Coherent combining of a 4 kW, eight-element fiber amplifier array

doi:10.1364/OL.36.002686

« Show journal navigation

Coherent combining of a 4 kW, eight-element fiber amplifier array

C. X. Yu, S. J. Augst, S. M. Redmond, K. C. Goldizen, D. V. Murphy, A. Sanchez, and T. Y. Fan »View Author Affiliations

Optics Letters, Vol. 36, Issue 14, pp. 2686-2688 (2011)
http://dx.doi.org/10.1364/OL.36.002686

View Full Text Article

Acrobat PDF (732 KB)

Journal Search
Article Lookup

Article Tools

Citations

Alert me when this article is cited
Export Citation/Save

Article Navigation

▸ Article Outline
– Abstract
– References and links

Article Tools
Full Text
Article Info
References
Cited By
Figures
Metrics
Download PDF

Viewing Equations in the
HTML Article

Optimal Viewing Recommendations

Full Text
Article Info
References (14)
Cited By
Figures (5)
Metrics

Abstract

Commercial $0.5 kW$ Yb-doped fiber amplifiers have been characterized and found to be suitable for coherent beam combining. Eight such fiber amplifiers have been coherently combined in a tiled-aperture configuration with 78% combining efficiency and total output power of $4 kW$ . The power-in-the-bucket vertical beam quality of the combined output is 1.25 times diffraction limited at full power. The beam-combining performance is independent of output power.

Coherent combining of Yb fiber amplifiers is a promising technique to increase the overall output power [1

C. X. Yu, J. E. Kansky, S. E. J. Shaw, D. V. Murphy, and C. Higgs, Electron. Lett. 42, 1024 (2006). [CrossRef]

, 2

S. J. Augst, T. Y. Fan, and A. Sanchez, Opt. Lett. 29, 474 (2004). [CrossRef] [PubMed]

, 3

T. Y. Fan, IEEE J. Quantum Electron. 11, 567 (2005). [CrossRef]

, 4

T. M. Shay, Opt. Express 14, 12188 (2006). [CrossRef] [PubMed]

, 5

J. Anderegg, S. Brosnan, E. Cheung, P. Epp, D. Hammons, H. Komine, M. Weber, and M. Wickham, Proc. SPIE 6102, 61020U (2006). [CrossRef]

, 6

J. Kansky, C. X. Yu, D. Murphy, S. Shaw, R. Lawrence, and C. Higgs, Proc. SPIE 6306, 63060G (2006). [CrossRef]

, 7

A. Flores, in CLEO:2011 - Laser Applications to Photonic Applications , OSA Technical Digest (CD) (Optical Society of America, 2011), paper CFE3.

, 8

Y. Ma, X. Wang, J. Leng, H. Xiao, X. Dong, J. Zhu, W. Du, P. Zhou, X. Xu, L. Si, Z. Liu, and Y. Zhao, Opt. Lett. 36, 951 (2011). [CrossRef] [PubMed]

, 9

R. Uberna, A. Bratcher, and B. G. Tiemann, IEEE J. Quantum Electron. 46, 1191 (2010). [CrossRef]

] from efficient laser sources. The most common coherent combining architecture with active phase control is to split the light from a common master oscillator (MO), feed it into multiple phase controllers and fiber amplifiers, and phase lock the outputs from all the amplifiers using an active control system. Although multikilowatt output power has been demonstrated from Yb fiber lasers whose bandwidth is several nanometers wide [10

Y. Yeong, J. K. Sahu, D. Payne, and J. Nilsson, Opt. Express 12, 6088 (2004). [CrossRef]

], narrower optical bandwidth fiber amplifiers are needed for beam- combining applications. Recently, the beam combinability of a

1.4 kW

Yb-doped non-polarization-maintaining fiber amplifier was demonstrated in which

25 GHz

linewidth was used to mitigate stimulated Brillouin scattering (SBS) effects [11

G. D. Goodno, S. J. McNaught, J. E. Rothenberg, T. S. McComb, P. A. Thielen, M. G. Wickham, and M. E. Weber, Opt. Lett. 35, 1542 (2010). [CrossRef] [PubMed]

]. Here, we report on the beam-combining characteristics of lower-power (

0.5 kW

) polarization-maintaining fiber amplifiers, which need less optical linewidth (

< 10 GHz

) to suppress SBS, easing path-length-matching requirements. Eight of these fiber amplifiers are coherently combined with high efficiency and excellent beam quality at

0.5 kW

per amplifier.

Commercial

0.5 kW

fiber amplifiers from two different manufacturers were characterized. The key measurements of interest for beam combining include output power, beam quality, preservation of spectrum, polarization purity, coherence, and phase noise. The amplifiers’ spatial outputs typically have

M^{2}

on the order of 1.1, with polarization extinction ratios of around

20 dB

To test for preservation of spectrum and coherence, these amplifiers have been seeded with one of two inputs. One is a filtered amplified-spontaneous-emission (ASE) noise source. The other is a single-frequency laser followed by an RF-noise-source-driven phase modulator [12

A. Hadjifotiou and G. A. Hill, IEEProc. J, Optoelectron. 133, 256 (1986). [CrossRef]

]. Significant power-dependent frequency generation is observed when the ASE source is used, as shown in Fig. 1a. This is because the effective refractive index

n_{eff}

is modulated by the optical intensity (I) via the nonlinear index

n_{2}

n_{eff} = n_{o} + n_{2} * I

, and causes effects such as four-wave mixing, self-phase modulation, and cross-phase modulation. Thus, any variation in optical intensity leads to a time-dependent phase and the gen eration of new frequency components. Since the phase- modulated seed has constant intensity, its nonlinear phase is also constant and does not lead to new frequen cy generation. Figure 1b shows the amplifier optical spectra at various output powers. No spectral broadening is observed. Because this amplifier exhibits strong

n_{2}

nonlinearity, we use the phase-modulated source for subsequent work described here.

To demonstrate the amplifiers’ applicability to coherent beam combining, we measure their coherence and phase noise. Figure 2a shows the coherence measurement results for one fiber amplifier. A

5 mW

10 GHz

phase-modulated optical source is used to seed the

0.5 kW

amplifier. The input and the output amplifier taps are interfered in a fiber coupler to ensure mode matching. Fringes are observed because of mechanical and thermal drifts. The 96% visibility shows that the amplifier keeps the same frequency content and preserves the signal coherence between its input and output. The remaining 4% is due to imperfect power equalization.

To ascertain that such fiber amplifiers are coherently beam combinable, one must also measure their phase noise to determine the required bandwidth of the control electronics. Because phase controllers such as electro-optic and acousto-optic modulators have low power-handling capabilities, they must be placed at the input of the high-power fiber amplifiers. Thus, any control loop will experience a loop delay that corresponds to the propagation time inside the high-power fiber amplifier. This delay sets an upper limit to the bandwidth of the feedback system for phase stabilization. The amplifier phase noise was characterized using a heterodyne measurement similar to those reported previously [2

S. J. Augst, T. Y. Fan, and A. Sanchez, Opt. Lett. 29, 474 (2004). [CrossRef] [PubMed]

]. Figure 2b shows the measured phase noise at

15 W

and at

0.5 kW

. The integrated phase noise falls off very sharply around

~ 200 Hz

. Thus, phase locking of such amplifiers requires feedback electronics with nominal kHz bandwidth and is not limited by the propagation delay inside the amplifiers themselves. Furthermore, there is virtually no difference in the phase noise between the two powers below

200 Hz

. Therefore, the phase noise is not dominated by fiber amplifier nonlinearity or by thermal effects for noise frequencies above a few Hz, which is similar to previous comparisons between low-power and high-power operation [2

S. J. Augst, T. Y. Fan, and A. Sanchez, Opt. Lett. 29, 474 (2004). [CrossRef] [PubMed]

Eight of these fiber amplifiers were beam combined in a linear-array tiled-aperture configuration, shown schematically in Fig. 3. The master oscillator output is split and sent through eight phase modulators used for phase control and adjustable delay lines used for path-length matching. The outputs from the delay lines are used as inputs to the high-power fiber amplifiers. The outputs from the fiber amplifiers are collimated by a microlens array to increase the fill factor. The output from the microlens array is sampled, with the sample going through a lens to transform to the far field. A slit is used to look at the on-axis intensity with a photodiode. The electrical signal from the photodiode is sent to the phase controller. A stochastic parallel gradient descent (SPGD)-based [13

M. A. Vorontsov and V. P. Sivokon, J. Opt. Soc. Am. A 15, 2745 (1998). [CrossRef]

] control system phases the individual fiber amplifiers with respect to each other by applying signals to the phase modulators [6

J. Kansky, C. X. Yu, D. Murphy, S. Shaw, R. Lawrence, and C. Higgs, Proc. SPIE 6306, 63060G (2006). [CrossRef]

]. SPGD is a hill-climbing algorithm that works to maximize the on-axis far-field intensity by applying appropriate sets of small dithers to the phase modulators.

The individual fiber outputs had free-space

1 mm

diam eter,

6 mm

long end-cap terminations fused to the output pigtails. These end caps had a

3 °

angle polish and an antireflection coating to reduce backreflection to the amplifier. The eight end-capped outputs were aggregated in a silicon V-groove array. This V-groove array had a pitch of

1.5 mm

. A microlens array does aperture filling to increase the fraction of power in the far-field central lobe. This microlens array has a focal length of

17.5 mm

, and an individual beam has a Gaussian beam radius of

0.5 mm

at this microlens array. The path lengths of the fiber amplifiers were matched to better than

0.1 cm

to accommodate the

10 GHz

linewidth of the system.

Figure 4a shows the far-field intensity patterns for combining the eight fibers at different power levels per fiber and shows that the far-field pattern is essentially independent of the output power. The less-than-unity fill factor leads to side lobes. At the highest power,

0.5 kW / fiber

, the fraction of power in the central lobe is 58%, while at

0.125 kW / fiber

this fraction is 57%. This compares with a calculated fractional power in the central lobe of 68%, for an output with uniform phase and the experimental near-field intensity profile. Figure 4b shows the far fields across the beam-combining direction with only a single fiber operating and with all eight fibers operating. Ideally the on-axis far-field intensity should increase as

N^{2}

, where N is the number of elements. The experimentally observed increase is 50, leading to a beam-combining efficiency of 78%. Nonidealities include power inequality among the fibers, imperfect polarization purity, imperfect path-length matching, errors in the arraying of the output fibers, causing differences in far-field pointing among the individual beams, differences in collimation caused by differences in focal length within the microlens array and in end-cap length, residual phase error, and higher-order-mode content. The estimated errors based on measurements are summarized in Table 1. An upper bound to the higher-order mode content is obtained by measuring the fiber amplifier’s wavefront quality. Some of these errors, such as polarization, are additive, while others, such as wavefront degradations, need to be rolled together in a root-sum-square (RSS) manner, so the total cannot be derived by simply either summing or RSS.

Another measure of the ideality of the beam, the fraction of the power as a function of the far-field diffraction angle, is shown in Fig. 5a. The aperture size, D, is defined as the width of the fiber array. The eight-fiber array spans nine V-grooves, with one V-groove unoccupied. Therefore, D is

1.5 mm \times 9 = 13.5 mm

. Furthermore, we assume that a beam director will be used to reshape the fiber array output into a unity-aspect-ratio beam. The three curves represent the far fields for a near field of an ideal top hat in a square aperture (uniform intensity and phase), a near field the same as our experiment but with uniform phase, and the experimental data. A common measure of beam quality in high-energy laser systems is the power-in-the-bucket (PIB) vertical beam quality (VBQ) [14

J. M. Slater and B. Edwards, Proc. SPIE 7686, 76860W (2010). [CrossRef]

], which is related to the fraction of the power within a given far-field angle compared with a reference ideal beam (the ideal top hat in this case). The PIB VBQ [14

J. M. Slater and B. Edwards, Proc. SPIE 7686, 76860W (2010). [CrossRef]

] is given by

(c / a)^{- 1 / 2}

[from Fig. 5a] at a far-field angle of

1.22 λ / D

, and for our experiment the VBQ is 1.25 times the diffraction limit. The best VBQ possible given the fill factor of the near field

(b / a)^{- 1 / 2}

is 1.10 times diffraction limited. It should be possible to improve the beam quality by going to a larger fill factor.

The final measure of performance is the dynamic response of the phase control system. This can be characterized by the rise time of the far-field on-axis intensity when the phase control system is activated, which is shown in Fig. 5b. In this experiment, the SPGD dither frequency is

300 kHz

; the rise time should be inversely proportional to the dither frequency and proportional to the number of fibers. For this eight-fiber system, the rise time is

240 μs

, which is clearly fast enough for robust phase control given the phase noise spectrum in Fig. 2b.

In summary, commercial

0.5 kW

fiber amplifiers have been characterized and shown to have performance useful for coherent beam combining. Eight such fiber amplifiers were coherently combined in a tiled-aperture linear-array configuration with close to ideal performance. The beam-combining efficiency was 78%, and the vertical beam quality was 1.25 times diffraction limited.

Acknowledgments

The authors would like to thank J. L. Daneu, A. Hare, and J. Manni for their work on this effort.

This work was sponsored by the Missile Defense Agency and Director, Defense Research and Engi neering, under Air Force Contract FA8721-05-C-0002. Opinions, interpretations, conclusions, and other recommendations are those of the authors and are not necessarily endorsed by the United States Government.

References and links

1.	C. X. Yu, J. E. Kansky, S. E. J. Shaw, D. V. Murphy, and C. Higgs, Electron. Lett. 42, 1024 (2006). [CrossRef]
2.	S. J. Augst, T. Y. Fan, and A. Sanchez, Opt. Lett. 29, 474 (2004). [CrossRef] [PubMed]
3.	T. Y. Fan, IEEE J. Quantum Electron. 11, 567 (2005). [CrossRef]
4.	T. M. Shay, Opt. Express 14, 12188 (2006). [CrossRef] [PubMed]
5.	J. Anderegg, S. Brosnan, E. Cheung, P. Epp, D. Hammons, H. Komine, M. Weber, and M. Wickham, Proc. SPIE 6102, 61020U (2006). [CrossRef]
6.	J. Kansky, C. X. Yu, D. Murphy, S. Shaw, R. Lawrence, and C. Higgs, Proc. SPIE 6306, 63060G (2006). [CrossRef]
7.	A. Flores, in CLEO:2011 - Laser Applications to Photonic Applications , OSA Technical Digest (CD) (Optical Society of America, 2011), paper CFE3.
8.	Y. Ma, X. Wang, J. Leng, H. Xiao, X. Dong, J. Zhu, W. Du, P. Zhou, X. Xu, L. Si, Z. Liu, and Y. Zhao, Opt. Lett. 36, 951 (2011). [CrossRef] [PubMed]
9.	R. Uberna, A. Bratcher, and B. G. Tiemann, IEEE J. Quantum Electron. 46, 1191 (2010). [CrossRef]
10.	Y. Yeong, J. K. Sahu, D. Payne, and J. Nilsson, Opt. Express 12, 6088 (2004). [CrossRef]
11.	G. D. Goodno, S. J. McNaught, J. E. Rothenberg, T. S. McComb, P. A. Thielen, M. G. Wickham, and M. E. Weber, Opt. Lett. 35, 1542 (2010). [CrossRef] [PubMed]
12.	A. Hadjifotiou and G. A. Hill, IEEProc. J, Optoelectron. 133, 256 (1986). [CrossRef]
13.	M. A. Vorontsov and V. P. Sivokon, J. Opt. Soc. Am. A 15, 2745 (1998). [CrossRef]
14.	J. M. Slater and B. Edwards, Proc. SPIE 7686, 76860W (2010). [CrossRef]

Table 1 Summary of Beam-Combining Inefficiencies

Nonideality	Error Magnitude	Efficiency Loss (%)
Power inequality	10% peak to peak (pp)	$< 1$
Polarization purity	$20 dB$ purity and $< 5 °$ pp fiber-to-fiber rotation	2
Path-length matching	$< 1 mm$ pp	1
Far-field pointing	13% rms relative to full-angle divergence of single fiber	10
Collimation	2% pp focal-length error	9
Residual phase	$λ / 40 rms$	2.5
Higher-order modes	$< 7 %$ fractional power	$< 7$
Total		22

Fig. 1 Optical spectra from fiber amplifiers operating at different powers (a) using ASE as the input source and (b) using the phase-modulated source as the input.

Fig. 2 (a) Fringe visibility from a fiber amplifier, (b) phase noise power spectral density and integrated phase noise at low power (

15 W

) and at full power for one of the

0.5 kW

fiber amplifiers.

Fig. 3 Schematic of fiber phasing demonstration.

Fig. 4 (a) Far-field intensity patterns at different output powers for the eight-amplifier combined array, (b) far-field intensity for a single amplifier array element and for all eight amplifier elements combined at

4 kW

Fig. 5 (a) Encircled far-field power as a function of far-field angle, with the vertical line at

1.22 λ / D

. The top-hat curve is for a uniform phase and intensity beam, the ideal curve is for a uniform phase with the experimental near-field pattern, and the measured array is the experimental data. (b) The on-axis intensity as a function of time when the phase controller is engaged.

OCIS Codes
(140.3290) Lasers and laser optics : Laser arrays
(140.3510) Lasers and laser optics : Lasers, fiber
(140.3298) Lasers and laser optics : Laser beam combining

ToC Category:
Lasers and Laser Optics

History
Original Manuscript: May 23, 2011
Manuscript Accepted: June 2, 2011
Published: July 13, 2011

Citation
C. X. Yu, S. J. Augst, S. M. Redmond, K. C. Goldizen, D. V. Murphy, A. Sanchez, and T. Y. Fan, "Coherent combining of a 4 kW, eight-element fiber amplifier array," Opt. Lett. 36, 2686-2688 (2011)
http://www.opticsinfobase.org/ol/abstract.cfm?URI=ol-36-14-2686

Sort: Author | Year | Journal | Reset

References

C. X. Yu, J. E. Kansky, S. E. J. Shaw, D. V. Murphy, and C. Higgs, Electron. Lett. 42, 1024 (2006). [CrossRef]
S. J. Augst, T. Y. Fan, and A. Sanchez, Opt. Lett. 29, 474(2004). [CrossRef] [PubMed]
T. Y. Fan, IEEE J. Quantum Electron. 11, 567 (2005). [CrossRef]
T. M. Shay, Opt. Express 14, 12188 (2006). [CrossRef] [PubMed]
J. Anderegg, S. Brosnan, E. Cheung, P. Epp, D. Hammons, H. Komine, M. Weber, and M. Wickham, Proc. SPIE 6102, 61020U (2006). [CrossRef]
J. Kansky, C. X. Yu, D. Murphy, S. Shaw, R. Lawrence, and C. Higgs, Proc. SPIE 6306, 63060G (2006). [CrossRef]
A. Flores, in CLEO:2011 - Laser Applications to Photonic Applications, OSA Technical Digest (CD) (Optical Society of America, 2011), paper CFE3.
Y. Ma, X. Wang, J. Leng, H. Xiao, X. Dong, J. Zhu, W. Du, P. Zhou, X. Xu, L. Si, Z. Liu, and Y. Zhao, Opt. Lett. 36, 951 (2011). [CrossRef] [PubMed]
R. Uberna, A. Bratcher, and B. G. Tiemann, IEEE J. Quantum Electron. 46, 1191 (2010). [CrossRef]
Y. Yeong, J. K. Sahu, D. Payne, and J. Nilsson, Opt. Express 12, 6088 (2004). [CrossRef]
G. D. Goodno, S. J. McNaught, J. E. Rothenberg, T. S. McComb, P. A. Thielen, M. G. Wickham, and M. E. Weber, Opt. Lett. 35, 1542 (2010). [CrossRef] [PubMed]
A. Hadjifotiou and G. A. Hill, IEE Proc. J, Optoelectron. 133, 256 (1986). [CrossRef]
M. A. Vorontsov and V. P. Sivokon, J. Opt. Soc. Am. A 15, 2745 (1998). [CrossRef]
J. M. Slater and B. Edwards, Proc. SPIE 7686, 76860W(2010). [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.

Click here to see a list of articles that cite this paper