ABSTRACT
In a mode-locked laser, all the longitudinal modes of the cavity are locked in phase and become in phase once every cavity round-trip time. The coherent interference of a large number of longitudinal modes spreading over a wide spectrum leads to a very short pulse every round-trip time. Mode locking was rst observed in Q-switched lasers [1]. It can be achieved either by amplitude or frequency modulation in an active mode-locked laser. An amplitude modulator forces pulse formation by opening a temporal gate once in a round-trip time. Both amplitude and frequency modulation can also be seen in the spectral domain to cause power coupling among longitudinal modes, which locks them in phase. A saturable absorber, which has low transmission at low intensity and high transmission at high intensity, is often used in passive mode-locked lasers. In this case, a pulse is formed by continuously attenuating the low intensity wings on each passage through the saturable absorber. Eventually nite gain bandwidth limits the pulse from further shortening and a steady state is reached. Pulse formation can also be achieved by nonlinear soliton dynamics in an anomalous dispersion ber in a passive mode-locking scheme. A saturable absorber is used in most passive mode-locked lasers for pulse stability against CW lasing. The pulse width in an active mode-locked laser is limited by the modulation speed. This is not the case in passive mode-locked lasers, where the temporal gate is effectively determined by the pulse itself and continuous pulse shaping takes place. Much shorter pulses can be obtained this way.
