JP5050003B2 - Light modulator - Google Patents

Description

  The present invention belongs to the field of optical modulators that are high-speed, low in driving voltage, low in bias voltage, and good in production yield.

An optical waveguide and a traveling wave electrode are provided on a substrate having a so-called electro-optic effect (hereinafter, the lithium niobate substrate is abbreviated as an LN substrate) such as lithium niobate (LiNbO ₃ ) whose refractive index is changed by applying an electric field. The formed traveling wave electrode type lithium niobate optical modulator (hereinafter abbreviated as LN optical modulator) is applied to a 2.5 Gbit / s, 10 Gbit / s large capacity optical transmission system because of its excellent chirping characteristics. Yes. Recently, application to an ultra large capacity optical transmission system of 40 Gbit / s is also being studied, and it is expected as a key device.

(First prior art)
In recent 40 Gbit / s, a phase modulation method such as a type using a z-cut substrate and a DQPSK or DPSK using an x-cut substrate (or a y-cut substrate) is being used.

  Consider DQPSK using a z-cut substrate as an example of an LN optical modulator. FIG. 12 shows the structure of a DQPSK LN optical modulator disclosed in Patent Document 1. As can be seen from FIG. 12, the DQPSK LN optical modulator has a structure in which two small Mach-Zehnders (or child Mach-Zehnders or child MZs) are incorporated in a large Mach-Zehnder (or parent Mach-Zehnder or parent MZ). Called).

Here, 1 is a z-cut LN substrate, 3 is an optical waveguide, 3a, 3b, 3c, and 3d are optical waveguides of an interaction portion where a high-frequency electrical signal and light interact (more precisely, an interaction for a high-frequency electrical signal) 4a, 4b, 4c, and 4d are the center conductors of traveling-wave electrodes that allow high-frequency electrical signals to propagate and allow high-frequency electrical signals and light to interact with each other. 5a, 5b, 5c, 5d, and 5e are ground conductors of traveling wave electrodes, and 6 is a bias unit (or π / 2 shift unit, phase shift unit, or parent) that shifts the phase of light as parent MZ by π / 2. 3e and 3f are optical waveguides of the π / 2 shift part (or an interaction optical waveguide for bias voltage). In order to simplify the description, the SiO ₂ buffer layer and the Si conductive layer for suppressing temperature drift deposited on the z-cut LN substrate 1 are omitted.

  Note that in an LN optical modulator using an actual z-cut LN substrate 1, since traveling wave electrodes and bias electrodes are formed above the optical waveguide, the optical waveguide cannot be seen from above these electrodes. In order to clarify the numbers assigned to the electrodes and the optical waveguide, in the present specification, a diagram is drawn so that the optical waveguide can be seen through the electrodes. Further, in order to simplify the explanation, in this specification, a traveling wave electrode is omitted in some drawings.

  FIG. 13 shows actual electrodes for the π / 2 shift unit of the DQPSK LN optical modulator of FIG. Here, 7a is a central conductor (or a bias voltage central conductor) of the π / 2 shift portion, and 7b is a ground conductor (or a bias voltage ground conductor) of the π / 2 shift portion. As described above, in the prior art, in order to realize the π / 2 shift as the parent MZ of the LN optical modulator for DQPSK, the optical waveguides 3e and 3f of the π / 2 shift portion immediately after the child MZ are used. .

  Normally, the length of the optical waveguides 3e and 3f of the π / 2 shift portion immediately after the child MZ (the length in which the applied bias and light interact) is as short as about 10 mm, and is necessary for the π / 2 shift. The voltage is as high as about 20V. In actual optical communication, there is a limit to the bias voltage that can be supplied, so there has been a problem of reliability over a long period of 20 years or more.

(Second prior art)
FIG. 14 shows electrodes for the π / 2 shift portion of the second prior art. Reference numerals 8a and 8b denote a central conductor and a ground conductor of the π / 2 shift portion, respectively. As can be seen from this figure, also in the second prior art, the central conductor (or the central conductor for bias voltage) 8a and the ground conductor (or the ground conductor for bias voltage) 8b of the π / 2 shift portion are The optical waveguides 3e and 3f, which are formed immediately after the child MZ and have a relatively short length of about 10 mm, are formed. Compared to the first prior art, in the second prior art, although the voltage required to shift the phase of the light emitted from the two child MZs by π / 2 can be reduced, it is still about 13V. There is also a problem in long-term reliability when used in an actual optical communication system.

  In addition, when the area | region which a high frequency electric signal propagates in the 2nd prior art shown in FIG. 14 is drawn in more detail, it will become like FIG. As described above, the center conductors 4a, 4b, 4c, and 4d of the traveling wave electrode that propagate the high-frequency electric signal and the ground conductors 5a, 5b, 5c, 5d, and 5e are formed up to the edge of the z-cut LN substrate 1. ing.

  In addition, I will explain a little more quantitatively for later discussion. FIG. 16 shows a simplified version of FIG. In order to make the drawing easier to see, the ground conductors 5a, 5b, 5c, 5d, and 5e for high-frequency electric signals are omitted. In addition, the central conductor and the ground conductor of the interaction unit for bias voltage are simplified as 8a ′ and 8b ′, respectively.

In FIG. 16, among the two child MZs, the high-frequency electric signal is such that the phase shift ΔΦ ₁ ′ of each light propagating through the optical waveguides 3a and 3b of the high-frequency electric signal interaction part of the upper child MZ is π. The voltages of the central conductors 4a and 4b of the interaction unit for use are defined as V _a ′ and V _b ′ (potential difference ΔV ₁ ′ = V _a ′ −V _b ′). In general, one of V _a ′ and V _b ′ is set to 0V. For example, when V _b ′ = 0V, ΔV ₁ ′ = V _a ′.

Similarly, the center of the high-frequency electrical signal interaction unit is set so that the phase shift ΔΦ ₂ ′ of each light propagating through the optical waveguides 3c and 3d of the high-frequency electrical signal interaction unit of the lower child MZ is π. The voltages of the conductors 4c and 4d are set as V _c ′ and V _d ′ (potential difference ΔV ₂ ′ = V _d ′ −V _c ′). In general, one of the voltages V _c ′ and V _d ′ is set to 0 V. Therefore, when V _c ′ = 0 V, ΔV ₂ ′ = V _d ′.

In this way, the light propagating through the child MZ in FIG. 16 can obtain a desired ΔV ₁ ′ and ΔV ₂ ′ so that the phase shift becomes π with respect to the interactive optical waveguides constituting each child MZ. The voltage V _a ′ of the center conductor 4 a and the voltage V _d ′ of the center conductor 4 d are applied to the traveling wave electrode of the high-frequency electrical signal interaction unit so that it can. Here, for the sake of simplicity, the voltage V _b ′ of the center conductor 4 _b and the voltage V _c ′ of the center conductor 4 _c are set to 0 V (that is, the voltage falls to the ground).

Then, the voltage of the bias voltage central conductor is set so that the phase shift ΔΦ ₃ ′ of each light propagating through the bias voltage interaction optical waveguides 3e and 3f of the π / 2 shift unit as the parent MZ is π / 2. The value of V _A ′ and the voltage V _B ′ of the bias voltage ground conductor is determined. Here, since V _B ′ = 0 normally, ΔV ₃ ′ = V _A ′ −V _B ′ = V _A ′. Since the length of the bias voltage interaction optical waveguides 3e and 3f of the π / 2 shift portion as the parent MZ is as short as about 10 mm, a desired light phase shift (or phase difference) ΔΦ ₃ ′ (= π / The potential difference ΔV ₃ ′ for obtaining 2) has increased.

(Third prior art)
FIG. 17 shows the third prior art shown for the DQPSK type transmitting apparatus proposed in Patent Document 2. In FIG. In FIG. 17, in addition to the LN optical modulator, the DQPSK signal source 700 that generates a high-frequency electrical signal, the LD 500 that emits light, the ABC circuit 100 that stabilizes the bias, and the phase difference of light between the two child MZs is π. A bias supply unit 120-2 that supplies a bias voltage for shifting, a phase shift control unit 110 that issues a phase shift control command, a bias supply unit 120-3 that supplies a bias voltage for π / 2 shift, etc. Electrical parts, electronic parts, and control circuits are also included, and FIG. 17 can be said to be a DQPSK transmitter.

  And in this prior art, the device for reducing the bias voltage necessary for the phase of the light between the child MZs to shift by π / 2 is devised. Hereinafter, the configuration of the conventional technology will be described.

  Here, 200 and 300 are child MZs, 200b and 300b are modulation electrodes (traveling wave electrodes), 200a and 300a are Y branch optical waveguides of the child MZ, and 200c and 300c are bias electrodes of the π shift unit. 500 is a light source composed of an LD (laser diode), 400 and 600 are parent MZ Y-branch optical waveguides, 700 is a DQPSK signal source, 800-1 and 800-2 are drivers, 900 is a PD (photodiode), 100 is ABC (Auto-bias control) circuit, 110 is a phase shift control unit, 120-1 and 120-2 are so-called π shift bias supplies for setting the phase shift between optical waveguides constituting each child MZ to π. The reference numeral 120-3 denotes a bias supply unit for π / 2 (or 2π / 4) shift between the child MZs (or between the parent MZs).

  In the third prior art, for example, in claim 1 of Patent Document 2, “the voltage for transfer in which the two signal lights combined in the multiplexing waveguide substantially have a phase shift difference of 2π / 4”. Is also applied together with the driving voltage signal ”, and the high-frequency electrical signal from the driver 800-2 is applied to the LN optical modulator in various places such as paragraphs [0030] and [0039]. Prior to application, the high-frequency electrical signal and the bias voltage from the bias supply unit 120-3 for π / 2 shift are superimposed by a bias T of 130-2, and this superimposed signal (or synthesized signal) is DQPSK type. It is described as being applied to an LN optical modulator. That is, in this prior art, the bias T130-2 outside the LN optical modulator uses the bias voltage supplied from the bias supply unit 120-3 for π / 2 shift between the high-frequency electrical signal from the driver 800-2 and the child MZ. It plays an important role of overlapping.

  However, in the third prior art in which an electrical signal for forming a π / 2 phase shift between the child MZs is superimposed on a high-frequency electrical signal and supplied to the LN optical modulator, two electrical signals are superimposed. Therefore, a new device such as a bias T is required, and the structure becomes complicated. Further, when the bias T is used, a high-frequency electric signal is attenuated and a waveform is sometimes deteriorated.

(Fourth prior art)
FIG. 18 shows the fourth prior art shown for the DQPSK type transmitting apparatus proposed in Patent Document 2. In FIG. In FIG. 18, the modulation electrodes (traveling wave electrodes) 200b and 300b apply only a high-frequency electric signal to the light, and have a function of providing a phase shift of π / 2 between the child MZs 200 and 300. Not. Therefore, as shown in FIG. 17, the bias T130-2 used when applying a bias for generating a phase shift of π / 2 using the traveling wave electrode, which is a feature in Patent Document 2, is unnecessary. It has become.

  Note that the bias voltage supplied from the ABC circuit 100 to shift the phase of light in each of the child MZs 200 and 300 is a phase shift π / 2 between the child MZs 200 and 300 (as can be seen from FIG. 18, Bias generated by the bias supply units 120-41 and 120-42 based on the command detected and sent by the phase shift control unit 110 to give π / 4-(− π / 4) = π / 2) Along with the voltage, the bias electrodes 200c and 300c are applied.

  In other words, in the fourth prior art, phase shifts π / in the positions of two parallel optical waveguides that can be used as interaction portions between high-frequency electrical signals and light in each of the child MZs 200 and 300. 2 is provided with bias electrodes 200c and 300c. The phase shift of π / 2 is completed in the two parallel optical waveguides.

  As a result, when the lengths of the bias electrodes 200c and 300c for the phase shift π / 2 that generate the phase shift π / 2 between the child MZs 200 and 300 are increased, the traveling wave electrodes 200b and 300b are shortened. As a result, the drive voltage as a high-frequency electrical signal increases. Conversely, if the length of the bias electrodes 200c and 300c is shortened in order to reduce the drive voltage as a high-frequency electric signal, the bias voltage for giving a phase shift π / 2 between the child MZs 200 and 300 increases. As a result, there is a problem of affecting long-term reliability as an optical modulator.

JP 2007-208472 A JP 2008-122786 A

  As described above, in the first and second prior arts, in order to form a phase shift π / 2 between the child MZs, only the parent MZ linear optical waveguide provided immediately after the child MZ is biased. Is applied, the bias voltage for setting the phase shift between the child MZs is high. In addition, since there is a limit to the bias voltage that can be supplied to the transponder used in the actual optical communication system, there has been a problem of reliability in the long-term use.

  Alternatively, in the third prior art, a bias voltage for forming a phase shift π / 2 between the high-frequency electric signal and the child MZ is superimposed outside the LN optical modulator, and the superimposed electric signal (or combined electric signal) Signal) to the LN optical modulator. A new device called a bias T is required to superimpose the high-frequency electric signal and the bias voltage, and the structure is complicated. Further, when the bias T is used, problems such as attenuation of the high-frequency electric signal or deterioration of the waveform occur.

  Furthermore, in the fourth prior art, a bias voltage for forming a phase shift π / 2 between the child MZs may be used as an interactive optical waveguide in which high-frequency electrical signals and light interact in the child MZ. A parallel linear optical waveguide is provided, and a phase shift of π / 2 is completed in the two parallel optical waveguides. Therefore, if the bias electrode for phase shift π / 2 between the child MZs is lengthened, the drive voltage of the high frequency electrical signal increases, and conversely, in order to reduce the drive voltage of the high frequency electrical signal, When the bias electrode of this is shortened, there is a problem that the bias voltage for forming the phase shift π / 2 increases.

In order to solve the above problems, an optical modulator according to claim 1 of the present invention includes a substrate having an electro-optic effect, an optical waveguide for guiding light formed on the substrate, and one of the substrates. A traveling wave electrode formed on the surface side and made up of a central conductor and a ground conductor for applying a high frequency electric signal for modulating the light; and an electric circuit electrically connected to the traveling wave electrode for attenuating the high frequency electric signal The optical waveguide is a nested optical waveguide having a child Mach-Zehnder optical waveguide on a branched optical waveguide of the parent Mach-Zehnder optical waveguide, and the electrical termination is DC include an electrical resistance which is not grounded, along said parent Mach-Zehnder optical waveguide Y-branch arms, and the bias electrode is formed, the child Mach-Zehnder Phase shift between the light between emitted from the waveguide to occur, with a bias voltage is applied to the bias electrode, said bias voltage to the center conductor of the traveling wave electrode through the electrical terminations It is characterized by being applied.

According to a second aspect of the present invention, there is provided an optical modulator comprising: a substrate having an electro-optic effect; an optical waveguide for guiding light formed on the substrate; and the optical modulator formed on one surface side of the substrate. A traveling wave electrode composed of a central conductor and a ground conductor for applying a high frequency electrical signal that modulates, and an electrical termination electrically connected to the traveling wave electrode for attenuating the high frequency electrical signal, The optical waveguide is an optical modulator composed of a nested optical waveguide having a child Mach-Zehnder optical waveguide on a branched optical waveguide of the parent Mach-Zehnder optical waveguide, and the electrical termination has an electrical resistance not grounded in DC. wherein, along said from Y branch arms of the parent Mach-Zehnder optical waveguide in the optical waveguide to the Y branch arms of the child Mach-Zehnder optical waveguide, the bias electrode shape Are, as the phase shift between the light between emitted from the child Mach-Zehnder optical waveguide is produced, with a bias voltage is applied to the bias electrode, wherein the traveling wave electrode through the electrical termination It said bias voltage is applied to the central conductor and said and Turkey.

An optical modulator according to a third aspect of the present invention is the optical modulator according to the first or second aspect , wherein the bias electrode is formed before and after the light is guided across the child Mach-Zehnder optical waveguide. The center conductors of the bias electrodes on both sides of each child Mach-Zehnder optical waveguide are electrically connected.

  In the present invention, unlike the first and second prior arts, for example, a bias for forming a phase shift between child MZs such as π / 2 is applied between the electrical resistance of the electrical termination and the capacitor. By electrically connecting the traveling wave electrodes, the child MZ is also applied to the traveling wave electrode having a long length. Thus, the substantial interaction length of the optical waveguide to which the bias voltage necessary to obtain the desired phase shift π / 2 is applied without using the bias T to superimpose the high-frequency electric signal and the DC bias voltage. Can be significantly lengthened. The voltage required to form the phase shift π / 2 is exactly inversely proportional to the substantial interaction length for this bias voltage, so it is necessary to form the phase shift π / 2 between the child MZs. The bias voltage can be significantly reduced. What is important in the present invention is that, unlike the third prior art, a bias voltage necessary to form a phase shift between the child MZs using the bias T and a high-frequency electric signal are superimposed, The superimposed signal is not applied to the traveling wave electrode. In general, in order to superimpose another electric signal or DC bias on a high-frequency electric signal, a special device such as a bias T is required, and the configuration of a transmission apparatus using an LN optical modulator becomes complicated. Furthermore, using such a special device not only increases the cost, but also inserts the device between the high-frequency electrical signal and the LN optical modulator, which attenuates the high-frequency electrical signal or degrades its waveform. End up. Therefore, in the present invention, a bias voltage that forms a phase shift between child MZs such as π / 2 is superimposed on a high-frequency electrical signal outside the LN optical modulator, and the superimposed voltage (or combined voltage) is LN light. The structure which does not apply to a modulator is taken. That is, in the present invention, the bias voltage is supplied to the optical waveguide completely independently of the high-frequency electric signal. Thus, in the present invention, since the bias T is not used in the present invention, the configuration is extremely simple. Furthermore, since the high frequency electrical signal can be almost completely attenuated by the electrical resistance in the electrical termination, by applying the present invention, the bias voltage for the phase shift π / 2 is defined as the high frequency electrical signal. It can be applied independently to the LN optical modulator independently of each other. This eliminates the need for a device such as a bias T that superimposes a bias voltage that forms a phase shift between the high-frequency electrical signal and the child MZ, which not only significantly reduces costs but also attenuates the high-frequency electrical signal. Neither does the waveform deteriorate. As a result, it is possible to improve long-term reliability in an actual optical communication system, to save power as a transmission device, and to improve the quality of the light waveform.

  The present invention also includes an embodiment in which a bias voltage is also applied to the optical waveguide of the Y branch arm (or simply the Y branch arm) constituting the parent MZ. And in those embodiments, not only the interaction portion between the high-frequency electrical signal and the light, but also the Y branch arm that multiplexes or demultiplexes the light of the parent MZ, the phase shift is generated, and as a result of combining them, The phase shift π / 2 as the parent MZ is generated. Therefore, not only the interaction part where the high-frequency electrical signal of child MZ interacts with light, but also the Y branch arm that multiplexes or demultiplexes the light of parent MZ as an interaction part for phase shift π / 2. Therefore, the substantial interaction length for obtaining the phase shift π / 2 due to the bias voltage is significantly increased, and the required bias voltage can be significantly reduced. Then, by making the length of the bias electrode as the parent MZ longer than the length of the bias electrode as the child MZ, the excellent advantage of the present invention is exhibited.

Schematic top view of the first embodiment of the optical modulator of the present invention. Partial enlarged view of the first embodiment of the optical modulator of the present invention. Top view of an optical waveguide for explaining the principle of the present invention Schematic top view for explaining the principle of the first embodiment of the present invention Schematic top view for explaining the principle of the first embodiment of the present invention Schematic top view for explaining the principle of the first embodiment of the present invention Schematic top view for explaining the principle of the first embodiment of the present invention The conceptual top view simplified in order to demonstrate the principle about 1st Embodiment of this invention Schematic top view of the second embodiment of the optical modulator of the present invention. Schematic top view of the third embodiment of the optical modulator of the present invention. Schematic top view of the fourth embodiment of the optical modulator of the present invention. Top view of the first prior art Top view of the first prior art Top view of second prior art Top view of second prior art Top view for explaining the operation of the second prior art Top view of third conventional DQPSK transmitter Top view of the fourth conventional DQPSK transmitter

  Hereinafter, embodiments of the present invention will be described. However, since the same reference numerals as those of the conventional embodiments shown in FIGS. 12 to 17 correspond to the same function units, description of the function units having the same numbers is omitted here. To do.

(First embodiment)
FIG. 1 shows a top view of a first embodiment in which the present invention is applied to a DQPSK type optical modulator. For reference, FIG. 2 shows a partially enlarged view of FIG. FIG. 3 shows the configuration of the optical waveguide.

1 to 3, reference numeral 1 denotes a z-cut LN substrate, 3 denotes an optical waveguide, 3a, 3b, 3c, and 3d denote optical waveguides of interaction portions where high-frequency electrical signals interact with light, 4a, 4b, and 4c. Reference numeral 4d denotes a central conductor of a traveling wave electrode that propagates a high-frequency electrical signal and causes the high-frequency electrical signal to interact with light. Here, in order to simplify the explanation, the input side of the high frequency electric signal is omitted. Further, the ground conductor, the SiO ₂ buffer layer deposited on the z-cut LN substrate 1, and the Si conductive layer for suppressing temperature drift are also omitted. Needless to say, a substrate other than the z-cut, such as an x-cut or a y-cut, may be used as the substrate.

  Here, the child MZ including the interaction optical waveguides 3a and 3b is referred to as a first child MZ, and the child MZ including the interaction optical waveguides 3c and 3d is referred to as a second child MZ.

  The parent MZ Y branch arms 3e ′, 3f ′, 3g ′, 3h ′ and the child MZ Y branch arms 3i ′, 3j ′, 3k ′, 3m ′ of FIG. Bias voltage center conductors (or bias electrodes) 20a, 20b, 20c, and 20d for shifting the phase are formed. 20a and 20b are electrically connected by an electric wiring 21a, and 20c and 20d are electrically connected by an electric wiring 21b. It is connected. Reference numerals 10a, 10b, 10c, and 10d are electrical terminations for high-frequency electrical signals.

  Here, the electrode pad 20e electrically connected to the bias electrodes 20a, 20b, and 21a and the electrical terminal 10a are electrically connected by, for example, a wire 22a. The electrode pad 20f electrically connected to the bias electrodes 20c, 20d, and 21b and the electrical terminal end 10c are electrically connected by, for example, a wire 22b. Here, 22c and 22d are wires for supplying a voltage for obtaining a phase shift as a parent MZ to the electrode pads 20e and 20f, respectively. Reference numerals 4a ', 4b', 4c ', and 4d' denote electrode pads that are electrically connected to the central conductors 4a, 4b, 4c, and 4d, respectively. Reference numeral 40 denotes an electrode wiring.

When a bias voltage _VA for causing a phase shift as the parent MZ is applied from the wire 22c to the electrode pad 20e, the voltage _VA is applied to the bias electrodes 20a and 20b electrically connected to the electrode pad 20e. The light refractive index change (phase change) occurs in the Y branch arms 3i ′ and 3j ′ of the first child MZ and the Y branch arms 3e ′ and 3f ′ of the parent MZ.

As will be described in detail later, as an aspect of the present invention, the electrode pad 20e and the electrical terminal 10a are electrically connected by, for example, a wire 22a. Further, the electrical terminal 10a and the electrode pad 4a 'of the central conductor 4a of the traveling wave electrode are electrically connected by, for example, a wire 22e. Therefore, since the bias voltage V _A is also applied to the central conductor 4a constituting the traveling wave electrode of the first child MZ, the phase shift π also occurs in the interaction optical waveguide 3a in which the high-frequency electrical signal and light interact. A refractive index change for / 2 occurs.

Further, when a bias voltage V _B for causing a phase shift as the parent MZ is applied from the wire 22d to the electrode pad 20f, the voltage V _B is applied to the bias electrodes 20c and 20d electrically connected to the electrode pad 20f. When applied, the refractive index change (light phase change) of the optical waveguide occurs in the Y branch arms 3k ′, 3m ′ of the second child MZ and the Y branch arms 3g ′, 3h ′ of the parent MZ.

In the same manner as described above, the electrode pad 20f and the electrical terminal 10c are electrically connected by, for example, a wire 22b. Further, the electrical terminal 10c and the electrode pad 4c 'of the central conductor 4c of the traveling wave electrode are electrically connected by, for example, a wire 22g. Therefore, since the bias voltage V _B is also applied to the central conductor 4c constituting the traveling wave electrode of the second child MZ, the phase shift π also occurs in the interaction optical waveguide 3c in which the high-frequency electrical signal and light interact. A refractive index change for / 2 occurs.

  In FIG. 1, the width of the electrode wiring 40 is broadly exaggerated for easy understanding, but in actuality, the width of each electrode wiring 40 in FIG. 1 is 10 to 50 μm. The length occupied by the total width in the longitudinal direction of the optical waveguide is almost negligible. Furthermore, in this embodiment, in order to obtain the phase shift π / 2 as the parent MZ, the optical waveguides 3a, 3b, 3c, and 3d of the child MZ and the Y branch arms 3i ′, 3j ′, 3k ′, 3m ′, and the parent MZ The change in refractive index of the Y branch arms 3e ′, 3f ′, 3g ′, 3h ′ is used.

  In this embodiment, a bias voltage for causing a phase shift as the parent MZ is superimposed on the high-frequency electric signal in advance by the bias T, and the superimposed signal is not applied to the LN optical modulator. That is, the bias voltage is applied to the LN optical modulator independently of the high frequency electric signal without being superimposed on the high frequency electric signal. To that end, the present invention utilizes 10a, 10b, 10c and 10d electrical terminations.

  First, consider the Y branch arms 3f ′ and 3e ′ of the parent MZ, the Y branch arms 3j ′ and 3i ′ of the first child MZ, and the interaction optical waveguides 3a and 3b in FIG.

In order to cause a phase shift as the parent MZ, the connection of the wires 22a and 22c is included for the electrical termination 10a to which the electrode pad 20e to which the bias voltage _VA is applied is connected via the wire 22c. As shown in FIG. Here, 12 is an electrical resistance, 13a, 13b and 13c are capacitors, 4a 'is an electrode pad connected to the center conductor 4a, and 5a' and 5a '' are electrode pads connected to a ground conductor (not shown) corresponding to the center conductor 4a. , 22e electrically connects the electrode pad 4a 'and the electrical resistor 12, for example, and wires 22e' and 22e "electrically connect the electrode pads 5a ', 5a" and the capacitors 13a, 13b, for example. It is a wire. As can be seen from FIG. 4, in the present invention, the wire 22c for applying the voltage of the parent MZ is electrically connected to the electrode pad 4a 'connected to the center conductor 4a of the first child MZ via the termination resistor 10a.

Since the electrode pad 4a 'and the electrical resistor 12 are electrically connected by the wire 22e, the bias voltage _VA applied to the electrode pad 20e through the wire 22c is an optical waveguide which is an interaction part for bias voltage. Not only applied to 3e 'and 3f' but also to the central conductor 4a of the traveling wave electrode, it is applied to the Y branch arms 3i 'and 3j' of the first child MZ and the interaction optical waveguide 3a of the high frequency electrical signal interaction unit. Also works. In FIG. 4, the high-frequency electrical signal that has propagated from the center conductor 4a to the electrode pad 4a 'has already been consumed by the electrical resistor 12 of the electrical termination 10a. That is, in the present invention, the bias voltage V _A can be supplied independently to the high frequency electric signal.

  Here are some things to note. The electrical terminal 10a is actually composed of a center conductor (not shown) and a ground conductor (not shown), like the traveling wave electrode of the LN optical modulator. An electric resistance 12 is provided between the center conductor (not shown) and the ground conductor (not shown). In the electric circuit diagram shown in FIG. 4, the wire 22a is connected directly below the electrical resistor 12, because the wire 22a is connected to a ground conductor (not shown) of the electrical terminal 10a. Now, if the connection point of the wire 22a is a central conductor (not shown) of the electrical terminal 10a, and is positioned so as to be past the electrical resistor 12 from the viewpoint of geometrical arrangement for the high-frequency electrical signal. The influence of the connection of the wire 22a to the center conductor (not shown) on the high-frequency electric signal can be made small enough to be ignored. In this case, the wire 22a is connected immediately above the electrical resistor 12 in the electrical circuit diagram of FIG.

  Note that the electric resistor 12 in FIG. 4 is not grounded in a direct current because of the presence of the capacitors 13a, 13b, and 13c.

As described above, in the present invention including this embodiment, the superimposed signal obtained by superimposing the bias voltage V _A and the high-frequency electrical signal for causing the phase shift as the parent MZ in advance using the bias T is the traveling wave electrode. The central conductor 4b is not applied.

On the other hand, the bias voltage _Vb of the first child MZ is applied to the electrode pad 4b 'and the electrical terminal 10b which belong to the first child MZ and to which the central conductor 4b which is a pair of the central conductor 4a is electrically connected. 5 including the wire 30e. Here, 12 'is an electrical resistance, 13a', 13b 'and 13c' are capacitors, 4b 'is an electrode pad connected to the center conductor 4b, and 5b' and 5b "are ground conductors (not shown) corresponding to the center conductor 4b. The electrode pads 22f are electrically connected to the electrode pads 4b 'and the electrical resistors 12', for example, wires, and 22f 'and 22f "are electrode pads 5b' and 5b" and capacitors 13a 'and 13b'. For example, a wire. Further, as described in FIG. 4, the wire 30e may be connected directly above the electrical resistor 12 'in the electric circuit diagram of FIG. Note that the electrical resistor 12 'in FIG. 5 is not grounded in a direct current manner because it has capacitors 13a', 13b ', and 13c', respectively.

  Next, consider the Y branch arms 3h ′ and 3g ′ of the parent MZ, the Y branch arms 3m ′ and 3k ′ of the second child MZ, and the interaction optical waveguides 3c and 3d in FIG.

To produce a phase shift as a parent MZ, the electrode pads 20f for applying a bias voltage V _B through a wire 22d is electrically connected to electrical termination 10c, including the state of the wire 22b, 22d of the connection As shown in FIG. Here, 14 is an electrical resistance, 15a, 15b and 15c are capacitors, 4c ′ is an electrode pad connected to the center conductor 4c, 5c ′ and 5c ″ are electrode pads of a ground conductor (not shown) corresponding to the center conductor 4c, 22g is, for example, a wire that electrically connects the electrode pad 4c 'and the electrical resistor 14, and 22g', 22g "is, for example, a wire that electrically connects the electrode pads 5c ', 5c" and the capacitors 14a, 14b. It is. As can be seen from FIG. 6, in the present invention, the wire 22d for applying the voltage of the parent MZ is electrically connected to the electrode pad 4c ′ connected to the center conductor 4c of the second child MZ via the termination resistor 10c. Since 4c ′ and the electrical resistance 14 are electrically connected by the wire 22g, the bias voltage V _B applied to the electrode pad 20f via the wire 22d is the optical waveguide 3h ′ which is an interaction portion for bias voltage. And 3g ′, and also applied to the Y branch arms 3k ′ and 3m ′ of the second child MZ and the central conductor 4c of the traveling wave electrode, and also to the interaction optical waveguide 3c of the high frequency electrical signal interaction unit. To do. In FIG. 6, the high-frequency electrical signal that has propagated from the central conductor 4c to the electrode pad 4c ′ is consumed by the electrical resistance 14 of the electrical termination 10c. That is, in the present invention, the bias voltage V _B is supplied independently to the high frequency electric signal. As described in FIG. 4, the wire 22 b may be connected immediately above the electrical resistor 14 in the electrical circuit diagram of FIG. 6.

  6 is not grounded in terms of direct current due to the presence of the capacitors 15a, 15b, and 15c.

As described above, in the present invention, the superimposed signal obtained by superimposing the bias voltage V _B and the high-frequency electric signal for causing the phase shift as the parent MZ in advance using the bias T is applied to the central conductor 4c of the traveling wave electrode. Not applying.

Further, the center conductor 4d is electrically connected to the electrode pads 4d' electrically terminating 10d in the pair of center conductors 4c, including the wires 30f to bias voltage V _d for the second child MZ is applied, As shown in FIG. Here, 14 'is an electrical resistance, 15a', 15b 'and 15c' are capacitors, 4d 'is an electrode pad connected to the center conductor 4d, and 5d' and 5d "are ground conductors (not shown) corresponding to the center conductor 4d. The electrode pad 30f is electrically connected to the electrode pad 4d 'and the electrical resistor 14', for example, a wire, and 22h 'and 22h "are electrode pads 5d' and 5d" and capacitors 15a 'and 15b'. For example, a wire. Further, as described in FIG. 5, the wire 30 f may be connected immediately above the electrical resistor 14 ′ in the electric circuit diagram of FIG. 7.

  It should be noted that the electrical resistor 14 ′ in FIG. 7 is not grounded in terms of direct current because there are capacitors 15 a ′, 15 b ′, and 15 c ′.

  FIG. 8 is a diagram schematically showing FIG. 1 for easy understanding of the principle of the present invention. 23a is a simulated wire indicating that the bias electrode 20a and the central conductor 4a are electrically connected, and 23b is a simulated wire indicating that the bias electrode 20c and the central conductor 4c are electrically connected. It is. The bias electrodes 20a and 20b and the center conductor 4a have substantially the same potential, and the bias electrodes 20c and 20d and the center conductor 4c have substantially the same potential.

  Next, the operation principle of the first embodiment will be described with reference to FIGS. The following control process is used for bias control of the DQPSKLN optical modulator.

a) First Control Step This step is a step of giving a phase shift π / 2 to the parent MZ. When the bias voltage _VA is applied to the bias electrode 20a through the wire 22c, the bias electrode 20b and the central conductor 4a are also at substantially the same potential V _A as described above. Then, the refractive index of the interaction optical waveguide 3a that forms the first child MZ with the Y branch arms 3e 'and 3f' of the parent MZ and the Y branch arms 3i 'and 3j' of the first child MZ and the first child MZ, that is, the optical waveguide The phase changes.

On the other hand, when through the wire 22d applies a bias voltage _{V B} to the bias electrode 20c, the bias electrode 20d, even more central conductor 4c is almost the same potential _{V B.} Then, the refractive index of the interaction optical waveguide 3c that constitutes the second child MZ with the Y branch arms 3g ′ and 3h ′ of the parent MZ and the Y branch arms 3k ′ and 3m ′ of the second child MZ and the second child MZ, that is, the optical waveguide The phase changes.

At this time, if the signs of the bias voltages V _A and V _B are different, the phase shift Δφ ₃ as the parent MZ is caused by the bias voltages V _A and V _B (potential difference ΔV ₃ = V _A −V _B ) having small absolute values. It can be set to π / 2.

b) Second control step This step is a step of giving a phase shift π to the two optical waveguides constituting the first child MZ. First child MZ center conductor 4b on the bias voltage V _b 2 pieces of interaction optical waveguides 3a constituting the first child MZ by applying the through wire 30e, the phase shift [Delta] [phi ₁ between 3b It can be set to π. The potential difference ΔV ₁ between the central conductors 4 a and 4 b in the interaction portion where the high-frequency electrical signal of the first child MZ and light interact is V _A −V _b .

c) Third control step This step is a step of giving a phase shift π to the two optical waveguides constituting the second child MZ. Two interaction optical waveguides 3c constituting the second child MZ by applying a bias voltage V _d to the center conductor 4d of the second child MZ through the wire 30f, the phase shift [Delta] [phi ₂ between the 3d It can be set to π. The potential difference ΔV ₂ between the central conductors 4c and 4d in the interaction portion where the high-frequency electrical signal of the second child MZ and light interact is V _B −V _d .

d) Performing a fourth control step the second control step and the third control step (i.e. [Delta] [phi ₁ [pi, the [Delta] V ₁ and [Delta] V ₂ sets the V _b and V _d to the a [Delta] [phi ₂ [pi The phase shift Δφ ₃ as the parent MZ set in the first control step deviates from π / 2. Then, although a 1st control process is performed again, the advantage of this invention can be exhibited as follows.

The phase shift [Delta] [phi ₃ as parent MZ to the [pi / 2, a bias voltage applied to the wire 22c from V _A of the first control step to V _{A ',} the bias voltage applied to the wire 22d first control step From V _B to V _B ′. As a potential difference between the result biasing electrodes 20a and 20c are changed from ΔV ₃ = _V A -V _B to _{ΔV 3 '= V A'-} V B '. Here, what is important is the length of the bias electrode that causes the phase shift Δφ ₃ as the parent MZ (the length of the electrodes 20a, 20b, 20c, and 20d formed on the Y branch arm of the parent MZ and the center conductor of the child MZ). 4a, the sum of the lengths of 4c) is longer than the central conductor 20b, 20d of the length of the for obtaining the phase shift [Delta] [phi ₁ and the phase shift [Delta] [phi ₂ as child MZ.

Therefore, the change ΔV ₃ '-ΔV ₃ newly bias voltage required to phase shift [Delta] [phi ₃ and [pi / 2 as parent MZ, to the [Delta] V ₁ and [Delta] V ₂ required as child MZ, Child MZ The length of the bias electrode (that is, the length of the central conductors 4b and 4d here) and the length of the bias electrode as the parent MZ can be reduced. Here, when this ratio is R, R <1 in the present invention.

Now, since the phase shift Δφ ₃ as the parent MZ is again set to π / 2, the phase shifts Δφ ₁ and Δφ ₂ of the child MZ shift from π. Therefore, ΔV ₁ and ΔV ₂ are changed so that Δφ ₁ and Δφ ₂ become π. As a result, the optimum bias voltage as the parent MZ changes again.

However, as can be easily understood, the required amount of change in the bias voltage of the parent MZ depends on the change in the bias voltages ΔV ₁ and ΔV ₂ of the child MZ and the length of the bias electrode as the child MZ and the length of the bias electrode as the parent MZ. It is sufficient to change only the voltage multiplied by the ratio R. Thus, since the bias electrode ratio R is smaller than 1, the voltages required for the child MZ and the parent MZ rapidly converge with the power of the length ratio R of the bias electrode for the child MZ and the parent MZ. . This is an instantaneous task for electronic circuits. On the other hand, since the optimum bias voltage drift phenomenon in the LN optical modulator is extremely slow in time, the use of the present invention causes no delay in bias voltage control.

  In addition, it cannot be overemphasized that said control process may change order. Further, after performing the second control step and the third control step, the phase difference between the light beams emitted from the first child MZ and the second child MZ affects the phase difference π / 2 of the parent MZ. In the case where a bias voltage is applied to the first child MZ and the second child MZ so as not to give the error, the fourth step can be omitted.

  The lengths of the central conductors 4a, 4b, 4c, and 4d are 40 to 70 mm. On the other hand, the length of the parent MZ is about 2 to 4 times that including the Y branch arm. Therefore, as compared with the third prior art shown in FIG. 17 and the fourth prior art shown in FIG. 18, the bias for realizing the phase shift π / 2 of the parent MZ by using this embodiment. Not only can the voltage be significantly reduced, but the optimum bias voltage can be set instantaneously by utilizing the fact that the ratio R of the lengths of the bias electrodes is smaller than 1.

  In addition, bias electrodes 20a, 20b, 20c, and 20d for realizing the phase shift π / 2 as the parent MZ are arranged before and after the central conductors 4a, 4b, 4c, and 4d for high-frequency electric signals, and they are electrically connected. However, only 20a and 20c, 20b and 20d, or only one of them may be arranged. And this can be said for all embodiments of the invention. At least one of the bias electrodes 20a, 20b, 20c, and 20d as the parent MZ may be disposed.

  In the above description, the embodiment in which the idea of the present invention is built in the chip of the LN modulator has been described. The key point is that a bias voltage for the parent MZ of the nested LN optical modulator, for example, is applied from the electrical terminal, and “the length of the bias electrode for generating the phase shift π / 2 as the parent MZ is determined by the child. The bias voltage required for the parent MZ and the child MZ is converged to a required value by setting the length longer than the length of the bias electrode in the child MZ for obtaining the phase shift (for example, π) as the MZ. That is. All embodiments that use this concept belong to the present invention.

  That is, even if the phase shift bias electrode as the parent MZ formed on the parent MZ arm and the traveling wave electrode of the child MZ are not electrically connected in the LN optical modulator chip, the LN optical modulation is performed. From the viewpoint of electrical length, the phase shift as parent MZ is obtained by electrically connecting the bias terminals of the parent MZ and the child MZ to the LN optical modulator module manufactured by incorporating the chip of the detector in the housing. 2 can be set longer than the length of the bias electrode in the child MZ for obtaining a phase shift (for example, π) as the child MZ ”. It can be realized. This concept can be applied to all embodiments except the third embodiment of the present invention.

(Second Embodiment)
FIG. 9 is a top view of the second embodiment of the present invention. In this embodiment, as can be seen by comparing the bias electrodes 20a, 20b, 20c, and 20d shown in FIG. 1 with 20a ′, 20b ′, 20c ′, and 20d ′ shown in FIG. 9, FIG. 9 shows the Y branch arm of the child MZ. An electrode is not formed on the substrate.

(Third embodiment)
FIG. 10 is a top view of the third embodiment of the present invention. In the present embodiment, in order to realize the phase shift π / 2 as the parent MZ, only the child MZ interaction optical waveguides 3a, 3b, 3c, and 3d shown in FIG. 3 are used without using the Y branch arm of the parent MZ. Is used. Also in the present invention, the bias voltage is applied to the central conductors 4a, 4c, 4b, 4d independently from the wires 22c 'and 22d' via the electrical terminations 10a, 10b, 10c, 10d without using the bias T. Is applied.

  Compared with the first embodiment shown in FIG. 1, the present embodiment has a shorter interaction length for generating the phase shift π / 2, so that the bias voltage necessary for generating the phase shift π / 2 is required. However, unlike the third prior art shown in FIG. 17, the cost reduction and the deterioration of the modulation waveform that can be obtained because the bias T is not used can be suppressed. Has the advantage.

  In order to operate the third embodiment of the present invention, first, a bias voltage is applied to the center conductors 4a and 4b, 4c and 4d, and the phase difference of the light as the first child MZ and the second child MZ. Realize π. Next, the phase difference π / 2 as the parent MZ may be realized by applying a bias voltage to at least one of the pair of center conductors 4a and 4b and the pair of 4c and 4d.

(Fourth embodiment)
FIG. 11 is a top view of the fourth embodiment of the present invention. In the figure, 50 is a domain-inverted region. In this embodiment, the bias electrodes 20a, 20b, 20c, and 20d are electrically connected by the electrode wiring 21b. 20f ″ is an electrode pad connected to these bias electrodes. For example, when a bias voltage is applied to the wire 22d ′, this bias voltage is applied not only to the bias electrodes 20a, 20b, 20c and 20d but also to the center conductors 4a and 4c of the child MZ. To be applied. In addition, in the application of a bias voltage as the parent MZ, a push-pull bias operation can be performed even if one bias power source is used.

(About each embodiment)
In the above description, DQPSK has been described as an example. However, the present invention can be applied to a structure having more than two child MZs, and is not a π / 2 shift, but an mπ / n shift where m and n are integers, and mπ / n + α. (Α is an arbitrary constant) This can also be applied to a shift.

  Further, although the CPW electrode has been described as an example of the traveling wave electrode, it goes without saying that various traveling wave electrodes such as an asymmetric coplanar strip (ACPS) and a symmetric coplanar strip (CPS), or a lumped constant electrode may be used.

  In the embodiment of the present invention, the bias electrode of the π / 2 shift unit has been described with a structure in which a ground conductor is provided on both sides of the center conductor, or a center conductor is provided on both sides of the ground conductor. It is only an example used for doing, and it is not restricted to this. The present invention can be applied to any bias electrode structure.

  Further, the bias voltage for π / 2 shift is applied to the traveling wave electrode for high-frequency electrical signals, but the bias separation type is used to shift the phase between the two optical waveguides constituting the child MZ by π. It goes without saying that the optical modulator adopting the above may be applied to the separated bias portion, the traveling wave electrode, or both of them.

  In the above embodiments, the z-cut substrate has been described, but the x-cut, y-cut or z-cut plane orientation, that is, the x-axis of the crystal in the direction perpendicular to the substrate surface (cut plane). It is applicable to a substrate having a y-axis or a z-axis, and the plane orientation in each of the embodiments described above may be used as a main plane orientation, and other plane orientations may be mixed as sub-plane orientations. . It goes without saying that not only the LN substrate but also other substrates such as lithium tantalate and semiconductors may be used.

1: z-cut LN substrate 3: optical waveguides 3a, 3b, 3c, 3d: high-frequency electric signal interaction optical waveguides 3e ', 3f', 3g ', 3h': parent MZ Y branch arm 3e, 3f: bias Voltage-interacting optical waveguides 3i ', 3j', 3k ', 3m': Child MZ Y-branch arms 4a, 4b, 4c, 4d: Center conductors of high-frequency electrical signal traveling wave electrodes 4a ', 4b', 4c ' 4d ′, 5a ′, 5a ″, 5b ′, 5b ″, 5c ′, 5c ″, 5d ′, 5d ″, 20e, 20f: electrode pads 5a, 5b, 5c, 5d, 5e: high frequency electricity Signal traveling wave electrode ground conductor 6: π / 2 shift portion 7a, 8a, 8a ′, 20a, 20a ′, 20b, 20b ′, 20c, 20c ′, 20d, 20d ′: bias voltage central conductors 21a, 21b 40: Electric wiring 7b, 8 8b ': Ground conductor for bias voltage 10a, 10b, 10c, 10d: Electrical termination 11, 22a, 22b, 22c, 22c', 22d, 22d ', 22e, 22e', 22e ", 22f, 22f ', 22f ″, 22g, 22g ′, 22g ″, 22h ′, 22h ″, 23a, 23b, 30e, 30f: Wire 12, 12 ′, 14, 14 ′: Electrical resistance 13a, 13b, 13c, 13a ′ , 13b ′, 13c ′, 15a, 15b, 15c, 15a ′, 15b ′, 15c ′: capacitor 50: polarization inversion region 100: ABC circuit 110: phase shift control unit 120-1, 120-2, 120-3, 120-41, 120-42: Bias supply unit 200, 300: Child MZ
200a, 300a: Y branch optical waveguide of child MZ 200b, 300b: Modulation electrode (traveling wave electrode)
200c, 300c: bias electrode 500: LD (laser diode)
400, 600: Parent MZ Y-branch optical waveguide 800-1, 800-2: Driver 700: DQPSK signal source 900: Photodiode (PD)

Claims (3)

A substrate having an electro-optic effect; an optical waveguide for guiding light formed on the substrate; a central conductor formed on one side of the substrate for applying a high-frequency electrical signal that modulates the light; A traveling wave electrode composed of a ground conductor; and an electrical termination electrically connected to the traveling wave electrode for attenuating the high frequency electrical signal;
The optical waveguide is an optical modulator composed of a nested optical waveguide each having a child Mach-Zehnder optical waveguide on a branched optical waveguide of a parent Mach-Zehnder optical waveguide,
The electrical termination includes an electrical resistance that is not DC grounded;
A bias electrode is formed along the Y branch arm of the parent Mach-Zehnder optical waveguide,
As the phase shift between the light between emitted from the child Mach-Zehnder optical waveguide is produced, with a bias voltage is applied to the bias electrode, the said central conductor of said traveling wave electrode through the electrical termination An optical modulator to which a bias voltage is applied. A substrate having an electro-optic effect; an optical waveguide for guiding light formed on the substrate; a central conductor formed on one side of the substrate for applying a high-frequency electrical signal that modulates the light; A traveling wave electrode composed of a ground conductor; and an electrical termination electrically connected to the traveling wave electrode for attenuating the high frequency electrical signal;
The optical waveguide is an optical modulator composed of a nested optical waveguide each having a child Mach-Zehnder optical waveguide on a branched optical waveguide of a parent Mach-Zehnder optical waveguide,
The electrical termination includes an electrical resistance that is not DC grounded;
A bias electrode is formed along the optical waveguide from the Y-branch arm of the parent Mach-Zehnder optical waveguide to the Y-branch arm of the child Mach-Zehnder optical waveguide ,
As the phase shift between the light between emitted from the child Mach-Zehnder optical waveguide is produced, with a bias voltage is applied to the bias electrode, the said central conductor of said traveling wave electrode through the electrical termination optical modulator characterized and Turkey bias voltage is applied. The bias electrode is formed before and after the light is guided across the Child Mach-Zehnder optical waveguide,
3. The optical modulator according to claim 1 , wherein central conductors of the bias electrodes on both sides of each child Mach-Zehnder optical waveguide are electrically connected .

Images