Leap Second GPS Observation

Discussion in 'Global Navigation Satellite Systems' started by Harold Rabbie, Jan 1, 2006.

  1. The NMEA-0183 sentence $GPRMC that my GPS receiver produced around the time
    of yesterday's leap second repeated the 235959 time stamp. In other words,
    it went:

    235958 311205
    235959 311205
    235959 311205 --> the leap second
    000000 010106

    I would have expected a unique value such as 235960 for the leap second
    sentence. On the other hand, I don't think that NMEA-0183 specifies this
    case. The receiver is a Navman, which contains a Rockwell chip set.
     
    Harold Rabbie, Jan 1, 2006
    #1
    1. Advertisements

  2. Hi Harold!

    I would expect that too. In my case Garmin V had shown 00:00:00 twice.
    I didn't hook it up to the computer, I photographed it on the balocony.

    http://members.cox.net/sylvek/leap/

    New Year cheers,

    Sylvek
     
    Szpieg z Krainy Deszczowców, Jan 1, 2006
    #2
    1. Advertisements

  3. Harold Rabbie

    Bob Greschke Guest

    My Garmin 76CS only put out NMEA stuff every 2 seconds, but it went from
    reporting every odd second (55,57,59) to even seconds after midnight
    starting at 00:00:02, so I'm not sure what it would have said at midnight.

    I had my Etrex Vista C recording a track point every second. No extra
    second shows up in that at all (57,58,59,00,01,02...).

    The display on both units "paused" at 00:00:00 during the extra second.

    Something like 23:59:60 in a string or integer variables in a program that
    uses time could break a lot of time error-checking routines. There just
    shouldn't be more than 60 seconds (0-59) in a minute. :) And to allow 61
    seconds all of the time would make for a bad error-checking routine.

    if Seconds < 0 or Seconds > 59:
    Catch errors most of the time except when there's a leap second

    if Seconds < 0 or Seconds > 60:
    Allow errors most of the time except when there is a leap second

    Bob
     
    Bob Greschke, Jan 1, 2006
    #3
  4. Hi,
    I have logs of my GPS for 31st night. I shall put them someplace
    soon. I observed, my receiver received the leap second at 235946. I was
    expecting the addition at 235959. I have GPRMC logs indicating the
    time.

    Any inputs ?

    Happy New Year,
    Akshay.
     
    Akshay Mishra, Jan 2, 2006
    #4
  5. I have put the reception logs below. The 235946 was sent twice.

    $GPRMC,235940.965,A,1907.4550,N,07254.9128,E,0.56,163.20,311205,,*0C
    $GPRMC,235941.965,A,1907.4548,N,07254.9129,E,0.49,158.81,311205,,*08
    $GPRMC,235942.965,A,1907.4546,N,07254.9131,E,0.43,159.00,311205,,*0E
    $GPRMC,235943.965,A,1907.4544,N,07254.9132,E,0.42,158.94,311205,,*03
    $GPRMC,235944.965,A,1907.4542,N,07254.9133,E,0.45,161.51,311205,,*07
    $GPRMC,235945.965,A,1907.4539,N,07254.9135,E,0.64,165.86,311205,,*01
    $GPRMC,235946.965,A,1907.4537,N,07254.9136,E,0.45,162.46,311205,,*07
    $GPRMC,235946.965,A,1907.4535,N,07254.9137,E,0.41,170.44,311205,,*01
    $GPRMC,235947.965,A,1907.4533,N,07254.9138,E,0.48,164.70,311205,,*02
    $GPRMC,235948.965,A,1907.4531,N,07254.9139,E,0.36,157.18,311205,,*09
    $GPRMC,235949.965,A,1907.4530,N,07254.9140,E,0.32,156.89,311205,,*0A
    $GPRMC,235950.965,A,1907.4528,N,07254.9141,E,0.36,158.76,311205,,*00
    $GPRMC,235951.965,A,1907.4526,N,07254.9142,E,0.30,157.16,311205,,*03
    $GPRMC,235952.965,A,1907.4524,N,07254.9142,E,0.31,157.71,311205,,*02
    $GPRMC,235953.964,A,1907.4522,N,07254.9143,E,0.34,158.91,311205,,*01
    $GPRMC,235954.964,A,1907.4519,N,07254.9144,E,0.38,161.32,311205,,*06
    $GPRMC,235955.964,A,1907.4516,N,07254.9145,E,0.34,162.59,311205,,*0B
    $GPRMC,235956.964,A,1907.4514,N,07254.9145,E,0.34,158.89,311205,,*0E
    $GPRMC,235957.964,A,1907.4511,N,07254.9146,E,0.33,160.12,311205,,*07
    $GPRMC,235958.964,A,1907.4508,N,07254.9147,E,0.39,163.15,311205,,*0F
    $GPRMC,235959.964,A,1907.4505,N,07254.9147,E,0.30,160.60,311205,,*0B
    $GPRMC,000000.964,A,1907.4502,N,07254.9147,E,0.26,156.86,010106,,*05
    $GPRMC,000001.964,A,1907.4500,N,07254.9148,E,0.21,150.15,010106,,*02


    Best Regards,
    Akshay
     
    Akshay Mishra, Jan 2, 2006
    #5
  6. Harold Rabbie

    Phil Wheeler Guest

    It did seem like a long night: Yawn :)
     
    Phil Wheeler, Jan 2, 2006
    #6
  7. Ha!
     
    Hans-Georg Michna, Jan 2, 2006
    #7
  8. Harold Rabbie

    Stichting ST Guest

    As Sam Wormley already mentioned in a post: the difference between GPS time
    and UTC is 14 s.

    Piet
     
    Stichting ST, Jan 2, 2006
    #8
  9. However the $GPRMC time format should be UTC, shouldn't it? And if it
    actually _did_ show GPS time, then it should not show the leap second at
    all.

    Best regards Preben
     
    Preben Mikael Bohn, Jan 2, 2006
    #9
  10. Harold Rabbie

    Roy Lewallen Guest

    That comment got my attention, too, so I checked: The time display shown
    on the main menu page of my 76CS agrees to the second with a
    self-setting, WWVB synchronized clock (which agrees with WWV). So the
    screen time display is UTC on that model GPSr at least. I didn't look at
    the data stream -- maybe it's 14 seconds off from the screen time display?

    Roy Lewallen
     
    Roy Lewallen, Jan 2, 2006
    #10
  11. Harold Rabbie

    Sam Wormley Guest

    Thanks Akshay.
     
    Sam Wormley, Jan 2, 2006
    #11
  12. Harold Rabbie

    Sam Wormley Guest

    See... time *doesn't* go faster when you get older! :)
     
    Sam Wormley, Jan 2, 2006
    #12
  13. Harold Rabbie

    Jack Erbes Guest

    So now we are all 1 second younger, right?

    Whoopee!

    Jack
     
    Jack Erbes, Jan 2, 2006
    #13
  14. Some of the Garmin technical manuals available online do specify the
    behavior of their receivers when a leap second occurs. According to these
    specs, the Garmin GPS 15/16/17/25/30/35 models generate a time output of
    235960 for the leap second.

    So I guess it's a function of the firmware in the actual receiver, and will
    vary from manufacturer to manufacturer. If the Garmin behavior is or
    becomes a de-facto standard, then other behaviors (such as repeating a
    sentence) become de-facto bugs.
     
    Harold Rabbie, Jan 2, 2006
    #14
  15. Garmin behavior is NOT defining anything in this matter. The standard
    is set by IERS in Paris, France.

    http://hpiers.obspm.fr/iers/bul/bulc/bulletinc.dat

    Any other behavior is actual bugs.
     
    Bjorn Gabrielsson, Jan 2, 2006
    #15
  16. Well there is no requirement for any consumer GPS system to obey this
    standard. Some GPS systems won't recognize a leap second until they are
    turned off and back on again since they only check the message at power
    on. Old Garmins work this way and it is by design, not a bug.

    Dale
     
    Dale DePriest, Jan 3, 2006
    #16
  17. You just need a special case that allows 23:59:60 only on the two days
    of the year when a leap second is allowed to occur. Any other day, or
    if hour != 23 or minute != 59, then a seconds field of "60" is an
    error.

    My Garmin GPS-25 (a receiver board, not a handheld unit) *does* output
    23:59:60 for a leap second.

    Dave
     
    Dave Martindale, Jan 3, 2006
    #17
  18. Hmmm... maybe then Garmin should state that the time in their NMEA
    messages usually is close to UTC, but not always. Does not the NMEA
    183 standard define certain timestamps to be shown in UTC? Many maybe
    even most consumer GPS units claim to follow one or another version of
    the NMEA standard. Are Garmin advertising their (old) units as non
    compliant?
     
    Bjorn Gabrielsson, Jan 3, 2006
    #18
  19. Harold Rabbie

    Bob Greschke Guest

    We had the Air Force come up from White Sands Missile Range with their GPS
    simulator to test our equipment back in September. A couple of months after
    that someone from higher up the AF food chain called to see what we had been
    testing and what we had found out like as if they were worried that a bunch
    of stuff produced over the last 7 years was not designed to handle leap
    seconds at all. Yipes? :)

    Bob
     
    Bob Greschke, Jan 3, 2006
    #19
  20. Harold Rabbie

    Jack Yeazel Guest

    Jack Yeazel, Jan 3, 2006
    #20
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.