We also measure the overall locking range of a FP semiconductor laser, which is determined
by both frequency detuning and the external injection power at a fixed test laser bias, shown in
Fig. 2. By fitting the data, we obtain the linewidth enhancement factor of the test laser is 1.8,
which is in agreement with a measurement using an independent method based on amplified
spontaneous emission spectroscopy [3]. We will use this value in our theoretical calculation of
the modulation response.
The small-signal amplitude-modulation response of the injection-locked test signal is
measured when a dc master laser signal is injected into the test laser biased above the threshold
with small-signal modulation. Fig. 3 shows the modulation response of the injection-locked test
laser under a constant test laser bias I=30 mA at different injection powers. We show the
improvement of 3 dB modulation bandwidth of an injection-locked FP laser, which is twice of its
free-running value. The relaxation frequency is 3.5 times of its free-running value. The injected
signal reduces more unwanted fluctuations and feedback, more stimulated emission than random
spontaneous emission occurs, and enhances the peaks. For injection locking laser system, the
photon density of the slave laser is coupled in phase, which enhances the bandwidth. Our
theoretical model also requires setup coupling between the photon density and its phase terms.
This important characteristic of injection-locked semiconductor lasers is called amplitude and
phase coupling. We also point out that the small-signal modulation of injection-locked lasers still
suffers low frequency roll-off, which comes from the carrier transport effect and parasitic effect
of the bias circuit. This phenomenon is also shown in the data of distributed feedback (DFB) laser
provided by Ref. [4]. Finally, we improve the existing small-signal model for injection locking by
adding the optical confinement factor of separate-confinement-heterostructure (SCH) QW lasers,
nonlinear gain saturation of the slave laser due to the master laser, and low frequency roll-off due
to carrier transport and parasitic effects. Our model includes all relevant phenomena, either
observed experimentally or predicted theoretically and shows good agreement with our
experimental results.
1
0
1
0
5
0
5
1
4
-3
Injection locking (Exper im ent)
No injection
Injection locking (Theory)
S
i
(x10 cm )=
0
0
.34 mW
.65 mW
5
0
(
No injection)
.005
0
1
1
.32 mW
.98 mW
0
0.2
1
2
2
.79 mW
-
-5
3.5
-
-
-
10
15
20
-10
Itest=30 mA
14
-3
-15 S0=6.2x10 cm
-20
0
3
6
9
12
15
18
0
3
6
9
12
15
18
Frequency (GHz)
b)
Frequency (GHz)
a)
(
(
Fig. 3 Experiment data (a) Theoretical calculation (b) of the small-signal modulation response of
an injection-locked test laser.
REFERENCES
[
1] Y. Hong and K. A. Shore, ``Locking characteristics of a side-mode injected semiconductor
laser," IEEE J. Quantum Electron., vol. 35, pp. 1713-1717, 1999.
2] J. Horer and E. Patzak, ``Large-signal analysis of all-optical wavelength conversion using
two-mode injection-locking in semiconductor lasers," IEEE J. Quantum Electron., vol. 33, pp.
96-608, 1997.
3] T. Keating, S. H. Park, J. Minch, and S. L. Chuang, ``Optimal refractive index changes for
cross-gain and cross-phase modulation," Proc. SPIE, vol. 3283, pp. 314-322, 1998.
4] X. J. Meng, T. Chau, and M. C. Wu, ``Experimental demonstration of modulation bandwidth
[
5
[
[
enhancement in distributed feedback lasers with external light injection," Electron. Lett., vol. 34,
pp. 2031-2032,1998.