Too much offset from GPS? - NTP

This is a discussion on Too much offset from GPS? - NTP ; Dear all I have problem with my stratum1 box. My box include: freebsd 5.2-RELEASE without PPS , ntp-4.2.0_1, Intel Pentium III, 40 G IDE, Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA. Freebsd box can get ...

+ Reply to Thread
Results 1 to 16 of 16

Thread: Too much offset from GPS?

  1. Too much offset from GPS?

    Dear all
    I have problem with my stratum1 box. My box include: freebsd
    5.2-RELEASE without PPS , ntp-4.2.0_1, Intel Pentium III, 40 G IDE,
    Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA.
    Freebsd box can get NMEA input string. My configuration of ntp.conf
    like this:

    server 127.127.20.0 mode 1 prefer minpoll 4
    server 127.127.1.0 minpoll 5

    server 69.25.96.14 minpoll 6
    server 192.43.244.18 minpoll 6
    server us.pool.ntp.org minpoll 6
    server 203.185.69.60 minpoll 6
    server 122.154.11.67 minpoll 6


    logfile /var/log/ntp
    logconfig =all
    driftfile /var/log/ntp.drift

    It's look OK. After I start my ntp. And I try to check with "ntpq
    -p", result like this

    remote refid st t when poll reach delay offset jitter
    ================================================== ============================
    *GPS_NMEA(0) .GPS. 0 l - 16 3 0.000 -7.127 7.982
    LOCAL(0) 73.78.73.84 5 l 22 32 1 0.000 0.000 0.002
    time.symmetrico .GPS. 1 u 21 64 1 248.084 167.016 0.002
    time.nist.gov .ACTS. 1 u 20 64 1 273.165 156.541 0.002
    andromeda.cs.pu .CDMA. 1 u 19 64 1 295.742 157.582 0.002
    203.185.69.60 .PPS. 1 u 18 64 1 270.912 292.165 0.002
    122.154.11.67 .INIT. 16 u - 64 0 0.000 0.000 4000.00


    My GPS have offset lower than other stratum 1 server ?? .......
    Please help me.


    Thanks you
    Chowalit

  2. Re: Too much offset from GPS?


    > I have problem with my stratum1 box. My box include: freebsd
    >5.2-RELEASE without PPS , ntp-4.2.0_1, Intel Pentium III, 40 G IDE,
    >Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA.
    >Freebsd box can get NMEA input string. My configuration of ntp.conf
    >like this:
    >
    >server 127.127.20.0 mode 1 prefer minpoll 4
    >server 127.127.1.0 minpoll 5


    Most NMEA devices (or at least all the ones I've worked with)
    send their text with a significant offset.

    You can correct that with
    fudge 27.127.20.0 time1 xxxx
    where xxxx is the offset to add in seconds.

    I forget the sign. You can work it out by trial and error.

    > remote refid st t when poll reach delay offset jitter
    >================================================== ============================
    >*GPS_NMEA(0) .GPS. 0 l - 16 3 0.000 -7.127 7.982
    > LOCAL(0) 73.78.73.84 5 l 22 32 1 0.000 0.000 0.002
    > time.symmetrico .GPS. 1 u 21 64 1 248.084 167.016 0.002
    > time.nist.gov .ACTS. 1 u 20 64 1 273.165 156.541 0.002
    > andromeda.cs.pu .CDMA. 1 u 19 64 1 295.742 157.582 0.002
    > 203.185.69.60 .PPS. 1 u 18 64 1 270.912 292.165 0.002
    > 122.154.11.67 .INIT. 16 u - 64 0 0.000 0.000 4000.00


    It looks like you are off about 160 ms.

    What I would do is use the "noselect" option on the server
    command on your NMEA line, and adjust the time1 fudge factor
    until it looks good. (Then turn off noselect.)

    If you like collecting data and makeing graphs, look into the
    statistics options.

    --
    These are my opinions, not necessarily my employer's. I hate spam.


  3. Re: Too much offset from GPS?

    chowalit.lab Chowalit Lab Linux wrote:
    > Dear all
    > I have problem with my stratum1 box. My box include: freebsd
    > 5.2-RELEASE without PPS , ntp-4.2.0_1, Intel Pentium III, 40 G IDE,
    > Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA.
    > Freebsd box can get NMEA input string. My configuration of ntp.conf
    > like this:
    >
    > server 127.127.20.0 mode 1 prefer minpoll 4
    > server 127.127.1.0 minpoll 5
    >
    > server 69.25.96.14 minpoll 6
    > server 192.43.244.18 minpoll 6
    > server us.pool.ntp.org minpoll 6
    > server 203.185.69.60 minpoll 6
    > server 122.154.11.67 minpoll 6
    >
    >
    > logfile /var/log/ntp
    > logconfig =all
    > driftfile /var/log/ntp.drift
    >
    > It's look OK. After I start my ntp. And I try to check with "ntpq
    > -p", result like this
    >
    > remote refid st t when poll reach delay offset jitter
    > ================================================== ============================
    > *GPS_NMEA(0) .GPS. 0 l - 16 3 0.000 -7.127 7.982
    > LOCAL(0) 73.78.73.84 5 l 22 32 1 0.000 0.000 0.002
    > time.symmetrico .GPS. 1 u 21 64 1 248.084 167.016 0.002
    > time.nist.gov .ACTS. 1 u 20 64 1 273.165 156.541 0.002
    > andromeda.cs.pu .CDMA. 1 u 19 64 1 295.742 157.582 0.002
    > 203.185.69.60 .PPS. 1 u 18 64 1 270.912 292.165 0.002
    > 122.154.11.67 .INIT. 16 u - 64 0 0.000 0.000 4000.00
    >
    >
    > My GPS have offset lower than other stratum 1 server ?? .......
    > Please help me.
    >
    >
    > Thanks you
    > Chowalit


    The delay values shown above suggest that you have chosen servers
    located many thousand miles away from your location! Since the
    potential error in transmitting time from server to client is one half
    of the round trip delay these would appear to be the worst possible
    choices unless they are the only available servers!


  4. Re: Too much offset from GPS?

    On 2008-03-07, Hal Murray wrote:

    >ATTRIBUTION MISSING wrote:
    >
    >>I have problem with my stratum1 box. My box include: freebsd
    >>5.2-RELEASE without PPS , ntp-4.2.0_1, Intel Pentium III, 40 G IDE,
    >>Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA.
    >>Freebsd box can get NMEA input string. My configuration of ntp.conf
    >>like this:
    >>
    >>server 127.127.20.0 mode 1 prefer minpoll 4
    >>server 127.127.1.0 minpoll 5

    >
    > Most NMEA devices (or at least all the ones I've worked with)
    > send their text with a significant offset.


    It gets even better ... NMEA sentences are output by the GPS device
    when it is not busy. This means that there is no fixed relationship
    between the top of the second and the NMEA sentence. And there will be
    significant jitter.

    Perhaps that doesn't matter in your application.

    --
    Steve Kostecke
    NTP Public Services Project - http://support.ntp.org/

  5. Re: Too much offset from GPS?

    chowalit.lab Chowalit Lab Linux wrote:

    > Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA.


    "map" suggests that it intended for navigation, not time transfer. You
    want a receiver intended for time transfer, and preferably with a 1
    pulse per second output.

    Also, you say connected with serial com. Can you confirm that that is
    not a USB serial adaptor, as USB will further compromise time measurements.

  6. Re: Too much offset from GPS?

    Steve Kostecke writes:

    >On 2008-03-07, Hal Murray wrote:


    >>ATTRIBUTION MISSING wrote:
    >>
    >>>I have problem with my stratum1 box. My box include: freebsd
    >>>5.2-RELEASE without PPS , ntp-4.2.0_1, Intel Pentium III, 40 G IDE,
    >>>Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA.
    >>>Freebsd box can get NMEA input string. My configuration of ntp.conf
    >>>like this:
    >>>
    >>>server 127.127.20.0 mode 1 prefer minpoll 4
    >>>server 127.127.1.0 minpoll 5

    >>
    >> Most NMEA devices (or at least all the ones I've worked with)
    >> send their text with a significant offset.


    >It gets even better ... NMEA sentences are output by the GPS device
    >when it is not busy. This means that there is no fixed relationship
    >between the top of the second and the NMEA sentence. And there will be
    >significant jitter.


    Ie, NMEA is good for getting the second, not for getting any better
    accuracy (well maybe 1/10 sec) If you want accurate time get a gps with PPS
    output.
    Otherwise there is no advantage of NMEA over a network ntp connection.



    >Perhaps that doesn't matter in your application.



  7. Re: Too much offset from GPS?

    chowalit.lab Chowalit Lab Linux wrote:
    > My GPS have offset lower than other stratum 1 server ?? .......


    This is, or should be, a FAQ.

    NMEA output is unusable as a time source.

    The time in an NMEA sentence tells you at what time the stated position
    etc. was measured, but not the current time. There is no time marker (no
    *now*) in NMEA.

    Experimentally finding an offset and fudging it away (hoping that the
    offset will be constant) just cannot be a substitute for a proper and
    well-defined time source.

    N

  8. Re: Too much offset from GPS?

    In article ,
    Unruh wrote:

    > Steve Kostecke writes:
    >
    > >On 2008-03-07, Hal Murray wrote:

    >
    > >>ATTRIBUTION MISSING wrote:
    > >>
    > >>>I have problem with my stratum1 box. My box include: freebsd
    > >>>5.2-RELEASE without PPS , ntp-4.2.0_1, Intel Pentium III, 40 G IDE,
    > >>>Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA.
    > >>>Freebsd box can get NMEA input string. My configuration of ntp.conf
    > >>>like this:
    > >>>
    > >>>server 127.127.20.0 mode 1 prefer minpoll 4
    > >>>server 127.127.1.0 minpoll 5
    > >>
    > >> Most NMEA devices (or at least all the ones I've worked with)
    > >> send their text with a significant offset.

    >
    > >It gets even better ... NMEA sentences are output by the GPS device
    > >when it is not busy. This means that there is no fixed relationship
    > >between the top of the second and the NMEA sentence. And there will be
    > >significant jitter.

    >
    > Ie, NMEA is good for getting the second, not for getting any better
    > accuracy (well maybe 1/10 sec) If you want accurate time get a gps with PPS
    > output.
    > Otherwise there is no advantage of NMEA over a network ntp connection.


    Well, except for the "no wires" part...

    Isaac

  9. Re: Too much offset from GPS?

    isw writes:

    >In article ,
    > Unruh wrote:


    >> Steve Kostecke writes:
    >>
    >> >On 2008-03-07, Hal Murray wrote:

    >>
    >> >>ATTRIBUTION MISSING wrote:
    >> >>
    >> >>>I have problem with my stratum1 box. My box include: freebsd
    >> >>>5.2-RELEASE without PPS , ntp-4.2.0_1, Intel Pentium III, 40 G IDE,
    >> >>>Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA.
    >> >>>Freebsd box can get NMEA input string. My configuration of ntp.conf
    >> >>>like this:
    >> >>>
    >> >>>server 127.127.20.0 mode 1 prefer minpoll 4
    >> >>>server 127.127.1.0 minpoll 5
    >> >>
    >> >> Most NMEA devices (or at least all the ones I've worked with)
    >> >> send their text with a significant offset.

    >>
    >> >It gets even better ... NMEA sentences are output by the GPS device
    >> >when it is not busy. This means that there is no fixed relationship
    >> >between the top of the second and the NMEA sentence. And there will be
    >> >significant jitter.

    >>
    >> Ie, NMEA is good for getting the second, not for getting any better
    >> accuracy (well maybe 1/10 sec) If you want accurate time get a gps with PPS
    >> output.
    >> Otherwise there is no advantage of NMEA over a network ntp connection.


    >Well, except for the "no wires" part...


    Then buy a gps with ppm. The GPS18LVC is only about $60. And you get
    microsecond accuracy rather than 1/10 sec.


    >Isaac


  10. Re: Too much offset from GPS?

    Thanks all,
    So It's wrong on my GPS (GPSmap60). Oh... So If I still use this
    GPS I should add offset with fudge option. Can I get other source from
    this GPS?

    Thank Again
    Chong

    On Sat, Mar 8, 2008 at 10:46 PM, Unruh wrote:
    > isw writes:
    >
    > >In article ,
    > > Unruh wrote:

    >
    > >> Steve Kostecke writes:
    > >>
    > >> >On 2008-03-07, Hal Murray wrote:
    > >>
    > >> >>ATTRIBUTION MISSING wrote:
    > >> >>
    > >> >>>I have problem with my stratum1 box. My box include: freebsd
    > >> >>>5.2-RELEASE without PPS , ntp-4.2.0_1, Intel Pentium III, 40 G IDE,
    > >> >>>Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA.
    > >> >>>Freebsd box can get NMEA input string. My configuration of ntp.conf
    > >> >>>like this:
    > >> >>>
    > >> >>>server 127.127.20.0 mode 1 prefer minpoll 4
    > >> >>>server 127.127.1.0 minpoll 5
    > >> >>
    > >> >> Most NMEA devices (or at least all the ones I've worked with)
    > >> >> send their text with a significant offset.
    > >>
    > >> >It gets even better ... NMEA sentences are output by the GPS device
    > >> >when it is not busy. This means that there is no fixed relationship
    > >> >between the top of the second and the NMEA sentence. And there will be
    > >> >significant jitter.
    > >>
    > >> Ie, NMEA is good for getting the second, not for getting any better
    > >> accuracy (well maybe 1/10 sec) If you want accurate time get a gps with PPS
    > >> output.
    > >> Otherwise there is no advantage of NMEA over a network ntp connection.

    >
    > >Well, except for the "no wires" part...

    >
    > Then buy a gps with ppm. The GPS18LVC is only about $60. And you get
    > microsecond accuracy rather than 1/10 sec.
    >
    >
    >
    > >Isaac

    >
    > _______________________________________________
    > questions mailing list
    > questions@lists.ntp.org
    > https://lists.ntp.org/mailman/listinfo/questions
    >


  11. Re: Too much offset from GPS?

    On 2008-03-11, chowalit.lab Chowalit Lab Linux wrote:

    > So It's wrong on my GPS (GPSmap60).


    The problem with the GPSMAP 60 is that it is the wrong type of GPS
    receiver for an application which needs highly stable time.

    There are two types of GPS receivers: mapping and timing.

    The key difference between these two type of receivers is that the
    timing receivers provide a Pulse Per Second (PPS) output which is
    typically aligned to within 1 microsecond (0.000001 sec) of UTC time.
    Higher quality timing receivers may be capable of closer (e.g.
    nanosecond) alignment.

    Both types of receivers are able to emit NMEA sentences via a serial or
    USB interface. These NMEA sentences are emitted when the receiver is
    not busy doing other things such as processing satellite data, updating
    the display, etc., and can occur at _any_ _time_ during second they are
    valid for. Time sources utilizing only NMEA sentences are not suitable
    for a Stratum 1 time server.

    It is up to you to decide if the GPSMAP 60 is good enough for your
    requirements.

    The USD249.99 GPSMAP 60 is a mapping receiver.

    The USD74.50 GPS18 LVC is a timing receiver.

    GPSMAP 60
    https://buy.garmin.com/shop/shop.do?...map60#specsTab

    GPS 18 OEM
    https://buy.garmin.com/shop/shop.do?cID=158&pID=223

    > Oh... So If I still use this GPS I should add offset with fudge
    > option.


    You may correct (i.e. fudge) the approximate offset but not the massive
    jitter inherent in the NMEA sentence(s).

    You also should enable only one of the NMEA sentences to reduce the
    receiver's workload.

    >Can I get other source from this GPS?


    No. The GPSMAP 60 is not a timing receiver and does not provide a PPS
    signal.

    --
    Steve Kostecke
    NTP Public Services Project - http://support.ntp.org/

  12. Re: Too much offset from GPS?

    chowalit.lab@gmail.com (chowalit.lab Chowalit Lab Linux) writes:

    >Thanks all,
    > So It's wrong on my GPS (GPSmap60). Oh... So If I still use this
    >GPS I should add offset with fudge option. Can I get other source from
    >this GPS?


    No idea what you are asking. GPS without a PPS (Pulse per Second) option
    CANNOT ever give very good timing accuracy. Ever. It sends the signal down
    a serial line ( which takes many many milliseconds) and does so somewhat
    randomly. Ie, it is good for maybe 1/10 second accuracy, but you should not
    trust it for more. IF you want better, get a gps with a PPS output (the
    standard is the Garmin 18LVC which needs a bit of soldering to set up for
    your system) and you can get microsecond accuracy.
    You gps is good for navigating with. Use it for what it is good for.



    >Thank Again
    >Chong


    >On Sat, Mar 8, 2008 at 10:46 PM, Unruh wrote:
    >> isw writes:
    >>
    >> >In article ,
    >> > Unruh wrote:

    >>
    >> >> Steve Kostecke writes:
    >> >>
    >> >> >On 2008-03-07, Hal Murray wrote:
    >> >>
    >> >> >>ATTRIBUTION MISSING wrote:
    >> >> >>
    >> >> >>>I have problem with my stratum1 box. My box include: freebsd
    >> >> >>>5.2-RELEASE without PPS , ntp-4.2.0_1, Intel Pentium III, 40 G IDE,
    >> >> >>>Germin GPSmap60 , connect with serial com. In GPSmap60 set as NMEA.
    >> >> >>>Freebsd box can get NMEA input string. My configuration of ntp.conf
    >> >> >>>like this:
    >> >> >>>
    >> >> >>>server 127.127.20.0 mode 1 prefer minpoll 4
    >> >> >>>server 127.127.1.0 minpoll 5
    >> >> >>
    >> >> >> Most NMEA devices (or at least all the ones I've worked with)
    >> >> >> send their text with a significant offset.
    >> >>
    >> >> >It gets even better ... NMEA sentences are output by the GPS device
    >> >> >when it is not busy. This means that there is no fixed relationship
    >> >> >between the top of the second and the NMEA sentence. And there will be
    >> >> >significant jitter.
    >> >>
    >> >> Ie, NMEA is good for getting the second, not for getting any better
    >> >> accuracy (well maybe 1/10 sec) If you want accurate time get a gps with PPS
    >> >> output.
    >> >> Otherwise there is no advantage of NMEA over a network ntp connection.

    >>
    >> >Well, except for the "no wires" part...

    >>
    >> Then buy a gps with ppm. The GPS18LVC is only about $60. And you get
    >> microsecond accuracy rather than 1/10 sec.
    >>
    >>
    >>
    >> >Isaac

    >>
    >> _______________________________________________
    >> questions mailing list
    >> questions@lists.ntp.org
    >> https://lists.ntp.org/mailman/listinfo/questions
    >>


  13. Re: Too much offset from GPS?

    Steve Kostecke writes:

    >On 2008-03-11, chowalit.lab Chowalit Lab Linux wrote:


    >> So It's wrong on my GPS (GPSmap60).


    >The problem with the GPSMAP 60 is that it is the wrong type of GPS
    >receiver for an application which needs highly stable time.


    >There are two types of GPS receivers: mapping and timing.


    >The key difference between these two type of receivers is that the
    >timing receivers provide a Pulse Per Second (PPS) output which is
    >typically aligned to within 1 microsecond (0.000001 sec) of UTC time.
    >Higher quality timing receivers may be capable of closer (e.g.
    >nanosecond) alignment.


    >Both types of receivers are able to emit NMEA sentences via a serial or
    >USB interface. These NMEA sentences are emitted when the receiver is
    >not busy doing other things such as processing satellite data, updating
    >the display, etc., and can occur at _any_ _time_ during second they are
    >valid for. Time sources utilizing only NMEA sentences are not suitable
    >for a Stratum 1 time server.


    >It is up to you to decide if the GPSMAP 60 is good enough for your
    >requirements.


    >The USD249.99 GPSMAP 60 is a mapping receiver.


    >The USD74.50 GPS18 LVC is a timing receiver.


    68.50 You are quoting the USB which is useless for timing.


    >GPSMAP 60
    >https://buy.garmin.com/shop/shop.do?...map60#specsTab


    >GPS 18 OEM
    >https://buy.garmin.com/shop/shop.do?cID=158&pID=223


    MAKE SURE you get the LVC, not the PC or USB version. Only the LVC is useful
    for timing.




  14. Re: Too much offset from GPS?

    On 2008-03-11, Unruh wrote:
    > Steve Kostecke writes:
    >
    >>The USD249.99 GPSMAP 60 is a mapping receiver.

    >
    >>The USD74.50 GPS18 LVC is a timing receiver.

    >
    > 68.50


    The exact price is irrelevant.

    The key detail here, which you overlooked in your haste, is the
    price differential between a ~$70 GPS suitable for timing and a ~$250
    suitable for geocaching.

    >You are quoting the USB which is useless for timing.


    I quite clearly wrote "GPS18 LVC". Again, the exact MSRP is not the
    point here.

    --
    Steve Kostecke
    NTP Public Services Project - http://support.ntp.org/

  15. Re: Too much offset from GPS?

    Steve Kostecke writes:

    >On 2008-03-11, Unruh wrote:
    >> Steve Kostecke writes:
    >>
    >>>The USD249.99 GPSMAP 60 is a mapping receiver.

    >>
    >>>The USD74.50 GPS18 LVC is a timing receiver.

    >>
    >> 68.50


    >The exact price is irrelevant.


    >The key detail here, which you overlooked in your haste, is the
    >price differential between a ~$70 GPS suitable for timing and a ~$250
    >suitable for geocaching.


    The key detail which you overlooked is that you quoted the price for the
    18USB, not the 18LVC, which could have confused the OP. The USB version is
    useless for timing. It is ONLY the LVC which can be used for timing.



    >>You are quoting the USB which is useless for timing.


    >I quite clearly wrote "GPS18 LVC". Again, the exact MSRP is not the
    >point here.


    No but the exact model is, and while you mentioned LVC your price was USB,
    and the OP could have thought that the USB was equivalent. It is not.



    >--
    >Steve Kostecke
    >NTP Public Services Project - http://support.ntp.org/


  16. Re: Too much offset from GPS?

    On 2008-03-11, Unruh wrote:

    > Steve Kostecke writes:
    >
    >>The key detail here, which you overlooked in your haste, is the price
    >>differential between a ~$70 GPS suitable for timing and a ~$250
    >>suitable for geocaching.

    >
    > The key detail which you overlooked is that you quoted the price for
    > the 18USB, not the 18LVC,


    That's the default price which appears on the GPS 18 OEM page.

    >which could have confused the OP.


    Really?

    > The USB version is useless for timing. It is ONLY the LVC which can be
    > used for timing.


    You're preaching to the choir. There's a reason why I own a GPS18 _LVC_.

    --
    Steve Kostecke
    NTP Public Services Project - http://support.ntp.org/

+ Reply to Thread