 shows a filter with an imaging waveguide 243, which allows radiations of a Wavelength Al and twice the wavelength 2 = # 2 | spatially separated from each other. To do this, the radiation mixture (# 1, # 2) through a diaphragm 244 in one half E0 of the first end face of the waveguide fed, which produces an inverse mapping for Al. The mapping equation (1) is thus satisfied with an odd h «(l). The radiation of the wavelength ti therefore emerges from the upper half of the other end face of the designated El Waveguide. For the radiation of wavelength 2 is the mapping equation (1) (2) also fulfilled, but with an even number hd²) = 2 h '). The illustration is therefore upright for A2, and the radiation of wavelength A2 emerges from the with E2, the lower half of the other end face of the waveguide. By A partition 245 can be used to separate the two radiations. In completely analogue Other radiation mixtures (Awl, A2) can also be separated if their Wavelengths behave like Al: A2 = hz (l): hz (2), where (h2 (l) + hz (2)) is odd have to be.

As already mentioned, the cross section of the one described here Waveguide generally be polygonal and z. B. also take the form of a regular Have triangle or hexagon. Also right triangles with certain aspect ratios is being brought up for consideration. In the case of an equilateral triangle, those are decisive for the illustration Transverse dimensions both equal to the height of the triangle, i.e. H. so Wy = Wz = 4fi3 / 4) 1/2, where s denotes the length of the side of the triangle. When the regular hexagon is the relevant transverse dimensions Wy = Wz equal to the distance between two opposite ones Pages. If the cross-section is a right triangle, then Wy and Wz are to identify with the cathetus of this triangle. The associated equivalents Thick Weq, y and Weq.z must meet the mapping conditions equations (1) and (4) fulfill.

Claims (1)

Claims: 1. Imaging system for imaging an object with a layered or prismatic or pyramidal waveguide with reflective Interfaces, characterized in that the length L is along the axis of the waveguide measured distance of the object (A) from the image (A) and at least one of a typical Transverse dimension of the waveguide corresponding size Wzq taking into account the The refractive index nl of the waveguide material and the wavelength A of the following condition in which h = 1,2,3 etc. is an integer: L = 4h n, W 2. imaging system according to claim 1, characterized in that a layered waveguide with essentially flat interfaces is provided, the layer thickness (Wz) the is the transverse dimension corresponding to the size Weq. 3. Imaging system according to claim 1, characterized in that a waveguide with essentially flat interfaces and with a rectangular cross-section is provided, the width Wz and the height or thickness Wz at least approximately the condition L,. = 4 hy nl Wy2 = 4 hw nl W2 are sufficient, where hy and hz are the same or are different integers. 4. Imaging system according to claim 3, characterized in that the two transverse dimensions (Wy, Wz) of the waveguide are like the square roots relate to one another from small whole numbers. 5. Imaging system according to one of the preceding claims, characterized characterized in that the length LY of the waveguide is greater than L - dn, ax = (nG + nB) / n1 with dwax L / 4h (M + 1), where nG or. ns the refractive indices of between the Waveguide and the object (A) or the image (A9 located media and M is the number of parallel nodal planes of the highest excited mode in the waveguide. 6. Imaging system according to claim 5, characterized in that the waveguide has the length L. 7. Imaging system according to one of the preceding claims, characterized characterized in that the light used for imaging comes from a monochromatic one Light source is generated and linearly polarized by a polarizer, whose plane of oscillation parallel or perpendicular to the interfaces of the waveguide. 8. Imaging system according to one of the preceding claims, characterized characterized in that the number (M) of parallel nodal planes of each in the waveguide excited modes is limited to a value that is the quotient of the transverse dimension Wzq and the desired spatial corresponds to the resolving power (r) of the system, however, it is much smaller than the number (Mo) of parallel nodal levels of the highest possible fashions. 9. Imaging system according to claim 8, characterized in that the aperture of the illumination optics to limit the number (M) of excited modes of the system is limited. 10. Imaging system according to one of the preceding claims, characterized characterized in that at a distance behind that imaged by the waveguide (93) Image (A9 a diaphragm (91) is arranged, which filters out undesirably high interfering modes. 11. Imaging system according to one of claims 1 to 9, characterized in that that at a distance behind the image (A ') shown by the waveguide (93) a Mask is arranged with transparency decreasing from the center to the edges, which filters out undesirably high interfering modes. 12. Imaging system according to one of the preceding claims, characterized characterized in that the waveguide at its interfaces with thin phase-correcting Layers (41) made of dielectric material and / or with partially reflective, absorbent Metal layers (101) is provided. 13. Imaging system according to one of the preceding claims, characterized characterized in that a value is chosen for the refractive index m of the waveguide material is, which can be determined according to the conditions a (n1 Weq) / @ n, = O or a (lli WILlla T = 0 where T is the variable temperature of the waveguide. 14. Imaging system according to one of the preceding claims, characterized characterized in that the refractive indices (no, n2, n3, n4) of the waveguide material adjacent media are the same size on two or more sides of the waveguide. 15. Imaging system according to one of the preceding claims, characterized characterized in that the cross section of the waveguide at its interfaces to Mode correction in a measure of the order of magnitude of the wavelength or is deformed into a pillow shape. 16. Imaging system according to one of the preceding claims, characterized characterized in that the waveguide consists of a stack of several layer conductors (51) consists, which by dielectric or metallic interlayers (52) from each other are separated and each have a thickness corresponding to the size Weq, and that the number h is even for every shift supervisor. 17. Imaging system according to one of claims 3 to 15, characterized in that that the waveguide consists of a bundle of several strip conductors (53) with a square, rectangular, triangular or hexagonal cross-section is made up by dielectric or metallic intermediate layers are separated from one another and each one have a thickness or width corresponding to the size Weq, and that the numbers hy and hz are straight for each stripline. 18. Imaging system according to one of the preceding claims, characterized characterized in that the end faces of the waveguide are each inclined at an angle (oc) are inclined to the plane perpendicular to the axis of the waveguide, the maximum is as large as the approximate value (M + I) 7ll Weq where M is the number parallel Node levels is the highest contributing mode to the mapping. 19. Imaging system according to one of the preceding claims, characterized characterized in that the cross section of the waveguide (113) extends in the direction of propagation continuously expanded or rejuvenated. 20. Imaging system according to claim 19, characterized in that the waveguide (123) is wedge-shaped and a liquid or gaseous conductor material contains and that its interfaces (124, 125) to change the magnification or Reduction ratio are movable relative to each other. 21. Imaging system according to one of the preceding claims, characterized in that characterized in that the waveguide (133) has a thickness on the order of the wavelength of light and on a substrate (132) between two layer conductors one integrated optical device is arranged. 22. Imaging system according to claim 1, characterized in that the waveguide (153) between a laser (151) and a consumer (154) of the laser light arranged metallic hollow tube with a polygonal cross-section is. 23. Imaging system according to claim 1, characterized in that the layered waveguide (163) together with a layer guide (162) a integrated optical arrangement on a common substrate and next to one Laser or optical amplifier (161) is arranged so that it the end face (169) of the layer conductor and the end face (168) of the active zone of the laser or amplifier mutually map each other. 24. Imaging system according to claim 23, characterized in that a cover plate (167) of the waveguide (163) consists of a material that the has the same refractive index as a substrate (165) of the optical layer conductor (162), and that the number h for the waveguide (163) is odd. 25. Imaging system according to one of the preceding claims, characterized characterized in that the waveguide (183) for coupling two on a substrate (185) located layer conductors (182, 182 ') of an integrated optical arrangement arranged between or on the ends of the two layer conductors and so dimensioned is that the ends of the waveguides are mutually mapped to one another in this way be that doing an image of the ends between the layer conductors at a distance from the substrate (185) arises. 26. Imaging system according to claim 1, characterized in that the waveguide (193, 203, 213) forms the resonator of a laser. 27. Imaging system according to claim 26, characterized in that at the ends of the waveguide (203) two mirrors (201, 202), one of which or both are dimensioned small according to the desired modes to be coupled, so arranged are that the waveguide alternately maps the mirrors onto one another. 28. Imaging system according to claim 26 or 27, characterized in that that the length (Ls) of the waveguide (213) is half of the length L. 29. Imaging system according to claim 26, characterized in that inside the waveguide located between two flat mirrors (Sp 1, Sp 2) (223) a perforated diaphragm (B) is arranged, which is supported by the shaft pieces on both sides is mapped onto itself. 30. Imaging system according to claim 1, characterized in that the waveguide (233) in an integrated optical amplifier thus on a layer conductor (232) attached to or inserted into an integrated optical assembly is that the light transmitted by this first enters the much thicker waveguide (233) and then again into the layer conductor (232), and that the waveguide (233) is exposed to pump radiation. 31. Imaging system according to claim 1, characterized in that the waveguide (243) between a diaphragm, the radiation of two given Supplying wavelengths, and one that spatially separates the radiations from one another Partition wall (245) is arranged. 32. Imaging system according to one of the preceding claims modified so that the waveguide (143) images sound waves and the refractive index (m) is replaced by the reciprocal speed of sound (1 / v). 33. Imaging system according to one of the preceding claims modified so that the axis of the waveguide is curved. The invention relates to an imaging system for imaging an object with a layered or prismatic waveguide with reflective interfaces. Especially in the miniaturization of optical components in shape the so-called "integrated optics" (see, for example, "Scientific American", April 1974, pp. 28 to 35) there is often the problem of optically imaging the smallest objects. The object can be, for example, the light exit surface of a GaAs laser (typical Dimensions 0.5 x 10 cm), which are placed on the face of a nearby, z. B. 0.1 to 1 mm distant optical layer conductor of the integrated optics shall be. By means of such an image, the light from the laser can enter the fiber are coupled. Conventional imaging systems come with lenses and mirrors hardly an issue for this figure, since it causes difficulties, lenses or mirrors to manufacture with the required small dimensions and tolerances.