Open Access
Add to BibSonomy
Abstract
A widely tunable V-cavity semiconductor laser with kilohertz linewidth is demonstrated using simple self-injection method. The tunable laser is packaged into a small-form-factor 9-pin TOSA and can be tuned over 22 channels at 100 GHz spacing with side mode suppression ratios over 36 dB. Self-injection method is used for linewidth reduction by applying optical feedback. Linewidth of the V-cavity laser can be compressed from several megahertz to tens of kilohertz. The linewidth reduction method is essentially channel independent and can be easily realized. It can meet the requirements of many applications such as high-order modulation, coherent communication and sensing.
© 2017 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Fast-growing internet traffic has resulted in the great demand for high speed optical transmission systems at bitrates of 100 Gbit/s and beyond. Coherent communication is a promising technique to increase the capacity and spectral efficiency of next-generation fiber-optical networks with optical phase modulation and demodulation [1], which requires narrow linewidth tunable lasers for signal sources and local oscillators. Besides, areas such as atomic clocks [2], and coherent LIDAR systems [3], narrow linewidth tunable lasers are also needed. Many efforts have been made to reduce the laser linewidth over the past several decades, which can be divided into two basic categories: electrical feedback and optical feedback. As for electrical feedback method, sub-kilohertz linewidth by using a high resolution Fabry-Perot interferometer (FPI) for a corrugation-pitch-modulated MQW-DFB Laser with multi-electrodes was realized, which has a flat FM characteristics and a sub-megahertz free-running spectral linewidth [4]. Optical feedback method is a simpler way to realize narrow linewidth compared to electrical feedback [5–7]. An all-fiber passive FBG FP etalon with very short cavity length together with a mode selecting device was used, which can get 5 kHz linewidth output [8]. By using two parallel fiber loops as the feedback cavities, the linewidth of a DBR laser can be reduced down to 3.2 kHz [9]. Narrow linewidth laser by using high-Q devices as optical feedback has also been investigated, such as whispering-gallery-mode-resonator and microsphere resonator [10–13].
Narrow linewidth tunable lasers are particularly desirable for coherent communication and sensing. A simple and compact tunable V-cavity laser has recently been reported [14, 15]. In this letter, we demonstrated a simple and effective way of linewidth reduction in the V-cavity tunable laser using self-injection with a coated fiber. The linewidth is narrowed from several megahertz at the free-running condition to several tens of kilohertz.
2. Device structure and experimental setup
Figure 1(a) gives the top view of a V-cavity laser (VCL) [14, 15]. It comprises a fixed gain cavity and a channel selector cavity with different optical path lengths, which form V-shaped branches with a reflective half-wave coupler. The half-wave coupler, in which the cross-coupling coefficient has a π phase relative to the bar-coupling coefficient, is important for achieving high side-mode-suppression ratio (SMSR). Wide-band wavelength tuning with high SMSR has been reported. In our experiment, the laser is packaged into a small-form-factor 9-pin TOSA with a thermo-electric cooler (TEC), and the electronic driver has been developed for the wavelength tuning and direct modulation. The advantages of the VCL in compactness, fabrication simplicity, and easy wavelength control offer great potential in low-cost network applications, as well as in portable devices for spectroscopic analysis. Figure 1(b) shows the picture of the small-form-factor 9-pin TOSA [15]. The TOSA box size is 8.4 × 5.4 × 5.2 mm3.
Fig. 1 (a) Top view of an InGaAsP/InP based V-cavity laser and (b) the picture of the small-form-factor 9pin TOSA.
A common DFB or DBR laser always has a linewidth of a few megahertz [16], so as for our V-cavity laser, which is not sufficiently narrow for applications in coherent optical communication or optical sensing. Here we use a simple self-injection method for linewidth reduction. Figure 2 shows the schematic diagram of the linewidth measurement system. The output of the TOSA is divided into two paths through a 2 × 2 coupler with different optical splitting ratios. One path of the light is injected into a coated fiber through a variable optical attenuator (VOA). The end facet of the fiber is coated with a gold film for total reflection feedback, and the other path is connected to the linewidth measurement system. Two cascaded optical isolators(>24 dB isolation with each isolator) are used to guarantee no feedback from the linewidth test system. Both the 2 × 2 coupler and the VOA are used for controlling the optical feedback strength. Different splitting ratios of 90:10, 80:20, 70:30, 50:50 are used in the experiment. We use the branch corresponding to the right number of the splitting ratio as the feedback branch. This convention is used throughout the paper. For example, for a 90:10 coupler, 10% of the TOSA output power reaches the feedback branch and 90% of the power is used for the linewidth measurement. The linewidth is measured by using delayed self-heterodyne method with a delay length of 10 km. An acoustic-optic modulator (AOM) with a 200 MHz frequency is used to shift the beat signal frequency in the experiment. Beat signals are detected by a photodetector (Anritsu MN4765A), which are displayed on an electrical spectrum analyzer (Agilent EXA Signal Analyzer-N9010A). Apart from the difference in the lasers, the linewidth compression system is much simpler as compared to Ref [9]. Only a 2 × 2 coupler and a short high-reflection coated fiber were used instead of two fiber loops with polarization controllers.
Fig. 2 Schematic diagram of the linewidth measurement system using optical self-injection method. VOA: variable optical attenuator; CF: coated fiber; PM: power meter; ISO: isolator; AOM: acousto-optic modulator; OSA: optical spectrum analyzer; EXA: Agilent EXA spectrum analyzer.
3. Measurement results and discussions
The optical spectrum of the V-cavity laser is firstly measured. The current of the half-wave coupler and the fixed gain cavity are biased at 30 mA and 28 mA, respectively. The operating temperature was set at 45°C by the TEC within the TOSA. The relatively high temperature was chosen in order to reduce the TEC power consumption at high ambient temperature up to 85°C. We adjust the current of the channel selector electrode for wavelength tuning. Twenty-two wavelength channels with SMSR above 36 dB were measured when the channel selector current changes from 30 mA to 107 mA. The output power is 1.6 mW in average. Figure 3(a) gives the overlapped optical spectra of the V-cavity laser. Figure 3(b) shows the measured wavelength tuning as a function of the current of the channel selector electrode.
Fig. 3 (a) Optical spectra of the V-cavity laser; (b) Measured wavelength tuning as a function of the current of the channel selector electrode
Figure 4 shows linewidth profile measured by delayed self-heterodyne method under different feedback strengths when the feedback path length is 5.44 m. The current of the coupler electrode, channel selector electrode and fixed gain electrode are biased at 30 mA, 30 mA and 28 mA, respectively. The linewidth of the V-cavity laser is measured to be 3.9 MHz when there is no feedback. The spectral linewidth is dramatically reduced from its original linewidth of 3.9 MHz down to 60 kHz when the coupler splitting ratio changes from 90:10 to 50:50. The linewidth is 280, 270, 80 and 60 kHz, respectively, when the splitting ratio is 90:10, 80:20, 70:30 and 50:50.
Fig. 4 Linewidth profile measured by delayed self-heterodyne method under different feedback strengths.
Figure 5(a) shows the linewidth of all the 22 channels without feedback. The linewidth ranges from 3.9 MHz to 7.5 MHz. The channels are numbered according to increasing current on the channel selector electrode, as indicated in Fig. 3(a). It is also shown that there is a linewidth jump between the 7th channel and 8th channel, which corresponds to the wavelength jump as a result of the change of a free spectral range (FSR) in the VCL tuning.
Fig. 5 (a) Linewidth of all 22 channels without feedback, (b) linewidth of all 22 channels with different feedback strengths.
The linewidth reduction under different coupling ratio was measured for all the 22 channels in Fig. 5(b). As we can see, the linewidth goes further compressed with increasing feedback. The black and red curves show the feedback condition of using couplers with splitting ratios of 90:10 and 80:20, respectively. We can see that the linewidth variation trends for different channels are almost the same as the case under no feedback. The blue and green curves show the feedback condition of using coupler with splitting ratio 70:30 and 50:50, respectively. The linewidth is significantly compressed. It also shows that the linewidth change due to the FSR change is reduced when the feedback strength is large enough.
We also measured the optical linewidth when the feedback power is varied by adjusting the VOA while the splitting ratio of the coupler is fixed at 50:50. The current of the coupler electrode, channel selector electrode and fixed gain electrode are 30 mA, 30 mA and 28 mA, respectively. Figure 6(a) gives the experimental results. The linewidth changes rapidly when the power meter shows low feedback power, but the change slows down when the feedback power is increased to beyond 40 μW. The linewidth matches very well with results of Fig. 5 when the power is reduced to the same level as in the case where we use the coupler with different splitting ratios. The inset in Fig. 6(a) gives the enlarged diagram when the feedback power changes from 0 to 1 μW. The linewith is little affected when the reflection phase is varied by adding a variable phase delay line. Figure 6(b) compared the linewidth compression for 4 channels when using 4 couplers of different splitting ratios: 80:20, 70:30, 50:50, and 30:70. The branch corresponding to the right number of the ratio is used as the feedback path. All the 4 channels show the same result: the linewidth cannot be further reduced when the feedback power is increased beyond about 300 μW or with the splitting ratio beyond 50:50 to increase the feedback strength.
Fig. 6 (a) Linewidth versus feedback power when adjusting the VOA; (b) Linewidth versus feedback power for 4 channels by using different splitters.
In this experiment, we also found that the length of feedback path could impact the linewidth significantly. In Fig. 7, we show different linewidths measured by using the coupler with splitting ratio 70:30 under five different lengths of feedback path: 1.61 m, 2.68 m, 5.44 m, 8.54 m, and 11.64 m. As the feedback length increases from 1.61 m to 5.44 m, the linewidth of each channel decreases from several hundreds of kilohertz to several tens of kilohertz. When the feedback path length increases further beyond ~6 meters, the linewidth reduction becomes saturated. Note that in Ref [9], two relatively long fiber-loops of 25 m and 55 m are used to obtain high Q-factor and to suppress the unwanted spur modes in the external cavity. The linewidth of a DBR laser is reduced from 12.5MHz to 3.2 kHz. Our experiments show that with our V-cavity laser, a much simpler compression system with a short high-reflection coated fiber is sufficient to reduce the linewidth from several MHz down to ~60 kHz.
Fig. 7 Linewidths of all channels for different lengths of the feedback path when the splitting ratio is fixed at 70:30 with the 30% branch as the feedback path.
We also measured the linewidth variation with feedback path length for two different couplers. The linewidth versus feedback path length for ch1 is plotted in Fig. 8 for coupler ratios of 70:30 and 50:50. It is obvious to see that the linewith decreases then saturates with increasing feedback path length and a narrower linewidth is achieved with a 50:50 coupler as compared to a 70:30 coupler. When the feedback path length is fixed, the main factor impacting the linewidth is the feedback power. When the feedback path length increases, the linewidth difference between the two cases using different couplers becomes smaller. That is to say, when the feedback path length is short, the feedback power plays an important role in linewidth compression. When the feedback path length goes longer, the influence of the feedback power is weakened.
The output power of the TOSA was also measured under different feedback conditions. Figure 9 shows the output power of all the channels when a 70:30 coupler is used as compared to the case of no feedback. The feedback path length is 5.44 m. We see that the output power also hops at the FSR jump. The output power of most channels is slightly higher with the feedback.
4. Conclusion
We have demonstrated a simple linewidth reduction method with the tunable V-cavity laser. By using a coupler and a high-reflection coated fiber as the feedback system, the linewidth of the V-cavity laser can be compressed from several megahertz to tens of kilohertz. The linewidth decreases with increasing feedback strength and increasing length of the feedback path, but saturates beyond a certain values. Linewidths as narrow as 60 kHz were obtained, which can meet the requirements of many applications such as high-order modulation, coherent communication and sensing. The linewidth reduction method is essentially channel independent and can be easily realized.
Acknowledgement
This work was supported by the National Science and Technology Major Project of China (No. 2015ZX03001021), the National Natural Science Foundation of China (grant No. 61535010), and the National High-Tech R&D Program of China (grant No. 2013AA014401).
References and links
1. M. Seimetz, “Laser linewidth limitations for optical systems with high-order modulation employing feed forward digital carrier phase estimation,” in Conference on Optical Fiber Communication/national Fiber Optic Engineers Conference (2008), pp. 1–3.
2. Y. Y. Jiang, A. D. Ludlow, N. D. Lemke, R. W. Fox, J. A. Sherman, L.-S. Ma, and C. W. Oates, “Making optical atomic clocks more stable with 10–16-level laser stabilization,” Nat. Photonics 5(3), 158–161 (2011).
3. E. Dale, W. Liang, D. Eliyahu, A. A. Savchenkov, V. S. Ilchenko, A. B. Matsko, D. Seidel, and L. Maleki, “Ultra-narrow line tunable semiconductor lasers for coherent LIDAR applications,” in Imaging Appl. Opt. (2014), paper JTu2C.3.
4. M. Kourogi, C. Shin, and M. Ohtsu, “A 250 Hz Spectral Linewidth 1.5 prn MQW-DFB Laser Diode with Negative-Electrical-Feedback,” IEEE Photonics Technol. Lett. 3(6), 496–498 (1991).
5. N. Wang, M. Feng, Z. Q. Feng, M. Y. Lam, L. Gao, B. Chen, A. Q. Liu, Y. H. Tsang, and X. M. Zhang, “Narrow-linewidth tunable lasers with retro-reflective external cavity,” IEEE Photonics Technol. Lett. 24(18), 1591–1593 (2012).
6. K. Aoyama, R. Yoshioka, N. Yokota, H. Yasaka, and W. Kobayashi, “Narrow-linewidth laser diode with compact optical feedback system,” in International Topical Meeting on Microwave Photonics (2014), pp. 79–81.
7. X. M. Zhang, N. Wang, L. Gao, M. Feng, B. Chen, Y. H. Tsang, and A. Q. Liu, “Narrow-Linewidth External-Cavity Tunable Lasers,” in International Conference on Optical Communications and Networks (2011), pp. 1–2.
8. Z. Y. Dai and X. X. Zhang, “Stable high power narrow linewidth single frequency fiber laser using a FBG F-P etalon and a fiber saturable absorber,” in Photonics Optoelectronic (2010), pp. 1–4.
9. L. Q. Yu, D. Lu, B. W. Pan, L. M. Zhang, L. Guo, Z. S. Li, and L. J. Zhao, “Widely tunable narrow-linewidth lasers using self-injection DBR lasers,” IEEE Photonics Technol. Lett. 27(1), 50–53 (2015).
10. V. S. Ilchenko, E. Dale, W. Liang, J. Byrd, D. Eliyahu, A. A. Savchenkov, A. B. Matsko, D. Seidel, and L. Maleki, “Compact tunable kHz-linewidth semiconductor laser stabilized with a whispering-gallery mode microresonator,” Proc. SPIE 7913(1), 79131G (2011).
11. B. Sprenger, H. G. L. Schwefel, and L. J. Wang, “Whispering-gallery-mode-resonator-stabilized narrow-linewidth fiber loop laser,” Opt. Lett. 34(21), 3370–3372 (2009). [PubMed]
12. K. Kieu and M. Mansuripur, “Fiber laser using a microsphere resonator as a feedback element,” Opt. Lett. 32(3), 244–246 (2007). [PubMed]
13. B. Sprenger, H. G. L. Schwefel, Z. H. Lu, S. Svitlov, and L. J. Wang, “CaF2 Whispering-Gallery-Mode-Resonator Stabilized-Narrow-Linewidth Laser,” Opt. Lett. 35(17), 2870–2872 (2010). [PubMed]
14. S. Zhang, J. Meng, S. Guo, L. Wang, and J.-J. He, “Simple and compact V-cavity semiconductor laser with 50×100 GHz wavelength tuning,” Opt. Express 21(11), 13564–13571 (2013). [PubMed]
15. J. J. Meng, X. H. Xiong, H. B. Xing, H. S. Jin, D. Zhong, L. Zou, J. S. Zhao, and J. J. He, “Full C-Band Tunable V-Cavity-Laser Based TOSA and SFP Transceiver Modules,” IEEE Photonics Technol. Lett. 29(12), 1035–1038 (2017).
16. K. Shi, D. A. Reid, L. P. Barry, Y. L. Yu, and F. Smyth, “Linewidth Calibration of SG-DBR Lasers,” IEEE Photonics Technol. Lett. 22(23), 1729–1731 (2010).
