Optical: systems and elements – Optical amplifier – Multiple pass
Reexamination Certificate
2001-04-27
2004-12-14
Black, Thomas G. (Department: 3663)
Optical: systems and elements
Optical amplifier
Multiple pass
C359S333000
Reexamination Certificate
active
06831779
ABSTRACT:
BACKGROUND OF THE INVENTION
Free-space optical transmitters are typically required to provide a high-power, single-polarization output. An optical amplifier capable of providing a high-power, single-polarization output is presented in a paper, Hakimi et al., “High-power Single-polarization EDFA with Wavelength Multiplexed pumps”, CLEO '98, CWK1, 1998.
High-power optical amplifiers, such as erbium-doped fiber amplifiers (EDFA's), are often employed in free-space communications. Single-polarization EDFA's provide additional capabilities for polarization diplexing and improved communication performance, since orthogonally polarized amplified spontaneous emission (ASE) can be eliminated at both receiver and transmitter. Such features can offer substantial benefit in free-space links, where an improvement in single-polarization receiver sensitivity directly reduces required transmitter power.
High gain enables a power amplifier to output its maximum saturated output power over an extended range of input power levels. In master oscillator power amplifier (MOPA) designs, this feature makes the transmitter less sensitive to insertion loss changes in the elements leading up to the power amplifier. Furthermore, saturated EDFA's are average power limited (APL), and, therefore, peak output power is inversely proportional to the duty-cycle of the input.
FIG. 1
is a schematic diagram of the erbium-doped fiber amplifier (EDFA)
100
a
by Hakimi et al., which is a two-stage, double-pass, polarization-maintaining EDFA with a high-power, saturated output composed of reliable, commercially available components.
FIGS. 2-5
further illustrate the EDFA of FIG.
1
.
Referring now to
FIG. 1
, a 1550 nm input optical signal
105
is coupled to a standard, single-mode, optical fiber
110
via a polarization beam splitter (PBS)
107
. The input signal
105
travels in a forward pass through the optical fiber
110
, among other optical media. The EDFA
100
a
of
FIG. 1
achieves a high output power level by multiplexing four pump lasers
125
together with polarization-insensitive, dense, WDM, fused tapered couplers
120
. See M. N. McLandrich et al., J. Lightwave Technol. 9, 442-447 (1991). The pump diode laser wavelengths range from 970 nm to 985 nm, with 5 nm of separation between each. According to Hakimi et al., at the recommended operating current, the nominal power of each pump before the WDM combiner is 90 mW, and the output pump power from each set of four combined pump lasers is 330 mW. The output from the dense WDM
120
is coupled to the optical medium, carrying the input signal
105
, by a standard WDM
115
.
In the forward pass, the output from the WDM
115
(i.e., the input signal
105
and signals from the diode pumps
125
) encounters an erbium-doped fiber
130
. The 970 nm-985 nm outputs from the four diode pumps
125
get absorbed by the erbium-doped fiber
130
, increasing the energy level in the erbium-doped fiber
130
. Unlike the 970 nm-985 nm outputs from the diode pumps
125
, the 1550 nm input signal
105
passes through the erbium-doped fiber
130
, increasing in power as a result of encountering the erbium-doped fiber
130
charged by the four diode pumps
125
. Because the input signal
105
is amplified by the erbium-doped fiber
130
, the erbium-doped fiber
130
is often referred to as an optical gain medium.
FIG. 2
is a schematic diagram representing how the erbium-doped fiber
130
, in combination with the output energy from the four diode pumps
125
, provides gain to amplify the input signal
105
. As stated above, the pump energy
205
is absorbed by the erbium-doped fiber
130
. The energy absorption representation
210
indicates that the energy in the erbium-doped fiber
130
reaches a peak level, Emax. The energy relaxes from its maximum potential energy level to a nominal potential energy level, Enom, as indicated by an energy relaxation representation
215
. In this manner, pump energy is used to transfer energy from the ground state, Emin, to the excited state Enom, creating an energy inversion. Enom corresponds to the energy of the input signal wavelength. Then, when the input signal
105
encounters the excited-state erbium-doped fiber
130
, stimulated emission occurs, as indicated by a stimulated emission representation
220
.
As a result of the release of optical energy resulting from the stimulated emission process, a signal photon
225
of 1.5 &mgr;m wavelength entering the energy-enhanced erbium-doped fiber
130
exits as multiple photons
230
of the same wavelength, thus amplifying the signal.
Referring again to
FIG. 1
, a second WDM
115
and erbium-doped fiber
130
are encountered by the input signal having been once amplified. Following the second erbium-doped fiber
130
, the twice amplified input signal encounters a Faraday mirror
135
, causing the twice amplified input signal to travel in a reverse pass with a 90 degree polarization rotation. The signal traveling in the reverse pass is amplified a third and a fourth time, as it traverses through the optical gain mediums
115
to the polarization beam splitter
107
. The output
140
a
is the saturated output from the EDFA
100
a.
The input port of the EDFA
100
a
, consisting of the fiber-pigtailed polarization beam splitter
110
, has a 0.4-dB port-to-port loss. The Er-fiber (i.e., erbium-doped fiber) splice losses, Faraday mirror loss, and 980/1550 WDM losses are 0.1 dB, 0.33 dB, and 0.1 dB, respectively. Each amplifier stage uses approximately 15 meters of conventional erbium-doped fiber.
FIG. 3
shows the output power verses the input signal power for the EDFA
100
a
measured after the polarization beam splitter
107
. The curve defined by the triangles is a result of the energy pumps providing 330 mW/stage (triangles). In this case, the output power of the EDFA
100
a
is just over 255 mW at 1556 nm, which is near the peak of the gain. When the pump lasers are turned up above the recommended operating point to 400 mW/stage (diamonds), the saturated output power is 315 mW.
FIG. 4
illustrates the wavelength dependence of the output of the EDFA
100
a
. A 1 mW input signal and 330 mW/stage pump power are held constant in generating the response curve. The output power is above 240 mW over a 30 nm range.
FIG. 5A
is a transfer function
500
of a typical amplifier, of which the EDFA
100
a
is a member. The transfer function has two regions: a small signal gain region
505
and a saturated output gain region
510
. The gain curve
515
in the small signal gain region
505
increases at a typical rate, where Pout equals g
o
*Pin, and g
o
is the small signal gain. Note that, in dB, the linear expression Pout=g
o
*Pin transforms to Pout [dB]=(g
o
+Pin) [dB]. In the saturated output gain region
510
, the curve
515
asymptotically increases to Psat_out. The EDFA
100
a
of
FIG. 1
operates entirely in the saturated output gain region
510
for input power levels about −15 dBm. However, for input power levels of below −15 dBm, the EDFA
100
a
becomes unstable, oscillating instead of outputting a constant power level for an input signal of constant power level. The EDFA
100
a
does not simply revert to a small signal gain amplifier, as might be suggested by the transfer function, because the EDFA
100
a
is designed only to operate stably for input power levels that drive the amplifier far into the saturated output gain region.
FIG. 5B
is a transfer function of gain versus output power corresponding to the transfer function of FIG.
5
A.
SUMMARY OF THE INVENTION
A fault-tolerant, loss-insensitive region in a two-stage, double-pass, polarization maintaining (PM) EDFA has been identified, in which assertion of optical elements can be used for, among other reasons, to improve stability, output power, and efficiency of the EDFA. Employing the principles of the present invention, stable, greater than 0.5 Watt output power levels are obtainable for input power levels to the EDFA spanning over a 30 dB dynamic range, and
Black Thomas G.
Cunningham Stephen
Hamilton Brook Smith & Reynolds P.C.
Massachusetts Institute of Technology
LandOfFree
Method and apparatus for stabilizing a high-gain, high-power... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for stabilizing a high-gain, high-power..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for stabilizing a high-gain, high-power... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3287664