OSA's Digital Library

Optics Express

Optics Express

  • Editor: C. Martijn de Sterke
  • Vol. 17, Iss. 19 — Sep. 14, 2009
  • pp: 16444–16449
« Show journal navigation

Observation of high-contrast, fast intensity noise of a continuous wave Raman fiber laser

Jochen Schröder and Stéphane Coen  »View Author Affiliations


Optics Express, Vol. 17, Issue 19, pp. 16444-16449 (2009)
http://dx.doi.org/10.1364/OE.17.016444


View Full Text Article

Acrobat PDF (574 KB)





Browse Journals / Lookup Meetings

Browse by Journal and Year


   


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools

Share
Citations

Abstract

We investigate the intensity noise of a continuous-wave Raman fiber laser based on a novel technique where the noise is sampled with a low repetition rate signal in a low walk-off Raman amplifier configuration. With this method, we experimentally demonstrate that continuous-wave Raman fiber lasers exhibit high contrast intensity fluctuations on a timescale of several tens to hundreds of GHz. In fact, the power of the Raman laser fluctuates between nearly zero and a high multiple of the average power. Numerical simulations are used to study the role of pump depletion on the noise transfer process. Our study validates models of pumps used in some recent continuous wave supercontinuum generation simulations.

© 2009 Optical Society of America

1. Introduction

Continuous wave (cw) Raman fiber lasers (RFLs) are key application components, e.g., for pumping amplifiers or for cw supercontinuum (SC) generation [1

1. A. V. Avdokhin, S. V. Popov, and J. R. Taylor, “Continuous-wave, high-power, Raman continuum generation in holey fibers,” Opt. Lett. 28, 1353–1355 (2003). [CrossRef] [PubMed]

,2

2. M. González-Herráez, S. Martín-López, P. Corredera, M. L. Hernanz, and P. R. Horche, “Supercontinuum generation using a continuous-wave Raman fiber laser,” Opt. Commun. 226, 323–328 (2003). [CrossRef]

]. They are readily available, offer wide tunability, high average output powers, and are generally considered as very stable and reliable. Their low intensity noise characteristics, however, has lately come into question. The present research was prompted by a number of recent publications on cw SC generation. In one of these, Vanholsbeeck et al found that the correct numerical modeling of the generated SC required that the cw RFL pump exhibited ultra-fast, high-contrast temporal intensity fluctuations [3

3. F. Vanholsbeeck, S. Martín-López, M. González-Herráez, and S. Coen, “The role of pump incoherence in continuous-wave supercontinuum generation.” Opt. Express 13, 6615–6625 (2005). [CrossRef] [PubMed]

]. These were modeled by considering the finite spectral width of the cw pump and by superimposing a random spectral phase on the measured spectrum. Since then, different strategies for modeling the spectrum of cw SC pump sources have been suggested, from a phase-diffusion model [4

4. M. H. Frosz, O. Bang, and A. Bjarklev, “Soliton collision and Raman gain regimes in continuous-wave pumped supercontinuum generation,” Opt. Express 14, 9391–9407 (2006). [CrossRef] [PubMed]

] to numerically simulating the actual pump laser [5

5. J. C. Travers, S. V. Popov, and J. R. Taylor, “A new model for CW supercontinuum generation,” in Conference on Lasers and Electro-Optics CLEO’2008 (The Optical Society of America, San Jose, 2008). Paper CMT3. [CrossRef]

]. These various models result however in significantly different temporal behavior of the pump source and only provide indirect indications of the true nature of the pump noise. As a matter of fact, even though existing RFL models predict relatively well the linewidth of RFLs [6

6. J.-C. Bouteiller, “Spectral modeling of Raman fiber lasers,” IEEE Photon. Technol. Lett. 15, 1698–1700 (2003). [CrossRef]

, 7

7. S. A. Babin, D. V. Churkin, A. E. Ismagulov, S. I. Kablukov, and E. V. Podivilov, “Spectral broadening in Raman fiber lasers.” Opt. Lett. 31, 3007–3009 (2006). [CrossRef] [PubMed]

], little is known about how this affects the temporal behavior of the RFL output. The linewidth of RFLs appears to be caused by four-wave mixing between the longitudinal modes of the laser, an effect similar to the beating between cavity modes in harmonically mode-locked lasers known as supermode noise [7

7. S. A. Babin, D. V. Churkin, A. E. Ismagulov, S. I. Kablukov, and E. V. Podivilov, “Spectral broadening in Raman fiber lasers.” Opt. Lett. 31, 3007–3009 (2006). [CrossRef] [PubMed]

]. Babin et al also experimentally observed mode beating on a megahertz timescale in the power spectra of some RFLs [8

8. S. A. Babin, D. V. Churkin, A. A. Fotiadi, S. I. Kablukov, O. I. Medvedkov, and E. V. Podivilov, “Relative intensity noise in cascaded-Raman fiber lasers,” IEEE Photon. Technol. Lett. 17, 2553–2555 (2005). [CrossRef]

, 9

9. S. A. Babin, V. Karalekas, P. Harper, E. V. Podivilov, V. K. Mezentsev, J. D. Ania-Castañón, and S. K. Turitsyn, “Experimental demonstration of mode structure in ultralong Raman fiber lasers,” Opt. Lett. 32, 1135 (2007). [CrossRef] [PubMed]

]. However it remains unclear if these observations apply to the cascaded cw RFLs used in the above-mentioned SC experiments where no such beating was observed at timescales below 10 GHz (see, e.g., Ref. [2

2. M. González-Herráez, S. Martín-López, P. Corredera, M. L. Hernanz, and P. R. Horche, “Supercontinuum generation using a continuous-wave Raman fiber laser,” Opt. Commun. 226, 323–328 (2003). [CrossRef]

]).

2. Experiments

Our experimental setup is as simple as a fiber Raman amplifier, similar to that described in Ref. [11

11. K. Hammani, C. Finot, J. M. Dudley, and G. Millot, “Optical rogue-wave-like extreme value fluctuations in fiber Raman amplifiers,” Opt. Express 16, 16467–16474 (2008). [CrossRef] [PubMed]

]. The cw RFL under test acts as the pump of that amplifier and is co-propagated with a weak pulsed signal from a mode-locked laser. When the walk-off between pump and signal is zero (i.e., when they propagate with equal group velocities), the individual signal pulses interact with different parts of the pump beam. If the pump exhibits intensity fluctuations, adjacent signal pulses experience different amount of gain and the pump noise is thus sampled and transferred onto the signal pulse train. By varying the pump-to-signal walk-off, we can measure the timescale of the pump fluctuations since for large velocity differences, the pump fluctuations will average out, leading to a measurable decrease in the reduced intensity noise (RIN) of the amplified signal pulses. To reliably sample the pump fluctuations, the signal should be weak enough not to suffer from nonlinear effects. Also, pump depletion must be avoided, i.e, the amplifier must be kept in the small signal gain regime, placing limitations on the signal and pump power as well as the fiber length. Finally, the temporal duration of the signal pulses must be shorter than the pump fluctuations, i.e., shorter than the coherence time of the pump.

In our experiment a passively mode-locked Er-doped fiber laser generated 4 ps signal pulses at 1534 nm with a 3.8 MHz repetition rate. The cw cascaded RFL under test is a commercially available RFL, tunable over 1430–1490 nm with 6 W maximum average power. Signal and pump were coupled together through a wavelength division multiplexer (WDM) into a large effective area fiber (LEAF). This fiber was chosen for its particular dispersion characteristics. With a zero-dispersion wavelength of about 1490 nm, group-velocity matching between pump and signal is obtained with a pump wavelength around 1444 nm. This is well within the tunability of our pump laser and also sets the peak gain of our Raman amplifier right at the signal wavelength. Additionally, the tunability of the RFL enables us to vary the pump-to-signal walk-off (a pump wavelength variation of ±10 nm induces a pump-signal walk-off of ⋍5 ps per 150 m of fiber length). A second WDM at the fiber output rejects most of the pump and allows for easy reconfiguration of the setup into a counter-propagating pump scheme. Following this, the signal is separated from the remaining pump using a prism in free space and is then measured with a fast photodiode and either an oscilloscope or an RF-spectrum analyzer.

Figures 1(a)–(b) depict the RF spectrum and temporal profile (solid black) of the signal after co-propagating through 150 m of LEAF with a 1444 nm pump, i.e., with no walk-off. Also shown are results without pump and with a counter-propagating pump. The signal had an average power of 1 mW and the pump power was 4 W. Considering a Raman gain coefficient gR ≈0.7W-1km-1, the small signal gain is 1.8 dB. The RF-spectra of the amplified signal, both in the co- and counter-propagating configurations, reflect this value: the peak at the 3.8 MHz signal repetition rate is about 2 dB larger than in the non-amplified case. The noise levels in the power spectra differ however markedly. The noise floor observed with a counter-propagating pump (dotted) is identical to that of the non-amplified signal (dashed), i.e., no noise was transferred from the pump to the signal. In contrast, with a co-propagating pump, the noise floor is raised by about 10 dB. Furthermore, apart from the two peaks spaced at ≈0.7 MHz around the main peak, which we attribute to relaxation oscillations of the RFL or possibly to spurious reflections, the noise floor has increased uniformly, i.e., the pulse-to-pulse fluctuations are completely random. The high-contrast nature of the fluctuations can be observed in Fig. 1(b) which shows 30 superimposed temporal traces of amplified (black) and non-amplified (grey) signal pulses. The gain of individual pulses varies between nearly 0 dB to >3 dB. Furthermore, as evidenced by the inset, this variation occurs between subsequent pulses. The RIN of the amplified signal can only be attributed to high-contrast intensity fluctuations of the pump. However, as these fluctuations are not evident when the RFL noise is measured using traditional methods (e.g., Ref. [3

3. F. Vanholsbeeck, S. Martín-López, M. González-Herráez, and S. Coen, “The role of pump incoherence in continuous-wave supercontinuum generation.” Opt. Express 13, 6615–6625 (2005). [CrossRef] [PubMed]

] reports a RIN of -110 dBc/Hz in the 0–1 GHz range), they must occur at much higher frequencies and we have to employ a different method to measure their timescale.

Fig. 1. (a) RF-spectra of the output signal with a co- (solid, zero walk-off) and counter-propagating (dotted) pump. (b) Overlay of 30 oscilloscope traces of pulses amplified with the co-propagating pump (black), with a single trace in inset. The dashed line in (a) and the gray lines in (b) are measured in the absence of pump. (c) Signal noise versus pump-to-signal walk-off. Fiber length is 150 m.

fc=12πDλLeff,
(1)

L eff=(1-exp(-αL))/α being the effective amplifier length, D the fiber dispersion (in ps/nm/km) and Δλ the pump-to-signal wavelength separation. Because DΔλ L eff corresponds to the walk-off time Δτ between pump and signal, we can use Fig. 1(c) to estimate the timescale of the pump fluctuations. The -3 dB walk-off is about 6 ps (half width) corresponding to a corner frequency and thus an estimate for the timescale of the fluctuations of about 25 GHz.

Care must be taken in interpreting these results. The walk-offs considered in Fig. 1(c) were obtained by varying the pump wavelength over as much as 35 nm. This leads to significant variation of the Raman gain coefficient experienced by the signal. Moreover, the transmittance of the WDM used to combine pump and signal is not uniform over that wavelength range. In addition, with 1 mWaverage signal power, pump depletion is significant. Also, we would like to investigate the noise transfer process for even larger walk-offs. We therefore conducted a second set of experiments with a longer LEAF of 1000 m. The longer fiber allows for a significantly larger walk-off for the same pump-to-signal separation. It also implies that we have to decrease the pulsed signal power to 6 µW to keep in check detrimental nonlinear effects such as self-phase-modulation. Simultaneously, the power of the cw RFL pump has to be reduced to 0.8 W to avoid spontaneous Raman scattering and keep the amplification in the small signal gain limit. As a matter of fact, increasing the fiber length is a delicate balancing act. It becomes increasingly difficult to avoid pump depletion and the very low signal power increases the influence of secondary noise sources, in particular in the free space pump-signal separation arm. For this experiment, the output signal was measured using a digital sampling oscilloscope with a 1 GHz bandwidth and a 5 GS/s sampling rate. Single traces containing trains of about 400 pulses were stored and later analyzed, allowing us to directly measure the pulse fluctuations.

The RIN measurements obtained with the 1000 m long LEAF are shown in Fig. 2(a). Although they are not immediately comparable with those of Fig. 1(c) because the amplifier has a different gain, the -3 dB walk-off of 7 ps is similar to the 6 ps measured previously. The small discrepancy can be explained by instantaneous depletion of the pump, which will be discussed in more details below. Note that here we have adjusted the pump power when varying the walk-off to keep the output signal power constant, thereby compensating the pump wavelength dependence of the gain coefficient. This compensation is very minor (<10%), however, and the good agreement between our two measures confirms that our RFL indeed exhibits high-contrast fluctuations over a bandwidth of several tens of GHz. To further illustrate the high-contrast nature of the noise and its dependence on the walk-off, Fig. 2(b) depicts the maximum peak power of our measured 400-pulse long trains normalized to the average peak power of those pulse trains versus walk-off. As expected, the highest peak power is obtained at zero walk-off, where individual pulses are amplified to a level up to 3 times that of the average amplification.

Fig. 2. (a) RIN versus walk-off with 1000 m of LEAF (circles). The result from the 150 m fiber [Fig. 1(c)], translated vertically for comparison, is superimposed (red crosses). (b) Maximum peak power normalized to the average peak power as a function of walk-off.

3. Numerical simulations

The noise transfer is modeled by two coupled equations taking into account absorption and Raman amplification and that depend on time t and position along the fiber z,

zps(z,t)=αPs(z,t)+gRPs(z,t)Pp(z,t),
(2)
[z(Δτ/L)t]Pp(z,t)=αPp(z,t)gRPs(z,t)Pp(z,t).
(3)

Here we consider that the absorption coefficient α is the same at the pump and signal wavelengths. Both the signal and pump powers, P s and P p, are functions of time t. The signal is modeled as a sech2-shaped pulse with a full-width-at-half-maximum (FWHM) of 4 ps corresponding to the mode-locked laser used in the experiment. The pump laser was modeled as a sech2-shaped spectral density with a 200 GHz FWHM matching that of our cw RFL (it changes by less than 10% in the power range we consider here), multiplied by a random phase with a uniform distribution as in Ref. [3

3. F. Vanholsbeeck, S. Martín-López, M. González-Herráez, and S. Coen, “The role of pump incoherence in continuous-wave supercontinuum generation.” Opt. Express 13, 6615–6625 (2005). [CrossRef] [PubMed]

]. The simulation time window was chosen to match the largest simulated walk-off time. We simulate the propagation of a single signal pulse at a time. The RIN is obtained by repeating the simulations 2000 times for each value of the walk-off parameter Δτ with different realizations of the random spectral phase.

Fig. 3. Signal RIN obtained by simulation of the noise transfer process in a 150 m long LEAF with parameters identical to that of the experiment for an average signal power Ps of 300 µW (solid) and 500 µW (dashed). The pump laser has a 200 GHz FWHM sech2 spectral linewidth with a random spectral phase.

4. Conclusion

We have investigated the intensity noise of a cw RFL based on a novel technique which samples the noise of the RFL under test with a low repetition rate signal in a zero walk-off Raman amplifier. Although we cannot draw a definitive conclusion as to the exact nature and timescale of the RFL fluctuations, we have shown that the RFL does exhibit high contrast intensity fluctuations faster than several tens of GHz. Our measurement is somewhat impaired by instantaneous pump depletion, which is unavoidable with high contrast pump fluctuations. Nevertheless, numerical simulations confirm our analysis and indicate that the high contrast intensity noise is probably present over a bandwidth as large as the laser linewidth. Our study further validates the pump laser model used in some recent cw SC generation simulations [3

3. F. Vanholsbeeck, S. Martín-López, M. González-Herráez, and S. Coen, “The role of pump incoherence in continuous-wave supercontinuum generation.” Opt. Express 13, 6615–6625 (2005). [CrossRef] [PubMed]

].

References and links

1.

A. V. Avdokhin, S. V. Popov, and J. R. Taylor, “Continuous-wave, high-power, Raman continuum generation in holey fibers,” Opt. Lett. 28, 1353–1355 (2003). [CrossRef] [PubMed]

2.

M. González-Herráez, S. Martín-López, P. Corredera, M. L. Hernanz, and P. R. Horche, “Supercontinuum generation using a continuous-wave Raman fiber laser,” Opt. Commun. 226, 323–328 (2003). [CrossRef]

3.

F. Vanholsbeeck, S. Martín-López, M. González-Herráez, and S. Coen, “The role of pump incoherence in continuous-wave supercontinuum generation.” Opt. Express 13, 6615–6625 (2005). [CrossRef] [PubMed]

4.

M. H. Frosz, O. Bang, and A. Bjarklev, “Soliton collision and Raman gain regimes in continuous-wave pumped supercontinuum generation,” Opt. Express 14, 9391–9407 (2006). [CrossRef] [PubMed]

5.

J. C. Travers, S. V. Popov, and J. R. Taylor, “A new model for CW supercontinuum generation,” in Conference on Lasers and Electro-Optics CLEO’2008 (The Optical Society of America, San Jose, 2008). Paper CMT3. [CrossRef]

6.

J.-C. Bouteiller, “Spectral modeling of Raman fiber lasers,” IEEE Photon. Technol. Lett. 15, 1698–1700 (2003). [CrossRef]

7.

S. A. Babin, D. V. Churkin, A. E. Ismagulov, S. I. Kablukov, and E. V. Podivilov, “Spectral broadening in Raman fiber lasers.” Opt. Lett. 31, 3007–3009 (2006). [CrossRef] [PubMed]

8.

S. A. Babin, D. V. Churkin, A. A. Fotiadi, S. I. Kablukov, O. I. Medvedkov, and E. V. Podivilov, “Relative intensity noise in cascaded-Raman fiber lasers,” IEEE Photon. Technol. Lett. 17, 2553–2555 (2005). [CrossRef]

9.

S. A. Babin, V. Karalekas, P. Harper, E. V. Podivilov, V. K. Mezentsev, J. D. Ania-Castañón, and S. K. Turitsyn, “Experimental demonstration of mode structure in ultralong Raman fiber lasers,” Opt. Lett. 32, 1135 (2007). [CrossRef] [PubMed]

10.

K. Keita, P. Delaye, R. Frey, and G. Roosen, “Relative intensity noise transfer of large-bandwidth pump lasers in Raman fiber amplifiers.” J. Opt. Soc. Am. B. 23, 2479–2485 (2006). [CrossRef]

11.

K. Hammani, C. Finot, J. M. Dudley, and G. Millot, “Optical rogue-wave-like extreme value fluctuations in fiber Raman amplifiers,” Opt. Express 16, 16467–16474 (2008). [CrossRef] [PubMed]

12.

P. Andrekson and M. Westlund, “Nonlinear optical fiber based high resolution all-optical waveform sampling,” Laser & Photonics Review , 1, 231–248 (2007). [CrossRef]

13.

C. R. S. Fludger, V. Handerek, and R. J. Mears, “Pump to signal RIN transfer in Raman fiber amplifiers.” J. Light-wave Technol. 19, 1140–1148 (2001). [CrossRef]

OCIS Codes
(060.2320) Fiber optics and optical communications : Fiber optics amplifiers and oscillators
(140.3550) Lasers and laser optics : Lasers, Raman
(270.2500) Quantum optics : Fluctuations, relaxations, and noise
(320.7100) Ultrafast optics : Ultrafast measurements
(320.6629) Ultrafast optics : Supercontinuum generation

ToC Category:
Lasers and Laser Optics

History
Original Manuscript: June 29, 2009
Revised Manuscript: August 19, 2009
Manuscript Accepted: August 20, 2009
Published: August 31, 2009

Citation
Jochen Schröder and Stéphane Coen, "Observation of high-contrast, fast intensity noise of a continuous wave Raman fiber laser," Opt. Express 17, 16444-16449 (2009)
http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-17-19-16444


Sort:  Author  |  Year  |  Journal  |  Reset  

References

  1. A. V. Avdokhin, S. V. Popov, and J. R. Taylor, "Continuous-wave, high-power, Raman continuum generation in holey fibers," Opt. Lett. 28, 1353-1355 (2003). [CrossRef] [PubMed]
  2. M. González-Herráez, S. Martín-López, P. Corredera, M. L. Hernanz, and P. R. Horche, "Supercontinuum generation using a continuous-wave Raman fiber laser," Opt. Commun. 226, 323-328 (2003). [CrossRef]
  3. F. Vanholsbeeck, S. Martín-López, M. González-Herráez, and S. Coen, "The role of pump incoherence in continuous-wave supercontinuum generation." Opt. Express 13, 6615-6625 (2005). [CrossRef] [PubMed]
  4. M. H. Frosz, O. Bang, and A. Bjarklev, "Soliton collision and Raman gain regimes in continuous-wave pumped supercontinuum generation," Opt. Express 14, 9391-9407 (2006). [CrossRef] [PubMed]
  5. J. C. Travers, S. V. Popov, and J. R. Taylor, "A new model for CW supercontinuum generation," in Conference on Lasers and Electro-Optics CLEO’2008 (The Optical Society of America, San Jose, 2008). Paper CMT3. [CrossRef]
  6. J.-C. Bouteiller, "Spectral modeling of Raman fiber lasers," IEEE Photon. Technol. Lett. 15, 1698-1700 (2003). [CrossRef]
  7. S. A. Babin, D. V. Churkin, A. E. Ismagulov, S. I. Kablukov, and E. V. Podivilov, "Spectral broadening in Raman fiber lasers." Opt. Lett. 31, 3007-3009 (2006). [CrossRef] [PubMed]
  8. S. A. Babin, D. V. Churkin, A. A. Fotiadi, S. I. Kablukov, O. I. Medvedkov, and E. V. Podivilov, "Relative intensity noise in cascaded-Raman fiber lasers," IEEE Photon. Technol. Lett. 17, 2553-2555 (2005). [CrossRef]
  9. S. A. Babin, V. Karalekas, P. Harper, E. V. Podivilov, V. K. Mezentsev, J. D. Ania-Castañón, and S. K. Turitsyn, "Experimental demonstration of mode structure in ultralong Raman fiber lasers," Opt. Lett. 32, 1135 (2007). [CrossRef] [PubMed]
  10. K. Keita, P. Delaye, R. Frey, and G. Roosen, "Relative intensity noise transfer of large-bandwidth pump lasers in Raman fiber amplifiers." J. Opt. Soc. Am. B. 23, 2479-2485 (2006). [CrossRef]
  11. K. Hammani, C. Finot, J. M. Dudley, and G. Millot, "Optical rogue-wave-like extreme value fluctuations in fiber Raman amplifiers," Opt. Express 16, 16467-16474 (2008). [CrossRef] [PubMed]
  12. P. Andrekson and M. Westlund, "Nonlinear optical fiber based high resolution all-optical waveform sampling," Laser & Photonics Review,  1, 231-248 (2007). [CrossRef]
  13. C. R. S. Fludger, V. Handerek, and R. J. Mears, "Pump to signal RIN transfer in Raman fiber amplifiers." J. Lightwave Technol. 19, 1140-1148 (2001). [CrossRef]

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.

Figures

Fig. 1. Fig. 2. Fig. 3.
 

« Previous Article  |  Next Article »

OSA is a member of CrossRef.

CrossCheck Deposited