Coupling of light from telescope into single mode fiber
Master thesis
Permanent lenke
http://hdl.handle.net/11250/2368943Utgivelsesdato
2006Metadata
Vis full innførselSamlinger
Sammendrag
A preliminary tip/tilt correction system was designed and implemented to demonstrate the principles of tip/tilt correction of a 633nm laser beam. Two mirrors mounted on one-axis rotation stages were used, one to simulate an angular deviation and the other to make the correction. A Si quad cell was used to detect the angular deviation of the incoming laser beam and give feedback to the correction mirror. The software program LabView was used to measure and process the quad cell signals and provide for both the distortion and correction of the beam by steering the mirrors. It was found that the preliminary sysem could correct low frequency signals of +/- 0.1Hz. Although the rotation stages has limitations in how fast they can correct, they are adequate for automatic alignment and slow beam drift compensation. Then a more complex fast tip/tilt correction system was designed and built. Fast steering mirrors replaced the rotation stage mounted mirrors to make corrections in the area of several hundred Hz possible. An IR laser of 1550nm was used, and the light from the laser was coupled into a single-mode fiber. For the measuring of angular detection the beam was divided in two by a beamsplitter. An InGaAs quad cell was placed in one path while the fiber in the other. The signals were processed in a LabView program similar to the one for the preliminary system. The fiber coupling output effect fell 3dB for an angular detection of ±125μrad in x-direction and ±147μrad in y-direction, emphasizing the importance of tip/tilt correction for a fiber coupling. The quad cell was found to have a linear relation between angular defection and lateral spot movement for at least ±1mrad, which facilitates the programmed feed-back algorithm. The fast tip/tilt system as a whole was endeavored to perform but unfortunately no effective tip/tilt corrections could be made due to difficulties with the LabView programming.