[ntp:questions] Advice about architecture

Chris Albertson albertson.chris at gmail.com
Thu Nov 10 19:14:01 UTC 2011


>>  With inexpensive
>> GPS devices, which I have, time does appear to vary about ±40 ms over the
>> day as satellites come into and go out of view.  But with stratum 1 network
>> time sources over the WAN, time accuracy appears to vary wildly in the
>> interval ±100 ms or more.

I think you bough the wrong inexpensive GPS.  I paid $18 for a GPS
receicer on eBay. (There are always many available. I did not find any
kind of rare deal.)
My $18 Moterola "UT" receiver runs with nanosecond level error and NTP
runs with microsecond level error.  I'd be a little shocked and
disapointed to see even a handfull of milliseconds or error in my
setup and with cables and other parts I might have only $50 invested
in that one system

I think your problem is a lack of a hardwired PPS.  Without that a
basic NMEA interfaced  GPS is only about as good as using a Internet
pool server and can be much worse.  NMEA can be as much as a full
second off and with some Garmin units I've seen even more than a
second.  One of my very old GPSes was continually "voted off the
island" as a false ticker.  It was useless.    The only GPS receivers
worth considering should be those with PPS that is spec's to have 1
sigma error better than 0.5 uS.   It is easy to find these for under
$20.

Chris Albertson
Redondo Beach, California


More information about the questions mailing list