Progress Towards Closed-Loop Control of Dynamic Spectrum Flattener

Abstract

Laser frequency combs (LFCs) can be used to calibrate high-resolution spectrographs to detect and characterize exoplanets. Spectral shaping serves an important role in increasing the signal-to-noise ratio (SNR) for effective detection. However, the large range of intensity across the LFC spectrum makes it infeasible to optimally use the entire comb; this leads to a reduced overall precision of calibration. To avoid this, setups are used to flatten the output spectrum before sending it to the spectrograph; these setups usually involve bulk optics. This research uses an all-photonic on-chip spectral shaper that flattens the spectrum of an LFC. The goal of my thesis is to improve the extraction, reduction, fitting, and subsequent control in an algorithmic manner with this spectral shaper to achieve a flat spectrum of an LFC. To flatten the power spectrum, I apply voltage changes to Mach-Zehnder Interferometers (MZIs) and Thermo-Optic Phase Modulators (TOPMs) at different spectral channels. I develop linear, logarithmic, and hybrid control-loop algorithms and test each model’s performance. I also test if the models are capable of converging to different reference power lines. I conduct field testing at the California Institute of Technology in Pasadena, CA. I find that the linear model performs the best since it reaches a defined power threshold and has sufficient convergence.