|The Web Newsletter||Front Page||Livingston|
LIGO Hanford Observatory NewsFirst Fringes Seen in World's Longest Optical Resonator!
Pivotal events continue to rack up here at the Hanford Observatory. In our "Got Beam!" story, we reported on the major achievement of first laser light being fired down the beam tube. Now we report another first...
Last December we observed the first fringes from the two-kilometer-long Fabry-Perot cavity along the Y-arm. This cavity--formed by two suspended, monolithic, synthetic fused-silica mirrors, each weighing approximately 11 kg (~23 lbs)--is currently the world's largest optical resonator. Unlike smaller optical resonators, such as those used for laser cavities or optical spectrum analyzers, this cavity does not use a rigid spacer. Instead, each mirror is suspended by a fiber sling, like a child on a swing, inside a suspension cage. These cages are clamped to separate optical tables that float on vibration-isolation systems. The entire system of mirrors, suspensions and isolation are enclosed in vacuum chambers separated by two kilometers (1-1/4 miles) of evacuated beam tube.
The only connection between the ends of this optical cavity is through a control system that senses interference between the laser light reflecting from the input mirror of the cavity (in the LIGO corner station), and any laser light that has leaked into the cavity and back out again. The control system applies feedback forces to the input mirror to make minor adjustments in the 2-km mirror separation. "Minor" in this sense means a few microns (1 micron = 1/1000th of a millimeter), just enough to match the swinging mirror length to the stabilized laser frequency. At resonance, the light wave leaking out of the cavity cancels the light wave reflecting from the input mirror. Under these conditions, laser light flows freely into the cavity and the stored optical power rises until equilibrium is established. At this point, the light circulating in the cavities is about 100 times stronger than incoming laser light. The left-and-right images shown here are taken from a camera view that is looking through the end mirror (in the Y-mid station), and through the beam tube toward the input mirror (in the corner station). In the left-hand photo, the field is dark except for some faint scattered light from other sources. In the photo at right, the control system has brought the cavity into resonance with the laser light. The cavity has now filled with light and a small fraction (about 10 parts per million) of this light leaks through the end mirror.
Even though the incoming laser light, filtered by the mode cleaner, was severely reduced in power prior to being injected into the cavity, the small leakage of resonating light through the cavity's end mirror drives the camera deeply into saturation. Nonetheless, a very low power "ghost" beam, reflected off the wedged surface of the vacuum chamber view port, provides an unsaturated image of the beam as a small circular disk to the upper right of the main beam. This so-called bright fringe at the end mirror occurs in concert with a slight loss of reflected light at the input mirror. The fringe shape tells us about the alignment of the cavity. The circular disk fringes--the TEM00 mode--are bright, while the more complicated-looking modes are dim when alignment is nearly perfect. You can watch a video clip of fringes as the cavity swings into and out of resonance, by downloading the file FPfringes.AVI (warning: 438 KB!) and using your media player.
Our initial experiments "locked on" to the fringes for periods as long as half a minute or so. This is much longer than the several milliseconds it takes for light circulating inside the cavity to come into equilibrium with incoming laser light, but far less than the tens of hours that we would like to see these arm cavities stay in resonance. Still, it was long enough to derive clues about the interaction between the stabilized laser light transmitted by our mode-cleaner filter cavity and the 2-km-long arm cavity. These clues helped us discover a small "bug" in the laser stabilization control system. We found a small frequency modulation of the laser light, causing the 300 teraHertz laser frequency to vary by approximately 450 Hz at a modulation frequency of 60 Hz--the telltale "hum" of electrical power distribution systems. Most people would consider a laser stabilized to 1.5 parts per trillion to be "pretty good" but that small modulation is still about 2.5 times the frequency resolution of the arm cavity. So the 60-Hz frequency modulation of the laser was sweeping the laser into and out of resonance with the cavity, overpowering feedback loop electronics which were not designed to fight such a large frequency excursion at this modulation frequency. As we entered the holiday season, work was ongoing to track down the electrical origins of this 60 Hz "bug" and eliminate this problem. That's part of our job, shaking down the first implementation of this interferometer design! By the time we commission the 4-km interferometers--first at Livingston and then at Hanford--we expect to have all these "bugs" eliminated.
[Editor's Note: As of press time, the 60 Hz "bug" noted above has been tracked down and squashed. "Lock" periods of about two hours have been achieved, and the commissioning experiments are continuing as planned.]