Slashdot Mirror
NASA To Test New Atomic Clock
edesio writes "Many satellites and spacecraft require accurate timing signals to ensure the proper operation of scientific instruments. In the case of GPS satellites, accurate timing is essential, otherwise anything relying on GPS signals to navigate could be misdirected. The third technology demonstration planned by NASA's Jet Propulsion Laboratory is the Deep Space Atomic Clock. The DSAC team plans to develop a small, low-mass atomic clock based on mercury-ion trap technology and demonstrate it in space."
... you might have heard about PHARAO, a caesium clock which is planned to fly on the ISS in 2013. Accuracy target is 1E-16.
Well, first, this is a NASA technology demonstration mission for a clock needed by JPL. That means that NASA is going to validate that this clock will work in space, so that JPL can use it where they operate (i.e., deep space, the Moon, Mars and beyond).
Second, yes, the very best modern clocks work differently with and without gravity. This is basically because the atoms used are so cold they are moving at human type velocities, and so gravity can't be ignored. The best terrestrial clocks are the fountains - take very cold atoms, moving at ~ 1 meter per second in a trap and shut off the trap. Some of the atoms (the ones that happen to be moving up) will ballistically go up, and then fall back down. (This is much like tossing your keys up 1 meter or so, and then catching them, except with single atoms.) The gravity is used to collimate the pulse of atoms going up and down, and (with timing the round trip) to select only the ultracold ones coming down. By timing the round trip, you can really select a particular set of velocities - the better constrained the velocity dispersion, the better constrained the clock read out.
NONE of that works in zero-G, and PHARAO (I am more familiar with this clock that the JPL Hg Ion one) is completely re-designed to use fountain-like ideas in a linear beam. I am not sure it would even work on the ground, and it definitely needs zero-g to meet its performance goals.