i've seen that myself, NTP's compensation can almost be used as a way to measure room temperature. You can see the AC cycles. For this reason I keep the NTP server in a walk in closet that lacks heating duct. The room temp cycles slowly.
I've also read but can't remember where about a project to replace the TTL can oscillator in a PC with a better quality one driven from a OCXO. That is a lot of work for small gain. One thing about that study. I notice the author had t remove the PPS reference clock. It looks like he was unabe to get the graph to move at all with PPS attached. I'd like to see the same study done with a good PPS reference clock let connected. Likely the results would be uninteresting, just a flat graph. On Wed, May 15, 2013 at 12:51 AM, Hal Murray <[email protected]>wrote: > > [email protected] said: > > Maybe not you, but I think some people think "Oh my, NTP is not checking > the > > time very often. It must not be accurate." That is thinking backwards, > > you need the long poll interval to see small errors in rate. But this > > assumes a stable local clock. NTP balances this. > > That "balance" is assuming a classic ADEV type V curve. That assumption > works better in old non-green systems that burned lots of power in an idle > loop when they didn't have any work to do. With modern systems, it's easy > to > change the temperature significantly by starting/finishing a big chunk of > work. > > External temperature changes can also be important. Every now and then, I > like to repost this link. This seems like a good opportunity. > NTP temperature compensation > http://www.ijs.si/time/temp-compensation/ > > > > -- > These are my opinions. I hate spam. > > > > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. > -- Chris Albertson Redondo Beach, California _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
