John,
This problem has come up many times. The clock selection algorithm
intersects the correctness interval to cleave the truechimers from the
falsetickers. Problem is, the two correctness intervals sometimes do not
overlap due to jitter. See the tos mindist command to set the minimum
distance to something larger than 5 ms, say 10 ms.
Dave
John Ackermann N8UR wrote:
Sorry, couldn't resist the subject line...
I am using a HP Z3801A GPS-disciplined oscillator with both the hpgps
and atom drivers in order to use the PPS signal as well as the serial
data stream.
I'm finding that the system keeps switching sync sources, like so:
10 Sep 15:11:41 ntpd[42461]: synchronized to PPS(0), stratum 0
10 Sep 15:13:51 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
10 Sep 15:14:54 ntpd[42461]: synchronized to PPS(0), stratum 0
10 Sep 15:15:58 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
10 Sep 15:17:03 ntpd[42461]: synchronized to PPS(0), stratum 0
10 Sep 15:19:10 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
I suspect this may be due to the (relatively) much greater jitter on the
serial data stream. Here's the current billboard from ntpq -p:
$ ntpq -p
remote refid st t when poll reach delay offset jitter
==============================================================================
+GPS_HP(0) .GPS. 0 l 11 64 377 0.000 -0.620 3.171
oPPS(0) .PPSZ. 0 l 9 64 255 0.000 0.001 0.004
Here is my current config file (with logging and other irrelevant stuff
pruned out):
$ cat /etc/ntp.conf
# /etc/ntp.conf for tock, using Z3801A and PPS
enable pps
# Z3801A
# note: "ttl 1" sets comm params for Z3801A
server 127.127.26.0 ttl 1 prefer
fudge 127.127.26.0 time1 -0.960
fudge 127.127.26.0 refid GPS
# PPS
server 127.127.22.0
fudge 127.127.22.0 refid PPSZ
fudge 127.127.22.0 flag3 1
fudge 127.127.22.0 true
Any suggestions on how I can force ntpd to lock to the PPS source and
avoid this ping-pong effect? Any other suggestions on improving the
configuration in general? (By the way -- this is on FreeBSD 5.4 and ntp
4.2.0a).
Thanks,
John
_______________________________________________
questions mailing list
[email protected]
https://lists.ntp.isc.org/mailman/listinfo/questions
_______________________________________________
questions mailing list
[email protected]
https://lists.ntp.isc.org/mailman/listinfo/questions
