Page 511 - High Power Laser Handbook
P. 511
478 Fi b er L a s er s Pulsed Fiber Lasers 479
DPSS sources exhibit an extremely compact form factor that is
very amenable to integration in a fiber-based system, as well as a
short cavity that naturally yields pulses of ~1 ns or shorter at multi-
kilohertz repetition rates, with pulse energies ~10 μJ, while pro-
viding support for single longitudinal mode operation. However,
being passively Q switched, microchip lasers do not exhibit a
dial-in electronically controllable repetition rate, and they incur
substantial temporal jitter, unless some form of injection seeding is
implemented.
In addition to pulse control, a fundamental advantage offered
by MOPA architectures is the possibility of staging the gain, which
is a key ASE mitigation strategy (see earlier discussion). This advan-
tage also permits such architectures to optimally deploy different
gain fibers for different segments of the amplifier, which can be lev-
eraged for simultaneous maximization of efficiency and minimiza-
tion of nonlinearities. However, gain-staged MOPAs require a
plurality of components, including interstage isolators and spectral
filters or active time gates, which may amount to high complexity
and cost.
To provide a simpler alternative, considerable research and devel-
opment has been devoted to the power scaling of actively Q-switched
fiber lasers, which can, in principle, replace a MOPA with a mere
power oscillator of significantly lower parts count. In these sources, a
linear or ring fiber laser cavity incorporates an intracavity electro-optic
(such as a Pockels cell) or acousto-optic switch, according to a design
reminiscent of bulk DPSS lasers. For millijoule-pulse-energy operation,
however, the intracavity irradiance usually exceeds the capabilities
of traditional fiber-coupled telecommunications-type components;
therefore, the laser cavity must include bulk free-space modulators
and coupling optics.
An important limitation of Q-switched fiber lasers is that the
fiber, and hence the laser cavity, length usually falls multi-meter
range. Because the pulse duration in Q-switched lasers is propor-
tional to the laser cavity length (for given cavity losses and gain
30
medium inversion above transparency ), pulses in Q-switched
fiber lasers are usually tens of nanoseconds long, which hampers
applications seeking high peak powers. This problem has recently
been addressed by the implementation of rod-type PCF as the gain
medium, which provides ample energy storage in a significantly
shorter length (see Sec. 16.4 for related results). Another issue with
actively Q-switched lasers is that even though the pulse repetition
rate is electronically controlled by the driving signal operating the
intracavity switch, pulse duration (and corresponding peak power)
and repetition rate are not independently settable. The pulse-PRF
coupling stems from the fact that for a given cavity length, the
pulse duration increases exponentially as the built-up population

