The characterization and matching algorithm of LDLS-based broadband spectrum matching light source

Abstract

The spectral wavelength range is expanded to 380nm -1100nm for the laser driven light sources (LDLS) based spectral matching light source. Unlike LED-based systems, which individually adjust the output of many LEDs with various center wavelengths, the new source programmatically changes the mirror on/off pattern of a digital micromirror device (DMD) onto which the broadband LDLS light is diffracted and imaged. The LDLS-based system offers higher matching accuracy with faster switching speed due to its superior spectral resolution, higher throughput over the entire interested spectral range as well as the high switching speed of the micromirrors. The DMD's light distribution data obtained through system characterization is fed into the matching algorithm to calculate the mirror fraction and generate the appropriate mirror pattern to match the target spectrum. Thus, the matching quality relies on whether the characterization data captures the light mapping features on the DMD both spectrally and spatially with acceptable S/N. This paper presents an automatic process to characterize the DMD's light distribution spectrally and spatially. Then, the data will be used in the matching algorithm based on linear least square optimization with constraint to calculate the mirror fraction and generate the appropriate mirror pattern to minimize the difference between the calculated and the target spectrum. The impact of characterization configurations such as row and column pixel widths on the matching accuracy for the various types of target spectrum will be discussed.