This thread mentioned puilsars and the best clocks. Here are some comments from those really in the know:
====== Latest Al+ clock comparision at NIST is at better than 10^-17 level see. http://arxiv.org/abs/0911.4527 ====== > >> rich...@karlquist.com said: > >>> I vaguely remember reading that pulsars have some fantastic stability > >>> like 1E-20. [snip] > > Pulsars decelerate due to > > electrodynamic drag -- the reaction torque from the pulsar's external > > magnetic field trying to spin the plasma surrounding the pulsar, and > > generating MHD waves in this plasma. Because the plasma density > > varies randomly, the drag torque varies randomly. Some pulsars are > > subject to other torques, for example because they have close binary > > companions. > > > > Even in the absence of external torque, some pulsars change angular > > velocity abruptly because their moments of inertia change when > > "starquakes" (like earthquakes, but in the neutron crystalline solid > > body of the star) occur. > > > In early 1969, shortly after the Crab Nebula pulsar (the first so- > > called "millisecond" pulsar, with a rotation rate of of about 30 revs > > per second) had been discovered, and (IIRC) before a starquake had > > been observed, we began observing this pulsar at Arecibo Observatory > in Puerto Rico. > > > In one daily observation that > > involved about an hour of averaging, we could determine the pulsar's > > rotation phase angle with precision of about 50 microradians. (This > > was after removing the time-varying propagation group-delay due to > > plasma between us and the pulsar. To distinguish the plasma delay, we > > observed the pulsar concurrently at radio frequencies ranging from 40 > > MHz to 430 or 611 MHz.) Within a few months, we found that the pulsar > > was a rotten clock relative to the Observatory's H-P Cesium-beam- > > referenced clock, which we checked daily against the Loran-C ground- > > wave (over sea water, actually) signal from Jupiter, FL. > > > Since then, other pulsars have turned out to be more stable (and > > others less so). AFAIK, none has yet beaten a good atomic clock. > > > > Pulsar PSR1937+21, discovered in 1982, attracted attention because its > > rotation rate was about 600 revs per second. (So it was more > > deserving of the title, "millisecond pulsar.) In its first two years > > of being observed, it did not have a "glitch" or starquake. In two > > years, about 2 x 10^11 rotations were observed, and the rotation phase > > had been determined within about about 1/160th of a revolution, so the > > average spin rate was known within about a part in 10^13. However, in > > the next twenty-five years, its spin rate on a time-scale of about one > > year turned out to be no more stable than about 1 in 10^13. AFAIK. > > It's been a few years since I read anything about this pulsar. > > > > Millisecond pulsars are more stable than slower pulsars, and hope has > > been expressed that, when thousands or tens of thousands of these > > pulsars have been discovered and can be observed nearly continuously, > > the ensemble of this many pulsars will allow establishment of a pulsar- > > based standard of time having stability, on a time-scale of about one > > year, of a few parts in 10^15. AFAIK, such a standard remains a > > dream, not a reality. ==== _______________________________________________ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.