Optical: systems and elements – Optical amplifier – Optical fiber
Reexamination Certificate
2002-03-11
2004-05-18
Black, Thomas G. (Department: 3663)
Optical: systems and elements
Optical amplifier
Optical fiber
Reexamination Certificate
active
06738186
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to fiber amplifiers and more specifically to multi-mode clad-pumped erbium micro fiber amplifiers (EMFAs).
2. Description of the Related Art
Significant and on-going efforts are being made to improve erbium-doped fiber amplifier (EDFA) characteristics such as gain, noise figure, saturation output power, form factor and cost. To date, most EDFAs have been deployed in the long haul or ultra-long markets, in which high gain and output power are critical parameters. The deployment into metro and access markets will require mid-gain amplifiers but at a much lower cost and smaller form factor.
To pump the laser, pump radiation at or near 980 nm must be coupled into the fiber and absorbed in the fiber core. Typically a single-mode pump laser is aligned with the fiber core to couple the pump radiation directly into the core. Absorption efficiencies are high; approximately 80% in conventional long silica fibers but the precision active alignment of the laser to the core is very expensive. More recent schemes propose coupling pump radiation from a multi-mode pump laser into the cladding surrounding the core of the silica fiber (see U.S. Pat. No. 3,808,549). Clad pumping is less expensive due to relaxed alignment tolerances but less efficient at absorbing pump power into the core than core pumping.
When multi-mode clad pumping was first considered, the belief was that the effective pump absorption per unit length would be approximately reduced by the ratio of the core/cladding area. The total pump absorption efficiency would scale linearly with &agr;L where &agr; is the absorption coefficient of the core and L is the length of the fiber. Hence, to effectively absorb pump light in a double clad fiber it appeared that L and/or &agr; need only be increased as long as the pump power is sufficient to invert the gain medium over the length of the fiber.
The absorption coefficient can be increased by co-doping the core with ions with large absorption cross-section, as in Er:Yb doped systems. However, doping concentrations of Er:Yb in silica fiber are low, on the order of 0.01 wt. %, and the efficiency of the energy transfer process from Yb to Er is limited due to the high average distance between Yb and Er ions in the glass. The latter limits the rate for energy transfer. In addition, the efficiency is also limited by back transfer reactions from Er to Yb. The latter is affected by the lifetime of the state I
11/2
of Er. Fast transfer from the level I
11/2
to the level I
13/2
(lowest excited state from which radiative recombination to the ground level I
15/2
is observed) prevents important back transfer. As a result, codoping with ytterbium has only a marginal effect on the absorption of silica fiber. Thus, investigators sought to increase absorption efficiency by simply increasing the length L of lightly doped silica fibers.
However, empirical testing of lightly doped silica fiber revealed that after some normalized distance (&agr;L) the pump light absorbed in circular symmetric fibers apparently saturates (see FIG.
1
). Geometric modeling explains this as follows: assuming a uniform pump intensity distribution at the input of the fiber, the pump light can be described by an ensemble of rays of differing input positions and incidence angles. Some fraction of these rays will cross the core as they propagate down the fiber and will thus be absorbed, whereas the majority of rays actually never cross the core as they orbit the core upon propagation down the fiber and are therefore not absorbed. In wave optical terms the crossing rays that intersect the core describe the lower order fiber modes that spatially overlap the core and are strongly absorbed, whereas the non-crossing rays describe the higher order modes, or whispering gallery modes, that are mainly localized in the cladding and experience relatively little absorption. Thus, from a geometrical optics perspective, once the crossing rays, or alternatively the lower order modes, are absorbed increasing &agr; or L further will not improve absorption efficiency. Furthermore, increasing &agr; or L beyond saturation may create a situation in which the full length of the fiber is not inverted, which reduces total gain significantly and increases noise figure.
As shown in
FIG. 1
, the absorption efficiency &eegr;
10
ramps up approximately linearly with &agr;L as the lower order modes are absorbed and then saturates at &agr;L~20. The absorption curve is produced by a ray tracing simulation based on the geometric model as described in Liu et al., Optics Comm. Vol. 132, p.511-518, (1996). Based on this simulation, a fiber with a core of 6 &mgr;m diameter and an inner cladding of 100 &mgr;m, corresponding to r
0
/R
0
of 0.06, has a maximum absorbed power is 7.6% where r
0
is the radius of the core and R
0
is the radius of the inner cladding. For small ratios of r
0
/R
0
, the maximum absorbed power can be approximated to (4r
0
/&pgr;R
0
). More elaborate wave optical approaches (Kouznetsov et al., J. Opt. Soc. Am. B 18, 743 (2001)) implemented to describe double-clad fibers with circular symmetry show that the fraction of the pump absorbed by the core remains close to that predicted by the ray optics approach. Thus, to maximize amplifier gain conventional wisdom would dictate selecting &agr;L just into saturation with large L and relatively small &agr;. Silica fibers can be made very long because their propagation losses are very small, on the order of 2 dB/km.
Subsequent empirical testing of lightly doped silica fibers tens of meters in length showed that actual power absorption was better than that predicted by the ideal model. This led to the discovery that the saturation effect discussed above can be overcome by perturbing the higher order modes thereby “mode coupling” them into lower order modes that are partially absorbed in the fiber core. Such processes can be intrinsic to the glass composition or extrinsic. Intrinsic processes include light scattering processes induced by the glass arising from random perturbation in the index as discussed for instance by Garito et al, Science 281, 962 (1998), which are often insufficient to generate substantial mode coupling. Extrinsic effects include bending of the fiber as proposed in U.S. Pat. No. 4,815,079, D-shaped fibers, or fibers with an eccentric core as described in U.S. Pat. Nos. 3,729,690 and 4,815,079. Extrinsic effects are difficult to control, quite sensitive to fiber management or packaging and expensive. But as shown in
FIG. 2
, these extrinsic length effects can produce enough mode coupling to substantially increase the power absorption efficiency &eegr;
12
for &agr;L>20 if the fiber is sufficiently long, in excess of several tens of meters. In fibers this long, the dopant levels must be low, e.g. <0.1 wt % erbium, to ensure inversion over the entire fiber length.
Amplifiers that use multi-component glass fibers have been limited to single-mode core pumped configurations (see T. Nishi et al., “The amplification properties of a highly Er
+3
doped phosphate fiber,”
Jpn. J. Appl. Phys
., Vol. 31 (1992), Pt. 2, 2B, pp. L177-L179 reaches moderate, up to 1.53 wt. %, erbium doping). As discussed, multi-mode pumping requires enhanced absorption to field commercially viable amplifiers. The propagation loss in the cladding in multi-component glass, generally non-silica fibers, is too high to use long fiber lengths and take advantage of these known mode coupling effects. Thus, to get sufficient pump power into the fiber core amplifiers have used the more expensive single-mode pumps.
SUMMARY OF THE INVENTION
In view of the above problems, the present invention provides a compact, low-cost mid-gain Erbium Micro-Fiber Amplifier (EMFA).
This is accomplished by multi-mode pumping a micro fiber formed from a specialty multi-component glass and highly co-doped with 0.5 to 5 wt. % erbium and 5 to 30 wt. % ytterbium. The specialty glass exhibits a much higher core absorption coefficient &agr; than standard g
Chavez-Pirson Arturo
Jiang Shibin
Kaneda Yushi
Mendes Sergio Brito
Nguyen Dan T.
Black Thomas G.
Gifford Eric A.
Hughes Deandra M.
NP Photonics Inc.
LandOfFree
Multi-mode erbium micro fiber amplifier (EMFA) does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Multi-mode erbium micro fiber amplifier (EMFA), we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multi-mode erbium micro fiber amplifier (EMFA) will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3234229