First Fringes for the Keck Interferometer

The two Keck telescopes were combined for the first time as an Interferometer on the night of Monday, March 12.  This first light attempt was a complete success.  At 10:40 pm HST the first fringe lock was obtained on the star HD 61294.

The path to the interferometer from each Keck consists of the telescope itself, an adaptive optics (AO) system on the Nasmyth platform of the telescope, a coude train to get the light to the basement, some transport optics, a long delay line (stationary for these observations) and a fast delay line.  The star images (at 1.25 um wavelength) from each telescope are incident on a fast readout IR camera (Keck Angle Tracker) which sends commands to the AO system to center the star.  The two star images (at 2.2 um wavelength) are thereby centered on a single mode fiber that feeds another fast readout IR camera (the fringe tracker camera).  The fast delay lines are pre-positioned at a position where the pathlengths from the two Keck telescopes are predicted to be matched.  The fast delay lines then begin a search to find where the pathlengths are exactly matched while simultaneously tracking at a rate to compensate for the earth's rotation.  This match occurs when intensity modulation is observed on the fringe tracker camera; a peak occurs when the paths are identically matched and a minimum occurs when the paths are different by 1/2 the wavelength of the light.  When this modulation exceeds a threshold then "fringe lock" is said to have occurred.

In the case of the first observed star, HD 61294, fringe lock was found only 4 cm from where it was predicted to be.  During the engineering run consisting of the first halves of the nights of March 12 to March 14 fringe lock was achieved on a total of ~20 stars.  As we refined our model for predicting where the fringes would be found, fringe lock was found only ~ 1 mm from where the search was begun.  On each star fringe lock would occur for up to ~ 10 sec at a time and overall fringe lock would occur for about 10% of the 10 to 30 minutes the star was being fringe tracked.

Why couldn't we stay fringe locked all the time?  Fringe lock would fail when the modulation fell below a certain threshold because insufficient light was incident on the fringe tracker fiber or vibrations or seeing caused the path error between the two Kecks to exceed the coherence length of the fringe tracker.  This will be improved with time.

Engineering tests during these three nights explored the magnitude limit for fringe tracking (currently a K magnitude of 4.2 when running the camera at 5 kHz), measurements of vibrations, measurements of wavefront quality, the performance of the angle tracker to AO loop, optimizing the images on the angle tracker, the performance of the fringe tracker, measurements to determine the baseline model, and measurements of visibilities on stars of known diameters.

This engineering run was an overwhelming success.  This is just the start for Interferometry with the Kecks and it was a wonderful start.  Happy members of the JPL and CARA team participating in this Interferometer First Light success are shown below.  On the summit: (from left toright) Gautam Vasisht, Rachel Akeson, Mark Colavita, Rick Kendrick, Gerard van Belle, Robert Ligon, Peter Wizinowich, Mark Swain & Andrew Booth.

The team in Remote Ops at headquarters (not shown) included: Scott Acton, Jim Beletic, Fred Chaffee, John Gathright, David Le Mignant, Barbara Schaefer, Chuck Sorenson, Paul Swanson, Udo Wehmeier & Meg Whittle.

Keck Interferometer: progress report  (pdf file, Adobe Acrobat Reader required)

Welcome Interference Science@NASA



  Copyright 2001 California Association for Research in Astronomy