Optical transmitter and bias control method

Abstract

To provide an optical transmitter with suppressed wavelength dependence of insertion loss, and a bias control method.SOLUTION: An optical transmitter includes: a Mach-Zehnder type optical modulator using InP-based material; a bias control unit that controls a DC bias applied to the optical modulator; a monitor unit that monitors output light of the optical modulator and generates a monitor signal; and a correction unit that corrects the gain of the monitor signal according to wavelength in the orientation of compensating the wavelength dependence of insertion loss of the optical modulator, at a stage before the bias control by the bias control unit.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] This disclosure relates to an optical transmitter and a bias control method. [Background technology]

[0002] To meet the recent demands for faster and higher-capacity communications, multi-level modulation using external modulators has been adopted. For a long time, core networks have used lithium niobate (LN) modulators. In recent years, however, the performance of InP-based modulators, which use compact InP-based materials, has improved compared to LN modulators, and they are now entering the commercialization stage. Examples of InP-based materials include INGaAsP and InAlGaAs, and modulators are formed using thick films or multiple quantum well structures of these materials.

[0003] In LN modulators and Mach-Zehnder modulators (MZMs) using InP-based materials, the DC bias that defines the operating point drifts due to temperature changes, aging, etc. Automatic bias control (ABC) is used to maintain the DC bias at the operating point of the optical modulator. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2021-67933 [Patent Document 2] Japanese Patent Publication No. 2009-265283 [Patent Document 3] Japanese Patent Publication No. 2012-257164 [Overview of the project] [Problems that the invention aims to solve]

[0005] In InP-based optical modulators, the application of an electric field changes the band gap, which alters the optical absorption rate and thus the speed of light propagation, i.e., the phase. In other words, the optical absorption characteristics of InP-based optical modulators are wavelength-dependent. Meanwhile, the transmission bandwidth of optical communications is being broadened, and the operating bandwidth of optical modulators is also being expanded. In the high-frequency range, it is difficult to broaden the bandwidth while maintaining the power characteristics, and bandwidth compensation is generally performed. Bandwidth compensation is a process that flattens the wavelength-to-power characteristics in the high-frequency range by reducing the gain in the low-frequency range to flatten the gain across the entire bandwidth. This bandwidth compensation not only reduces the output level, but also causes variations in loss depending on the amount of compensation. Loss variations due to high-frequency bandwidth compensation have not been a problem in LN modulators. However, in optical modulators using InP-based materials, the loss variations due to high-frequency bandwidth compensation, combined with the wavelength dependence of the optical absorption characteristics, result in a wavelength dependence of insertion loss that cannot be ignored. The wavelength dependence of insertion loss in InP-based optical modulators also affects ABC control, making it difficult to maintain the DC bias at the correct operating point.

[0006] One objective of this disclosure is to provide an optical transmitter and a bias control method that suppress the wavelength dependence of insertion loss. [Means for solving the problem]

[0007] In the embodiment, the optical transmitter is A Mach-Zehnder type optical modulator using InP-based materials, A bias control unit for controlling the DC bias applied to the optical modulator, A monitoring unit that monitors the output light of the optical modulator and generates a monitor signal, Prior to the bias control by the bias control unit, a correction unit corrects the gain of the monitor signal in a direction that compensates for the wavelength dependence of the insertion loss of the optical modulator, according to the wavelength. It is equipped with. [Effects of the Invention]

[0008] An optical transmitter and a bias control method that suppress the wavelength dependence of insertion loss are implemented. [BRIEF DESCRIPTION OF THE DRAWINGS]

[0009] [Figure 1] It is a diagram showing the wavelength dependence of the insertion loss of an MZM. [Figure 2] It is a schematic diagram of an optical transceiver module using the optical transmitter according to the embodiment. [Figure 3] It is a schematic diagram of the optical transmitter according to the embodiment. [Figure 4] It is a diagram showing the voltage versus optical power characteristics of an MZM. [Figure 5] It is a diagram showing an example of channel correspondence information. [Figure 6] It is a flowchart of bias control according to the embodiment. [Figure 7] It is a flowchart of wavelength dependence correction. [Figure 8] It is a diagram showing an example of correction value determination. [Figure 9] It is a diagram showing a correction example for compensating the wavelength dependence of insertion loss. [Figure 10] It is a schematic diagram of a modified example of the optical transmitter. [Figure 11] It is a schematic diagram of another modified example of the optical transmitter. [MODE FOR CARRYING OUT THE INVENTION]

[0010] Before describing the configuration and the bias control method of the optical transmitter according to the embodiment, the wavelength dependence of insertion loss, which is a problem in an optical modulator using an InP-based material, will be described with reference to Fig. 1. In Fig. 1, the horizontal axis represents wavelength, and the vertical axis represents the relative value (%) of the light detection level with respect to the center wavelength. The insertion loss that occurs when a carrier wave is modulated by an InP-based optical modulator varies greatly depending on the wavelength. In the wavelength range from 1528 nm to 1566 nm including the C-band, assuming that the light detection level at the center wavelength of 1547 nm is 100%, at 1528 nm on the short wavelength side, the light detection level is +10% relative to the center wavelength, while at 1565 nm on the long wavelength side, it is -30% relative to the center wavelength, resulting in a large insertion loss.

[0011] An insertion loss of -30% affects the accuracy or convergence time of ABC control to the optimal bias point. The characteristics shown in Figure 1 are an example of measurements from a single sample, but insertion loss can vary between elements or between manufacturing lots. Considering variations between elements or manufacturing lots, the wavelength-dependent bias in insertion loss is likely to be even greater.

[0012] In the embodiment, the gain of the monitor value used for bias control is adjusted according to the wavelength to cancel out the wavelength dependence of the insertion loss of the MZM. The following embodiments are examples for realizing the technical concept of this disclosure and do not limit the scope of the disclosure. The size, positional relationships, etc., of the components shown in each drawing may be exaggerated to facilitate understanding of the invention. The same components or functions may be given the same name or reference numeral, and redundant explanations may be omitted.

[0013] <Configuration of optical transceiver module and optical transmitter> Figure 2 is a schematic diagram of an optical transceiver module 1 having an optical transmitter 10 according to an embodiment, and Figure 3 is a schematic diagram of the optical transmitter 10. Solid arrows indicate electrical signals, and thick lines indicate the optical path. The optical transceiver module 1 is a digital coherent optical transceiver that combines coherent reception and digital signal processing. The optical transceiver module 1 includes a digital signal processor (DSP) 2, a light source (indicated as "LD" in the figure) 5, an optical transmitter 10, and an optical receiver 30. The optical transmitter 10 includes an optical modulator 13 and a microprocessor 20 that controls the DC bias of the optical modulator 13. The optical modulator 13 is a modulator using an InP-based material.

[0014] The optical transceiver module 1 is connected to optical fibers 6 and 7. The optical transmitter 10 is connected to optical fiber 6 and transmits an optical signal to the transmission path. The optical receiver is connected to optical fiber 7 and receives an optical signal from the transmission path. Light emitted from the light source 5 is supplied to the optical transmitter 10 and the optical receiver 30. The light supplied to the optical transmitter 10 is incident on the optical modulator 13 as a carrier wave, and the optical signal modulated by the optical modulator 13 is output to the optical fiber 6. The light supplied to the optical receiver 30 is used as a local light emitter to detect the optical signal received from optical fiber 7.

[0015] Referring to Figure 3, the optical transmitter 10 includes a digital-to-analog converter (DAC) 12, an optical modulator 13, a bias control unit 200 that controls the DC bias applied to the optical modulator 13, and a monitor unit 19 that monitors the output light of the optical modulator 13. The bias control unit 200 is implemented by the functions of a microprocessor 20. The monitor unit 19 includes a photodetector (PD) 14, a transimpedance amplifier (TIA) 15, a bandpass filter (BPF) 16, and an electrical amplifier 17. The PD 14 detects the output light from the optical modulator 13. The TIA 15 converts the photocurrent output from the PD 14 into a voltage signal. The BPF 16 passes through a predetermined frequency variation component contained in the voltage signal output from the TIA 15. The electrical amplifier 17 amplifies the frequency variation component extracted by the BPF 16.

[0016] The optical modulator 13 is a Mach-Zehnder type modulator, with two child MZM131 and 132 connected in parallel to form a parent MZM133. A 90-degree phase difference is given between the light passing through child MZM131 and the light passing through child MZM132. One of the child MZM131 and 132 acts as an in-phase (I) channel, and the other acts as an orthogonal (Q) channel, forming a phase-modulated IP modulator.

[0017] Electrodes 135, 136, and 137 are provided on the child MZM131 and 132, and the parent MZM133, respectively. Data signals generated by DSP2 are input as high-speed drive signals to the high-frequency (RF) terminals of electrodes 135 and 136. Light incident from the light source 5 into the I channel and Q channel of the optical modulator 13 is modulated by the data signals in each channel. The light modulated by child MZM131 and the light modulated by child MZM132 are combined with a 90-degree phase difference from each other and output from the optical modulator 13 as an optical signal with four phase states.

[0018] A DC bias is applied to the DC terminals of electrodes 135, 136, and 137. A bias voltage V1 is applied to the DC terminal of electrode 135 of child MZM131, and a bias voltage V2 is applied to the DC terminal of electrode 136 of child MZM132. A bias voltage V3 is applied to the DC terminal of electrode 137 of parent MZM133. The bias voltages V1, V2, and V3 are controlled to the appropriate operating point by the bias control unit 200. In the optical modulator 13 that modulates the optical phase, the bias voltages V1 and V2 of child MZM131 and 132 are controlled to the Null point where the output power of the optical modulator is minimized when there is no data signal input. The bias voltage V3 of parent MZM133 is set to give a 90-degree phase difference between the light passing through child MZM131 and the light passing through child MZM132.

[0019] Figure 4 shows the voltage-to-optical power characteristic of the MZM. This voltage-to-optical power characteristic is also called the extinction curve, with the horizontal axis representing the applied bias voltage and the vertical axis representing the output power of the MZM. The bias voltages V1 and V2 are set so that, when there is no data signal input, the phases of the light propagating through the two arms of the MZM are 180 degrees out of phase. By giving a 180-degree phase difference between the two arms, when the light passing through the two arms is combined, they cancel each other out, and the output of the MZM is minimized. The point where the output power of the MZM is minimized is called the Null point. When the light passing through the two arms is in the same phase, when they are combined, they reinforce each other, and the output of the MZM is maximized (peak). The driving voltage required to change the output of the MZM from minimum to maximum is called the half-wavelength voltage Vπ.

[0020] The InP-based optical modulator 13 exhibits wavelength dependence, and especially when the transmission bandwidth is widened, the relationship between voltage and phase is not necessarily linear. Therefore, the bias voltage of the optical modulator 13 is adjusted so that the half-wavelength voltage Vπ remains as constant as possible across the bandwidth, but the wavelength dependence of insertion loss still remains. That is, the optical power output from the optical modulator 13 fluctuates depending on the wavelength, and this fluctuation in optical power affects the accuracy and convergence time of ABC control. Therefore, ABC control is performed to cancel out this wavelength dependence of insertion loss.

[0021] The bias control itself uses a low-frequency dither signal, as in conventional methods. A dither signal that varies in frequency from tens of Hz to hundreds of Hz is superimposed on the bias voltages V1, V2, and V3. The superimposed dither signal appears as a minute fluctuation in the amplitude of the high-speed optical signal generated by each MZM. As shown in Figure 4, if the bias voltage is shifted from the operating point (e.g., the Null point) of the optical modulator 13, a dither component that fluctuates at the same frequency as the dither signal is observed in the output light of the optical modulator 13. Depending on the direction of the bias voltage shift, the phase of the observed dither component will be in phase with or out of phase with the dither signal.

[0022] To observe the dither component, a portion of the output light from the optical modulator 13 is branched by the tap 138 and detected by the PD 14. The photocurrent output from the PD 14 is converted into a voltage signal by the TIA 15, and the component that fluctuates at the same frequency as the dither signal is extracted by the BPF 16. The fluctuating component that changes at the same frequency as the dither signal (this is called the "dither component") is amplified by the electric amplifier 17. The gain of the electric amplifier 17 is adjusted by the digital potentiometer (DPOT) 18.

[0023] When the bias voltage of the child MZM is at the null point, i.e., the bottom of the voltage-to-optical power characteristic, the phase of the dither component is inverted 180 degrees around the null point, and a frequency twice the dither frequency is observed. If the bias voltage is shifted from the null point, a dither component with the same frequency as the dither signal is observed, and a phase difference occurs between the dither component and the dither signal depending on the degree of bias shift. ABC control controls the bias voltage to make the phase difference between the dither component and the dither signal zero or minimal.

[0024] However, if the insertion loss of the optical modulator 13 is highly wavelength-dependent, the power of the monitor light detected by PD14 will vary greatly between wavelengths with high and low loss, causing the level of the monitor signal input to the bias control unit 200 to fluctuate with wavelength. Fluctuations in the input level to the bias control unit 200 affect the accuracy and speed of convergence of the ABC control loop, which corrects for phase lead or lag of the optical signal. For example, if a wavelength near 1560 nm is used and the input level to the bias control unit 200 drops significantly, the phase shift of the dither component relative to the dither signal cannot be accurately detected, and the ABC control will not converge. Alternatively, the extinction point in the ABC control loop may not be determined, resulting in a very long convergence time. On the other hand, if a wavelength near 1530 nm is used and the optical power incident on PD14 is too high, there is a concern that a dead zone may occur due to output saturation of the electric amplifier 17, or abnormal oscillation may occur due to excessive control gain.

[0025] To suppress such degradation of bias control, in this embodiment, the gain of the monitor signal generated by the monitor unit 19 is adjusted according to the wavelength before the ABC control. In the configuration example shown in Figure 3, the DPOT 18 of the electric amplifier 17 is used to correct the gain of the monitor signal before the input to the bias control unit 200 in a manner that cancels out the wavelength dependence of the insertion loss of the optical modulator 13.

[0026] The bias control unit 200 includes a dither generation unit 21, a control error calculation unit 22, a bias value calculation unit 23, an analog-to-digital converter (ADC) 24, a wavelength-dependent correction unit 25, and a wavelength channel data acquisition unit 26. The dither generation unit 21 generates a dither signal that is superimposed on the bias voltages V1, V2, and V3. As an example, a 300Hz dither signal is generated. The dither signal, along with the DC bias value of each MZM determined by the bias value calculation unit 23, is supplied to the DAC 12, converted into an analog electrical signal, and applied to the electrodes 135, 136, and 137 of the corresponding MZM. At the start of ABC control, V1, V2, and V3 may be set to predetermined voltage values ​​as initial bias values.

[0027] With the dither signal superimposed, the output light from the optical modulator 13 is monitored by the monitor unit 19. The monitor signal input to the bias control unit 200 is converted into a digital signal by the ADC 24 and input to the control error calculation unit 22. The ADC 24 may be located outside the bias control unit 200. In this case, the digital monitor signal is input to the bias control unit 200.

[0028] The control error calculation unit 22 uses the dither signal generated by the dither generation unit 21 to synchronously detect the dither component included in the monitor signal and calculates the DC bias control error from the phase lead or lag. The calculated control error is supplied to the bias value calculation unit 23 and added to the current bias value, updating the bias value in a direction that reduces the control error. For example, when the phase of the fluctuation (dither) component included in the monitor signal is in phase with the dither signal superimposed on the DC bias, the bias value is updated in a direction that decreases the DC bias toward the Null point (the phase point of π / 2 in the case of the parent MZM). When the phase of the fluctuation (dither) component included in the monitor signal is out of phase with the dither signal superimposed on the DC bias, the bias value is updated in a direction that increases the DC bias toward the Null point (the phase point of π / 2 in the case of the parent MZM).

[0029] The bias voltages V1, V2, and V3, whose bias values ​​have been updated, are applied to the optical modulator 13, and the ABC control loop is repeatedly executed until the control error (phase difference) is minimized. The ABC control of the child MZMs 131 and 132 and the parent MZM 133 may be performed in time division, or a bias control unit 200 may be provided for each MZM and executed simultaneously in parallel.

[0030] A feature of this embodiment is that information on the currently used wavelength channel is supplied to the light source 5 and input to the bias control unit 200. When the wavelength channel data acquisition unit 26 acquires the current wavelength channel information, it supplies the wavelength channel information to the wavelength-dependent correction unit 25. The wavelength-dependent correction unit 25 refers to the channel correspondence information 251 and corrects the gain of the monitor signal in a way that compensates for the insertion loss that occurs according to the current wavelength. For example, the correction value determined by the wavelength-dependent correction unit 25 is supplied to the electric amplifier 17, and the gain of the electric amplifier 17 is adjusted according to the wavelength.

[0031] The channel correspondence information 251 used in the wavelength-dependent correction unit 25 is stored in or in external memory of the microprocessor 20 (see Figure 2). The channel correspondence information may be in the form of a lookup table (LUT) that describes the compensation value corresponding to each channel (wavelength), or it may be a function that defines the compensation value as a function of wavelength.

[0032] Figure 5 shows an example of channel correspondence information 251. This channel correspondence information 251 is stored as a LUT that associates the wavelength setting channel with the insertion loss compensation value. In the optical transmitter 10 in Figure 3, the insertion loss is compensated for using the variable resistor of DPOT 18 of the electric amplifier 17, so the DPOT setting data is described as the insertion loss compensation value. By changing the variable resistor value of DPOT 18 depending on the channel, the gain of the electric amplifier 17 is changed to compensate for the insertion loss that fluctuates with wavelength.

[0033] In the LUT of Figure 5, the DPOT 18 setting value is described for every 50 channels. If the wavelength dependence of the pass loss of each MZM forming the optical modulator 13 can be simulated, the DPOT setting value data may be decimated at predetermined channel intervals. The DPOT setting values ​​of the decimated channels are calculated, for example, by linear interpolation. Different linear interpolation formulas may be used depending on the channel range, i.e., the bandwidth range. For example, the slope of the linear interpolation may be steeper in the region on the longer wavelength side of the bandwidth than in the region on the shorter wavelength side. Instead of the LUT of Figure 5, all channels may be divided into multiple channel groups, and a function formula may be maintained for each channel group.

[0034] In either configuration, the insertion loss of the optical modulator 13, which varies with wavelength, is compensated by the gain of the electric amplifier 17, thereby compensating for the wavelength dependence of the insertion loss before ABC control begins. This suppresses the deterioration of the focusing accuracy of ABC control and the increase in focusing time.

[0035] <Bias control method and correction of wavelength dependence of insertion loss> Figure 6 is a flowchart of the bias control in the embodiment. This control flow is performed during initial startup or when the wavelength is changed. Bias control is performed for each of the child MZM131 and 132, and the parent MZM133. In Figure 6, we focus on the bias control for child MZM131. First, the input of the dither signal to child MZM132 is turned off, and a bias voltage V1 superimposed with the dither signal is applied to the DC terminal of electrode 135 of child MZM131 (S11). Due to the superposition of the dither signal, the output light of the optical modulator 13 contains a dither component that fluctuates at the same frequency as the dither signal. The dither component has a phase difference with respect to the dither signal that corresponds to the deviation of the bias voltage V1 from the null point.

[0036] The monitor unit 19 monitors the output light of the optical modulator 13 and generates a monitor signal (S12). As described above, if the insertion loss of the optical modulator 13 is constant across the transmission bandwidth, the monitor signal output from the electric amplifier 17 may be supplied directly to ABC control. However, in the InP-based optical modulator 13, the insertion loss varies greatly depending on the wavelength, so the wavelength dependence of the insertion loss is corrected in the preceding stage of ABC control (S13). Specifically, the gain of the monitor signal is adjusted according to the wavelength.

[0037] ABC control is performed using the gain-adjusted monitor signal (S14). ABC control controls the bias voltage V1 of interest in a direction that minimizes the phase shift of the dither component from the dither signal. As a result, the DC bias is controlled to the operating point or near it.

[0038] Figure 7 is a flowchart of the wavelength-dependent correction in S13 of Figure 6. The correction of the wavelength dependence of the insertion loss is performed by the wavelength-dependent correction unit 25, i.e., the microprocessor 20. First, wavelength channel setting information is acquired (S21). The wavelength channel setting information may be the wavelength channel selected by the user, or, in the case of wavelength division multiplexing (WDM) optical communication, it may be the current set wavelength of the tunable laser array of the light source 5.

[0039] Based on the current wavelength, a correction value is determined to compensate for the variation in insertion loss (S22). For example, the variable resistance value of DPOT18 or the gain value of the electric amplifier 17 becomes the correction value. The correction value is not limited to the setting of DPOT18; as described later, the gain of TIA15 may be adjusted, or the digital gain may be adjusted in the stage after ADC24.

[0040] Compensate for the wavelength dependence of insertion loss based on the determined correction value (S23). For example, when a wavelength near 1565 nm is used, the variable resistance value of DPOT18 is set to increase the gain of the electric amplifier 17. When a wavelength near 1528 nm is used, the variable resistance value of DPOT18 is set to decrease the gain of the electric amplifier 17. This compensates for the wavelength dependence of insertion loss.

[0041] FIG. 8 shows an example of correction value determination. With the center of the band as a boundary, the correction value for the wavelength dependence of insertion loss is calculated by linear interpolation on the long wavelength side and the short wavelength side. In this example, the correction target is the set value of DPOT18. First, the variable resistance value of DPOT18 is set so that a specified ABC control gain can be obtained at the center wavelength λref (1547 nm in this example). Let the DPOT set value at this time be DPOTref. When the set value of DPOT18 is set to DPOTref, let the light detection level at the center wavelength λref be Pref = 100%.

[0042] For a wavelength λ shorter than the center wavelength λref S , let the light detection level be P S , and for a wavelength λ longer than the center wavelength λref L , let the light detection level be P L . In the wavelength versus light detection level characteristic (or wavelength versus insertion loss characteristic) of FIG. 6, when Pref is 100%, P at 1528 nm S is 112%, and P at 1566 nm L is 69%.

[0043] Correction coefficients based on the center wavelength λref are calculated on the short wavelength side and the long wavelength side, respectively. Correction coefficient for wavelength 1528 nm: Pref / P S = 100 / 112 = 0.89 Correction coefficient for wavelength 1566 nm: Pref / P L = 100 / 69 = 1.45

[0044] Using the above correction coefficients, for the target wavelength λ on the short wavelength side and the long wavelength side respectively T , the set value DPOT(λ of DPOT18 atT ) is interpolated. Short wavelength side: DPOT(λ T ) =DPOTref×[(0.89-1.00) / (λ S -λref)(λ T -λref)+1] =DPOTref×[(0.89-1.00) / (1528-1547)(λ T -1547)+1] Short wavelength side: DPOT(λ T ) =DPOTref×[(1.45-1.00) / (λ L -λref)(λ T -λref)+1] =DPOTref×[(0.89-1.00) / (1566-1547)(λ T -1566)+1]

[0045] In this way, the gain of the monitor signal of the optical modulator 13, which operates at an arbitrary wavelength, can be corrected in a way that cancels out the wavelength dependence of the insertion loss before it is input to the ABC control. This prevents situations where the power of the monitor signal becomes excessive or significantly reduced depending on the wavelength, and allows the convergence accuracy or convergence time of the ABC control to be maintained within an acceptable range.

[0046] Figure 9 shows an example of a correction to compensate for the wavelength dependence of insertion loss. The horizontal axis represents wavelength, the left vertical axis represents the maximum error detection amount, and the right vertical axis represents the amplification factor of the electrical amplifier 17. The amplification factor at the center wavelength λref is set to 1. Black circles represent the error amount, and triangle marks represent the gain (amplification factor) of the electrical amplifier 17. Square marks represent the center wavelength λref and the short wavelength (λ S The error amounts are measured at three points: ), and at a long wavelength (λL). The dotted lines are fitting lines obtained by linear interpolation on the short wavelength side and the long wavelength side, respectively, based on the three measurement points.

[0047] The measured error amount and the fitting line show good agreement. The DPOT setting value is determined such that the maximum error detection amount remains almost constant across the entire bandwidth, with the gain of the electrical amplifier 17 at the center wavelength λref set to 1. The gain characteristic of the electrical amplifier 17 at this time decreases nonlinearly toward longer wavelengths, as indicated by the triangular mark. As a result, in the pre-ABC control stage, the gain of the monitor signal representing the dither component is corrected in a way that eliminates the wavelength dependence of the insertion loss.

[0048] <Variation> Figure 10 is a schematic diagram of a modified optical transmitter 10A. In the optical transmitter 10A, the gain of the TIA 15 is adjusted before ABC control to correct the wavelength dependence of insertion loss. The wavelength dependence correction unit 25 of the bias control unit 200A has channel correspondence information 251A, which includes a TIA gain setting value corresponding to each channel as a correction value. The TIA gain setting values ​​may be described corresponding to channels at predetermined intervals, as shown in Figure 5. The TIA gain between channels can be calculated by linear interpolation. Based on the correction value determined for each wavelength by the wavelength dependence correction unit 25, the amplification factor of the TIA 15 is adjusted, and a voltage signal with the wavelength dependence of insertion loss corrected across the transmission bandwidth is input to the BPS 16.

[0049] Since the photocurrent of PD14 input to TIA15 is minute, it is desirable to take noise countermeasures, such as minimizing the electrical wiring on the printed circuit board connecting the output of PD14 to the input of TIA15. The other configurations of the optical transmitter 10A and bias control unit 200A are as described with reference to Figure 3. Even with the configuration of the optical transmitter 10A, the wavelength dependence of the insertion loss of the optical modulator 13 can be corrected in the preceding stage of ABC control to suppress deterioration of the focusing accuracy of ABC control or an increase in the focusing time.

[0050] Figure 11 is a schematic diagram of the optical transmitter 10B as another modified example. The bias control unit 200B of the optical transmitter 10B has a digital gain adjustment unit 27 connected to the output of the ADC 24. By digitally adjusting the gain of the monitor signal, the wavelength dependence of the insertion loss is corrected in the preceding stage of ABC control. The wavelength dependence correction unit 25 has channel correspondence information 251B in which the corresponding digital gain setting value (coefficient) for each channel is described. The digital gain setting values ​​may be described corresponding to channels at predetermined intervals, as shown in Figure 5. The digital gain values ​​between channels can be linearly interpolated. Based on the digital gain coefficient determined for each wavelength by the wavelength dependence correction unit 25, the digital monitor signal is amplified. The digital monitor signal, with the wavelength dependence of the insertion loss corrected across the transmission bandwidth, is input to the control error calculation unit 22.

[0051] Gain adjustment via digital processing is straightforward. When using the configuration shown in Figure 11, it is desirable to set the analog circuit gain of the electric amplifier 17 to the maximum level that does not saturate at the input to the ADC 24. Limiting the analog circuit gain of the electric amplifier 17 slightly restricts the dynamic range of the ADC 24, but by optimizing the signal-to-noise characteristics, the wavelength dependence of the insertion loss can be compensated for, thereby suppressing deterioration of the convergence accuracy or increase in the convergence time of ABC control.

[0052] Although this disclosure has been described above based on a specific configuration example, this disclosure is not limited to the above-described configuration example. The light source 5 may be provided inside the optical transmitter 10. Each correction value described in the channel correspondence information 251 may be described as a coefficient for adjusting the amplification factor of the monitor signal. The frequency of the dither signal can be any frequency, such as 100Hz, 200Hz, etc., as long as it is sufficiently lower than the analog drive signal that drives the optical modulator 13. When the microprocessor 20 implements the functions of the bias control unit 200 corresponding to the child MZMs 131, 132, and the parent MZM 133, dither signals of different frequencies may be applied to the child MZMs 131, 132, and the parent MZM 133. In that case as well, by adjusting the gain of the monitor signal according to the wavelength in the stage preceding the DC bias control of the MZM of interest, a monitor signal with the wavelength dependence of insertion loss corrected over the transmission bandwidth can be supplied to the ABC control.

[0053] As the modulation method for the optical modulator 13, quadrature phase shift keying (QPSK) or quadrature amplitude modulation (QAM) may be used. Alternatively, two parent MZM133s may be connected in parallel to perform dual polarization (DP) QPSK modulation. In this case, ABC control is performed for both the optical modulator for X polarization and the optical modulator for Y polarization, and prior to the ABC control, the gain of the monitor signal of the output light of each polarization optical modulator is adjusted according to the wavelength. [Explanation of Symbols]

[0054] 1 Optical transceiver module 2 DSP 5 light source 6, 7 Optical fiber 10, 10A, 10B Optical Transmitters 12 DAC 13 Optical modulator 131, 132 Child MZM 133 Parent MZM 135, 136, 137 electrode 14 PD 15 TIA 17 Electric Amplifier 18 DPOT 19. Monitor section 20 microprocessors 21 Dither generation unit 22 Control Error Calculation Unit 23. Bias Value Calculation Unit 24 ADC 25 Wavelength dependence correction section 26 Wavelength channel data acquisition unit 27 Digital Gain Adjustment Section 200, 200A, 200B Bias Control Unit Channel compatibility information for 251, 251A, and 251B

Claims

1. A Mach-Zehnder type optical modulator using an InP-based material for modulating incoming light, A bias control unit that controls the DC bias applied to the optical modulator based on a monitor signal, A monitoring unit that monitors the output light of the optical modulator and generates the monitoring signal, Prior to the bias control by the bias control unit, a correction unit corrects the gain of the monitor signal based on the wavelength of the input light, compensating for the dependence of the insertion loss of the optical modulator on the wavelength, and correcting the gain at a first wavelength to be greater than the gain at a second wavelength shorter than the first wavelength. A light transmitter.

2. The monitoring unit includes a photodetector that detects a portion of the output light of the optical modulator, a transimpedance amplifier that converts the photocurrent obtained by the photodetector into a voltage signal, a bandpass filter that extracts a predetermined frequency component included in the voltage signal output from the transimpedance amplifier, and an electric amplifier that amplifies the predetermined frequency component extracted by the bandpass filter. The correction unit corrects the gain of the electric amplifier based on the wavelength. The optical transmitter according to claim 1.

3. The monitoring unit includes a photodetector that detects a portion of the output light of the optical modulator, a transimpedance amplifier that converts the photocurrent obtained by the photodetector into a voltage signal, a bandpass filter that extracts a predetermined frequency component included in the voltage signal output from the transimpedance amplifier, and an electric amplifier that amplifies the predetermined frequency component extracted by the bandpass filter. The correction unit corrects the gain of the transimpedance amplifier based on the wavelength. The optical transmitter according to claim 1.

4. The bias control unit has a digital gain adjustment unit that digitally adjusts the gain of the monitor signal. The correction unit corrects the digital gain of the digital gain adjustment unit based on the wavelength. The optical transmitter according to claim 1.

5. The bias control unit has a wavelength channel data acquisition unit that acquires the current wavelength channel setting information, The correction unit determines a correction value to adjust the gain of the monitor signal based on the setting information acquired by the wavelength channel data acquisition unit. The optical transmitter according to any one of claims 1 to 4.

6. The bias control unit superimposes a dither signal onto the DC bias applied to the optical modulator and controls the voltage value of the DC bias in a direction that reduces the phase error between the dither component included in the gain-corrected monitor signal and the dither signal. The optical transmitter according to any one of claims 1 to 4.

7. A dither signal is superimposed on the DC bias of a Mach-Zehnder type optical modulator using an InP-based material that modulates the input light. The output light of the optical modulator is monitored, and a monitor signal is generated that includes a dither component that fluctuates at the same frequency as the dither signal. Based on the wavelength of the input light, the gain of the monitor signal is corrected to compensate for the dependence of the insertion loss of the optical modulator on the wavelength, and to make the gain at the first wavelength greater than the gain at the second wavelength which is shorter than the first wavelength. The bias of the optical modulator is controlled using the corrected monitor signal. Bias control method.

8. A portion of the output light from the optical modulator is detected by a photodetector, and the gain of a transimpedance amplifier that converts the photocurrent output from the photodetector into a voltage signal is corrected based on the wavelength. The bias control method according to claim 7.

9. A dither component that fluctuates at the same frequency as the dither signal is extracted from the output light of the optical modulator, and the gain of the electric amplifier that amplifies the dither component is corrected based on the wavelength. The bias control method according to claim 7.

10. The monitor signal is digitally converted, and the digital gain of the digital monitor signal is corrected based on the wavelength. The bias control method according to claim 7.

Images