folkert wrote:
Hi,
Something odd is happening.
I've connected a garmin 18 lvc to a pc.
The system has a kernel with pps support and i verified it comes in:
root@hoei:/home/folkert# cat /sys/devices/virtual/pps/pps0/{assert,clear}
1372528130.997131540#54
1372528131.097120120#54
I'm running gpsd and it says the gps has a fix and it also has all
appropriate devices open:
gpsd 27762 nobody 7u CHR 4,65 0t0
1144 /dev/ttyS1
gpsd 27762 nobody 9r CHR 251,0 0t0
82440 /dev/pps0
gps is started with '-n'
ntpd is configurated to listen to the shared memory segments:
server 127.127.28.0 minpoll 4
fudge 127.127.28.0 refid NMEA
server 127.127.28.1 minpoll 4 prefer
fudge 127.127.28.1 refid PPS
Both nptd and gpsd look at the first 2 segments:
------ Shared Memory Segments --------
key shmid owner perms bytes nattch status
0x4e545030 0 root 600 96 2
0x4e545031 32769 root 600 96 2
0x4e545032 65538 root 666 96 1
0x4e545033 98307 root 666 96 1
0x47505344 131076 root 666 8240 1
The PPS seems to be seen by gpsd:
root@hoei:/home/folkert# strace -fp 27872 2>&1 | grep ioctl
...
[pid 27873] ioctl(9, PPS_FETCH, 0x7f1e63cc9d00) = 0
and the NMEA sentences are processed correctly:
x127.127.28.0 .NMEA. 0 l 10 16 377 0.000 -275.55 4.421
but the pps isn't:
127.127.28.1 .PPS. 0 l - 16 0 0.000 0.000 0.000
Anyone any idea?
For my "Sure" gps I have:
server 127.127.20.2 mode 18 prefer
# NB the server line is "prefer"
fudge 127.127.20.2 stratum 4 time2 0.417 refid GPSb
server 127.127.22.2
fudge 127.127.22.2 flag2 0 flag3 1 refid PPSb
The time2 value was obtained by running with noselect for a
few days and taking an average from the peer_summary mean for
the GPS. That value has since been adjusted to give a mean
closer to zero.
My MSF derived refclock uses radioclkd2 and that has:
server 127.127.28.0
fudge 127.127.28.0 stratum 1 time 1 0.024000 refid MSFa
with prefer being the pps synced server.
MSF server is about same as my other systems with rms offset of
about 0.4ms vs 3us for GPS/PPS. Also note that minpoll 4 seems
to introduces an occasional spike of >> 10us so I'm back to
not having minpoll or maxpoll specified.
David
Folkert van Heusden
_______________________________________________
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
