[Lowell Observer, Fall 1994, pg. 3]
The Navy Prototype Optical Interferometer at Lowell Observatory (NPOI for short) is nearly ready to begin stellar observations, according to Project Manager Nat White. The instrument was formally dedicated October 19.
Basic tests of the system have been completed successfully using monochromatic lasers and later white light.
The interferometer makes its measurements by interfering beams of light from several small telescopes, called siderostats, each looking at the same star at the same time. In order for the interference to take place, the optical path length from each siderostat to the detector must be exactly the same. To adjust the path length, the light is reflected down evacuated pipes containing rolling carts with various optical parts on them. Dr. White's tests indicate that these carts can be positioned to within ten nanometers (about 100 times the dimension of a typical atom).
The ability of a telescope to gather light (that is, how faint an object it can see) is determined by its collecting area — the more reflecting surface it has, the fainter it sees.
The resolving capability of a telescope (how small the details it can see), however, depends on the width of the reflecting surface. If we took a saw and began slicing on a large telescope mirror, removing the reflective surface from the center outward, we would lose light-collecting area and would not be able to see such faint objects, but we could see just as much detail, in principle, as we could when the center of the mirror was still present. The interferometer allows several separate mirrors to simulate the resolving power of a telescope having a mirror as big as the separation between the small mirrors. The fact that the "mirror" is mostly not there (the space between the small mirrors doesn't reflect) only means that the instrument can't detect very faint objects. In essense, the interferometer has a huge diameter but very little area.
in theory, we could have a few small telescopes spread around Coconino County, simulating a telescope with a main mirror miles in diameter and capable of incredible resolution. In practice, combining the light from several small telescopes coherently into a single signal is extremely difficult. It is this recombination of the light from pairs of siderostats that requires absolutely equal path lengths to the central detector. While this problem has long been solved for radio telescopes, where wavelengths are enormously longer, to do it for visible light on such a scale is a new challenge. The new interferometer used 55 lasers to measure path lengths and control the optical carts.