Richard Eich wrote: > [EMAIL PROTECTED] wrote... >> Richard, >> >> Try clicking on this link: >> http://tf.nist.gov/timefreq/time/gps.htm >> >> You can achieve precision anywhere between a few milliseconds and a few >> nanoseconds, >> depending on how much you are willing to invest. Have you asked >> yourself how much >> precision the application needs? You can almost certainly achieve it! > > Great link, thank you. > > The app needs 500ns precision. We can set up GPS at each locality, > but there was a concern about the relative accuracy of GPS receiver > timestamps if they're sync'ing to different physical sats.
Using a proper timing receiver, and a good kernel, and some hw to let you measure interrupt latencies, would allow you to get about an order of magnitude better, i.e. on the order of 50 ns. Skipping just one of those prerequisites, i.e. the interrupt latency timer, might make it very hard to stay within your 500 ns limit. Terje -- - <[EMAIL PROTECTED]> "almost all programming can be viewed as an exercise in caching" _______________________________________________ questions mailing list [email protected] https://lists.ntp.isc.org/mailman/listinfo/questions
