Wavelength Stabilized Laser Module
THE BACKGROUND OF THE INVENTION AND PRIOR ART
The present invention relates generally to the production of laser signals which are suitable for wavelength division multiplex optical transmission systems. More particularly the invention relates to a wavelength stabilized laser module according to the preamble of claim 1.
In an optical transmission system information is represented by light pulses which are transmitted over optical fibers. Further- more, a single fiber may be used to transmit light of multiple wavelengths in order to obtain a high capacity and a high flexibility. Typically, semiconductor lasers are used to produce this light. In its most simple form, a wavelength division multiplexed (WDM) optical transmission system includes a number of transmission channels, where each channel transmits at a well- defined wavelength. Figure 1 schematically illustrates a WDM transmission system, where a number of different channels Ch1 , Ch2, ... , Chn are combined into a compound signal by means of an optical multiplexer (MUX) 1 10. The compound signal is then transmitted through an optical fiber 130, and at one or more end locations the individual channels Ch1 , Ch2, ... , Chn are separated out by means of an optical demultiplexer (DEMUX) 120. Normally, the channels Ch1 , Ch2, ... , Chn are placed at standardized wavelengths λ f λ2 λn with a particular wavelength separation Δλ between each channel. In a dense WDM (DWDM) system the channel separation is 100 GHz, which corresponds to a wavelength separation Δλ of approximately O.δnm, see figure 2. The maximal allowed wavelength error is, of course considerably smaller, say 100 pm. This places very strict requirements on the accuracy and stability of the laser's wavelength. The relative accuracy needed is in the order of 0.005 percent or better. Typical semiconductor lasers incor-
porate a grating to obtain a stable single wavelength. Nevertheless, it is difficult to guarantee a particular wavelength over the lifetime of the component. Therefore it is important that mechanisms be provided, which monitor the laser wavelength and force it to its nominal value. These mechanisms normally include at least one feedback control loop that monitors and stabilizes the laser wavelength by locking it to a stable reference.
The prior art includes many different examples as how to ac- complish a so-called wavelength locker for stabilizing a laser. One common element to obtain a wavelength reference is to use a Fabry-Perot etalon. This element has two parallel and partially reflecting surfaces. A common implementation is to use quarts (or a similar material) where a reflecting layer is deposited on one or both of the two surfaces. Light that is transmitted through the etalon has a periodic transmission dependence on the wavelength. The classical way to make use of a Fabry-Perot etalon is to direct a collimated, or nearly collimated, beam towards the etalon and detect the light which is transmitted through the etalon. By adjusting the reflectance of the mirrors it is possible to change the finesse of the etalon. The diagram in figure 5 illustrates the definition of the term finesse for a relatively high value. The horizontal axis shows fringe orders FO: m, m+1 , m+2 etc. for monochromatic light passing through the etalon. The vertical axis shows the relative light transmission T in percent. For higher values, the finesse may be defined as the ratio of the separation of adjacent maxima, (the free spectral range, FSR), to the full width at half the maximum (FWHM), i.e. at T=50%. Thus, narrow and well-separated peaks are indicative of a high finesse, whereas a low finesse value is equivalent to much less well-defined peaks. In fact, more strictly, the finesse parameter is defined as the ratio R/CI - R) , where R denotes the reflectance of the two parallel surfaces. This definition is applicable also to low finesse values.
In addition to the front facet, light may normally also be taken from the back facet of a semiconductor laser. This back-facet light is readily accessible, and may be used to measure and monitor the light produced by the laser. For instance, the back- facet light can be utilized to adjust the laser's drive current, such that the laser generates a constant output power. Moreover, the back-facet light may be used to accomplish a wavelength locking function.
The patent document EP 1 133 034 describes a wavelength sta- bilized laser module where light from the back facet is collimated by a lens. One part of the collimated beam is passed through an etalon-type filter to a first photoelectric converter, and another part of the collimated beam is transmitted directly to a second photoelectric converter. Hence, the first photoelectric converter measures light which has been transmitted through the etalon, whereas the second photoelectric converter samples unfiltered light from the laser so as to enable a power control. This solution presumes a parallel luminous flux, and therefore requires a lens or the like. However, such a lens arrangement is relatively expensive and uses considerable space. Moreover, the lens may cause reflections from the etalon and the photo- detector to be reflected back and focused into the laser. This in turn, risks seriously degrading the laser's performance. According to the design of this document, the transmitted light shows the traditional Airy function dependence with the wavelength. It is therefore clear that a low-finesse etalon is required in order to obtain a simple locking to the function.
The U.S. patent 5,825,792 discloses a wavelength monitoring and control assembly for WDM optical transmission systems. Also here a lens is proposed to control the divergence of the light from the laser incident on a filter element in the form of a Fabry-Perot etalon. However, in this case the beam need not be completely collimated. The lens is therefore optional. The balance between two detectors is used to control the laser. Also here a relatively low-finesse etalon is needed for a proper
function. Moreover, the detectors are placed symmetrical to the output light from the laser and both detectors receive light which has passed through the etalon. The laser wavelength is tuned by adjusting the angle of the etalon. Naturally, this renders the design very sensitive to any uncontrolled tilting of the etalon.
SUMMARY OF THE INVENTION
The object of the present invention is therefore to provide a cost efficient, compact and robust wavelength stabilization for a laser module, which alleviates the above problems, and thus enables a reliable and efficient laser function over a broad range of filter angles.
According to one aspect of the invention the object is achieved by the laser module as initially described, wherein the first sensor is arranged such that the first light signal directly illumi- nates a light-sensing surface of the sensor, i.e. without passing through the filter element.
An important advantage attained by means of this laser module is that a filter element may be used, which both has steep slopes (i.e. a feature associated with a high-finesse etalon) and a wide symmetric capture range (i.e. a feature normally associated with low-finesse etalons). The design is also relatively insensitive to variations in the relative angle between the filter element and the laser.
According to a preferred embodiment of this aspect of the inven- tion, the second control loop includes a thermo-electric cooling unit which is associated with the laser. The thermo-electric cooling unit is adapted to receive the wavelength control signal and in response thereto adjust the temperature of the laser. This adjustment, in turn, results in a wavelength change. According to another preferred embodiment of this aspect of the invention, the laser module includes a division unit, which is
adapted to receive the first and the second electric signals. In response to these signals the division unit produces a composite control signal that is indicative of a ratio between the first electric signal and the second electric signal. The wavelength-monitoring module is adapted to receive the composite control signal, and in response thereto produce the wavelength control signal. By means of this design, the control current ratio is constant and the laser temperature will be tuned to the wavelength set value irrespective of the first electric signal, i.e. the power control current level.
According to yet another preferred embodiment of this aspect of the invention, the laser is adapted to produce the first and second light signals in the form of an uncollimated light flux. More preferably, the first light signal constitutes a first fraction of a divergent light beam from the laser while the second light signal constitutes a second fraction of this light beam. Thereby, a very compact and simple overall design is attained.
According to yet another preferred embodiment of this aspect of the invention, the second sensor and the filter element are arranged relative to the laser, such that a light ray from the laser passes perpendicularly through the filter element to the light- sensing surface of the second sensor. Namely, thereby, a light ray emitted radially from the laser may travel a shortest way through the filter element to the second sensor. This, in turn, results in comparatively good transmission efficiency from the filter element to the second sensor.
According to another preferred embodiment of this aspect of the invention, the filter element may have a non-zero filter tilt angle towards a laser central axis of the laser, and the light-sensing surface of the second sensor has a tilt angle towards the laser central axis, which is essentially equal to the filter tilt angle. This enhances the efficiency of the second sensor to some extent.
According to still another preferred embodiment of this aspect of
the invention, the filter element has a filter tilt angle of seven degrees or less. Although a small filter tilt angle is generally preferable, the invention performs well over the entire range of angles between zero and seven degrees, and consequently allows a relatively large tolerance to any uncontrolled tilting of the filter element.
According to another preferred embodiment of this aspect of the invention, the laser is adapted to produce light, which has a wavelength in the interval 1530 nm to 1610 nm, e.g. in the C- band (1530-1565 nm) or in the L-band (1570-1610 nm). Moreover, the filter element is arranged, such that its light incident surface is located within 1600 μm from the laser. Hence, the invention provides a large degree of flexibility in terms of wavelength and geometry. According to yet another preferred embodiment of this aspect of the invention, the light-sensing surface of the second sensor has an essentially rectangular outline with dimensions, which lie in the interval 50 μm to 400 μm, preferably above 200 μm. Namely, this sensor size strikes good balance between cost effectiveness/ yield and angular sensitivity.
According to still another preferred embodiment of this aspect of the invention, a central axis of the second sensor is displaced at an offset distance with respect to the laser central axis, which is less than or equal to 350 μm. Thus, a relatively large offset of the second sensor can be accepted before the electric signal there from deteriorates to an unacceptably low quality.
According to another preferred embodiment of this aspect of the invention, the filter element has a thickness ranging from 300 μm to 1000 μm. Moreover, the filter element is arranged with the light exit surface within 1000 μm from the second sensor. Thus, the invention also allows relatively large geometry flexibility on the light-exit side.
To sum up, the invention provides compact and cost efficient
laser wavelength stabilization over a broad tuning range. A design according to the proposed solution is also reliable, since there is no lens (or equivalent collimating element) between the laser and the filter element, and therefore the risk of back reflec- tions from the sensor via the filter being focused into the laser is eliminated.
Further advantages, advantageous features and applications of the present invention will be apparent from the following description and the dependent claims.
BRIEF DESCRIPTION OF THE DRAWINGS
The present invention is now to be explained more closely by means of preferred embodiments, which are disclosed as examples, and with reference to the attached drawings.
Figure 1 shows a schematic block diagram over a per se known WDM optical transmission system,
Figure 2 shows a diagram which illustrates the wavelength separation between the optical channels of the system in figure 1 ,
Figures 3a-b show diagrams illustrating the differences in angular sensitivity for a collimated light beam used in the prior art and an uncollimated beam used according to one embodiment of the invention,
Figure 4 illustrates, by means of another diagram, the angular dependence of the locking wavelength for an uncollimated light beam and a collimated light beam respectively,
Figure 5 defines the finesse parameter, which is used as a figure of the merit for the transmission bandwidth of a Fabry-Perot interferometer,
Figure 6 show diagrams over per se known wavelength dependencies, with reference to which the finesse
parameter is further elaborated upon,
Figure 7 shows a diagram illustrating the capture range according to one embodiment of the invention,
Figures 8a-b show diagrams which elucidate the differences in the tuning ranges attainable by means of a collimated light beam used according to the prior art respective an uncollimated light beam used according to one embodiment of the invention,
Figure 9 shows a block diagram over a wavelength stabi- lized laser module according to a first embodiment of the invention,
Figure 10 shows a diagram illustrating how an appropriate set value is selected for the wavelength control signal according to one embodiment of the inven- tion,
Figure 1 1 schematically illustrates the geometry of the wavelength stabilized laser module according to the invention,
Figure 12 shows a block diagram over a wavelength stabi- lized laser module according, to a second embodiment of the invention,
Figure 13 shows a diagram which illustrates the sensitivity of the displacement of the second sensor according to one embodiment of the. invention, Figures 14-16 show diagrams which illustrate the sensitivity of the distance between the laser and the filter element according to one embodiment of the invention,
Figures 17-19 show diagrams which illustrate the sensitivity of the dimensions of the second sensor according to one embodiment of the invention,
Figure 20 shows a diagram representing measurements of the light transmission prior to assembly respective
after assembly of a filter element in an arrangement according to one embodiment of the invention, and
Figure 21 shows a diagram which illustrates the relationship between an actual laser wavelength and the laser's case temperature according to one embodiment of the invention.
DESCRIPTION OF PREFERRED EMBODIMENTS OF THE INVENTION Figures 3a shows a diagram over the relationship between the light transmission for different tilt angles α of a particular narrow bandpass wavelength selective transmission filter element that has a Fabry-Perot structure, where a collimated light beam is used according to a known design. The horizontal axis of the diagram indicates the wavelength λ and the intensity lAU of a sensor signal registered after the filter is represented along the vertical axis. As can be seen, the wavelength at which the transmission peaks occur here varies substantially depending on the filter tilt angle αF. For example, a filter tilt angle αF equal to 0° results in a first transmission peak around the wavelength λ=1 ,5499 μm, whereas for a filter angle αF of 3,0°, this peak occurs around λ=1 ,5492 μm. Thus, the light transmission has a relatively strong angular dependency.
Figures 3b shows another diagram over the relationship bet- ween the light transmission for different tilt angles αF of the same filter element as in the figure 3a. Here, the filter element is arranged according to the invention, such that the filter constitutes an exclusive intermediary element between a laser and a light sensor. Hence, the light signal passes directly from the laser, through the filter element and further directly to the sensor. Preferably, the light from the laser represents an uncollimated flux. The filter element is tilted in angular steps of 0,5°, from 0° to 3,0°; however, there is no noticeable difference in the
transmission profile. Instead, an accentuated transmission peak remained stable at approximately λ=1 ,5498 μm for all filter angles αF. Consequently, the invention is insensitive for filter tilting, at least for relatively small tilt angles.
Figure 4 illustrates, by means of another diagram, the angular dependence of the locking wavelength for an uncollimated light beam UCB and a collimated light beam CB respectively. Here, the horizontal axis indicates the filter tilt angle αF, and the vertical axis represents a center crossing wavelength shift δλ. For the collimated light beam CB a wavelength shift δλ of 50 pm occurs already at a filter tilt angle αF of 0,6°. According to one embodiment of the invention, where an uncollimated light beam UCB is used, a tilt angle of αF=3,5° merely results in a wavelength shift δλ of 3 pm. In fact, when UCB is used, a filter tilt angle as large as αF=7,0° can be allowed before a wavelength shift δλ of 50 pm is reached. The corresponding value for the collimated light beam CB is 1350 pm (i.e. at αF=7,0°).
Returning now to the figure 5, we find that the diagram shows an example of a Fabry-Perot type of filter element with a relatively high finesse, i.e. the ratio FSR/FWHM is rather large.
Figure 6 shows diagram where the horizontal axis indicates a wavelength λ and the vertical axis represents a light intensity lAu detected after an etalon filter element of Fabry-Perot structure. A top-most diagram shows a light transmission graph for a conventional collimated light beam directed towards a filter with a finesse of 0,7, i.e. a relatively low value; and a bottom-most diagram shows a light transmission graph for a filter with a finesse of 30, i.e. a relatively high value. In both cases, a set point, set0, and set30 respectively, around half the amplitude is used. The high finesse filter has a relatively steep slope around the set point set30. This is preferable because thereby the wavelength λ can be adjusted very precisely without an extremely precise power measurement being required. However, a capture range CR30, i.e. the interval within which the wave-
length may be tuned to the set point set30, becomes very asymmetric. This means that, in practice, only rather a small initial wavelength deviation can be handled , namely a deviation equal to the shortest fraction of the asymmetric capture range CR30 from the set point set30. As is apparent from the top-most diagram, the low finesse etalon has a much more symmetric capture range CR0] , and thus a wider effective tuning range. However here, the slope is less steep and therefore the wavelength precision less accurate.
In some applications it is more important to have a broad tuning range of the wavelength locker than to obtain a relatively high wavelength precision. One example is when the etalon thickness, and hence resonance wavelength, is not well controlled. In this case it is desirable with a characteristic were the absolute value of the derivative is reasonable high, while the regions close to a maximum can be avoided. Provided that a collimated light beam is used, this means that a low finesse filter element is required. However, for this type of filter there are large regions in the characteristics were the locking slope is very low. Moreover, the sign of the slope varies between the regions within the capture range.
Figure 7 shows a diagram where again the horizontal axis indicates a wavelength λ and the vertical axis represents a light intensity lAu detected after an etalon filter element of Fabry- Perot structure. Here, however, the design according to the invention is used, where an uncollimated light beam is directed towards the filter element. As can be seen from the graph, if a set point setucB is selected approximately around half the amplitude, the attainable capture range CRUCB is relatively symmetric around this set point setucβ- At the same time, the slope around the set point setucβ is very steep, and consequently the wavelength λ may be adjusted very precisely without an exceedingly precise power measurement.
Figures 8a-b further elucidate the differences in the characteristics of the prior art strategy and the proposed design. The figure 8a shows a light transmission diagram where the horizontal axis indicates a wavelength λ and the vertical axis represents a light intensity lAu of a collimated light beam, which is detected after an etalon filter element of Fabry-Perot structure with a relatively low finesse. Regions to be avoided around the maxima and minima are designated by means of broken-line boxes. Hence, a tuning range TRCB is available between two of these neighboring boxes.
The figure 8b shows a light transmission diagram according to one embodiment of the invention where an uncollimated light beam is transmitted through an etalon filter element of Fabry- Perot structure. The horizontal axis indicates a wavelength λ and the vertical axis represents a light intensity lAu- Again, regions to be avoided around the maxima and minima are designated by means of broken-line boxes. Here however, the available continuous tuning range TRUCB between two neighboring boxes is nearly twice as broad as in the example of the figure 8a, i.e. TRUCB«2TRCB.
The general design of a laser module according to the invention will now be discussed with reference to figure 9. Here, a block diagram shows a first embodiment of a wavelength stabilized laser module. The laser module includes a semiconductor laser 900 and a thermo-electric cooling unit 905 associated therewith to control the wavelength of the light produced. However, according to the invention, alternative mechanisms are conceivable to adjust the wavelength, such as by means of an electrically tunable laser. In any case, the design also includes a first control loop, a second control loop and a narrow bandpass wavelength selective transmission filter element 930 of Fabry- Perot structure.
It is presumed that the spectrum characteristics of the light L produced by the laser 900 depends on both a power control
signal lbias and a temperature of the laser 900. The temperature is controlled by means of a wavelength control signal ltec to the thermo-electric cooling unit 905. The light L primarily exits the laser 900 via a front facet and a lens arrangement 940 and continues, for example into an optical fiber. However, the light L is also tapped off via a back facet in the form of a first and a second light signal Li and L2 respectively, and is directed towards a first and a second sensor 910 and 920. Preferably, the laser 900 is adapted to produce the first and second light signals L-i , and L2 as an uncollimated light flux. For example, the first light signal L-i constitutes a first fraction of a divergent light beam from the laser's 900 back facet, and the second light signal L2 constitutes a second fraction of the divergent light beam from laser's 900 back facet.
The first control loop, in turn, includes the first sensor 910, which is adapted to produce a first electric signal lP in response to the first light signal L-i from the laser 900. A power monitoring module 915 in the loop receives the first electric signal lP and a power set value lP.set representing a desired power control signal lbias to be delivered to the laser 900. The power monitoring module 915 produces the power control signal lbias in response to the first electric signal lP and the power set value lP-set. For instance, the power monitoring module 915 may implement a comparator function, such that if the first electric signal lP is smaller than the set value lP-set the power control signal l ias is increased, and vice versa.
Correspondingly, the second control loop includes the second sensor 920, which is adapted to produce a second electric signal l in response to the second light signal L2 from the laser. Nevertheless, the second light signal L2 has first passed through the filter element 930, which is arranged to constitute an exclusive intermediary element between the laser 900 and the second sensor 920. A wavelength-monitoring module 925 in the second control loop receives the second electric signal l and a wave- length set value lw-set- Based on the second electric signal lw,
value relative to the wavelength set value lw-set, the wavelength monitoring module 925 produces the wavelength control signal Itec- Analogous to the above-described power-monitoring module 915, the wavelength-monitoring module 925 may represent a comparator function.
Of course, an appropriate wavelength set value lw-set for the laser 900, which is adapted to transmit on a particular channel Chn (see the figure 1 ) is a current value lw that corresponds to a desired wavelength λn for this channel. Figure 10 shows a diagram that illustrates the relationship between the second electric signal l and the wavelength λ of the laser light L. Preferably, the wavelength set value lw-set also represents approximately half the peak value lWp of the second electric signal lw.
Thus, returning now to figure 9, according to the invention the second light signal L2 passes through the filter element 930 to a light-sensing surface of the second sensor 920, and the first light signal L-i directly illuminates a light-sensing surface of the first sensor 910. Thereby, a robust wavelength stabilization of the laser 900 is accomplished, such that the light L has a well- defined wavelength over the entire life of the laser 900.
Figure 1 1 schematically illustrates the geometric relationships between the laser 900, the filter element 930 and the second sensor 920 of the wavelength stabilized laser module according to the invention. The filter element 930 is presumed to have a filter tilt angle αF towards a laser central axis AL of the laser 900 (i.e. the main direction along which light is emitted). In order to obtain an optimal performance, the filter tilt angle αF should be as small as possible. Hence, the ideal filter tilt angle αF is zero. Nevertheless, a filter tilt angle αF which is smaller than or equal to 7 degrees is acceptable according to the invention.
According to a preferred embodiment of the invention, the second sensor 920 and the filter element 930 are arranged relative to the laser 900, such that a light ray lr from the laser
900 may pass perpendicularly NF to the light incident and exit surfaces of the filter element 930 through the filter element 930 to the light-sensing surface S2 of the second sensor 920. Namely, this orientation of the filter element 930 and the second sensor 920 renders it possible for a radial light ray lr from the laser 900 (which may be regarded as a point source) to pass a shortest distance through the filter element 930 and strike the light-sensing surface S2. This, potentially, produces a strong electric signal (which is the second electric signal lw in the figure 9).
The second sensor 920 has a light-sensing surface S2, which may have a tilt angle α2 towards the laser central axis AL. According to another preferred embodiment of the invention, the tilt angle α2 is essentially equal to the filter tilt angle F, such that the light-sensing surface S2 is parallel to the light incident and exit surfaces of the filter element 930. Thereby, a light ray passing perpendicularly through the filter element 930 will also strike the light-sensing surface S2 at a right angle, which vouches for an even stronger resulting electric signal.
Moreover, it is preferable if a central axis A2 of the light-sensing surface S2 is also essentially aligned with a line NF from a light emitting point of the laser 900, which is perpendicular to the light incident and exit surfaces of the filter element 930. Thereby, the light ray lr will strike the center of the light-sensing surface S2 l and thus further improve the quality of the second electric signal lw-
Measurements and simulations (see figures 14, 15 and 16) have confirmed that the filter element 930 should be arranged, such that a first distance D-i between the laser 900 and the light inci- dent surface of the filter element 930 is less than, or equal to, 1600 μm. Furthermore, other measurements and simulations (see figure 13) have shown that the central axis A2 of the second sensor 920 should not be displaced, i.e. offset, by more than a distance x2 equal to 350 μm with respect to the laser cen-
tral axis AL. Of course, the longest acceptable offset distance x2 depends on the dimensions of the second sensor's 920 light- sensing surface S2, such that a larger surface S2, tolerates a longer offset distance x2, and vice versa. Generally, a large monitor size (i.e. a light-sensing surface S2, with a large area) is preferable because this enables a strong resulting electric signal. A large monitor is also less sensitive to angular mismatches than a smaller ditto. However, large monitors are expensive, i.a. due to a lower yield. Additionally, an excessive light-sensing surface decreases the relative amplitude of the registered characteristic. Therefore, the monitor size is a parameter which must be optimized. According to measurements and simulations (see figures 17, 18 and 19), it is advantageous if the second sensor's 920 light sensing surface S2 has an outline with dimensions in the interval 50 μm to 400 μm. Moreover, the surface should preferably be essentially rectangular and have a width w2, which is approximately two times the height, say 400 μm wide and 200 μm high. (The width w2 is here measured along the direction of any offset distance x2 with respect to the laser central axis AL). A sensor of these dimensions is namely relatively un problematic to fabricate. Furthermore, a sensor of this size has characteristics that are favorable with respect to angular dependence and relative amplitude.
A distance D2 between the light exit surface of the filter element 930 and the second sensor 920 is less critical. However, the invention has proven to be efficient up to a distance D2 of 1000 μm. Likewise, a thickness d of the filter element 930 may be selected relatively freely, say in the interval 300 μm to 1000 μm.
The dimensions and measures discussed above with reference to the figure 11 presume that the semiconductor laser 900 produces light, which has a wavelength in the interval 1 ,5480 μm to 1 ,5525 μm. Naturally, if light is used, which has different spectral properties, this affects the dimensions and measures, so that a shorter wavelength generally requires a reduction of the
geometry, and vice versa. In any case, the invention is at least applicable over the entire C- and L-bands, i.e. 1530 nm to 1565 nm and 1570 to 1610 nm respectively.
Figure 12 shows a block diagram over a wavelength stabilized laser module according to a second embodiment of the invention. The units and components whose reference signs occur also in the figure 9 designate the same units as discussed above with reference to this figure. Thus, the only difference between this embodiment of the invention and the first embodi- ment is that the second control loop includes a division unit 1240 and an alternative wavelength-monitoring module 1225.
The division unit 1240 receives the first electric signal lP and the second electric signal lw. In response to these signals, the unit 1240 produces a composite control signal RWP, which is indicative of a ratio between the first electric signal lP and the second electric signal lw, i.e. RWP,=c-lP/l , where c is a scale • factor, say c=1 . The wavelength-monitoring module 1225 receives the composite control signal RWP and a composite wavelength set value lw-set- In response to these signals, the wave- length-monitoring module 1225 produces the wavelength control signal ltec- An advantage attained by means of this design is that the laser's 900 temperature will be controlled towards a value given by the composite wavelength set value lw-set irrespective of the first electric signal's lP level. Returning briefly now to the figure 13, a diagram here illustrates the sensitivity of the displacement of the second sensor 920 according to one embodiment of the invention. The horizontal axis indicates the wavelength λ of the light from the laser 900, and the vertical axis shows an electric signal SSAU produced by the second sensor 920. The distance D-i between the laser 900 and the filter element 930 is 700 μm. As can be seen, offset distances x2 of 250 μm or less result in clearly distinguishable characteristics. However, the design is operable at least up to offset distances x2=350 μm for larger distances D-i .
Returning also to the figures 14-16, we see that the invention is relatively insensitive to filter tilt angles αF up to seven degrees for distances D-i between the laser 900 and the filter element 930 up to 1600 μm. Each of the diagrams in the figures 14-16 indicate the wavelength λ along the horizontal axis and show an amplitude SSALι of the second electric signal l produced by the second sensor 920 along the vertical axis. In the figure 14 the distance D-, between the laser 900 and the filter element 930 is 100 μm, and in the figures 15 and 16 this distance is 700 μm and 1600 μm respectively.
Similarly, by studying the figures 17-19 we find that the second electric signal lw produced by the second sensor 920 has a clearly distinguishable characteristics for widths w2 of the second sensor's 920 light-sensing surface S2 in the interval 50 μm to 400 μm. Each of the diagrams also here indicate the wavelength λ along the horizontal axis and show amplitude SSAy of the electric signal lw along the vertical axis. In the figure 17 the width w2 is 50 μm, and in the figures 18 and 19, this measure is 200 μm and 400 μm respectively.
Figure 20 shows a diagram, where measurements of the filter element's 930 light transmission have been registered prior to assembly as well as after assembly in an arrangement according to one embodiment of the invention. The horizontal axis indicates a wavelength λ, a left-most vertical axis shows an ampli- tude SSAU of the second electric signal lw produced by the second sensor 920 and a right-most vertical axis shows a filter reflectivity FR in dB (which represents an inverted measure of the transmission through the filter element 930).
A dashed graph represents a relationship between the filter reflectivity FR and the wavelength λ for the filter element 930 prior to the assembly. A solid graph represents a relationship between the second electric signal l and the wavelength λ, i.e. when the filter element has been arranged according to one embodiment of the invention. Clearly, the solid graph shows a
transmission peak around the wavelength λ=1546,6 nm. It is also apparent from the diagram that the unassembled filter element 930 has a transmission peak (i.e. a reflectance dip) approximately at this wavelength λ. Consequently, the arrange- ment is well tuned.
Finally, figure 21 shows a diagram, which illustrates the relationship between an actual laser wavelength λ and the laser's case temperature T according to one embodiment of the invention. Here, the horizontal axis indicates the temperature T, and the vertical axis shows the wavelength λ. The wavelength λ is plotted for each individual temperature value T as a thin solid graph, and an average temperature T is drawn as a bold solid graph. The latter illustrates the general relationship between the laser temperature T and the wavelength λ of the light produced by the laser. Apparently, the wavelength λincreases with an increasing temperature T.
The term "comprises/comprising" when used in this specification is taken to specify the presence of stated features, integers, steps or components. However, the term does not preclude the presence or addition of one or more additional features, integers, steps or components or groups thereof.
The invention is not restricted to the described embodiments in the figures, but may be varied freely within the scope of the claims.