Abstract

Pulse train instabilities have been a plaguing problem in the era of dye lasers, when coherent artifacts have frequently been misinterpreted as a repetitive train of coherent and short pulses. While the coherent artifact has often been considered an obsolete problem after dye lasers had been replaced with solid-state lasers, recent debates about self-mode-locked semiconductor lasers clearly indicate the opposite. Moreover, numerical investigations of several instability scenarios indicated that autocorrelation based pulse characterization techniques including FROG have a clear edge over those that are based on spectral interferometry [1]. In fact, it has been claimed that spectral phase interferometry for direct electric-field reconstruction (SPIDER), as the prototypical example of a spectral-interferometry based technique, measures only the coherent artifact. In the following, we show that SPIDER measurements hold some hidden information on pulse train instability, which is normally discarded in the interpretation of measured SPIDER traces. In particular, we address the case of an instability of the chirp of the pulse, i.e., fluctuations of the group delay dispersion (GDD). For this case, SPIDER shows a remarkable sensitivity whereas autocorrelation-based techniques do not appear overly susceptible to this instability.

© 2019 IEEE