Open Access
Add to BibSonomy
Abstract
Optical measurements are closely related to the optical signal-to-noise ratio (OSNR) of the laser, which can be improved using a tunable optical filter (TOF) to suppress frequency noise. For an external-cavity tunable laser with a tuning range larger than the TOF bandwidth, the wavelength at the center of the TOF passband must be varied based on the laser tuning. This study proposes a tunable-laser OSNR-enhancement method based on the Fabry-Pérot (FP) interferometer. The FP signal contains the wavelength information of the swept laser, which can be used to determine the real-time driving voltage of the TOF. Notably, the laser needs to be continuously tunable without mode hopping, and the free spectral range of the FP interferometer must be smaller than the TOF bandwidth.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
An external-cavity tunable laser (ECTL) can continuously change its output wavelength within a particular range, and it has been widely used in many cutting-edge technologies and key fields such as human brain monitoring [1,2], coherent spectral detection [3–6], laser ranging [7–9], and three-dimensional topography measurement [10–12]. The noise of the laser has a significant influence on the precision of the measurement results. For example, in a scanning intrinsic-oscillation coherent spectral-analysis system, a tunable laser is the reference for obtaining information about the signal to be measured. Its noise affects the power spectral density of the beat-frequency signal, limiting the sensitivity of detection. In frequency-modulated continuous-wave ranging [7], laser noise causes over-zero jitter in the outlier signal, limiting the accuracy of the distance measurement.
As the core laser source of the application system, the optical signal-to-noise ratio (OSNR) of the laser affects the final measurement result. Many methods for suppressing the laser noise have been developed, such as self-injection locking [13–16], optoelectronic feedback [17–19], and mode cleaning [20–22].
The principle of self-injection locking is to suppress laser noise by returning the initial signal from the laser itself to the resonant cavity through a coupler. In 2016, Hou et al. [14] used two optical couplers to construct a simple passive optical-feedback loop that did not introduce any additional noise and was insensitive to external perturbations. The OSNR of the laser improved from 60 dB to 70 dB, which effectively reduced the single-frequency noise of the fiber laser. In 2017, Zhang et al. [15] used a complex optical-feedback loop comprising erbium-doped fiber amplifiers combined with self-injection locking technology to achieve laser-noise suppression. The OSNR of the laser increased from 52 dB to 60 dB.
The optical-feedback method converts a laser-noise signal into an electrical signal and compares it with a stable reference-voltage source to obtain an error signal. Subsequently, the pump-source drive current is modulated by the feedback circuit to suppress the laser noise. In 2012, Zhang et al. [18] analyzed the factors affecting the intensity noise of an erbium-doped fiber laser and optimized the parameters of the feedback circuit through a circuit simulation. The middle- and low-frequency noises of the laser were well suppressed. In 2023, Liu et al. [19] analyzed the transfer function from the pumping disturbance of the fiber lasers to the laser-output power fluctuation. They designed and built a feedback-control system based on an existing commercial proportional integral-control circuit that significantly suppressed the intensity noise of the laser.
A laser-mode cleaner constructed using a resonant cavity can be used to improve the mode and spatial directivity of a laser beam and enhance its quality. The mode cleaner is a three-mirror annular cavity, which belongs to the passive cavity. It can filter high-frequency noise, which can reduce laser noise. In 1998, Willke et al. [20] used a fixed-spaced triangular annular cavity as a mode cleaner to filter laser noise. In 2022, Wang et al. [22] demonstrated that a mode cleaner can realize the interconversion of the amplitude and phase noises of semiconductor lasers and that the high-frequency noise in the transmitted optical field is directly reflected.
Currently, methods for suppressing laser noise by external means are mainly used for single-frequency lasers. Continuous mode-hopping-free tunable lasers also exhibit frequency noise. However, the challenge in suppressing laser noise is heightened due to the variable wavelength of the output laser, and external noise suppression for continuous mode-hopping-free tunable lasers has not been reported. A micro-electromechanical-system (MEMS) filter [23–25] is a tunable optical filter (TOF) based on MEMS technology that allows the laser in the passband range to pass through the filter and suppresses the noise caused by other wavelengths. Additionally, it incorporates a variable passband center. The TOF can serve as an independent frequency-selective element for tunable lasers. Balaswamy et al. [26] demonstrated the application of TOF in a laser system, achieving continuous tunability with a wavelength ranging from 1050 to 1100 nm. Moreover, enhancing the OSNR of the laser by employing TOF for secondary filtering in a swept source not only contributes to improved laser measurement accuracy but also facilitates human brain monitoring [1,2]. The TOF can be used to improve the OSNR of lasers. However, the continuous mode-hopping-free tuning range of the ECTL can reach several tens of nanometers, which is significantly larger than the bandwidth of the TOF. Thus, the passband-center wavelength of the TOF must be synchronized with the wavelength of the tunable laser.
Therefore, a method for laser OSNR enhancement based on a FP interferometer, which is suitable for continuous mode-hopping-free tunable lasers, is proposed in this study. The wavelength of the laser is calibrated in real time using the initial wavelength of the laser and the FP signal; thus, each FP contains the absolute-wavelength information. The resonant wavelengths of FP interferometer can drift due to temperature instability, which we will need to evaluate for a specific FP interferometer in the subsequent content. Then, the driving voltage of the TOF is adjusted based on the time of appearance of the FP signals, which enables the passband center of the TOF to match the laser wavelength in real time. This method realizes filtering of the entire wavelength scanning range and reduces laser noise, which improves laser accuracy in the field of precision measurement.
2. Methodology
The wavelength-passband characteristics of the TOF at a specific voltage are shown in Fig. 1(a), and its filtering characteristics are equivalent to those of the bandpass filter. The relationship between the voltage and wavelength curves of the TOF is shown in Fig. 1(b). The wavelength of the passband center of the filter can be adjusted using the drive voltage to achieve a tunable function. For single-frequency lasers, the TOF can be maintained at a fixed voltage. For a tunable laser, the drive voltage of the TOF must be synchronized to adjust the passband-center wavelength.
Fig. 1. (a) Bandpass characteristics of the TOF. (b) Corresponding voltage and wavelength curve of TOF.
The wavelength of the laser is set to 1520 nm. The laser is directly connected to the spectrometer to record the laser spectrum when the TOF is not used. The laser is then connected to the TOF. The filter drive voltage is adjusted such that the passband-center wavelength is also 1520 nm, and the spectrum is recorded at this point using the TOF. The spectral-line patterns before and after the TOF was used are shown in Fig. 2. The laser spectra directly reflect the relationship between the signal and noise. The noise was substantially suppressed by the use of the TOF. The OSNR can be calculated from the spectrum using the following formula:
where Psignal and Pnoise are the powers of the useful signal and the noise, respectively. The power of the useful signal is the average power within the bandwidth of ±0.05 nm. The noise power is the average power of the noise 1 nm away from the center wavelength of the signal in the bandwidth range of ±0.05 nm. The OSNR of the laser without and with the filter were calculated as 60 dB and 74 dB, respectively. The TOF increased the OSNR of the laser by more than 10 dB.The principle of OSNR enhancement based on the FP interferometer is shown in Fig. 3. The laser beam from the tunable laser passes through a fiber beam splitter. Part of the laser enters the FP interferometer, and the other part enters the TOF. The tunable laser transmitted via the FP interferometer produces transmission peaks at the end of the interferometer and is detected by the FP controller. The FP interferometer is configured as a resonant cavity. The laser entering the resonator is either enhanced or cancelled, and the optical-frequency change between the transmission peaks formed by the interference enhancement in the cavity is the free spectral range (FSR). This information is processed by a single chip that outputs the corresponding filter drive voltage to determine the central wavelength of the filter. At this time, the center wavelength of the filter and wavelength of the laser are the same. The laser entering the optical filter is filtered and output, and the output laser is then connected to the photodetector to monitor the optical-power signal in real time to determine whether synchronous filtering of the laser can be realized.
The laser produces a series of FP signals over time during the tuning process. The process of converting the FP signal into a filter-driving voltage is shown in Fig. 4. The FP signal contains information regarding the relative wavelength change of the laser; as long as the wavelength corresponding to the first FP signal is obtained, the wavelength corresponding to each FP signal can be known. The wavelength corresponding to the first FP peak is λ1, and the wavelength corresponding to the kth FP signal generated subsequently is calculated as follows:
Thus, the wavelength profile of the laser can be obtained from the FP signal generated during laser tuning. Combined with the voltage–wavelength correspondence curve of the TOF in Fig. 1(b), we can obtain the filter driving voltage corresponding to the tunable laser. Finally, we obtain the filter driving voltage corresponding to each FP signal.
For the passband-center wavelength of the filter to synchronously change with the laser wavelength, the laser wavelength must always be within the passband of the filter. The following constraints must be satisfied:
where λFilter is the passband-center wavelength, λECTL is the laser wavelength, and BWFilter is the 3-dB bandwidth of the TOF. The wavelength deviation B can also be expressed in terms of the FSR of the FP interferometer:Combining Eqs. (3) and (4), the FSR of the FP interferometer must satisfy the following conditions:
To obtain the center wavelength corresponding to each FP signal, only the center wavelength corresponding to the first FP signal needed to be determined. The specific method is illustrated in Fig. 5. A broadband optical source with a wavelength range covering the C-band is connected to the spectrometer after the FP interferometer. The spectrometer displays the center wavelength corresponding to each FP transmission peak. The initial scanning wavelength of the laser is determined using a spectrometer in the stationary state. The center wavelength of the FP peak can be combined with the initial wavelength of the laser to determine the laser wavelength corresponding to the first FP peak on the spectrometer. For example, in the rising section of the wavelength scan, the laser wavelength corresponding to the first FP peak is located at the center of the FP transmission peak adjacent to the initial wavelength, and the wavelength at the center of the FP transmission peak is larger than the initial wavelength of the laser.
The output voltage of a single chip is generally less than 10 V, whereas the driving voltage of the TOF needs to reach a maximum of 50 V. To satisfy the experimental requirements, the system used a voltage amplifier to amplify the voltage of the microcontroller.
In a synchronous filtering system, the deviation between the laser wavelength of the ECTL and the wavelength at the center of the TOF passband is influenced by the response time of devices such as voltage amplifier and TOF. To ensure that the wavelength deviation falls within an acceptable range, the laser tuning speed cannot be excessively fast. The minimum allowable time, denoted as Ta, between adjacent FP signals generated by the ECTL can be calculated using the following equation:
3. Experiments and discussion
The experimental setup is built according to the principle of tunable-laser OSNR enhancement, as shown in Fig. 6. The laser used in this study is a Littman-type ECTL with a continuous mode-free range of 1519.533–1545.637 nm. The TOF (H03, OE Photonics, China) operates in the wavelength range of 1517–1547 nm, with a 3-dB bandwidth of approximately 0.4 nm and a typical response time of 1 ms. The voltage amplifier can output up to 130 V with a rise time of 5 µs. The FP interferometer (SA200-12B, Throlabs, USA) had an FSR of 1.5 GHz. Here, we need to evaluate the thermal instability of the SA200-12B. The drift in resonant wavelength is attributed to variations in ambient temperature, specified at 0.2 GHz/°C. In the C-band range, this corresponds to a wavelength fluctuation of approximately 1.6 pm/°C. For this experiment, a deviation of the laser wavelength from the center of the TOF passband of less than 200 pm is considered sufficient. The wavelength drift caused by thermal instability has minimal effect on the experiment. Additionally, it is recommended to place the FP interferometer on vibration isolation pads to minimize the impact of vibration on its stability.
The tunable laser is connected to the FP interferometer. A set of FP signals is acquired using a digital acquisition card with the signal-sampling rate set to 2 MHz. Peak-point extraction is performed on the FP signals, and the number of FP signals is determined based on the peak points. The acquired FP signals and peak points are shown in Fig. 7(a). The laser-wavelength and filter-voltage curves obtained based on the FP signal are shown in Fig. 7(b).
The minimum time interval between adjacent FP signals can be calculated based on the center position of the FP signal as follows:
where Δnmin is the number of sampling points between the center positions of adjacent FP signal, and fs is the sampling rate of the digital acquisition card.The shortest time interval between adjacent FP signals is calculated to be 100 µs, which is greater than the minimum allowable time Ta (60 µs) and can meet the demand of the experimental program.
The synchronized filtering experiment is initialized after obtaining the filter driving voltage corresponding to the FP signal. The spectrometer is unable to perform real-time spectral measurements; however, based on the bandpass characteristics of the filter, the laser power can be monitored online using a photodetector. The laser power can be used to determine the synchronization between the tunable laser and TOF.
Before opening the synchronous filtering system, the FP and power signals of the laser in the wavelength-scanning process are collected and compared with the final synchronization results. The optical power of the frequency-scanned laser after the filter, without synchronous filtering, is shown in Fig. 8. Owing to the mismatch between the TOF passband and laser wavelength, the laser beam cannot pass through the TOF, and the laser power of the entire wavelength-scanning section is zero.
The synchronized filter system is turned on; however, the voltage amplifier and TOF are not connected. First, the voltage input to the filter is monitored via the microcontroller and amplifier to ensure that the range of this voltage is the same as that calibrated ahead of time; thus, the safe voltage of the TOF is not exceeded. The actual and theoretical voltages are shown in Fig. 9, which indicate that the system can output the desired voltage.
The voltage amplifier is then connected to the TOF, and the FP and laser-power signals of the photodetector are acquired simultaneously. The collected data are shown in Fig. 10. A comparison of the laser power with that in Fig. 8 (without using the synchronized filtering system) shows that after using the synchronized filtering system, the optical power is available for the entire wavelength-scanning section of the laser. This indicates that synchronized filtering of the laser was realized.
4. Summary
In conclusion, this study proposed an OSNR-enhancement method based on an FP interferometer that is applicable to lasers with continuous mode-hopping-free tuning. The laser wavelength is calibrated in real time using the initial wavelength of the laser and FP peak signals, and each FP peak contains absolute wavelength information. The driving voltage of the TOF at the current wavelength is output according to the time of appearance of the FP signals. This enables the center wavelength of the passband of the filter to match the laser wavelength in real time and realizes filtering of the entire wavelength-scanning range. The OSNR of the laser is increased by approximately 10 dB, which helps improve laser accuracy in the field of precision measurement.
Funding
National Natural Science Foundation of China (51875447).
Acknowledgment
We appreciate the reviewers for their constructive comments to improve the manuscript.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. M. Zhao, W. Zhou, S. Aparanji, et al., “Interferometric diffusing wave spectroscopy imaging with an electronically variable time-of-flight filter,” Optica 10(1), 42–52 (2023). [CrossRef]
2. S. Aparanji, M. Zhao, and V. J. Srinivasan, “Decoding diffuse light scattering dynamics in layered tissues: path length versus fluctuation time scale,” Opt. Lett. 48(22), 6056–6059 (2023). [CrossRef]
3. B. Szafraniec, J. Y. Law, and D. M. Barney, “Frequency resolution and amplitude accuracy of the coherent optical spectrum analyzer with a swept local oscillator,” Opt. Lett. 27(21), 1896–1898 (2002). [CrossRef]
4. B. Szafraniec, A. Lee, J. Y. Law, et al., “Swept coherent optical spectrum analysis,” IEEE Trans. Instrum. Meas. 53(1), 203–215 (2004). [CrossRef]
5. K. Feng, J. Cui, H. Dang, et al., “Investigation and development of a high spectral resolution coherent optical spectrum analysis system,” Opt. Express 24(22), 25389–25402 (2016). [CrossRef]
6. J. Cui, S. Zhang, H. Dang, et al., “High resolution coherent spectral analysis method based on Brillouin scattering in an optical fiber,” Opt. Express 29(15), 23258–23271 (2021). [CrossRef]
7. T. H. Ebben, D. L. Begley, and R. G. Marshalek, “Phase-noise-limited accuracy of distance measurements in a frequency-modulated continuous-wave LIDAR with a tunable twin-guide laser diode,” Opt. Eng. 34(3), 896–903 (1995). [CrossRef]
8. A. P. Cabral, M. A. Abreu, and J. M. Rebordão, “Dual-frequency sweeping interferometry for absolute metrology of long distances,” Opt. Eng. 49(8), 085601 (2010). [CrossRef]
9. Z. W. Deng, Z. G. Liu, X. Y. Jia, et al., “Dynamic cascade-model-based frequency scanning interferometry for real-time and rapid absolute optical ranging,” Opt. Express 27(15), 21929–21945 (2019). [CrossRef]
10. H. Yu, C. Aleksoff, and J. Ni, “Accuracy of a multiple height-transfer interferometric technique for absolute distance metrology,” Appl. Opt. 51(21), 5283–5294 (2012). [CrossRef]
11. E. Baumann, F. R. Giorgetta, J.- D. Deschênes, et al., “Comb-calibrated laser ranging for three-dimensional surface profiling with micrometer-level precision at a distance,” Opt. Express 22(21), 24914–24928 (2014). [CrossRef]
12. M. F. Shirazi, P. Kimb, M. Jeon, et al., “Free space broad-bandwidth tunable laser diode based on Littman configuration for 3D profile measurement,” Opt. Laser Technol. 101, 462–467 (2018). [CrossRef]
13. C. Li, S. Xu, X. Huang, et al., “All-optical frequency and intensity noise suppression of single-frequency fiber laser,” Opt. Lett. 40(9), 1964–1967 (2015). [CrossRef]
14. Y. Hou, Q. Zhang, and P. Wang, “Frequency-and intensity-noise suppression in Yb3+-doped single-frequency fiber laser by a passive optical-feedback loop,” Opt. Express 24(12), 12991–12999 (2016). [CrossRef]
15. Q. Zhang, Y. Hou, S. Qi, et al., “Low-noise single-frequency 1.5um fiber laser with a complex optical-feedback loop,” IEEE Photonics Technol. Lett. 29(2), 193–196 (2017). [CrossRef]
16. Y. Li, W. Liu, Z. Pan, et al., “Self-injection locking technique for resonant micro-optical gyroscope,” Opt. Eng. 61(1), 014103 (2022). [CrossRef]
17. G. A. Ball, G. Hull-Allen, C. Holton, et al., “Low noise single frequency linear fibre laser,” Electron. Lett. 29(18), 1623–1625 (1993). [CrossRef]
18. F. Zhang, J. Zhu, H. Wang, et al., “Intensity noise of erbium doped fiber laser at low frequency suppression through optoelectronic feedback,” Chinese Journal of Quantum Electronics 29(3), 311–315 (2012).
19. Y. Liu, S. Wei, Y. Chen, et al., “Relaxation oscillation noise suppression in a DBR single-longitudinal-mode fiber laser using optoelectronic feedback,” Chinese Journal of Quantum Electronics 40(5), 677–683 (2023). [CrossRef]
20. B. Willke, N. Uehara, E. K. Gustafson, et al., “Spatial and temporal filtering of a 10W Nd:YAG laser with a Fabry-Pérot ring-cavity premode cleaner,” Opt. Lett. 23(21), 1704–1706 (1998). [CrossRef]
21. Y. Chen, J. Zhang, Y. Li, et al., “Reduction of intensity noise of single-frequency Nd∶YVO4 laser using mode cleaner,” Chinese Journal of Lasers 28(3), 197–200 (2001).
22. X. Wang, H. Yan, W. Yang, et al., “Noise conversion of a mode-cleaner cavity and noise analysis of output fields,” Acta Photonica Sin. 51(4), 0414002 (2022). [CrossRef]
23. K. Yu and N. Park, “Characterization of MEMS optical bandpass filters with narrow transition bands,” Optical Transmission, Switching, and Subsystems III.60212R, 709–716 (2005). [CrossRef]
24. H. Omran, Y. M. Sabry, M. Sadek, et al., “MEMS optical tunable filter based on freestanding subwavelength silicon layers,” MOEMS and Miniaturized Systems XIII.89770V, 229–236 (2014). [CrossRef]
25. P. L. Uppalapati, B. Malayappan, N. Krishnaswamy, et al., “Electrostatically tunable MOEMS waveguide Bragg grating-based DWDM optical filter,” J. Micro Nanolithogr. MEMS MOEMS 18(01), 1 (2019). [CrossRef]
26. V. Balaswamy, S. Aparanji, G. Chayran, et al., “High-Power, Independently Wavelength, Power, and Linewidth Tunable Ytterbium Fiber Laser,” IEEE Photonics Technol. Lett. 31(8), 583–586 (2019). [CrossRef]
