FR2808093A1 - METHOD FOR ALIGNING LIGHT BEAMS - Google Patents

Abstract

The method for aligning a light beam (1) with a waveguide, or an optical fiber, or another component used to transport light (2) and comprising a core (3) and a coating (4) , includes the steps of: surrounding the end of the waveguide with a material (8) whose refractive index is close to that of the coating (4) of the waveguide, directing the light beam ( 1) towards said end of the waveguide (2), detecting and evaluating the light energy propagating in the coating (4) of the waveguide, and controlling the positioning and, or else, the orientation of the beam of light (1) relating to the waveguide (2) in order to minimize the energy lost in the coating (4) of the waveguide.

Description

The present invention relates to the field of alignment of a beam of

  light with a waveguide and, or, an optical fiber or any other

similar optical component.

  In the field of optoelectronics it is often necessary to align a beam of light with an optical component, such as a waveguide, an optical fiber, or a guided optics component comprising a waveguide, in which, or which, the light is transported. Normally this light beam comes from another optical component, which can be a free optical component or a guided optical component such as for example another optical fiber or a waveguide, so that the method also relates to the alignment of two waveguides, or of two optical fibers, or of an optical fiber with a waveguide. It should be understood that the present invention applies to all these kinds of processes. In what follows we will call "waveguide" any optical element in which light propagates and we will generally refer to a "light beam" without limitation with respect to

the source of it.

  Often the alignment process takes place "once and for all" during the fabrication of an optoelectronic circuit, the optical components being fixed in place once alignment is achieved. However, there are also applications where the alignment of the beam and the waveguide takes place continuously actively during

  of the use of the concerned optical components.

  A conventional alignment method is the use of a pair of mounting locations in which the waveguide and the source of the light beam are fixed, respectively. The alignment of the light beam and the waveguide involves the alignment of these mounting blocks (which have markings which must be aligned). We can characterize this process of

"<mechanical>.

  The mechanical alignment process is only suitable for certain applications. In addition, according to this method, the alignment accuracy depends on the accuracy of the positioning of the optical components with respect to the mounting blocks and on the alignment accuracy of the mounting blocks. There are, therefore, two

  quite significant sources of alignment errors.

  There is also another conventional alignment method, this time optical.

  According to this optical method, the positioning of the light beam and of the waveguide is first of all carried out in a coarse manner, so that the light beam is approximately parallel to the optical axis of the waveguide and that this beam is incident on the central part (the heart) of the latter. Precise alignment involves applying the output signal from the waveguide to a measuring instrument and detecting the maximum value of the output signal during the relative movement of the light beam and the waveguide.

  The conventional optical method has several drawbacks.

  First, only one value, that of the overall power of the output signal, must be used to colrrigrate positioning and orientation errors which are linked to 6 parameters (positioning of the light beam in x, y, z and orientation of that (here in 0.4, M). Second, this alignment method uses at least part of the energy carried by the waveguide. In a continuous alignment process this therefore results in a loss of energy which must be compensated for, either by using an amplifier, or by increasing the power of the input signal. Thirdly, this method requires the taking of part of the light conveyed by the waveguide, which is not necessarily simple depending on the configuration concerned. Finally, when trying to align a series of waveguides end to end according to the conventional method, the alignment of the last elements of the chain requires the measurement of the light having traveled

the whole system.

  The present invention makes it possible to remedy the drawbacks mentioned above. More particularly, the preferred embodiments of the present invention provide a method allowing the alignment of an optical beam with a waveguide according to which there are several sources of information serving to determine the optimal alignment and, when optimal alignment is achieved, the

energy losses are minimized.

  A waveguide has a core surrounded by a material of lower index, called "cladding" in the case of an optical fiber and "sublayer and covering layer" in the case of a waveguide planar. In what follows, we generally call "coating" the low index material surrounding the

heart.

  The present invention provides a method of aligning a beam of light with the end of a waveguide comprising a core and a coating, characterized in that it comprises the steps consisting in: directing the beam of light towards said end of the waveguide, detecting and evaluating the light energy coming from said light beam and propagating in the coating of the waveguide, and controlling the positioning and, or else, the orientation of the relative light beam to the waveguide, depending on the energy detected in the step of detecting and evaluating the light energy, in order to minimize the light energy propagating in the coating of the waveguide. According to a preferred embodiment of the invention, the method comprises the step of surrounding the end of the waveguide with a material whose refractive index is close to that of the coating of the waveguide. Thus, when the light beam is aligned not with the core of the waveguide but with the coating of the latter, the light energy propagating in the coating is not reflected towards the interior of the guide. wave when it reaches the interface between the waveguide and the environment (which would normally be the case). On the contrary, this energy is transported in the material surrounding the end of the waveguide and having a refractive index similar to that of the coating. It can, therefore, be picked up and it gives an indication that the light beam is not correctly aligned with the core of the waveguide. The overall amount of energy propagating in the coating of the waveguide is a minimum (theoretically zero) when the light beam is

  aligned with the core of the waveguide.

  The method according to the invention makes it possible to detect alignment errors of the order of less than 0.1 ltm. In addition, it causes minimal energy losses when optimal alignment is achieved and. therefore, this process is suitable

  oarticularly to the continuous alignment of optical components.

  In practice, provision may be made to carry out a preliminary step of approximate alignment of the beam with the waveguide according to a mechanical or optical method, followed by fine alignment of the relative position of the beam of light by a method of enslavement. The control method can be classic, for example of the PID (Proportional, Integral, Derivative) type or of the genetic algorithm, fuzzy logic, network type.

of neuron, simulated annealing ..

  In the preferred embodiments of the invention, light sensors (such as photodiodes) are arranged around the waveguide at a certain distance from the end of the latter in order to receive the light energy propagating in the coating of the waveguide and in the material surrounding the end of the latter. The sensitive parts of the sensors can be oriented either perpendicular to the optical axis of the waveguide, or parallel to it. Preferably, at least three light sensors are used, arranged around the waveguide so as to be approximately equidistant from the optical axis thereof and uniformly distributed angularly. However, for reasons which will be explained later, there is a certain margin in the positioning of these sensors. The use of signals from several sensors makes the alignment method according to the invention faster than the

  conventional processes (which are based on a single source of information).

  The method according to the preferred embodiments of the invention can be applied to the alignment of a beam of light with a waveguide mounted in a mounting block. In this case, it is advantageous to fix the

light on the mounting block.

  Similarly, the method according to the preferred embodiments of the invention can be applied to the alignment of a group of optical fibers.

  mounted in a pigtailing block with other optoelectronic components.

  In this case also, it is advantageous to fix the light sensors on the

pigtailing block.

  We can also foresee the case where the light sensors are

  deposited directly on the coating of the waveguide, by an appropriate method.

  Other characteristics and advantages of the present invention will appear

  in light of the following description of a preferred embodiment given to

  by way of example, and illustrated by the accompanying drawings, in which: - Figure 1 is a diagram illustrating the implementation of the method according to the present invention, in which the figure shows the case where the light beam is not not aligned with the core of the waveguide and FIG. 1b shows the case of good alignment - FIG. 2 is a diagram illustrating the implementation of the method according to an embodiment of the invention in which 4 photodiodes are employed, - Figure 3 is a diagram illustrating the implementation of the method according to an embodiment of the invention in which 6 photodiodes are used, - Figure 4 is a diagram showing the case where the photodiodes are not centered by with respect to the optical fiber, in which FIG. 4a shows the case where the light beam is not aligned with the core of the optical fiber and FIG. 4b shows the case of good alignment; - Figure 5 shows a modified arrangement of photodiodes with respect to an optical fiber, in which Figure 5a shows the case where the light beam is not aligned with the core of the optical fiber and Figure 5b shows the case of 'a good alignment, - Figure 6 is a diagram similar to Figure lb explaining the positioning of photodiodes relative to the waveguide, and - Figure 7 is a diagram similar to the figure allowing it to

  understand the calculation of the energy received by a photodiode.

  A preferred embodiment of the present invention will now be described with reference to FIGS. 1 and 2, which show the case of the alignment of a beam of light 1 with an optical fiber 2. More precisely, it is sought to focus the beam of light! on the core 3 of the optical fiber 2 and not on

coating you 4 thereof.

  In this embodiment of the invention, four photodiodes 10 are mounted on a base S surrounding the end of the optical fiber. However, it is also conceivable to use only three photodiodes. The photodiodes are distributed around the fiber 2 in a quasi-uniform manner. Thus, the photodiodes 10 are approximately equidistant from the axis of the fiber 2 and are distributed angularly in a uniform manner. The sensitive parts 6 of the photodiodes are oriented approximately perpendicular to the optical axis of the fiber 2. A drop 8 of an index-matching material, having the same refractive index as the coating 4 of the optical fiber, or a neighboring refractive index, is placed at the end of the optical fiber 2 so that light energy, transported in the coating 4, can leave the optical fiber 2 and reach the photodiodes 10. However, this drop 8 does not cover the exposed end of the core 3 of the fiber 2. The drop 8 can be

  consisting of a drop of glue, for example.

  Light energy passes through the coating 4 of the optical fiber 2

  dl in the case of misalignment of the light beam 1 (see Figure la).

  If the light beam is properly aligned with the core 3 of the optical fiber there is

  no (or almost no) energy lost in coating 4 (see Figure 1b).

  The output signals from the photodiodes 10 provide a measure of the alignment of the light beam and the optical fiber. These signals are minimal when alignment is the best possible. Fine displacements of the light beam (and, or, of the optical fiber) will be carried out by monitoring the output signals of the photodiodes 10. When the optimal alignment is reached, the sum

  photodiodes output signals is a minimum.

  We will give below with reference to FIG. 7, an example of calculation of the light energy received by the sensitive part 6 of a photodiode 10 disposed on a base 5 near an optical fiber 2 on which a beam arrives. light 1, the sensitive part 6 being set back at a distance z from the end of the fiber 2 at which the light beam i has a

throttle of diameter w.

  The parameters appearing in the funnels (1) to (3) are defined as follows: w = diameter of the beam constriction at the fiber end 2, W (z) = diameter of the beam at the level from the sensitive part 6 at a distance z from the end of the fiber 2 in the longitudinal direction, i 5 Po = power of the incident light beam, n = refractive index of the glue drop 8, / = offset of the light beam with respect to the optical axis of fiber 2, = wavelength of the light of the beam, dfiber = diameter of fiber 2, dmount = diameter of the mounting support (base 5),

  dactive = diameter of the active surface of the photodiode.

  The diameter of the beam W (z) at the sensitive part of the photodiode is given by the following formula: W (z) = w I) 1 11 The input power Pin ,, of the light beam in the fiber is given by the following formula: - Pin = 'eó The power received by the photodiode is given by the following formula: chwit -chftx + c7Lbr dhnui -de + iiL cdtve + o + -' P5f i2 2 - E (1 22 2 As can be see it in FIG. 6, the photodiodes are separated from the optical fiber 2 laterally and they are positioned recessed at a certain distance x from the end of the fiber along the axis of the fiber in order to receive part of the divergent beam leaving the coating. This distance x depends on the size t of the fiber, the size of the opening o through which the fiber passes, the size e of the photodiodes as well as the sensitivity of these, the minimum level d energy that must be captured by each photodiode in order to be greater than

  electronic noise, and of the aperture y of the light beam as well as the power of this

  this. A modeling, carried out with an optical fiber of the SMF28 type (coating diameter 125.m, core diameter 8.3 pm) and photodiodes of the EPITAXX ETX75 type, commonly available, gives an optimal value of 4.6 mm for the distance x. This i 5 configuration makes it possible to detect displacements of 0.01 pim of a light beam 1 relative to the core 3 of the optical fiber, by using an input power of ImW (the value

  minimum signal that can be detected will be -60dB).

  It should be noted that the present invention is not limited to the use of three or four photodiodes (or other light sensors, or other direct or indirect measurement devices such as for example a photoluminescent material observed from a distance by a camera. ). It is even advantageous to use more than four sensors in order to reach optimal alignment even faster. FIG. 3 shows an embodiment of the invention in which six photodiodes 10 are arranged around an optical fiber. The group of photodiodes can be manufactured directly on the base 5 or else assembled on the base 5 according to any standard procedure, typically chosen from hybrid integration methods such as the so-called "MCM" method, that is to say the "Multi Chip Modules" method which is an encapsulation technology combining several chips and electronic components in a

  same housing or the so-called "flip-chip" method.

  According to the method of the invention, precise positioning of the light sensors relative to the optical axis of the fiber (or else of the waveguide) makes it possible to obtain an increased measurement sensitivity. However, the method according to the invention can tolerate a difference (typically of up to + 50 gm) between the optimal positions of the light sensors and their actual positions. As can be seen in Figures 4a and 4b, even if the group of photodiodes is not centered with respect to the optical axis, the photodiodes always receive light energy when the light beam is focused on the coating 4, and the sum of the photodiodes output signals is always a minimum when the

  1 0 beam 1 is focused on the core 3.

  In practice, when using an approximate arrangement of the photodiodes 10 with respect to the optical axis of the waveguide / fiber 2, it is preferable to follow a preliminary procedure of coarse alignment, in order to center the light beam I on the center of the group of photodiodes, before proceeding with fine alignment. This fine alignment implies the realization of fine displacements of the light beam I (and, or, of the waveguide / fiber 2) as a function of the output signals of the photodiodes 10, according to so-called "genetic" algorithms, or else according to any other method which makes it possible to avoid reaching a local minimum of the sum of the output signals from the photodiodes 1 0. The dependence of the output signals from the photodiodes 10 as a function of the position of the beam 1 is complex, especially when the photodiodes are not centered on the optical axis of the waveguide / fiber 2. The use of methods, such as so-called algorithms

  "Genetic", neural networks, or simulated annealing, is therefore indicated.

  The method of genetic algorithms consists in initially creating a population of sets of parameters making it possible to calculate a result. For example, this population can include a set of positions for the incident light that we want to introduce into the fiber. We then proceed to an iteration cycle during which all the parameter sets are evaluated and we select a fraction of the best parameter sets. These sets of parameters are then crossed between them to create a new population. For the present application, the crossing can be done for example by taking the position located halfway between two points randomly selected from the best. Each new generation starts the iteration cycle again and performs a test on each cycle to stop iterations when a result

satisfactory is achieved.

  In the neural network method, at least three layers of neurons are defined in hardware or software, namely an input layer, hidden layers and an output layer. Each neuron has several inputs and several outputs. The value provided at the output of each neuron is a weighted sum of the inputs. The weighting coefficients are specific to each neuron. All the neurons are interconnected with one another in a more or less complete mesh, but the outputs of a neuron are always connected to the inputs of the neurons of the next layer. In the present application, the neurons of the input layer have for input the signals supplied by the photodiodes while the neurons of the output layer have their outputs.

  connected to the actuators allowing the input beam to be moved.

  Network learning involves setting all the weights of each neuron so that the output neurons provide a response

  given for each case of input signal presented to the input neurons.

  The simulated annealing method is an optimization method in which an initial parameter is chosen, which in this case may be an initial position of the incident beam taken at random. A system energy function is established which can be, in the present application, the sum of the signals of the photodiodes. We try to minimize this function. To do this, we randomly modify the parameters (position) and reassess the energy function. If it is better than the previous one, we keep it. Otherwise we do a statistical test where the energy difference between the new energy obtained and the best old energy comes into play, as well as a value called temperature. The higher the temperature, the greater the difference in energy is likely to be retained. The process is iterated by slowly varying the temperature downwards until a

optimal position.

  In the methods mentioned above, in order to detect the alignment for which a minimum of light energy propagates in the coating of the fiber, it is possible to measure the sum of the output signals of the photodiodes, or any other appropriate function of the signals, like the difference two by two, the product, the weighted sum, etc. In the case of alignment using genetic algorithms, neural networks, fuzzy logic or the like, the function of the signals that is used to align can be very complex. Finally, each signal or pair of signals can be used to control one of the alignment parameters: for example, a difference in the output signals from two sensors placed on the X axis can control the position along X, same for Y, the sum of the output signals can be used for Z, etc. Although the present invention has been described above with reference to a particular preferred embodiment, it will be understood that numerous modifications and adaptations can be made thereto without, however, departing from the

part of the invention.

  In particular, although the embodiment shown in FIGS. I and 2 uses photodiodes whose sensitive surfaces are oriented perpendicular to the optical axis of the optical fiber, the photodiodes can be oriented in such a way that the sensitive parts of these are oriented

  parallel to the optical axis of the fiber (see Figure 5).

  Furthermore, an embodiment can be provided according to which the

  Light sensors are deposited directly on the coating of the optical fiber.

  In addition, in the case where the invention applies to the alignment of a light beam 1 5 with the waveguide of an optoelectronic component, the distribution of the photodiodes around the waveguide may not be uniform if the latter

  is part of a planar component.

  The alignment method according to the invention can be applied to the alignment of a waveguide mounted in a mounting block. In this case, the light sensors should be mounted on the mounting block in positions that will surround the waveguide after it has been placed in the mounting block. In this way the mounting block serves both to fix the waveguide in position and to provide information allowing the alignment of the latter with a beam of light coming from another optical component. This allows for faster alignment than conventional "mechanical" methods which use only one

  information source (the total power of a signal leaving the waveguide).

  Once the waveguide is aligned, it will be connected to the other optical component and

  the assembly will be removed from the mounting block.

  Similarly, the method according to the present invention can be applied to the alignment of optical fibers positioned in a so-called “pigtailing” block, that is to say of connection of interface fibers (which serves to collect and align several optical fibers with optoelectronic components, or other optical fibers, waveguides, etc.). In such a case too, it

  light sensors should be positioned on the pigtailing block.

Claims (6)

CLAIMS l. Method for aligning a beam of light (1) with the end of a waveguide (2) comprising a core (3) and a coating (4), characterized in that it comprises the stages consisting in: - directing the light beam towards said end of the waveguide (2), - detecting and evaluating the light energy coming from said light beam and propagating in the coating (4) of the waveguide (2) , and - control the positioning and, or else, the orientation of the light beam (1) relative to the waveguide (2), as a function of the energy detected in the step of detecting and evaluating the light energy, in order to minimize the light energy propagating in the coating (4) of the waveguide (2).   2. Method according to claim 1, characterized in that it comprises the step consisting in surrounding the end of the waveguide (2) with a material (8) of which   the refractive index is close to that of the coating (4) of the waveguide.   Method according to claim 2, characterized in that the step consisting in capturing the energy propagating in the coating (4) of the waveguide (2) consists in: - placing a plurality of light sensors (10) around waveguide, cn withdrawal at a certain distance (x) from said end thereof, along the optical axis thereof   - And measure the output signals of said plurality of sensors.   4. Method according to claim 3, characterized in that the sensors (10) are arranged around the waveguide (2) so that the sensitive parts (6) of the sensors are oriented approximately parallel to the optical axis of the waveguide.   Method according to claim 3, characterized in that the sensors (10) are arranged around the waveguide (2) in such a way that the sensitive parts (6) of the sensors are oriented approximately perpendicular to the axis optical waveguide.   6. Method according to claim 3, characterized in that at least three   light sensors (10) are arranged around the waveguide (2).   7. Method according to claim 3, characterized in that the light sensors (10) are mounted directly on the outside of the waveguide (2). 8. Method according to claim 3, characterized in that the light sensors (10) are mounted on a waveguide mounting block in positions which will surround the waveguide (2) once it has been mounted in the block mounting.   9. Method according to claim 1, characterized in that it comprises a prior alignment step serving to align the light beam (1) and the   waveguide (2) approximately.   I10. Method according to claim 1, characterized in that the light sensors (10) are mounted on a pigtailing block intended to bring together a plurality of waveguides (2), the light sensors (10) being positioned in such a way that they collect light energy leaving the coating (4) of said plurality of waveguides (2) when these are mounted in the pigtailing block. I 11. Method according to claim 1, characterized in that the method alignment takes place continuously.   12. Method according to claim 1, characterized in that the beam of   light (1) comes from another waveguide or from another optical device.