[ntp:questions] Performance?

dhavey dhavey at gmail.com
Mon Dec 29 20:02:24 UTC 2008


On Dec 24, 8:09 am, k... at uiowa.edu (David R. Andersen) wrote:
> I have a Garmin 18LVC GPS set up at 9600 baud and emitting the GPGGA
> sentence.  PPS is connected to pin 1 of the DB9 (DCD pin).  I typically
> get 1-2 microsecond jitter with this setup and FreeBSD O/S.  ntp.conf
> statements to enable this clock are:
>
> server 127.127.20.1 mode 18 prefer # mode 2 enables $GPGGA operation,
> mode 16 enables 9600 baud operation
> fudge 127.127.20.1 time1 0.000 flag2 1 # flag2 0=pps edge on assert,
> 1=pps edge on clear
>
> The server mode 18 parameter does two things - enable GPGGA operation
> and enable 9600 baud operation.  Fudge flag2 should be set or cleared
> depending on your PPS signal.  time1 should be set as appropriate for
> your system.  Also, be sure to set symlinks as appropriate in /dev so
> that the clock driver can find the serial port.  With FreeBSD I have
> /dev/gps1 symlinked to /dev/cuad1.  With linux, /dev/gps1 should be
> symlinked to /dev/ttyS# where # is the number of the serial port you are
> using.
>
> Dave
>
>
>
> dhavey wrote:
> > On Dec 22, 5:41 pm, hal-use... at ip-64-139-1-69.sjc.megapath.net (Hal
> > Murray) wrote:
>
> >>> I'm using Garmin LVC's.  I used the windows config program on one of
> >>> them and changed the baud rate to 9600, turned the PPS on, and enabled
> >>> NEMA 3.x.
>
> >> The NMEA driver in ntpd is not expecting you to change the baud rate.
> >> There may be some way to configure it to use other baud rates, but I
> >> don't remember anything like that.
>
> >>> The other one I configured using cu.  I turned off all of the output
> >>> sentences except for GPGG and turned the PPS on.
>
> >> The default setup for the NMEA driver is to use the GPRMC sentence.
> >> It can use a few others.  Check the documentation or fix it to send
> >> (only) the GPRMC sentence.
>
> >> --
> >> These are my opinions, not necessarily my employer's.  I hate spam.
>
> > The performance is still really bad:
> >      remote           refid      st t when poll reach   delay
> > offset  jitter
> > ==============================================================================
> > x192.168.0.2     69.25.96.13      2 u    1   16  377    0.130
> > 4.307   0.417
> >  192.168.0.26    192.168.0.2      2 u   47   16  210    0.000    0.000
> > 130.357
> >  192.168.0.27    .STEP.          16 u    -   16    0    0.000
> > 0.000   0.000
> > x192.168.0.29    .GPS.            1 u   11   16  376    0.086
> > -199.29   0.172
> > xGPS_NMEA(0)     .GPS.            0 l    4   16  377    0.000
> > -220.73   2.039
>
> >     remote           refid      st t when poll reach   delay   offset
> > jitter
> > ==============================================================================
> >  192.168.0.26    192.168.0.2      2 u   13   16   42    0.007
> > 15.204   0.409
> >  192.168.0.27    192.168.0.29     2 u 1214   16    0    0.001
> > 40.722   0.000
> >  192.168.0.28    71.80.83.0       3 u    2   16  377    0.018
> > 199.327   0.123
> > *GPS_NMEA(0)     .GPS.            0 l    3   16  377    0.000
> > 0.605   3.320
>
> > _______________________________________________
> > questions mailing list
> > questi... at lists.ntp.org
> >https://lists.ntp.org/mailman/listinfo/questions
>
> --
>
> David R. Andersen
> Professor
> Department of Electrical and Computer Engineering
> Department of Physics and Astronomy
> The University of Iowa
> Iowa City, IA 52242
>
> Email: k... at uiowa.edu <mailto:k... at uiowa.edu>
> Tel: (319) 335-2529
> Fax: (319) 353-1115
> Web:http://dx.eng.uiowa.edu/dave/


The sym links are ok.
NTP doesn't seem to like 9600 baud with GPGGA.  Reach is still zero.
     remote           refid      st t when poll reach   delay
offset  jitter
==============================================================================
*192.168.0.29    .GPS.            1 u    2   16  376    0.085
-0.457   0.021
 GPS_NMEA(0)     .GPS.            0 l    -   64    0    0.000
0.000   0.001




More information about the questions mailing list