This paper proposes and demonstrates a novel method to produce the narrow-bandwidth, narrow-pulse-width and high-repetition-rate pulses with actively Q-switched ring-cavity all-fiber lasers. By using a specially designed low-reflectivity cladding power stripper in the cavity, and inserting a length-optimized ytterbium-doped single-cladding fiber self-pumped by the backward amplified spontaneous emission (ASE) from the YDF to improve the amplification of the initial weak ASE feedback by the narrowband filter, the ASE gain self-saturation can be suppressed efficiently, and the lasing pulses can be established quickly within the opening time of Q-switch even operating for very high repetition-rate. With the proposed technique, watt-level Q-switched pulses with bandwidth and pulse width narrowed to 0.15 nm and 9 ns, and repetition rate up to 175 kHz are achieved.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Large-energy nanosecond laser pulses are very important for many applications, including precision machining , laser radar , and non-linear frequency conversion . Such laser pulses can be produced by a master oscillator power amplifier (MOPA) system consisting of a pulsed laser seed source and power amplifiers [4,5]. Common pulse seed sources include directly modulated semiconductor laser , Q-switched solid-state laser , and Q-switched fiber laser [8–10]. Nowadays, the Q-switched fiber laser has gradually become an important seed source because of its good beam quality, compact structure, high conversion efficiency and ability to resist the reflection influence of subsequent power amplifiers . Especially, actively Q-switched fiber lasers with acousto-optic modulators (AOMs) have attracted considerable attention because an AOM offers the advantages of controllable modulation waveform and frequency .
To date, both single- and double-cladding gain fibers have been used for constructing actively Q-switched fiber lasers with AOMs [10–14]. However, owing to the limited total rare-earth-doped ions available, the single mode gain fiber is very prone to the gain saturation . The gain saturation can decrease the gain value actually provided by the gain fiber, resulting in more round-trips (i.e., longer time) required for the pulse establishment. Moreover, the low gain also causes slow increase of the pulse power per round-trip, leading to a low peak power for the built-up pulse, which in turn can not completely deplete photons in the cavity within a few round-trips. Finally, the gain saturation will widen the pulse width and lower the peak power for actively Q-switched single-cladding fiber lasers . To generate the Q-switched pulses with a narrower pulse width and a higher peak power, double-clad gain fibers have to be used because, under a strong pump condition, they can offer very high gain and the gains are not easy to be saturated . Many actively Q-switched linear cavity double-clad fiber lasers with different discrete devices have already been proposed and demonstrated to achieve narrow-pulse-width and high-peak-power Q-switched pulses [17,18]. For example, Boullet et al. achieved a Q-switched fiber laser, which can produce 12 ns pulses at 10 kHz repetition rate, or 190 kHz pulses with width of 32 ns . With the similar motivation, the photonic crystal gain fiber has also been used, and 10 ns pulse-width and 100 kHz repetition-rate Q-switched pulses have also been obtained with linear-cavity configuration based on discrete devices . However, the Q-switched fiber lasers with discrete devices have poor stability. It is necessary to develop actively Q-switched all-fiber lasers. Unfortunately, the photothermal effect leads to a very low damage threshold for the fiber-pigtailed AOM. When an actively Q-switched linear cavity double-clad all-fiber laser is constructed with the fiber-pigtailed AOM, it is no longer allowed to give full play to the advantage of high saturation gain of the double-clad gain fiber by increasing the pump power, resulting in undesirable performance for the Q-switched double-clad fiber laser. For instance, it is very difficult to achieve Q-switched pulses with pulse width narrower than 100 ns and repetition rate up to 50 kHz for an actively linear-cavity Q-switched double-clad all-fiber laser [19,20]. To overcome the difficulty of vulnerability to the photothermal damage of fiber-pigtailed AOM, Lecourt et al. proposed a skillful scheme for the Q-switched double-clad all-fiber laser that changes the linear cavity to ring cavity, and places the fiber-pigtailed AOM after an inserted fiber coupler to extract the pulse energy in the cavity, so that the pump is still allowed to be increased considerably to give full play to the advantage of high saturation gain of the double-clad gain fiber, while the fiber-pigtailed AOM is just right located at the position where the intracavity laser power is sufficiently low [21,22]. Using this technique, combined with the method of reducing the modulation duty cycle of the Q-switch to increase the gain recovery time for the gain fiber , they demonstrated an actively Q-switched ring-cavity double-clad all-fiber laser with a pulse width of 10 ns, and the repetition rate is also higher than 100kHz [21,22]. However, the bandwidth of the Q-switched pulses obtained in  is as wide as 1 nm. If narrower bandwidth Q-switched pulses are needed, a sub-nano narrowband filter has to be used, which inevitably weakens the cavity feedback intensity. In other words, the initial amplified spontaneous emission (ASE) feedback into the cavity extracted by the filter decreases with the narrowing of the Q-switched pulse bandwidth, resulting in a delay in the building time of the Q-switched pulses. This is obviously in contradiction with decreasing the modulation duty cycle of the Q-switch to increase the gain recovery time of the gain fiber for high repetition-rate Q-switched pulses. On the other hand, the spontaneous emission at any position in the gain fiber can easily be collected into the fiber to form guided modes propagating along the fiber, thus, large forward and backward ASEs are inevitably generated by the gain fiber under a strong pump. The excessive ASEs lead to a strong gain self-saturation in the gain fiber [23–25], thereby affecting the amplification of the initial weak ASE feedback to the cavity originating from the narrowband filtering, and extending the building time of the Q-switched pulse in the cavity, which is harmful to the generation of high-repetition-rate Q-switched pulses. Consequently, narrow-bandwidth, narrow-pulse-width, and high-repetition-rate pulses are still difficult to produce for actively Q-switched all-fiber lasers with AOM so far. This goal may be achieved with the injection seeding technique , but it would engender extremely complex laser systems. Therefore, how to produce narrow-bandwidth, narrow-pulse-width, and high-repetition-rate pulses with actively Q-switched all-fiber laser must be further studied.
In this paper, we propose and demonstrate a method to generate narrow-bandwidth, narrow-pulse-width and high-repetition-rate pulses with an actively Q-switched ring-cavity all-fiber laser. A specially-designed low-reflectivity cladding power stripper (CPS) is used in the cavity; a length-optimized ytterbium-doped single-cladding fiber (YSF) self-pumped by the backward ASE from the ytterbium-doped double-cladding fiber (YDF) is inserted into the cavity to improve the amplification of the initial weak ASE feedback extracted by the narrowband filter. This enables the ASE gain self-saturation to be suppressed efficiently and then the narrow-bandwidth Q-switched laser pulse to be established quickly. The experimental results demonstrate that with the proposed technique our laser can produce watt-level Q-switched laser pulses with bandwidth and pulse width narrowed to 0.15 nm and 9 ns, respectively, and the repetition rate reaches to 175 kHz. To our knowledge, they may be the narrowest bandwidth and pulse width generated with such Q-switched all-fiber lasers to date.
2. Experimental setup
Figure 1 illustrates the schematic of the proposed actively Q-switched all-fiber laser with AOM. A segment of the 3 m YDF (Nufern, LMA-YDF-10/130-M) with an absorption of 4.2 dB/m at 975 nm, and core and inner cladding numerical aperture of 0.08 and 0.46, respectively, serves as the gain fiber. The YDF is pumped by a 975 nm multimode laser diode via a (2 + 1)×1 multimode pump combiner (MPC). The output fiber of the MPC is the matching fiber (Nufern, LMA-GDF-10/130-M) for the YDF. The input and output ports of a fiber circulator (CIR) are connected to the cavity to ensure that lights propagate clockwise in the cavity, and its common port is connected to a 1064.5 nm single-mode fiber Bragg grating (FBG) with 3 dB bandwidth of 0.2 nm and reflectivity of 99%, which forms the narrowband filter. After amplification by the YDF, 80% of the light power is extracted by an optical coupler (OC) with a single-mode pigtailed fiber (Nufern, LMA-GDF-10/125-M), and the rest of the light power enters into the fiber-pigtailed AOM (Gooch & Housego, T-M200-0.1C2G-3-F2S). This ensures that the AOM is in a position where the laser power is low enough to avoid possible photothermal damage. The measured insertion loss of the AOM is approximately 3 dB.
Unlike the case of single mode fiber fusions, the discontinuities of the fusion points between the YDF and its matching fibers are very difficult to eliminate efficiently. Such discontinuities may convert the forward (backward) ASE propagating along the inner cladding into the initial backward (forward) ASE at the fusion points, resulting in a significant increase in the backward (forward) ASE owing to the extremely high gain of the YDF. These excessive ASEs aggravate the ASE gain self-saturation in the gain fiber, which is harmful to the quick formation of narrow-bandwidth Q-switched pulses. Considering that a forward ASE with a moderately high spectral intensity is helpful in extracting a relatively high initial narrowband ASE feedback into the cavity to quicken the formation of Q-switched pulses in the cavity, this study focuses on the decrease of the reflection of the forward ASE at the YDF output end. For this purpose, a specially designed CPS with the YDF as its input fiber directly is used for omitting the fusion point between YDF and its matching fiber of a common CPS at the YDF output end. The output fiber of the CPS is the same single-mode fiber with the OC, which can also omit the fusion point between the output double-cladding matching fiber and the single-mode fiber. Thus the specially designed CPS has advantageous in reducing the initial backward ASE of the YDF caused by the reflections of excessive fusion points in the fiber link from the YDF output end to the single-mode fiber. The YDF is directly spliced with the single-mode fiber after the precision alignment of the two fiber cores, the loss and reflection of the fusion point are reduced by repeatedly optimizing the splicing discharge parameters. A high-refractive index resin adhesive is coated on the inner cladding surface of the YDF in the CPS. Thus, the residual pump and forward ASE propagating in the inner cladding can be effectively removed before reaching the fusion point. Finally, the encapsulation and heat dissipation of the CPS is achieved with the standard technology by using a metal shell. Taking the elimination of two fusion points into account, initial backward ASE reflected back to the gain fiber core in the link from the YDF to single-mode fiber will be greatly reduced by using the specially-designed CPS. The specially designed CPS can not only alleviate the ASE gain self-saturation efficiently, but can also be conducive to achieving short cavity length that is essential for the quick formation of Q-switched pulses (depending on the cavity round-trip time) and the achievement of high repetition frequency.
On the other hand, the narrow bandwidth of the FBG leads to a weak cavity feedback and affects the building time of the Q-switched pulse in the cavity. Hence, a high doping-concentration YSF (CorActive, Yb198) with an absorption of 840 dB/m at 975 nm is connected to the cavity near the input port of the MPC and self-pumped by the backward ASE from the YDF, which can amplify the initial weak ASE extracted by the narrowband FBG. This ensures that the initial weak ASE feedback can quickly achieve efficient amplification before being amplified by the YDF to shorten the building time of the Q-switched pulses. Note that the high doping-concentration YSF ensures that the cavity length is not significantly increased after adding it into the cavity.
The pigtailed fibers of the AOM and CIR, and the input fiber of MPC are all HI 1060 fibers. To shorten the cavity length and facilitate the establishment of Q-switched laser pulses, all the fiber lengths are reduced as much as possible. The total cavity length is finally controlled at approximately 6 m, and the total cavity loss is approximately 12 dB. The properties of the Q-switched pulses exited from the output isolator are measured using an optical power meter (Ophir, Vega7Z01560), a spectrum analyzer (Yokogawa, AQ6370D), and a 1.5 GHz photodetector followed by an oscilloscope (Agilent, MSO8064A) or a frequency analyzer (Agilent, E4402B).
3. Results and discussion
Firstly, the output Q-switched pulse properties are investigated when the YSF was not inserted in the laser as shown in Fig. 1. It was observed that regular Q-switched pulses are easily produced when the square wave modulation frequency for the AOM is lower than 120 kHz, even with a duty cycle of 1:1. However, the pulse width is relatively wide. If the duty cycle is suitably reduced, a pulse width below 10 ns can be obtained. However, when the modulation frequency is between 120 and 150 kHz, irregular pulses with the pulse missing phenomenon were observed in the output pulses with a duty cycle of 1:1. By reducing the duty cycle similarly, regular Q-switched pulses can also be generated.
Figures 2(a)–2(c) show the Q-switched pulse waveforms measured at AOM modulation frequencies of 10, 120, and 150 kHz while the duty cycles reduced to 1:10, 1:52, and 1:53, respectively when the pump power is 7 W. The figures show that the pulse widths in the three cases are 10.1, 10.3, and 10.4 ns respectively, and the time delays for the pulse establishment with respect to the instant turning on the Q-switch is 180, 150, and 120 ns, corresponding to 6, 5, and 4 round-trip times in the cavity (the cavity round-trip time is approximately 30 ns), respectively. Note that the pulse waveforms shown in Figs. 2(a)–2(c) can be adjusted to bell-shaped by suitably increasing the rising times of their modulation square waves. The variations of the Q-switched laser pulse with the modulation frequency and duty cycle in Fig. 2(a)–2(c) can be explained as follows. With the increase in the AOM modulation frequency, the given gain recovery time for the gain fiber is shortened. Thus, it is necessary to reduce the duty cycle to relatively promote the recovery of the population inversions  for generating the regular Q-switched laser pulses. However, with the increase in the repetition rate and decrease in duty cycle, the time in the opening state of the Q-switch is shortened unavoidably, and the giving time for building-up the pulse or the opening time of Q-switch has to be reduced accordingly. Moreover, since the average population inversion of the YDF during the Q-switch’s opening state is reduced with the increase in the repetition rate and decrease in duty cycle, the amplitude of the output Q-switched pulses also decreases accordingly as the repetition rate increases.
However, when the modulation frequency exceeds 150 kHz, the pulse missing can still be eliminated by reducing the duty cycle. Figure 2(d) shows the measured pulse waveform with a modulation frequency of 155 kHz and duty cycle of 1:64. It can be seen that the pulse amplitude considerably decreases, the pulse width increases to 25 ns, and the time delay of the pulse establishment with respect to the moment turning on the Q-switch is 90 ns which only corresponds to a 3 cavity circulation cycles. This may be because of the ASE gain self-saturation effect. Therefore, the spectra at repetition rates of 150 kHz and 155 kHz were measured at the idle pump port of MPC, as shown in Fig. 2(e). This figure shows that the measured spectrum for 150 kHz is composed of the reflected Q-switched pulse spectral peak caused by the discontinuity points (including Rayleigh scattering in fibers) superimposed on a smooth, low-amplitude ASE background. However, the ASE background considerably increases when the repetition rate increases to 155 kHz, there is an ASE peak at 1030 nm, and its amplitude is even comparable to that of the reflected Q-switched pulse, i.e., the backward ASE for the 155 kHz is stronger than that for 150 kHz. In fact, to obtain high-repetition-rate regular pulses, the modulation duty cycle should be reduced to promote the recovery of the population inversion of the gain fiber. However, because the population inversion decreases with the increase in the modulation frequency caused by the reduction of the absolute time for gain recovery, the feedback initial ASE extracted by the narrowband FBG also decreases with the increase in the modulation frequency. When the modulation frequency is higher than 150 kHz, the inversion populations in YDF accumulated when the switch is off are mainly used to amplify the forward and backward ASEs produced by itself because the feedback initial ASE is too weak. This results in the gain self-saturation caused by the excessive ASE. The ASE gain self-saturation restricts the efficient amplification for the feedback initial narrowband ASE, and the output pulse is an ASE-like pulse without lasing oscillation characteristics during the opening time of Q-switch. In fact, the 3 dB bandwidth of the output pulses at 150 kHz is only 0.16 nm, as shown in Fig. 2(f), while it increases to 0.19 nm for the ASE-like pulses at a 155 kHz. This indicates that the lasing oscillation characteristics of the ASE-like pulses are weakened owing to the lower cavity gain caused by the ASE gain self-saturation. Because of the lower coherence and peak power which are disadvantageous to subsequent power boosts, this type of ASE-like pulse is generally not concerned.
The above results indicate that, when trying to obtain a very narrow-bandwidth pulses (below 0.2 nm), the regular Q-switched laser pulses can no longer be obtained by reducing the modulation duty cycle due to the ASE gain self-saturation if the modulation frequency is too high. The reason is that the feedback initial narrowband ASE decreases with the increase in the modulation frequency. Moreover, because of the ASE gain self-saturation, the repetition rate of the output Q-switched pulses cannot be increased by increasing the pump power either. When the pump power was further increased to 8 W for the 155 kHz repetition rate, a few self-excited oscillations were observed in our experiment near the 1030 nm ASE peak owing to the ASE gain self-saturation effect, which may result in the damage of the MPC or other devices in the cavity. Therefore, suppressing the reflection of discontinuity points in the cavity to reduce the ASE accumulation is essential to obtain high-repetition-rate Q-switched pulses, particularly for narrow bandwidth pulses. The backward ASE is effectively suppressed thanks to the specially-designed CPS, which allows a pump power up to 7 W so that a repetition rate of Q-switched pulses up to 150 kHz is obtained. When replacing the special CPS with a common CPS in the experiment, with the other cavity parameters unchanged, the ASE gain self-saturation is evidently aggravated because the repetition rate of the Q-switched laser pulses generated is only below 130 kHz in the experiment.
When the modulation frequency exceeds 150 kHz, the backward ASE power measured at the idle pump port of the MPC is 3 mW. With the help of a 99:1 OC inserted into the cavity at the position of YSF shown in Fig. 1, the backward ASE power measured at the input port of the MPC reaches 87 mW, and the spectrum is shown in the inset of Fig. 2(e). The spectral shape is similar to that measured at the idle pump port of the MPC, and the ASE peak is also near 1030 nm. Therefore, a YSF (Yb198) is inserted into the cavity shown in Fig. 1 and self-pumped by the backward ASE from the YDF, which can pre-amplify the feedback weak initial ASE extracted by the narrowband FBG, engendering the suppression of the ASE gain self-saturation to increase the repetition rate of Q-switched pulses.
Figures 3(a)–3(c) show the output pulse waveforms obtained by reducing duty cycle for regular pulses at different AOM modulation frequencies when an 8 cm YSF is inserted into the cavity. Compared with the results shown in Fig. 2(c), the round-trip times required for pulse establishment is also 4 at 150 kHz. However, with the YSF inserted, the amplitude of the output Q-switched pulsed is considerably increase (Fig. 3(a)), and the pulse width is narrowed from 10 ns to approximately 8.7 ns, indicating that the gain supplied by the YDF is more effectively utilized for the formation of the Q-switched pulse because ASE gain self-saturation of YDF has been efficiently suppressed because the weak initial ASE extracted by the narrowband FBG has been pre-amplified by the inserted YSF before entering the YDF. Moreover, the original ASE-like pulse at 155 kHz is varied to the Q-switched laser pulse with the insertion of the YSF (Fig. 3(b)), showing that the ASE gain self-saturation has indeed be efficiently suppressed. The repetition rate can actually reach 175 kHz with the YSF inserted, as shown in Fig. 3(c). When the repetition rate is higher than 175 kHz, the ASE-like pulses appear again (Fig. 2(d)). Thus, the repetition rate of the Q-switched pulses is increased by 25 kHz. Figure 3(d) shows the measured RF spectrum of the Q-switched pulses at the repetition rate of 175 kHz. It can be seen that the signal-to-noise ratio is only 52 dB and the 3 dB bandwidth of the fundamental frequency spectral peak is only 100 Hz (the resolution bandwidth is 50 Hz), indicating regular and stable pulses with small timing jitter are achieved. Thus, the ASE gain self-saturation can be suppressed and the repetition rate of the Q-switched pulses can be increased evidently by amplifying the feedback weak initial ASE extracted by the narrowband FBG with a length-optimized YSF inserted into the cavity and self-pumped with the backward ASE from the YDF itself. In addition, because of the short length of the inserted YSF, there is no significant difference in the pulse building time before and after inserting the YSF. However, the pulse width of the Q-switched pulses is narrowed because the insertion of YSF is equivalent to the effect that improves the cavity feedback intensity. Note that the YSF length must be reasonably optimized. It was found that, the repetition rate of the Q-switched pulses cannot be effectively increased as the gain provided by the YSF is too low when the inserted YSF is too short, or as the backward ASE from the YDF is not enough to pump the YSF completely when the inserted YSF is too long.
Figure 4(a) shows the output pulse width and energy of the laser as functions of the repetition rate when the pump power is 7 W, in which the duty cycle is reduced to the value that can eliminate the pulse missing and also deplete the photons in the cavity within a single round-trip of the built pulse, thus, the output pulse widths shown in Fig. 4(a) should be the achievable narrowest pulse width for the laser. As seen, the lower the repetition rate, the greater is the output pulse energy. The pulse energy and peak power of the 10 kHz pulses reaches 130 μJ and 15.7 kW, respectively, and the continuous wave component can be basically ignored by testing with a chopper in the experiment. The average output powers are approximately 1.3 W for the repetition rate ranging from 10 to 175 kHz. Moreover, when the repetition rate is tuned from 10 to 175 kHz, the pulse width is always between 8 and 9 ns, with the narrowest pulse of 8.3 ns. To our best knowledge, this may be the narrowest pulses achieved by an actively Q-switched all-fiber laser with AOM so far. Figure 4(b) shows the measured Q-switched pulse spectrum at the repetition rate of 175 kHz. The 3 dB bandwidth is only 0.15 nm and the signal-to-noise ratio is 50 dB. Consequently, the pulse bandwidth slightly widens owing to the nonlinearities in fiber as the repetition rate decreases . However, the 3 dB bandwidth is still below 0.16 nm in the range of 10 -175 kHz. To our knowledge, this may also be the narrowest bandwidth for the Q-switched pulses produced by an actively Q-switched all-fiber laser with AOM.
We have proposed and demonstrated a method to obtain narrow-bandwidth, narrow-pulse-width and high-repetition-rate pulses with actively Q-switched all-fiber lasers. Owing to the ASE gain self-saturation, the narrow-bandwidth Q-switched pulses are difficult to build in a short duration as the YDF cannot efficiently amplify the initial ASE feedback by a narrowband filter, resulting that narrow-pulse-width and high-repetition-rate Q-switched pulses cannot be obtained. By using a specially-designed low-reflectivity CPS in the cavity, and inserting a length-optimized YSF self-pumped by the backward ASE from the YDF to improve the amplification of the initial weak narrowband ASE, the ASE gain self-saturation can be suppressed efficiently, and thus, the Q-switched laser pulse can be established quickly, giving a narrow pulse width and high repetition rate for the Q-switched pulses. With our proposed technique, the Q-switched pulses with bandwidth of 0.15 nm, pulse width of 9 ns, and repetition rate of 175 kHz have been produced by an actively Q-switched all-fiber laser with AOM. To our knowledge, they may be the narrowest bandwidth and pulse width generated with such Q-switched all-fiber lasers so far. To produce narrower bandwidth and higher repetition-rate Q-switched pulses, it is still necessary to further suppress the ASE self-saturation.
National Natural Science Foundation of China (61377044, 61805258); National Basic Research Program of China (973 Program) (2017YFB0405100, 2017YFB0405200); Chinese Academy of Sciences (XDB21010300); State Key Laboratory of Pulsed Power Laser Technology (SKL2017KF03).
The authors acknowledge helpful discussions with Qiao Lu and Yingqiu Mao.
1. W. O’Neill and K. Li, “High-quality micromachining of silicon at 1064 nm using a high-brightness MOPA-based 20-W Yb fiber laser,” IEEE J. Sel. Top. Quantum Electron. 15(2), 462–470 (2009). [CrossRef]
2. V. Philippov, C. Codemard, Y. Jeong, C. Alegria, J. K. Sahu, J. Nilsson, and G. N. Pearson, “High-energy in-fiber pulse amplification for coherent lidar applications,” Opt. Lett. 29(22), 2590–2592 (2004). [CrossRef]
3. W. Shi, M. Leigh, J. Zong, and S. Jiang, “Single-frequency terahertz source pumped by Q-switched fiber lasers based on difference-frequency generation in GaSe crystal,” Opt. Lett. 32(8), 949–951 (2007). [CrossRef]
4. M. Siebold, J. Hein, C. Wandt, S. Klingebiel, F. Krausz, and S. Karsch, “High-energy, diode-pumped, nanosecond Yb: YAG MOPA system,” Opt. Express 16(6), 3674–3679 (2008). [CrossRef]
5. H. Zhang, X. Wang, P. Zhou, Z. Gong, and X. Xu, “6 mJ, high-average-power, all-fiberized Q-switched fiber master oscillator power amplifier with low repetition rate,” Appl. Opt. 51(29), 6933 (2012). [CrossRef]
6. K. Meng, H. Zhang, M. Liu, D. Li, P. Yan, and M. Gong, “670 kW nanosecond all-fiber super-irradiation pulsed amplifiers at high repetition rates,” J. Opt. 16(10), 105202 (2014). [CrossRef]
7. T. Ogawa, T. Imai, K. Onodera, H. Machida, M. Higuchi, Y. Urata, and S. Wada, “Efficient pulse operation of Nd:GdVO4 laser with AO Q-switch,” Appl. Phys. B: Lasers Opt. 81(4), 521–524 (2005). [CrossRef]
8. J. Liu, S. Wu, Q. H. Yang, and P. Wang, “Stable nanosecond pulse generation from a graphene-based passively Q-switched Yb-doped fiber laser,” Opt. Lett. 36(20), 4008–4010 (2011). [CrossRef]
9. U. Chakravarty, S. Gurram, A. Kuruvilla, B. N. Upadhyaya, and K. S. Bindra, “Short pulse generation in active Q-switched Yb-doped all fiber laser and its amplification,” Opt. Laser Technol. 109, 186–192 (2019). [CrossRef]
10. Y. Wang and C. Q. Xu, “Actively Q-switched fiber lasers: Switching dynamics and nonlinear processes,” Prog. Quantum Electron. 31(3-5), 131–216 (2007). [CrossRef]
11. W. C. Li, H. W. Liu, J. Zhang, B. Yao, S. J. Feng, L. Wei, and Q. H. Mao, “Mode-hopping-free single-longitudinal-mode actively Q-switched ring cavity fiber laser with an injection seeding technique,” IEEE Photonics J. 9(1), 1–7 (2017). [CrossRef]
12. Y. Zhang, C. Yang, C. Li, Z. Feng, S. Xu, H. Deng, and Z. Yang, “Linearly frequency-modulated pulsed single-frequency fiber laser at 1083 nm,” Opt. Express 24(4), 3162–3167 (2016). [CrossRef]
13. G. Ramírez-Meléndez, M. Bello-Jiménez, O. Pottiez, L. Escalante-Zarate, M. Bello-Jiménez, R. López-Estopier, B. Ibarra-Escamilla, M. Durán-Sánchez, E. A. Kuzin, and M. V. Andrés, “Q-switching of an all-fiber ring laser based on in-fiber acousto-optic bandpass modulator,” Appl. Phys. B: Lasers Opt. 123(9), 249 (2017). [CrossRef]
14. J. Swiderski, A. Zajac, M. Skorczakowski, and P. Konieczny, “Q-switched Nd-doped double-clad fiber laser,” Opto-Electron. Rev. 13(3), 187 (2005).
15. E. Desurvire, “Analysis of transient gain saturation and recovery in erbium-doped fiber amplifiers,” IEEE Photonics Technol. Lett. 1(8), 196–199 (1989). [CrossRef]
16. L. Zenteno, “High-power double-clad fiber lasers,” J. Lightwave Technol. 11(9), 1435–1446 (1993). [CrossRef]
17. J. Boullet, R. Dubrasquet, C. Médina, R. Bello-Doua, N. Traynor, and E. Cormier, “Millijoule-class Yb-doped pulsed fiber laser operatingat 977 nm,” Opt. Lett. 35(10), 1650–1652 (2010). [CrossRef]
18. J. Limpert, N. Deguil-Robin, S. Petit, I. Manek-Hönninger, F. Salin, P. Rigail, C. Hönninger, and E. Mottay, “High power Q-switched Yb-doped photonic crystal fiber laser producing sub-10 ns pulses,” Appl. Phys. B: Lasers Opt. 81(1), 19–21 (2005). [CrossRef]
19. J. Wang, S. Cui, L. Si, J. Chen, and Y. Feng, “All-fiber single-mode actively Q-switched laser at 1120 nm,” Opt. Express 21(1), 289–294 (2013). [CrossRef]
20. Q. Fang, Y. Qin, B. Wang, and W. Shi, “230 W average-power all-fiber-based actively Q-switched fiber master oscillator–power amplifier,” Appl. Opt. 52(27), 6744–6747 (2013). [CrossRef]
21. J. B. Lecourt, A. Bertrand, Y. Hernandez, and D. Giannone, “Short pulse and high repetition rate actively Q-switched all-in-fibre laser,” Electron. Lett. 46(5), 365–367 (2010). [CrossRef]
22. J. B. Lecourt, A. Bertrand, S. Guillemet, Y. Hernandez, and D. Giannone, “High repetition rate, high energy, actively Q-switched all-in-fiber laser,” Proc. SPIE 7721, 772108 (2010). [CrossRef]
23. M. A. Mahdi, S. Thirumeni, P. Poopalan, S. Selvakennedy, F. M. Adikan, W. Y. Chan, and H. Ahmad, “Effects of self-saturation in an Erbium-doped fiber amplifier,” Opt. Fiber Technol. 6(3), 265–274 (2000). [CrossRef]
24. K. Y. Ko, M. S. Demokan, and H. Y. Tam, “Transient analysis of erbium-doped fiber amplifiers,” IEEE Photonics Technol. Lett. 6(12), 1436–1438 (1994). [CrossRef]
25. E. Desurvire, Erbium-doped fiber amplifiers (Wiley-Interscience, 2002).
26. W. C. Li, H. W. Liu, J. Zhang, H. Long, S. J. Feng, and Q. H. Mao, “Q-switched fiber laser based on an acousto-optic modulator with injection seeding technique,” Appl. Opt. 55(17), 4584–4588 (2016). [CrossRef]
27. S. C. Pinault and M. J. Potasek, “Frequency broadening by self-phase modulation in optical fibers,” J. Opt. Soc. Am. B 2(8), 1318–1319 (1985). [CrossRef]