Light source unit

Abstract

To provide a light source unit that can reduce wavelength dependence while achieving miniaturization. [Solution] The light source unit 1 comprises a semiconductor optical amplifier 10, a filter unit 30, a Michelson interferometer 40, and a detector 44. The Michelson interferometer 40 includes a coupler 41 that splits the light from the filter unit 30 input to the first port 41a into a waveguide 51 connected to the second port 41b and a waveguide 52 connected to the third port 41c and having a different optical path length from waveguide 51, at a predetermined branching ratio; a first mirror 42 that returns at least a portion of the first branched light Lb1 output from the second port 41b to the coupler 41; and a second mirror 43 that returns at least a portion of the second branched light Lb2 output from the third port 41c to the coupler 41. The detector 44 detects the interference light IL of the light returned from the first mirror 42 and the second mirror 43 to the coupler 41. The branching ratio of the coupler 41 is set to an unequal ratio.

Description

[Technical Field]

[0001] This disclosure relates to a light source unit. [Background technology]

[0002] Patent Document 1 discloses a system in which a wavelength-tunable light source using an optical integrated circuit (PIC) is provided with a wavelength monitor using an asymmetric Mach-Zehnder interferometer, separate from the components including a mirror section for returning light to a semiconductor optical amplifier (SOA). [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] U.S. Patent Application Publication No. 2023 / 0036048 [Overview of the Initiative] [Problems that the invention aims to solve]

[0004] In the configuration disclosed in Patent Document 1, a wavelength monitor is provided as a separate component from the mirror section, resulting in a large number of components in the system. As a result, variations in the manufacturing of each component may increase the system's wavelength dependence (the property that the system's characteristics change in response to changes in the wavelength of the output light). Furthermore, a large number of components leads to a larger system size.

[0005] Therefore, the purpose of this disclosure is to provide a light source unit that can reduce wavelength dependence while achieving miniaturization. [Means for solving the problem]

[0006] This disclosure includes the following light source units [1] to

[18] .

[0007] [1] A semiconductor optical amplifier and A wavelength adjustment unit is provided on the substrate, optically coupled to the semiconductor optical amplifier, and changes the wavelength of the output light. A Michelson interferometer is optically coupled to the end of the wavelength adjustment section opposite to the end coupled to the semiconductor optical amplifier, The system comprises a detector connected to the Michelson interferometer, The aforementioned Michelson interferometer is A coupler having a first port, a second port, a third port, and a fourth port, which splits the light from the wavelength adjustment unit input to the first port into a first waveguide connected to the second port and a second waveguide connected to the third port and having a different optical path length from the first waveguide, at a predetermined branching ratio. A first mirror connected to the first waveguide and configured to return at least a portion of the first branched light output from the second port to the first waveguide back to the coupler, The device includes a second mirror connected to the second waveguide and configured to return at least a portion of the second branched light output from the third port to the second waveguide back to the coupler, The detector is connected to the fourth port and detects the first interference light of the first return light returned from the first mirror to the second port and the second return light returned from the second mirror to the third port. The coupler is configured such that, in response to the input of the first and second return beams, a second interference beam is output from the first port to the wavelength adjustment unit. The aforementioned branching ratio is set to an unequal ratio in the light source unit.

[0008] According to the light source unit described in [1] above, a Michelson interferometer including a mirror section (first mirror and second mirror) can be used to return light to a semiconductor optical amplifier while also functioning as a wavelength monitor. More specifically, the wavelength of the output light (wavelength adjusted by the wavelength adjustment section) can be determined based on the power of the first interference light detected by the detector. This eliminates the need for a separate wavelength monitor (e.g., an asymmetric Mach-Zehnder interferometer) in addition to the configuration including the mirror section (Michelson interferometer), thus reducing the number of components. As a result, the entire light source unit can be miniaturized, and the problem of high wavelength dependence due to manufacturing variations in each component can be avoided. Furthermore, by setting the branching ratio of the coupler to be unequal, it is possible to monitor the wavelength of the output light by the detector connected to the fourth port while reducing the wavelength dependence (ripple) of the second interference light compared to when the branching ratio of the coupler is equal. In addition, by setting the branching ratio to be unequal, it is possible to use a coupler with less manufacturing variation and less wavelength dependence than a coupler set to an equal ratio. Therefore, the above-mentioned light source unit makes it possible to reduce wavelength dependence while miniaturizing the unit.

[0009] [2] The light source unit of [1], wherein the branching ratio is set such that the ratio of the power of the first branched light to the power of the input light to the first port is 0.75 or more.

[0010] According to the configuration described in [2] above, by setting the branching ratio to a value that is a certain distance away from the equal division ratio, it becomes possible to use a coupler with less wavelength dependence while effectively reducing the ripple of the second interference light returned from the first port to the semiconductor optical amplifier. As a result, the wavelength dependence of the light source unit can be suitably reduced.

[0011] [3] The light source unit of [1] or [2], wherein the branching ratio is set such that the ratio of the power of the first branched light to the power of the input light to the first port is 0.85 or more.

[0012] According to the configuration of [3] above, in all wavelength ranges (that is, no matter what value the wavelength of the output light is set to), the power of the second interference light can be made larger than the power of the first interference light. That is, since the power of the light returned to the semiconductor optical amplifier (the second interference light) can be made relatively large, the power of the output light (laser output) can be increased.

[0013] [4] The branching ratio is set such that the ratio of the power of the first branched light to the power of the input light to the first port is 0.90 or more, and the light source unit according to any one of [1] to [3].

[0014] According to the configuration of [4] above, by setting the branching ratio to a value further away from the equal division ratio, the effects of the configurations of [2] and [3] above can be further improved.

[0015] [5] The branching ratio is set such that the ratio of the power of the first branched light to the power of the input light to the first port is 0.99 or more, and the light source unit according to any one of [1] to [4].

[0016] According to the configuration of [4] above, by setting the branching ratio to a value further away from the equal division ratio, the effect of the configuration of [4] above can be further improved.

[0017] [6] When the side where the semiconductor optical amplifier is located with respect to the wavelength adjustment unit is the upstream side, a first measuring instrument is further provided, which is arranged on the downstream side of the wavelength adjustment unit and measures the power of the light that has passed through the wavelength adjustment unit, and the light source unit according to any one of [1] to [5].

[0018] According to the configuration of [6] above, based on the detection result of the detector (the power of the first interference light) and the measurement result of the first measuring instrument (the power of the light that has passed through the wavelength adjustment unit), the wavelength of the output light can be estimated more accurately.

[0019] [7] A light source unit of any of [1] to [6], further comprising a second measuring instrument disposed between the first port of the coupler and the wavelength adjustment unit for measuring the power of the second interference light.

[0020] According to the configuration described in [7] above, the mirror characteristics (e.g., ripple of the second interference light) can be determined based on the measurement results (power of the second interference light) from the second measuring instrument. The mirror characteristics thus determined can be used for various purposes such as correction and feedback to the detection results (wavelength estimation results) of the detector.

[0021] [8] A second coupler having a fifth port, which is positioned between the first port of the coupler and the wavelength adjustment unit and into which light from the wavelength adjustment unit is input, a sixth port, a seventh port, and an eighth port connected to the first port, A first measuring instrument connected to the seventh port measures the power of light that has passed through the wavelength adjustment section, The system further comprises a second measuring instrument connected to the eighth port for measuring the power of the second interfering light, The second coupler is configured to split the light input from the fifth port to the sixth and seventh ports, and to split the light input from the sixth port to the fifth and eighth ports, and is one of the light source units [1] to [7].

[0022] According to the configuration described in [8] above, the effects of the configurations described in [6] and [7] above can be obtained by providing a first measuring instrument and a second measuring instrument. Furthermore, by connecting the first measuring instrument and the second measuring instrument to a second coupler positioned between the wavelength adjustment unit and the coupler, a configuration (i.e., a configuration comprising a first measuring instrument and a second measuring instrument) that can measure both the power of the light that has passed through the wavelength adjustment unit and the power of the second interference light can be easily and compactly realized.

[0023] [9] The branching ratio is set such that the ripple of the second interference light is 10 dB or less, and is one of the light source units from [1] to [8].

[0024] According to the configuration described in [9] above, the wavelength dependence of the light source unit can be suitably reduced by sufficiently reducing the ripple of the second interference light.

[0025]

[10] The branching ratio is set such that the ripple of the second interference light is 6 dB or less, and is one of the light source units from [1] to [9].

[0026] The configuration described in

[10] above can further improve the effectiveness of the configuration described in [9] above.

[0027]

[11] The branching ratio is set such that the ripple of the second interference light is 2 dB or less, and is one of the light source units from [1] to

[10] .

[0028] The configuration described in

[11] above can further improve the effectiveness of the configuration described in

[10] above.

[0029]

[12] The branching ratio is set such that the ripple of the second interference light is 0.18 dB or less, and is one of the light source units from [1] to

[11] .

[0030] The configuration described in

[12] above can further improve the effectiveness of the configuration described in

[11] above.

[0031]

[13] The coupler is a light source unit of any of [1] to

[12] , which is composed of a directional coupler.

[0032] According to the configuration described in

[13] above, couplers with unequal branching ratios can be manufactured easily and accurately, thereby effectively suppressing the large wavelength dependence caused by manufacturing variations in the couplers.

[0033]

[14] The semiconductor optical amplifier is configured to output the output light from the end opposite to the end coupled to the wavelength adjustment unit, any of the light source units from [1] to

[13] .

[0034] According to the configuration described in

[14] above, there is no need to place circuit components on the substrate to guide the output light, thus simplifying the circuit configuration on the substrate. In addition, the output light can be extracted externally without affecting the characteristics of the Michelson interferometer.

[0035]

[15] The branching ratio of the coupler is set such that the power of the first branched light is greater than the power of the second branched light. A light source unit, one of [1] to

[14] , from which a portion of the first branched light is extracted as the output light.

[0036] According to the configuration described in

[15] above, by extracting the output light from the first branched light, which has higher power, among the light branches by the coupler, the power of the extractable output light can be increased compared to the case where a portion of the second branched light is extracted as output light.

[0037]

[16] The branching ratio of the coupler is set such that the power of the first branched light is greater than the power of the second branched light. A light source unit, one of [1] to

[15] , from which a portion of the second branched light is extracted as the output light.

[0038] According to the configuration described in

[16] above, by extracting the output light from the second branched light, which has less power than the light branched by the coupler, the power difference between the first and second return light returning to the coupler can be increased compared to the case where a portion of the first branched light is extracted as the output light, and consequently, the wavelength dependence (ripple) of the second interference light can be reduced.

[0039]

[17] A light source unit, one of [1] to

[16] , wherein, between the semiconductor optical amplifier and the wavelength adjustment unit, a portion of the light traveling from the semiconductor optical amplifier to the wavelength adjustment unit or the light returning from the wavelength adjustment unit to the semiconductor optical amplifier is extracted as the output light.

[0040] The configuration described in

[17] above makes it possible to output light to other circuits on the substrate without affecting the characteristics of the Michelson interferometer.

[0041]

[18] The wavelength adjustment section is composed of a ring resonator having a ring waveguide and a linear waveguide optically coupled to the ring waveguide, A light source unit from which the output light is extracted from the linear waveguide, one of the [1] to

[17] .

[0042] According to the configuration described in

[18] above, the output light can be output to other circuits on the substrate by extending the linear waveguide of the ring resonator. Furthermore, by utilizing the linear waveguide originally provided in the wavelength adjustment section (ring resonator), the output light can be extracted without increasing the number of components from the original configuration. [Effects of the Invention]

[0043] According to one aspect of this disclosure, it is possible to provide a light source unit that can reduce wavelength dependence while achieving miniaturization. [Brief explanation of the drawing]

[0044] [Figure 1] Figure 1 shows an example of the configuration of a light source unit 1 according to one embodiment. [Figure 2] Figure 2 shows an example of the characteristics of the filter section 30. [Figure 3] Figure 3 shows an example of the characteristics of the Michelson interferometer 40. [Figure 4] Figure 4 shows an example of the characteristics of detector 44 (wavelength monitor). [Figure 5] Figure 5 shows an example of the wavelength dependence (mirror characteristics) of the reflected light RL. [Figure 6] Figure 6 shows an example of the relationship between the coupling coefficient of coupler 41, the ripple of reflected light RL, and the maximum power of interfering light IL. [Figure 7]Figure 7 illustrates an example of determining the minimum coupling coefficient of coupler 41. [Figure 8] Figure 8 illustrates an example of determining the minimum coupling coefficient of coupler 41. [Figure 9] Figure 9 shows an example of the characteristics of the filter section 30. [Figure 10] Figure 10 is a diagram illustrating the effect of the ripple of the reflected light RL on the SMSR of the filter section 30. [Figure 11] Figure 11 shows the simulation results for the ripple of the reflected light RL, the coupling coefficient of coupler 41, and the maximum reduction amount of SMSR. [Figure 12] Figure 12 shows an example configuration of the first modified light source unit 1A. [Figure 13] Figure 13 shows an example of a measurement using detector 46 (power monitor). [Figure 14] Figure 14 shows examples of measurements using detector 44 (wavelength monitor) and detector 46 (power monitor). [Figure 15] Figure 15 shows the ratio (PIL / PL3) of the measured value PIL from detector 44 to the measured value PL3 from detector 46 in Figure 14. [Figure 16] Figure 16 shows an example configuration of the second modified light source unit 1B. [Figure 17] Figure 17 shows an example configuration of the third modified light source unit 1C. [Figure 18] Figure 18 shows an example configuration of the fourth modified light source unit 1D. [Figure 19] Figure 19 shows an example configuration of the fifth modified light source unit 1E. [Figure 20] Figure 20 shows an example of the configuration of the sixth modified light source unit 1F. [Figure 21] Figure 21 shows an example configuration of the seventh modified light source unit 1G. [Figure 22] Figure 22 shows an example configuration of the eighth modified light source unit 1H. [Modes for carrying out the invention]

[0045] Hereinafter, one embodiment of the present disclosure will be described in detail with reference to the drawings. In the following description, the same or equivalent elements will be denoted by the same reference numerals, and redundant descriptions will be omitted. [Embodiment]

[0046] Referring to Figure 1, the light source unit 1 of this embodiment will be described. As shown in Figure 1, the light source unit 1 comprises a semiconductor optical amplifier 10 and an optical integrated circuit 20. The optical integrated circuit 20 comprises a substrate 21 and various optical elements including optical waveguides provided on the substrate 21. The substrate 21 is, for example, a rectangular silicon substrate. In this embodiment, a filter unit 30, a Michelson interferometer 40, and a detector 44 are mounted on one main surface (the front surface in Figure 1) of the substrate 21 constituting the optical integrated circuit 20. In the following description, the side where the semiconductor optical amplifier 10 is located relative to the filter unit 30 will be referred to as "upstream," and the side where the filter unit 30 is located relative to the semiconductor optical amplifier 10 will be referred to as "downstream."

[0047] The semiconductor optical amplifier 10 is composed of, for example, a quantum cascade laser (QCL). The semiconductor optical amplifier 10 outputs light L1 from one end face 10a to the optical integrated circuit 20. Subsequently, a return light of a predetermined wavelength (wavelength determined by the filter section 30) (light returned from the filter section 30 to the end face 10a due to reflected light RL, described later) is input from the optical integrated circuit 20 to the end face 10a, and the light of that wavelength is amplified inside the semiconductor optical amplifier 10. A portion of the light amplified inside the semiconductor optical amplifier 10 is output again from the end face 10a to the optical integrated circuit 20, and similarly, its return light is returned to the semiconductor optical amplifier 10, further amplified inside the semiconductor optical amplifier 10. A portion of the light amplified inside the semiconductor optical amplifier 10 in this way is output to the outside as output light Lout from the other end face 10b of the semiconductor optical amplifier 10.

[0048] The filter section 30 (wavelength adjustment section) is provided on the substrate 21 and is optically coupled to the semiconductor optical amplifier 10. The filter section 30 changes the wavelength of the output light Lout by the mechanism described above. In other words, the filter section 30 is capable of setting the wavelength of the output light Lout to be variable. As an example, the filter section 30 is configured as a ring resonator. The filter section 30 has five optical waveguides (waveguides 31, 33, 35 and ring waveguides 32, 34) that are independent of each other. These optical waveguides may be formed from the same material as the substrate 21 (e.g., silicon), or from a different material than the substrate 21 (e.g., silicon nitride, germanium, etc.). The same applies to the other optical waveguides described later.

[0049] Waveguide 31 is optically coupled to the end face 10a of the semiconductor optical amplifier 10. Waveguide 33 is coupled to waveguide 31 via ring waveguide 32. Specifically, when the circumference of ring waveguide 32 is an integer multiple of the wavelength of light passing through waveguide 31 in direction A in Figure 1, light passing through waveguide 33 in direction B in Figure 1 is generated.

[0050] Waveguide 35 is coupled to waveguide 33 via ring waveguide 34. Specifically, when the circumference of ring waveguide 34 is an integer multiple of the wavelength of light passing through waveguide 33 in direction B in Figure 1, light is generated passing through waveguide 35 in direction C in Figure 1. The end of waveguide 35 in direction C in Figure 1 is connected to Michelson interferometer 40 (coupler 41). As a result, light L2 transmitted through filter section 30 (i.e., light of a specific wavelength generated in waveguide 35 based on light L1 passing through waveguide 31) is input to coupler 41 of Michelson interferometer 40.

[0051] Curve T1 shown in Figure 2 represents the filter characteristics (transmittance for each wavelength) of the ring waveguide 32. Curve T2 shown in Figure 2 represents the filter characteristics of the ring waveguide 34. In the graph in Figure 2, the horizontal axis represents wavelength, and the vertical axis represents transmittance (dB). Curve T1 has periodically occurring peaks in transmittance. The wavelengths of each peak in curve T1 correspond to the wavelengths when the circumference of the ring waveguide 32 is an integer multiple of the wavelength of light passing through waveguide 31. Similarly, curve T2 has periodically occurring peaks in transmittance. The wavelengths of each peak in curve T2 correspond to the wavelengths when the circumference of the ring waveguide 34 is an integer multiple of the wavelength of light passing through waveguide 33.

[0052] As shown in Figure 2, in this embodiment, the circumference lengths of ring waveguide 32 and ring waveguide 34 are different from each other. That is, the peak spacing of curve T1 and the peak spacing of curve T2 are different from each other. As a result, the Vernier effect makes it possible to sweep the wavelength of transmitted light (light L2) over a wider wavelength range than when using a ring resonator consisting of a single ring (ring waveguide 32 or 34) (i.e., sweep the wavelength of output light Lout over a wider wavelength range). The filter characteristics of the filter section 30 are the sum of curve T1 (filter characteristics of ring waveguide 32) and curve T2 (filter characteristics of ring waveguide 34), and are therefore represented as curve T. In this way, the filter section 30 has a larger FSR (Free Spectrum Range; F1 in Figure 2) than when the filter section 30 is constructed with a single ring waveguide, because it has two ring waveguides 32 and 34 with different circumference lengths.

[0053] The filter section 30 includes a phase shifter 36 for filter control arranged along the ring waveguide 32, a phase shifter 37 for filter control arranged along the ring waveguide 34, and a phase shifter 38 for longitudinal mode control arranged along the waveguide 35. The position of the phase shifter 38 is not limited to that along the waveguide 35. For example, the phase shifter 38 may be positioned along the waveguide 31 or waveguide 33.

[0054] The phase shifter 36 is, for example, a thermal phase shifter that generates heat when power is supplied. The phase shifter 36 is composed of, for example, a thin-film heater. By generating such heat, the phase shifter 36 changes (increases) the refractive index of the portion of the ring waveguide 32 along the phase shifter 36, thereby shifting the phase of the light orbiting the ring waveguide 32. By controlling the power supplied to the phase shifter 36, the amount of phase adjustment (phase shift amount) of the ring waveguide 32 can be controlled.

[0055] Phase shifter 37 is a thermal phase shifter, similar to, for example, phase shifter 36. That is, phase shifter 37 generates heat to change (increase) the refractive index of the portion of the ring waveguide 34 along the phase shifter 37, thereby shifting the phase of the light orbiting the ring waveguide 34. The amount of phase adjustment of the ring waveguide 34 can be controlled by controlling the power supplied to phase shifter 37.

[0056] By controlling the phase adjustment amount in the ring waveguides 32 and 34 using phase shifters 36 and 37, the filter characteristics of the filter section 30 described above can be changed. Specifically, by controlling the phase adjustment amount of the ring waveguide 32 with phase shifter 36, the curve T1 in Figure 2 can be changed. Similarly, by controlling the phase adjustment amount of the ring waveguide 34 with phase shifter 37, the curve T2 in Figure 2 can be changed. In other words, by controlling the phase adjustment amount of the ring waveguides 32 and 34 with phase shifters 36 and 37, the resonator wavelength can be shifted. As a result, the filter characteristics of the filter section 30 (curve T in Figure 2) can be changed.

[0057] The phase shifter 38 is a thermal phase shifter, similar to, for example, phase shifters 36 and 37. By generating heat, the phase shifter 38 changes the refractive index of the portion of the waveguide 35 along the phase shifter 38, thereby shifting the longitudinal modes of the resonator formed by the semiconductor optical amplifier 10 and the optical integrated circuit 20 (filter section 30 and Michelson interferometer 40). This allows the wavelength of light L2 passing through the waveguide 35 to be changed. In other words, the phase shifter 38 allows for fine-tuning of the wavelength of light L2.

[0058] The Michelson interferometer 40 is optically coupled to the end of the filter section 30 opposite to the end of the waveguide 31 that is coupled to the semiconductor optical amplifier 10 (in this embodiment, the end of the waveguide 35 in direction C) (in this embodiment, the end of the waveguide 35 in direction C).

[0059] The Michelson interferometer 40 includes a coupler 41, a first mirror 42, and a second mirror 43. As an example, each component constituting the Michelson interferometer 40 is provided on a substrate 21. However, some components constituting the Michelson interferometer 40 (for example, the first mirror 42 or the second mirror 43) may be arranged on the side or outside of the substrate 21. The Michelson interferometer 40 splits the light L2 that has passed through the filter section 30 and been input to the coupler 41 into a first branched light Lb1 that goes toward the first mirror 42 and a second branched light Lb2 that goes toward the second mirror 43. The optical path length of the waveguide 51 (first waveguide) from the coupler 41 to the first mirror 42 is different from the optical path length of the waveguide 52 (second waveguide) from the coupler 41 to the second mirror 43. The Michelson interferometer 40 is configured to reflect the first branched light Lb1 and the second branched light Lb2, which are branched into two paths with different optical path lengths, using the first mirror 42 and the second mirror 43, and then reconnect them using the coupler 41.

[0060] In this embodiment, as an example, the optical path length of waveguide 52 is made longer than that of waveguide 51. The optical path length difference between waveguide 51 and waveguide 52 may be set, for example, by making the physical lengths (the lengths d1 and d2 shown in FIG. 3) of waveguides 51 and 52 different from each other, or by making the refractive indices of waveguides 51 and 52 different from each other, or by both of the above. The refractive index difference between waveguides 51 and 52 may be set, for example, by making the materials of waveguides 51 and 52 different from each other, or by phase shifters 71 arranged along waveguide 51 and phase shifters 72 arranged along waveguide 52. Phase shifters 71 and 72 are, for example, thermal type phase shifters similar to the above-described phase shifters 36 to 38. Phase shifter 71 changes the refractive index of the portion along phase shifter 71 in waveguide 51 by generating heat. Phase shifter 72 changes the refractive index of the portion along phase shifter 72 in waveguide 52 by generating heat. By controlling such phase shifters 71 and 72, the optical path length difference between waveguides 51 and 52 can be adjusted. When it is not necessary to change the optical path length difference between waveguides 51 and 52, etc., phase shifters 71 and 72 may be omitted.

[0061] As shown in FIG. 3, coupler 41 is, for example, constituted by a directional coupler having 4 (2×2) ports. Examples of the directional coupler constituting coupler 41 include an asymmetric directional coupler, a bent directional coupler, etc. Coupler 41 has two upstream ports (first port 41a and fourth port 41d) and two downstream ports (second port 41b and third port 41c). First port 41a is connected to the downstream end of waveguide 35. That is, light L2 that has passed through waveguide 35 is input to first port 41a. Second port 41b is connected to waveguide 51. Third port 41c is connected to waveguide 52. Coupler 41 branches the light L2 from filter section 30 input to first port 41a into waveguide 51 connected to second port 41b and waveguide 52 connected to third port 41c at a predetermined branching ratio “x:1−x” (0<x<1).

[0062] The first mirror 42 is connected to the end of the waveguide 51 opposite to the side connected to the second port 41b. The first mirror 42 is configured to return at least a portion of the first branched light Lb1 output from the second port 41b to the waveguide 51 to the coupler 41. In this embodiment, the first mirror 42 is configured to return the first branched light Lb1 to the second port 41b of the coupler 41 with 100% reflectivity.

[0063] The second mirror 43 is connected to the end of the waveguide 52 opposite to the side connected to the third port 41c. The second mirror 43 is configured to return at least a portion of the second branched light Lb2 output from the third port 41c to the waveguide 52 back to the coupler 41. In this embodiment, the second mirror 43 is configured to return the second branched light Lb2 to the third port 41c of the coupler 41 with 100% reflectivity.

[0064] The first mirror 42 and the second mirror 43 are configured, for example, as loop mirrors in which the two output ports of a 1x2 coupler are connected by a loop-shaped waveguide. However, the configuration of the first mirror 42 and the second mirror 43 is not limited to the above configuration. For example, at least one of the first mirror 42 and the second mirror 43 may be made of a highly reflective film provided on the side surface of the substrate 21.

[0065] As shown in Figures 1 and 3, a detector 44 is connected to the fourth port 41d of the coupler 41 via a waveguide 53. The detector 44 is composed of, for example, a photodiode, and detects the interference light IL (first interference light) between the first return light Lr1 returned from the first mirror 42 to the second port 41b and the second return light Lr2 returned from the second mirror 43 to the third port 41c.

[0066] The branching ratio "x:1 - x" of the coupler 41 is set to an unequal ratio. That is, in the above branching ratio, x is set to satisfy the condition of "0 < x < 1, x ≠ 0.5". In this embodiment, "x = 0.99", and the branching ratio of the coupler 41 is set to "99:1". That is, as an example in this embodiment, the branching ratio of the coupler 41 is set such that the ratio of the power of the first branched light Lb1 to the power of the input light (light L2) to the first port 41a is 0.99.

[0067] As shown in FIG. 3, if the power of the light L2 input from the waveguide 35 to the first port 41a is represented as "P" in ", the power of the first branched light Lb1 output from the second port 41b is "xP" in ", and the power of the second branched light Lb2 output from the third port 41c is "(1 - x)P" in ". As described above, since the reflectivities of the first mirror 42 and the second mirror 43 are both 100%, the power of the first return light Lr1 is the same as the power of the first branched light Lb1, and the power of the second return light Lr2 is the same as the power of the second branched light Lb2.

[0068] The coupler 41 branches the first return light Lr1 input to the second port 41b into the light Lr1a output from the first port 41a and the light Lr1b output from the fourth port 41d. At this time, due to the nature of the coupler 41, the branching ratio between the light Lr1a and the light Lr1b is "x:1 - x". Therefore, the power of the light Lr1a is "x" 2 P in ", and the power of the light Lr1b is "x(1 - x)P" in ".

[0069] The coupler 41 branches the second return light Lr2 input to the third port 41c into the light Lr2a output from the first port 41a and the light Lr2b output from the fourth port 41d. At this time, due to the nature of the coupler 41, the branching ratio between the light Lr2a and the light Lr2b is "1 - x:x". Therefore, the power of the light Lr2b is "(1 - x)" 2 P in」, and the power of the light Lr2b is "x(1 - x)P in 」.

[0070] The reflected light RL (the second interference light) returning from the first port 41a to the filter unit 30 is the light combining the lights Lr1a and Lr2a. The interference light IL traveling from the fourth port 41d to the detector 44 is the light combining the lights Lr1b and Lr2b. Both the reflected light RL and the interference light IL are interference lights generated by the interference between the first return light Lr1 and the second return light Lr2.

[0071] When the reflectivities of the first mirror 42 and the second mirror 43 are 100% (that is, when the first branched light Lb1 is directly returned to the coupler 41 as the first return light Lr1 and the second branched light Lb2 is directly returned to the coupler 41 as the second return light Lr2), the powers of the two lights Lr1b and Lr2b constituting the interference light IL are equal to each other regardless of the branching ratio of the coupler 41. Therefore, the interference light IL becomes an interference light with a high degree of interference of each light (lights Lr1b and Lr2b) passing through the waveguides 51 and 52 having different optical path lengths. That is, the interference light IL becomes a light with a large wavelength dependence (that is, a change in power (detection result of the detector 44) corresponding to a change in wavelength). By utilizing such a property of the interference light IL, the wavelength of the light L2 (that is, the wavelength set by the filter unit 30) can be monitored based on the detection result (power of the interference light IL) by the detector 44.

[0072] Note that the powers of the optical signals Lr1b and Lr2b that constitute the interference optical signal IL do not need to be exactly equal. For example, if the reflectivity of the first mirror 42 or the second mirror 43 is not 100%, or if a portion of the first branched optical signal Lb1 or the second branched optical signal Lb2 is taken out externally, the powers of optical signals Lr1b and Lr2b will not be equal. Even in this case, by returning most of the first branched optical signal Lb1 to the coupler 41 as the first return optical signal Lr1, and returning most of the second branched optical signal Lb2 to the coupler 41 as the second return optical signal Lr2, the power difference between optical signals Lr1b and Lr2b can be made sufficiently small, and the degree of interference can be made sufficiently high, so that wavelength monitoring can be performed based on the interference optical signal IL.

[0073] Figure 4 is a graph showing an example of the power characteristics (wavelength dependence) of the interference light IL detected by the detector 44. The horizontal axis represents the wavelength of light L2 (i.e., the wavelength set by the filter unit 30), and the vertical axis represents the power of the interference light IL (dB). The FSR of the interference light IL (F2 in Figure 4) is determined by the difference in optical path lengths between waveguides 51 and 52. Specifically, the larger the difference in optical path lengths between waveguides 51 and 52, the smaller the FSR of the interference light IL (F2 in Figure 4). Range M is a region where the wavelength and the power of the interference light IL correspond in a 1:1 ratio (i.e., a region where the power of the interference light IL changes monotonically (increases or decreases) in response to wavelength changes), and is an example of a range used for wavelength monitoring. In other words, the light source unit 1 can variably adjust the wavelength of the output light Lout while monitoring it within range M. Note that range M is not limited to the example in Figure 4, and only needs to be set to a region where the wavelength and the power of the interference light IL correspond in a 1:1 ratio.

[0074] The characteristics of the wavelength monitor shown in Figure 4 (power characteristics of the interference light IL detected by the detector 44) can be determined in advance, for example, by pre-measuring the power of the interference light IL corresponding to each wavelength while changing the wavelength transmitted through the filter section 30 (wavelength of light L2) under predetermined driving conditions (e.g., laser output power). Then, when actually using the light source unit 1 to output output light Lout of a predetermined wavelength included in the range M, the predetermined wavelength can be identified based on the pre-determined power characteristics of the interference light IL (Figure 4) and the power of the interference light IL detected by the detector 44. That is, as shown in Figure 4, when a detected value Vp is obtained by inputting interference light IL to the detector 44, the wavelength of light L2 (wavelength λ corresponding to the detected value Vp) can be estimated based on the detected value Vp.

[0075] On the other hand, the power of the light Lr1a that constitutes the reflected light RL is "x 2 P in " and the power of Light Lr2a "(1-x) 2 P inBecause they are different from each other, the reflected light RL becomes interference light with a relatively low degree of interference between light Lr1a and light Lr2a. In other words, the reflected light RL becomes light with relatively little wavelength dependence. As a result, the reflected light RL can be used as mirror light to return to the semiconductor optical amplifier 10. That is, the influence of the reflected light RL on the filter section 30 can be suppressed while the reflected light RL can be transmitted through the filter section 30 and returned to the semiconductor optical amplifier 10. More specifically, as the reflected light RL flows upstream in the waveguide 35, light is generated that transmits through the filter section 30 in the opposite direction to the light L1 described above. That is, as the reflected light RL flows upstream in the waveguide 35 (opposite direction C in Figure 1), light is generated that flows upstream (opposite direction B in Figure 1) in the waveguide 33 which is optically coupled to the waveguide 35 via the ring waveguide 34. Next, light is generated flowing upstream (in the opposite direction to direction A in Figure 1) through the waveguide 31, which is optically coupled to the waveguide 33 via the ring waveguide 32, and this light is input to the semiconductor optical amplifier 10 from the end face 10a. As a result, light of a predetermined wavelength set by the filter unit 30 (reflected light RL that has passed through the filter unit 30 as described above) is amplified by the semiconductor optical amplifier 10, and output light Lout of the predetermined wavelength is output to the outside from the end face 10b.

[0076] Figure 5 is a graph showing an example of the characteristics (wavelength dependence) of reflected light RL. The horizontal axis represents the wavelength of light L2 (i.e., the wavelength set by the filter section 30), and the vertical axis represents the power (dB) of the reflected light RL. In this example, the ripple R of the reflected light RL is a very small value of about 0.18 dB.

[0077] As described above, it is preferable that the reflected light RL returning to the filter section 30 has low wavelength dependence. In other words, it is preferable that the ripple R of the reflected light RL be as small as possible. The wavelength dependence (ripple R) of the reflected light RL decreases as the coupling coefficient of the coupler 41 decreases (i.e., as the branching ratio of the coupler 41 moves away from "50:50"). Here, the coupling coefficient is the power of one of the two beams of light branched by the coupler 41 (in this embodiment, the second branched beam Lb2 output from the third port 41c) when the power of the input light to the coupler 41 is set to 1. Therefore, in this embodiment, the coupling coefficient is 0.5 when the branching ratio of the coupler 41 (power of the first branched beam Lb1:power of the second branched beam Lb2) is "50:50", and the coupling coefficient is 0.01 (=10) when the branching ratio is "99:1". -2 )

[0078] On the other hand, the smaller the coupling coefficient, the smaller the maximum power of the interference light IL output from the coupler 41. When the maximum power of the interference light IL is small, the range of wavelengths that can be monitored based on the detection results of the detector 44 may be narrowed. For example, if the maximum power of the interference light IL output from the coupler 41 is smaller than the maximum power value in the range M shown in Figure 4 (in the example of Figure 4, the power value corresponding to the right end of range M), the range of monitorable wavelengths (i.e., the range in which wavelength sweeping is possible by the filter unit 30 while monitoring the wavelength) becomes smaller than the range M that can be monitored. For this reason, it is preferable to set the coupling coefficient of the coupler 41 to a certain size or larger so as not to cause the above-mentioned problems.

[0079] As shown in Figure 6, there is a trade-off relationship between the wavelength dependence (ripple) of the reflected light RL and the maximum power of the interference light IL output from the coupler 41. That is, as described above, reducing the coupling coefficient reduces the ripple of the reflected light RL, but at the same time, it reduces the maximum power of the interference light IL output from the coupler 41. For this reason, the maximum value of the coupling coefficient of the coupler 41 may be determined based on the maximum acceptable value of the reflected light RL ripple. Alternatively, the minimum value of the coupling coefficient of the coupler 41 may be determined based on the minimum maximum power of the interference light IL that does not cause the problems described above.

[0080] First, with reference to Figures 7 and 8, an example of how to determine the minimum coupling coefficient of the coupler 41 will be explained. Figure 7 shows a simplified power characteristic (see Figure 4) of the interference light IL detected by the detector 44. In Figure 7, Pmin is the minimum power detectable by the detector 44 (hereinafter referred to as "minimum detectable power"). The minimum detectable power Pmin is determined by the type and performance of the detector 44. Here, as an example, the minimum detectable power Pmin of the detector 44 is assumed to be 10 nW. Also, as an example, the output power of the semiconductor optical amplifier 10 is set to 10 mW and the extinction ratio is assumed to be 30 dB. In this case, the maximum power Pmax detected by the detector 44 (i.e., the peak value of the waveform showing the characteristics of the wavelength monitor as shown in Figures 4 and 7) is -20 dBm (= 10 μW). Therefore, in this case, in order to monitor the wavelength over the entire range M, the coupling coefficient should be set so that the maximum power of the interference light IL output from the coupler 41 is greater than or equal to the maximum power Pmax. Figure 8 shows the relationship between the coupling coefficient and the maximum power of the interferential light IL output from the coupler 41 when the output power of the semiconductor optical amplifier 10 is set to 10 mW. In this example, from the graph in Figure 8, in order to satisfy the condition that the maximum power of the interferential light IL output from the coupler 41 is greater than or equal to the maximum power Pmax, the coupling coefficient must be 2.5 × 10 -4 You can see that this is all you need to do.

[0081] Next, with reference to Figures 9 to 11, an example of how to determine the maximum value of the coupling coefficient of the coupler 41 will be explained. Figure 9 shows an example of the filter characteristics (transmittance for each wavelength) of the filter section 30. Peak p1 in Figure 9 indicates the peak with the maximum transmittance, and peak p2 indicates the peak with the next largest transmittance after peak p1 (the peak adjacent to peak p1). Here, in order to cause single-mode oscillation of light at the wavelength corresponding to peak p1, it is generally necessary to make the SMSR (Side Mode Suppression Ratio), which is the difference in transmittance (intensity ratio) between peak p1 and peak p2, greater than 3 dB. That is, if the SMSR is less than 3 dB, laser oscillation may occur not only at the wavelength corresponding to peak p1 but also at the wavelength corresponding to peak p2, and there is a risk that single-mode light cannot be obtained.

[0082] Here, the filter characteristics of the filter section 30 shown in Figure 9 ignore the effect of the wavelength dependence of the reflected light RL. However, the actual filter characteristics of the filter section 30 are affected by the wavelength dependence (ripple) of the reflected light RL. More specifically, the filter characteristics of the filter section 30 are the product of the transmittance of the reflected light RL (corresponding to the power shown on the vertical axis in Figure 5) and the characteristics of the filter section 30 shown in Figure 9.

[0083] Figure 10 shows an example of the wavelength dependence (transmittance for each wavelength) of reflected light RL. The dots in Figure 10 represent the local peaks in the filter characteristics of the filter section 30. dmax in Figure 10 is the maximum value of the transmittance difference of reflected light RL between these local peaks. In the example in Figure 10, the SMSR may be smallest when the two peaks where the transmittance difference of reflected light RL is dmax correspond to peaks p1 and p2.

[0084] Figure 11 shows the ripple, coupling coefficient, and maximum SMSR reduction (dB) obtained by simulation, assuming that there are 10 localized peaks within the FSR (F1 in Figure 2) of the filter section 30. From the simulation results in Figure 11, it can be seen that as the coupling coefficient increases, the ripple of the reflected light RL increases, and consequently, the maximum SMSR reduction (see Figure 9) increases. From the above, it can be concluded that the coupling coefficient can be determined based on the acceptable maximum SMSR reduction. That is, the coupling coefficient can be determined within a range where the maximum SMSR reduction is an acceptable value for maintaining single-mode oscillation.

[0085] Based on the above, the branching ratio of the coupler 41 may be set such that the ratio of the power of the first branched light Lb1 to the power of the input light to the first port 41a is 0.75 or more (i.e., the coupling coefficient is 0.25 or less). With the above configuration, by setting the branching ratio to a value that is a certain distance away from the equal division ratio "50:50" (coupling coefficient = 0.5), it becomes possible to use a coupler 41 with less wavelength dependence while effectively reducing the ripple R of the reflected light RL returned to the upstream side from the first port 41a. As a result, the wavelength dependence of the light source unit 1 can be suitably reduced. Specifically, as shown in the simulation results in Figure 11, by setting the coupling coefficient to 0.25 or less, the ripple R can be suppressed to about 6.0 dB or less.

[0086] The branching ratio of the coupler 41 may be set such that the ratio of the power of the first branched light Lb1 to the power of the input light to the first port 41a is 0.85 or greater (i.e., the coupling coefficient is 0.15 or less). With the above configuration, the power of the reflected light RL can be made greater than the power of the interference light IL in the entire wavelength range (i.e., regardless of the value set for the wavelength of the output light Lout). In other words, the power of the light returned to the semiconductor optical amplifier 10 (reflected light RL) can be made relatively large, and therefore the power of the output light Lout (laser output) can be made large.

[0087] The branching ratio of the coupler 41 may be set such that the ratio of the power of the first branched light Lb1 to the power of the input light to the first port 41a is 0.90 or greater (i.e., the coupling coefficient is 0.1 or less). With the above configuration, the above-mentioned effects can be further improved by setting the branching ratio to a value further away from the equal division ratio. Specifically, as shown in the simulation results of Figure 11, by setting the coupling coefficient to 0.1 or less, the ripple R can be suppressed to about 2.0 dB or less. In addition, the effect of increasing the laser output by increasing the power of the reflected light RL as described above can be improved.

[0088] The branching ratio of the coupler 41 may be set such that the ratio of the power of the first branched light Lb1 to the power of the input light to the first port 41a is 0.99 or greater (i.e., the coupling coefficient is 0.01 or less). With the above configuration, the above-mentioned effects can be further improved by setting the branching ratio to a value further away from the equal division ratio. Specifically, by setting the branching ratio to "99:1" (coupling coefficient = 0.01) as in this embodiment, the ripple R can be suppressed to about 0.18 dB or less (see Figure 5). In addition, the effect of increasing the laser output by increasing the power of the reflected light RL as described above can be further improved.

[0089] Furthermore, the branching ratio of the coupler 41 may be set by focusing on the magnitude of the ripple R of the reflected light RL. For example, the branching ratio of the coupler 41 may be set so that the ripple R of the reflected light RL is 10 dB or less. With the above configuration, by sufficiently reducing the ripple R of the reflected light RL, the wavelength dependence of the light source unit 1 can be suitably reduced. For example, the maximum decrease in the SMSR (see Figure 9) of the filter characteristics of the filter section 30 due to the effect of ripple R can be reduced, and single-mode laser oscillation can be suitably realized. The upper limit of the above branching ratio (branching ratio when ripple R is 10 dB) is approximately "66:34" (coupling coefficient = 0.342). Also, as an example, in the simulation results shown in Figure 11, the maximum decrease in SMSR can be kept to approximately 1.87 dB or less.

[0090] The branching ratio of coupler 41 may be set such that the ripple R of the reflected light RL is 6 dB or less. With the above configuration, by further reducing the ripple R of the reflected light RL, the wavelength dependence of the light source unit 1 can be further reduced. The upper limit of the above branching ratio (the branching ratio when the ripple R is 6 dB) is approximately "75:25" (coupling coefficient = 0.249). Also, as an example, in the simulation results shown in Figure 11, the maximum reduction in SMSR can be kept to approximately 0.99 dB or less.

[0091] The branching ratio of coupler 41 may be set such that the ripple R of the reflected light RL is 2 dB or less. With the above configuration, by further reducing the ripple R of the reflected light RL, the wavelength dependence of the light source unit 1 can be reduced even more favorably. The upper limit of the above branching ratio (the branching ratio when the ripple R is 2 dB) is approximately "90:10" (coupling coefficient = 0.102). Also, as an example, in the simulation results shown in Figure 11, the maximum reduction in SMSR can be kept to approximately 0.31 dB or less.

[0092] The branching ratio of coupler 41 may be set such that the ripple R of the reflected light RL is 0.18 dB or less. With the above configuration, by further reducing the ripple R of the reflected light RL, the wavelength dependence of the light source unit 1 can be further reduced. The upper limit of the above branching ratio (the branching ratio when the ripple R is 0.18 dB) is approximately "99:1" (coupling coefficient = 0.01). Also, as an example, in the simulation results shown in Figure 11, the maximum reduction in SMSR can be kept to approximately 0.03 dB or less. [Effects and Effects]

[0093] According to the light source unit 1, the Michelson interferometer 40, including the mirror section (first mirror 42 and second mirror 43), can be used to return light to the semiconductor optical amplifier 10 while also functioning as a wavelength monitor. More specifically, as shown in Figure 4, the wavelength λ of the output light Lout (wavelength adjusted by the filter section 30) can be determined (estimated) based on the power (detected value Vp) of the interference light IL detected by the detector 44. This eliminates the need to provide a separate wavelength monitor (e.g., an asymmetric Mach-Zehnder interferometer) from the configuration including the mirror section (Michelson interferometer 40), thus reducing the number of components. As a result, the entire light source unit 1 can be miniaturized, and the problem of large wavelength dependence due to manufacturing variations in each component can be avoided. Furthermore, by setting the branching ratio of the coupler 41 to be unequal (for example, "99:1"), it becomes possible to monitor the wavelength of the output light Lout using the detector 44 connected to the fourth port 41d, while reducing the wavelength dependence of the reflected light RL (ripple R shown in Figure 5) compared to when the branching ratio of the coupler 41 is set to an equal ratio of "50:50". In addition, by setting the branching ratio to be unequal, it becomes possible to use a coupler 41 that has less manufacturing variation and less wavelength dependence than a coupler set to an equal ratio. As a result, the light source unit 1 can be miniaturized while reducing wavelength dependence.

[0094] As shown in Figure 3, in the light source unit 1, the coupler 41 is configured as a directional coupler. Here, the coupler 41 may be configured as a coupler other than a directional coupler (for example, a multimode interference coupler (MMI), etc.), but by configuring the coupler 41 as a directional coupler (for example, by adopting an asymmetric directional coupler or a bendable directional coupler as the coupler 41), a coupler 41 with an unequal branching ratio can be easily and accurately manufactured. As a result, the large wavelength dependence caused by manufacturing variations in the coupler 41 can be suitably suppressed. In other words, in the light source unit 1, in order to obtain both the reflected light RL and the interference light IL as described above, by deliberately setting the branching ratio of the coupler 41 to be unequal, a synergistic effect can be obtained in which a coupler 41 with less wavelength dependence than a coupler with an equal branching ratio can be realized.

[0095] As shown in Figure 1, in the light source unit 1, the semiconductor optical amplifier 10 is configured to output output light Lout from the end face 10b opposite to the end face 10a that is coupled to the filter section 30. With this configuration, there is no need to arrange circuit members on the substrate 21 to guide the output light Lout, thus simplifying the circuit configuration on the substrate 21. In addition, the output light Lout can be extracted externally without affecting the characteristics of the Michelson interferometer 40 shown in Figure 3. [Differentiation]

[0096] This disclosure is not limited to the configurations shown in the embodiments described above. The materials and shapes of each component of the light source unit are not limited to the specific materials and shapes described above, but can be made from a variety of other materials and shapes. Furthermore, some of the components included in the embodiments described above may be omitted or modified as appropriate, and can be arbitrarily combined with additional elements. Several variations of the light source unit (light source units 1A to 1H) are described below. In light source units 1A to 1H described below, the phase shifters 38, 71, and 72 provided in light source unit 1 are omitted, but light source units 1A to 1H may be equipped with these phase shifters 38, 71, and 72, just like light source unit 1. [First variation]

[0097] As shown in Figure 12, the light source unit 1A according to the first modified example differs from the light source unit 1 in that it has a Michelson interferometer 40A with a power monitor instead of the Michelson interferometer 40. The Michelson interferometer 40A further differs from the Michelson interferometer 40 in that it has a coupler 45, a power monitor 46 (first measuring instrument), a waveguide 54, and a waveguide 55. The coupler 45 has at least one upstream port 45a and two downstream ports 45b and 45c. In the Michelson interferometer 40A, the downstream end of the waveguide 51 is connected to the upstream port 45a of the coupler 45 instead of the first mirror 42. Also, one downstream port 45b of the coupler 45 is connected to the first mirror 42 via the waveguide 54. The other downstream port 45c of the coupler 45 is connected to the power monitor 46 via the waveguide 55. The power monitor 46 is composed of, for example, a photodiode similar to the detector 44, and detects (measures) the power of the input light.

[0098] The coupler 45, for example, splits the light input from the upstream port 45a to the downstream ports 45b and 45c at a splitting ratio (for example, "99:1") such that the power of the light output from the downstream port 45b is greater than the power of the light output from the downstream port 45c. The light output from the downstream port 45b to the first mirror 42 is reflected by the first mirror 42, and the first return light Lr1, similar to that of the light source unit 1, is input again to the downstream port 45b of the coupler 45. The first return light Lr1 input to the downstream port 45b is input from the upstream port 45a to the second port 41b of the coupler 41 via the waveguide 51.

[0099] In the light source unit 1A, the branching ratio of the coupler 45 is set so that most of the light input from the upstream port 45a of the coupler 45 is output from the downstream port 45b connected to the first mirror 42. This allows for the generation of reflected light RL and interference light IL, similar to the light source unit 1. In addition, the power of the light L3 output from the downstream port 45c of the coupler 45 can be measured by the power monitor 46 in the light source unit 1A. Figure 13 shows an example of the power of light L3 measured for each wavelength by the power monitor 46. In the example in Figure 13, the difference between the maximum and minimum values ​​of the power of light L3 is about 0.002 dB, and the amount of change in power when the wavelength is changed is relatively small, however, the output power from the light source (semiconductor optical amplifier 10) may fluctuate over time. Furthermore, the characteristics of the wavelength monitor (power characteristics of interference light IL detected by the detector 44) as shown in Figure 4 are affected by the output power from the light source. That is, the power of interference light IL changes by the same amount as the output power from the light source.

[0100] Based on the above, the light source unit 1A acquires both the detection result of the detector 44 (power PIL of the interference light IL for each wavelength) and the measurement result of the power monitor 46 (power PL3 of light L3), as shown in Figure 14. This allows the characteristics of the ratio of power PIL to power PL3 (PIL / PL3) (i.e., the value of the above ratio for each wavelength) as shown in Figure 15 to be determined in advance. When actually using the light source unit 1A to output output light Lout of a predetermined wavelength included in the range M, the ratio Vc of the power PIL of the interference light IL detected by the detector 44 and the power PL3 of light L3 measured by the power monitor 46 can be calculated. Based on this ratio Vc, the wavelength of light L2 (wavelength λ corresponding to the ratio Vc) can be estimated.

[0101] The light source unit 1A is positioned downstream of the filter unit 30 and further includes a power monitor 46 that measures the power of the light that has passed through the filter unit 30 (in the example in Figure 12, the light L3 output to the waveguide 55 via couplers 41 and 45). With the above configuration, the wavelength of the output light Lout (i.e., the wavelength set by the filter unit 30) can be estimated more accurately based on the detection result of the detector 44 (power PIL of the interfering light IL) and the measurement result of the power monitor 46 (power PL3 of the light L3 that has passed through the filter unit 30). More specifically, as described above, by using the ratio of the power PIL of the interfering light IL to the light source power (power PL3 of light L3), even if the light source power fluctuates, the effect of such fluctuation (change in the power PIL of the interfering light IL) can be canceled out, so that the wavelength of the output light Lout can be estimated with high accuracy.

[0102] In the light source unit 1A, the power monitor 46 is configured to detect a portion of the light output from the second port 41b of the coupler 41 (first branched light Lb1). However, the power monitor 46 may also be configured to detect a portion of the light output from the third port 41c of the coupler 41 (second branched light Lb2). However, if the branching ratio of the coupler 41 is significantly deviated from an equal division ratio (e.g., 99:1), as in this embodiment, the power of the second branched light Lb2 becomes very small, making it difficult to extract light for power monitoring from the second branched light Lb2. In such cases, as in this modified example, it is preferable to extract and measure a portion of the first branched light Lb1, which constitutes the majority of the light L2 (in this modified example, 1 / 100th of the light of the first branched light Lb1), for power monitoring purposes. Furthermore, as in the light source unit 1B of the second modified example described later, the power monitor 46 may be configured to detect a portion of the light L2 passing through the waveguide 35. In this case, since it is not necessary to extract a portion of the first branched light Lb1 or the second branched light Lb2 for power monitoring, the light source power can be monitored without affecting the characteristics of the Michelson interferometer 40 shown in Figure 3. [Second variation]

[0103] As shown in Figure 16, the light source unit 1B according to the second modification mainly differs from the light source unit 1 in that it has a Michelson interferometer 40B with a power monitor instead of the Michelson interferometer 40. The Michelson interferometer 40B mainly differs from the Michelson interferometer 40 in that it further has a power monitor 46 (first measuring instrument) similar to the first modification, a coupler 47 (second coupler), a reflected light monitor 48 (second measuring instrument), a waveguide 56, a waveguide 57, and a waveguide 58. The coupler 47 has two upstream ports 47a, 47b and two downstream ports 47c, 47d. In the Michelson interferometer 40B, the downstream end of the waveguide 35 is connected to one of the upstream ports 47a of the coupler 47, rather than the first port 41a of the coupler 41. Furthermore, one downstream port 47c of coupler 47 is connected to the first port 41a of coupler 41 via waveguide 56. The other downstream port 47d of coupler 47 is connected to power monitor 46 via waveguide 57. The other upstream port 47b of coupler 47 is connected to reflected light monitor 48 via waveguide 58. The reflected light monitor 48 is composed of, for example, a photodiode similar to detector 44, and detects (measures) the power of the input light.

[0104] Coupler 47 splits the light L2 input from upstream port 47a to downstream port 47c and downstream port 47d at a splitting ratio (for example, "99:1") such that the power of the light output from downstream port 47c is greater than the power of the light output from downstream port 47d. Power monitor 46 measures the power of the light L4 (in this modified example, 1 / 100th of the light L2) output from downstream port 47d to waveguide 57. Coupler 47 also splits the reflected light RL input to downstream port 47c to upstream port 47a and upstream port 47b at the above splitting ratio. Reflected light monitor 48 measures the power of the light L5 (in this modified example, 1 / 100th of the reflected light RL) output from upstream port 47b to waveguide 58.

[0105] In the light source unit 1B, the branching ratio of the coupler 47 is set so that most of the light L2 input from the upstream port 47a of the coupler 47 (99% in this modified example) is output from the downstream port 47c connected to the first port 41a of the coupler 41. This allows for the generation of reflected light RL and interference light IL, similar to the light source unit 1. Furthermore, in the light source unit 1B, the power of the light L4 output from the downstream port 47d of the coupler 47 can be measured by the power monitor 46. This allows for the same effect as the light source unit 1A described above. In addition, in the light source unit 1B, the power of the light L5 output from the upstream port 47b of the coupler 47 can be measured by the reflected light monitor 48. The reflected light monitor 48 makes it possible to measure the wavelength dependence of the reflected light RL as shown in Figure 5.

[0106] As described above, the light source unit 1B is positioned between the first port 41a of the coupler 41 and the filter section 30, and further includes a reflected light monitor 48 that measures the power of the reflected light RL (in this modified example, the power of a portion of the reflected light L5). With the above configuration, based on the measurement results from the reflected light monitor 48, the mirror characteristics (for example, the ripple R of the reflected light RL, etc.) as shown in Figure 5 can be determined. The mirror characteristics thus determined can be used for various purposes such as correction and feedback to the detection results (wavelength estimation results) of the detector 44.

[0107] The light source unit 1B is positioned between the first port 41a of the coupler 41 and the filter unit 30, and includes a coupler 47 having an upstream port 47a (fifth port) into which light L2 from the filter unit 30 is input, a downstream port 47c (sixth port) connected to the first port 41a, a downstream port 47d (seventh port), and an upstream port 47b (eighth port). The light source unit 1B also includes a power monitor 46 connected to the downstream port 47d, which measures the power of light L2 that has passed through the filter unit 30 (in this modified example, the power of light L4, which is 1 / 100 of light L2), and a reflected light monitor 48 connected to the upstream port 47b, which measures the power of reflected light RL (in this modified example, the power of light L5, which is 1 / 100 of reflected light RL). The coupler 47 is configured to split the light input from the upstream port 47a to the downstream ports 47c and 47d, and to split the light input from the downstream port 47c to the upstream ports 47a and 47b. With the above configuration, by providing a power monitor 46 and a reflected light monitor 48, both the effects of the power monitor 46 and the reflected light monitor 48 described above can be obtained. Furthermore, by connecting the power monitor 46 and the reflected light monitor 48 to the coupler 47 placed between the filter unit 30 and the coupler 41, a configuration (i.e., a configuration equipped with a power monitor 46 and a reflected light monitor 48) that can measure both the power of the light L2 that has passed through the filter unit 30 (light source power) and the power of the reflected light RL can be easily and compactly realized.

[0108] Furthermore, with the light source unit 1B, there is no need to extract monitoring light from either the first branched light Lb1 or the second branched light Lb2. Therefore, the power of the light source and the power of the reflected light RL can be monitored without affecting the characteristics of the Michelson interferometer 40 shown in Figure 3.

[0109] As shown in the third to eighth modified examples in Figures 17 to 22, the output light Lout (i.e., light of the wavelength set by the filter unit 30) is not limited to being output to the outside from the end face 10b of the semiconductor optical amplifier 10, but may also be output to other circuits mounted on the substrate 21 of the optical integrated circuit 20. [Third variation]

[0110] As shown in Figure 17, the third modified light source unit 1C differs from the light source unit 1 mainly in that it includes a first mirror 42C instead of the first mirror 42, and further includes a waveguide 59 connected to the first mirror 42C. In the light source unit 1C, as in the light source unit 1, the branching ratio of the coupler 41 is set so that the power of the first branched light Lb1 is greater than the power of the second branched light Lb2. For example, the branching ratio of the coupler 41 is "99:1". A portion of the first branched light Lb1 is then taken out as output light Lout. For example, the first mirror 42C is configured to output a portion of the first branched light Lb1 to the waveguide 59. The first mirror 42C is configured, for example, as a Sagnac loop mirror using a 2x2 coupler. The end of the waveguide 59 opposite to the side connected to the first mirror 42C is connected to another circuit (not shown) provided on the substrate 21.

[0111] In the light source unit 1C, the power of the first return light Lr1 decreases by the amount that output light Lout is extracted from the first branched light Lb1. As a result, the ripple R of the reflected light RL returning to the filter section 30 increases compared to the case where output light Lout is not extracted from the first branched light Lb1. In addition, the extinction ratio of the interference light IL input to the detector 44 decreases. On the other hand, the light source unit 1C allows for a relatively simple configuration in which output light Lout can be output to other circuits on the substrate 21. Furthermore, by extracting output light Lout from the first branched light Lb1, which has the higher power among the light branched by the coupler 41, the power of the extractable output light Lout can be increased compared to the case where a portion of the second branched light Lb2 is extracted as output light Lout. [Fourth variation]

[0112] As shown in Figure 18, the light source unit 1D according to the fourth modification mainly differs from the light source unit 1 in that it includes a second mirror 43D instead of the second mirror 43, and further includes a waveguide 60 connected to the second mirror 43D. In the light source unit 1D, similar to the light source unit 1, the branching ratio of the coupler 41 is set such that the power of the first branched light Lb1 is greater than the power of the second branched light Lb2. For example, the branching ratio of the coupler 41 is "99:1". A portion of the second branched light Lb2 is then taken out as output light Lout. For example, the second mirror 43D is configured to output a portion of the second branched light Lb2 to the waveguide 60. The second mirror 43D is configured as a Sagnac loop mirror using a 2x2 coupler, similar to the first mirror 42C in the third modification, for example. The end of the waveguide 60 opposite to the side connected to the second mirror 43D is connected to another circuit (not shown) provided on the substrate 21.

[0113] In the light source unit 1D, the output light Lout is extracted from the second branched light Lb2, which has lower power. Therefore, the power of the extractable output light Lout is lower compared to the light source unit 1C. Also, similar to the light source unit 1C, the extinction ratio of the interference light IL input to the detector 44 is reduced. On the other hand, the light source unit 1D allows for a relatively simple configuration in which the output light Lout can be output to other circuits on the substrate 21. Furthermore, by extracting the output light Lout from the second branched light Lb2, which has lower power, compared to the case where a portion of the first branched light Lb1 is extracted as the output light Lout, the power difference between the first return light Lr1 and the second return light Lr2 returning to the coupler 41 can be increased, and consequently, the wavelength dependence (ripple R) of the reflected light RL can be reduced. [Fifth and sixth variations]

[0114] As shown in Figure 19, the light source unit 1E according to the fifth modification is configured such that a portion of the light returning from the filter section 30 to the semiconductor optical amplifier 10 is extracted as output light Lout between the semiconductor optical amplifier 10 and the filter section 30. The light source unit 1E mainly differs from the light source unit 1 in that it further includes a coupler 61 arranged on the waveguide 31 and a waveguide 62 connected to the coupler 61. The coupler 61 guides the light L1 output from the end face 10a of the semiconductor optical amplifier 10 directly to the filter section 30 (the portion of the waveguide 31 coupled with the ring waveguide 32). On the other hand, the coupler 61 is configured to branch a portion of the light returning from the filter section 30 to the semiconductor optical amplifier 10 as output light Lout to the waveguide 62. The end of the waveguide 62 opposite to the side connected to the coupler 61 is connected to another circuit (not shown) provided on the substrate 21.

[0115] As shown in Figure 20, the light source unit 1F according to the sixth modified example is configured such that a portion of the light traveling from the semiconductor optical amplifier 10 to the filter unit 30 is extracted as output light Lout between the semiconductor optical amplifier 10 and the filter unit 30. The light source unit 1F mainly differs from the light source unit 1 in that it further includes a coupler 61 placed on the waveguide 31 and a waveguide 63 connected to the coupler 61. The coupler 61 guides the return light traveling from the filter unit 30 to the semiconductor optical amplifier 10 directly to the end face 10a of the semiconductor optical amplifier 10. On the other hand, the coupler 61 is configured to branch a portion of the light L1 output from the end face 10a of the semiconductor optical amplifier 10 to the waveguide 63 as output light Lout. The end of the waveguide 63 opposite to the side connected to the coupler 61 is connected to another circuit (not shown) provided on the substrate 21.

[0116] With light source units 1E and 1F, the power of the light returning to the semiconductor optical amplifier 10 (i.e., light of the wavelength set by the filter unit 30) is reduced, which may inhibit laser oscillation in the semiconductor optical amplifier 10. However, it is possible to output output light Lout to other circuits on the substrate 21 without affecting the characteristics of the Michelson interferometer 40 shown in Figure 3. [Variations 7 and 8]

[0117] As shown in Figure 21, the light source unit 1G according to the seventh modification differs from the light source unit 1 mainly in that it has a waveguide 31G instead of waveguide 31. Waveguide 31G is configured as a linear waveguide optically coupled to the ring waveguide 32, similar to waveguide 31. On the other hand, while the downstream end of waveguide 31 was closed, the downstream end of waveguide 31G is connected to other circuits on the substrate 21 that are not shown. That is, in the light source unit 1G, the output light Lout is taken out to other circuits on the substrate 21 via waveguide 31G.

[0118] As shown in Figure 22, the light source unit 1H according to the eighth modification differs from the light source unit 1 mainly in that it has a waveguide 33H instead of waveguide 33. Waveguide 33H is configured as a linear waveguide optically coupled to ring waveguides 32 and 34, similar to waveguide 33. On the other hand, while the downstream end of waveguide 33 was closed, the downstream end of waveguide 33H is connected to other circuits on the substrate 21 that are not shown. That is, in the light source unit 1H, the output light Lout is taken out to other circuits on the substrate 21 via waveguide 33H.

[0119] As described above, in the light source unit 1G, the filter section 30 is composed of a ring resonator having a ring waveguide 32 and a waveguide 31G which is a linear waveguide optically coupled to the ring waveguide 32, and the output light Lout is taken out from waveguide 31G. In the light source unit 1H, the filter section 30 is composed of a ring resonator having ring waveguides 32, 34 and a waveguide 33H which is a linear waveguide optically coupled to the ring waveguides 32, 34, and the output light Lout is taken out from waveguide 33H.

[0120] With the above configuration, by extending the linear waveguide (waveguide 31G or waveguide 33H) of the ring resonator, it becomes possible to output the output light Lout to other circuits on the substrate 21. Furthermore, by utilizing the linear waveguide originally provided in the filter section 30 (ring resonator), the output light Lout can be extracted without increasing the number of components from the original configuration. However, in the above configuration, there is a disadvantage in that the power of the output light Lout is determined by the coupling efficiency of the filter section 30 (ring resonator), so there is no flexibility in adjustment. Nevertheless, it is possible to effectively utilize the light that is not being used in the light source unit 1 (light flowing through waveguide 31 or 33, which is a linear waveguide) as the output light Lout. [Other variations]

[0121] The layout of the circuit elements included in the optical integrated circuit 20 of the light source units 1,1A to 1H described above may be changed as appropriate, and some circuit elements may be omitted. For example, in the above embodiment, the filter section 30 was configured as a ring resonator, but the filter section 30 may be configured to transmit light of a specific wavelength between the semiconductor optical amplifier 10 and the mirror section (first mirror 42 and second mirror 43). For example, the filter section 30 may be configured as a filter other than a ring resonator, such as a Bragg grating filter or an asymmetric Mach-Zehnder interferometer.

[0122] Furthermore, in light source unit 1, phase shifters 38, 71, and 72 may be omitted as appropriate. Conversely, light source units 1A to 1H may be equipped with phase shifters 38, 71, and 72, similar to light source unit 1.

[0123] Furthermore, light source units 1C to 1H, which are configured to output output light Lout (light of a wavelength set by the filter unit 30) to other circuits on the substrate 21, may be configured to output output light Lout from the end face 10b of the semiconductor optical amplifier 10, similar to light source unit 1. Alternatively, if it is not necessary to output output light Lout from the end face 10b of the semiconductor optical amplifier 10, a reflective film may be provided on the end face 10b. In addition, the output light Lout extraction configurations shown for light source units 1, 1A to 1H may be combined as appropriate. That is, output light Lout may be extracted from multiple locations as described in the above embodiment and each of its modifications.

[0124] Furthermore, in the above embodiment, some of the optical elements (e.g., detector 44, power monitor 46, reflected light monitor 48, etc.) arranged on the substrate 21 may be arranged in locations other than on the substrate 21 (e.g., the side of the substrate 21, the outside of the substrate 21, etc.). [Explanation of symbols]

[0125] 1,1A~1H…Light source unit, 10…Semiconductor optical amplifier, 30…Filter section (wavelength adjustment section), 31G,33H…Waveguide (linear waveguide), 32,34…Ring waveguide, 40,40A,40B…Michelson interferometer, 41…Coupler, 41a…First port, 41b…Second port, 41c…Third port, 41d…Fourth port, 42,42C…First mirror, 43,43D…Second mirror, 44…Detector, 46…Power monitor (first measuring instrument), 47…Coupler (second Coupler), 47a...Upstream port (5th port), 47b...Upstream port (8th port), 47c...Downstream port (6th port), 47d...Downstream port (7th port), 48...Reflected light monitor (2nd measuring instrument), 51...Waveguide (1st waveguide), 52...Waveguide (2nd waveguide), IL...Interfering light (1st interfering light), Lout...Output light, Lb1...1st branched light, Lb2...2nd branched light, Lr1...1st reflected light, Lr2...2nd reflected light, RL...Reflected light (2nd interfering light), R...Ripple.

Claims

1. Semiconductor optical amplifier and A wavelength adjustment unit is provided on the substrate, optically coupled to the semiconductor optical amplifier, and changes the wavelength of the output light. A Michelson interferometer is optically coupled to the end of the wavelength adjustment section opposite to the end coupled to the semiconductor optical amplifier, The system comprises a detector connected to the Michelson interferometer, The aforementioned Michelson interferometer is A coupler having a first port, a second port, a third port, and a fourth port, which splits the light from the wavelength adjustment unit input to the first port into a first waveguide connected to the second port and a second waveguide connected to the third port and having a different optical path length from the first waveguide, at a predetermined branching ratio. A first mirror connected to the first waveguide and configured to return at least a portion of the first branched light output from the second port to the first waveguide back to the coupler, The device includes a second mirror connected to the second waveguide and configured to return at least a portion of the second branched light output from the third port to the second waveguide back to the coupler, The detector is connected to the fourth port and detects the first interference light of the first return light returned from the first mirror to the second port and the second return light returned from the second mirror to the third port. The coupler is configured such that, in response to the input of the first and second return beams, a second interference beam is output from the first port to the wavelength adjustment unit. The aforementioned branching ratio is set to an unequal ratio in the light source unit.

2. The light source unit according to claim 1, wherein the branching ratio is set such that the ratio of the power of the first branched light to the power of the input light to the first port is 0.75 or more.

3. The light source unit according to claim 1, wherein the branching ratio is set such that the ratio of the power of the first branched light to the power of the input light to the first port is 0.85 or more.

4. The light source unit according to claim 1, wherein the branching ratio is set such that the ratio of the power of the first branched light to the power of the input light to the first port is 0.90 or more.

5. The light source unit according to claim 1, wherein the branching ratio is set such that the ratio of the power of the first branched light to the power of the input light to the first port is 0.99 or more.

6. The light source unit according to claim 1, further comprising a first measuring instrument disposed downstream of the wavelength adjustment unit, where the semiconductor optical amplifier is located relative to the wavelength adjustment unit, and which measures the power of the light that has passed through the wavelength adjustment unit, with respect to the wavelength adjustment unit, the side on which the semiconductor optical amplifier is located being considered the upstream side.

7. The light source unit according to claim 1, further comprising a second measuring instrument disposed between the first port of the coupler and the wavelength adjustment unit for measuring the power of the second interference light.

8. A second coupler having a fifth port, which is positioned between the first port of the coupler and the wavelength adjustment unit and into which light from the wavelength adjustment unit is input, a sixth port, a seventh port, and an eighth port connected to the first port, A first measuring instrument connected to the seventh port measures the power of light that has passed through the wavelength adjustment section, The system further comprises a second measuring instrument connected to the eighth port for measuring the power of the second interference light, The light source unit according to claim 1, wherein the second coupler is configured to split the light input from the fifth port to the sixth port and the seventh port, and to split the light input from the sixth port to the fifth port and the eighth port.

9. The light source unit according to claim 1, wherein the branching ratio is set such that the ripple of the second interference light is 10 dB or less.

10. The light source unit according to claim 1, wherein the branching ratio is set such that the ripple of the second interference light is 6 dB or less.

11. The light source unit according to claim 1, wherein the branching ratio is set such that the ripple of the second interference light is 2 dB or less.

12. The light source unit according to claim 1, wherein the branching ratio is set such that the ripple of the second interference light is 0.18 dB or less.

13. The light source unit according to claim 1, wherein the coupler is composed of a directional coupler.

14. The light source unit according to claim 1, wherein the semiconductor optical amplifier is configured to output the output light from the end opposite to the end coupled to the wavelength adjustment unit.

15. The branching ratio of the coupler is set such that the power of the first branched light is greater than the power of the second branched light. The light source unit according to claim 1, wherein a portion of the first branched light is extracted as the output light.

16. The branching ratio of the coupler is set such that the power of the first branched light is greater than the power of the second branched light. The light source unit according to claim 1, wherein a portion of the second branched light is extracted as the output light.

17. The light source unit according to claim 1, wherein, between the semiconductor optical amplifier and the wavelength adjustment unit, a portion of the light traveling from the semiconductor optical amplifier to the wavelength adjustment unit or the light returning from the wavelength adjustment unit to the semiconductor optical amplifier is extracted as the output light.

18. The wavelength adjustment section is composed of a ring resonator having a ring waveguide and a linear waveguide optically coupled to the ring waveguide. The light source unit according to claim 1, wherein the output light is extracted from the linear waveguide.

Images