The Web Newsletter Front Page Hanford Livingston Caltech

LIGO MIT News

Phase Noise Interferometer Nearing Its Goal!

Phase Noise Interferometer Nearing Its Goal!

- Contributed by David Shoemaker

Using the five meter facility at MIT, the phase noise research effort is designed to develop and demonstrate the technology for the shot-noise limited interferometer operation at initial LIGO power levels to achieve the required phase sensitivity. We have closely followed its development in our LIGO Newsletters (please see Volume 1, Numbers 1, 3, 6 and Volume 2, Numbers 2 and 4). The research is now coming to a close with very encouraging results for LIGO. Brian Lantz will write up the experiment for his PhD thesis at MIT; other important participants are Peter Fritschel, Gabriela Gonzalez, and Haisheng Rong.

Figure 1. Frequency Noise in the PNI The first phase of the research using the Phase Noise Interferometer with Nd:YAG Infrared light was performed on a simplified configuration to allow laser tests. The system consisted of a 700 mW pre-stabilized laser (closely resembling the configuration planned for the LIGO Pre-Stabilized Laser), and two suspended in-vacuum masses forming a quiet reference Fabry-Perot. The performance of the laser, and thus the overall system, is very good. Figure 1 at right shows the frequency noise of the laser measured using the separate suspended cavity. This is a level which is satisfactory for the final phase of the Phase Noise Interferometer, enabling the next phase of research to begin: measurements of the phase noise, at the LIGO sensitivity, on a recycled suspended Michelson interferometer. It also shows that the stabilization technology is in place for LIGO.

The additional optical components (the beamsplitter and second Michelson mirror) were prepared for in-vacuum suspension (attachments, suspension balancing, and vacuum baking), with the suspension controllers accurately diagonalized, ensuring that both sensors indicate and actuators deliver pure translations. A modest data acquisition system was set up. Sensors and feedback for additional degrees of alignment freedom were installed. An Indium Gallium Arsenide photodetector-amplifier was prepared which will be used to sense the light in an early prototype test for the LIGO Length Sensing/Control photodetector.

Figure 2. PNI Spectra This final stage of the experiment was commissioned in September and has made rapid progress since that time. The complete power-recycled Michelson has reached a new sensitivity limit in phase of less than 2x10-10rad/ sqrt(Hz), as shown in Figure 2. Shot noise dominates down to 400 Hz, and corresponds to a circulating power of ~150 watts. This is to be compared to the LIGO requirement of 8x10-11rad/ sqrt(Hz), and a planned circulating power of ~300 w. The present experiment is limited by thermal focusing in the beamsplitter (a lower loss material is planned for LIGO). A modification in the geometry of the layout may allow higher sensitivities to be reached but the present sensitivity limit suffices to research problems of photodetection and light scatter at LIGO-like phase sensitivities.

At lower frequencies, research has led to significant reductions in the noise level. One large group of resonances in the best Argon-laser phase of the experiment was identified as being due to the "fins" which are attached to the optics as part of the suspension control system. These fins were changed and the resonances have moved up in frequency and down in amplitude as expected. Improvements in the frequency stabilization have reduced the contribution from that coupling. The present spectrum is limited by a noise source which fluctuates with time in a way consistent with a parasitic interferometer, and present research is targeted at identifying the source of scatter and eliminating it. We do, however, already feel that our understanding of the system is such that the LIGO design has been effectively validated.

The research will conclude by the end of February to make way for a last use of the interferometer: to help characterize the digital feedback system to be used in LIGO. In an interesting (and pleasing!) coincidence, the dynamic range, absolute light intensities, and modulation characteristics for the Phase Noise Interferometer and for the full-scale LIGO interferometer are very similar, allowing an end-to-end system test of the photodetection, conversion, processing, and all-critical digital-to-analog conversion systems. That research will run until Building 20 comes down around our ears in mid-May, and then the instrument will be decommissioned.