This site uses cookies to offer you a better browsing experience. Find out more about how to manage cookies, or dismiss this message and continue to use cookies.

How accurate is the auto time?

Just curious, My 260 and 255w both have the Auto Time Set enabled. How accurate is this time?
Does it get the time from the satellites? If so, I would guess it is within a few seconds of actual time?

Comments

  • gatorguy 328 Points
    Pretty accurate. Within nanoseconds.
  • Cool, thanks, figured it might be, but was not positive.
    Now I have a quick reference for the correct time.
  • Tim 1497 Points
    Pretty accurate. Within nanoseconds.
    The GPS satellites themselves are pretty darn accurate, but I doubt the GPS which is just syncing the time is itself that accurate. Within a second, sure... but I don't think it is within a few 0.000000001 seconds. :)
  • gatorguy 328 Points
    Perhaps semantics, but from what I've read, and a very recent Space Command video, the internal clock is within nanaseconds. I suppose by the time it gets displayed on screen it could be 1/2-1 sec.

    http://gpsinformation.net/main/gpstime.htm
  • Boyd 2027 Points
    As long as we're nitpicking :) I think the seconds aren't displayed on the screen at all, are they? (don't see any display of seconds on my nuvi 205) Therefore it would only be accurate to the minute from the user's perspective.
  • gatorguy 328 Points
    Touche' :lol:
  • sviking 141 Points
    As long as we're nitpicking :) I think the seconds aren't displayed on the screen at all, are they? (don't see any display of seconds on my nuvi 205) Therefore it would only be accurate to the minute from the user's perspective.
    I know...BFD...but my 660 displays seconds, in both digital and analog presentations, on the world clock page.
  • It shows seconds on the diagnostic screen (when you hold down on the battery meter). That is what got me wondering.

    And as said, the world clock shows the seconds also.


    Like I said, I was just wanting to know if it was accurate within a few seconds, within 10 is plenty for me. I like having 1 clock to set everything else off of.
  • Like I said, I was just wanting to know if it was accurate within a few seconds, within 10 is plenty for me. I like having 1 clock to set everything else off of.
    Use your cell phone then. Those clocks are highly accurate. Probably not as accurate as the GPS but certainly good enough for setting the other clocks. And the phone will display seconds.
  • Like I said, I was just wanting to know if it was accurate within a few seconds, within 10 is plenty for me. I like having 1 clock to set everything else off of.
    Use your cell phone then. Those clocks are highly accurate. Probably not as accurate as the GPS but certainly good enough for setting the other clocks. And the phone will display seconds.
    My cell phone was off 2 minutes from the GPS (and does not show seconds).
    Also, as said, the GPS does show seconds.
  • Cell phone time is only as accurate as the carrier's network. Maybe it's improved (doubtful), but a few years ago when I had the misfortune to have an AT&T phone, the time shown on the phone was off by a good minute. Verizon has been better, but I still use the GPS when I need an accurate time, like when changing to/from DST. Since each GPS bird has an on-board atomic clock, and the whole premise of GPS location-finding is based on very precise timing, I'm pretty sure the GPS time display is good enough for me. :wink:
  • a few years ago when I had the misfortune to have an AT&T phone, the time shown on the phone was off by a good minute. Verizon has been better, but I still use the GPS when I need an accurate time, like when changing to/from DST
    Really! I am surprised. I was under the impression that the cell phone network used NIST clocks to synchronize the entire network.

    Thanks :D for correcting my erroneous :oops: belief.
  • cuc tu 91 Points
    GPS time is the actual time reference. The 1PPS is used in PLLs to reduce time errors, as well as other error correcting routines. It depends entirely on the GPS device, but at the least should be within sub-second accuracy in the minute or hour.

    The best I've been able to determine is from an older rockwell 12-channel receiver that features a phase-locked 10kHz output.
  • jonnop 101 Points
    If accurate time is an issue thenperhaps it would be a good idea to purchase one of those radio controlled clocks that picks up a time signal from an atomic clock. I know there is one in the UK and Germany. I would assume that the US has a similar facility.
  • If accurate time is an issue thenperhaps it would be a good idea to purchase one of those radio controlled clocks that picks up a time signal from an atomic clock. I know there is one in the UK and Germany. I would assume that the US has a similar facility.
    Yep; Boulder Colorado facility transmits time information a few times a day; these clocks can pick it up usually within 6 hours or so.
  • alanb 557 Points
    I have a Casio Wave Ceptor digital watch that I love. It does daily time updates from the radio signal ... also solar charged battery that never needs replacement (at least not in my lifetime), and huge digital time readout that I can read even without my glasses. That said, I suspect the time on the GPS updated from the satellite is just as accurate as the radio updated Casio. When I hold them side by side, they always show the exact same time down to the second. Of course, it is impossible to visually detect any sub-second variation between the two.
  • But it's important to note:

    GPS Time signals do NOT ACCOUNT for LEAP-SECONDS.

    Universal Coordinated Time (UTC) does.

    Right now, GPS clocks are 15 or 16 seconds AHEAD of UTC.

    LEAP SECOND data is broadcast separately from the GPS time signals, but are broadcast relatively infrequently.

    The NIST Radio broadcasts are UTC. Most GPS receivers these days DO listen for the Leap Second message and adjust accordingly, but they are NOT required to. And if they are unable to receive the Leap Second broadcasts over a long enough period of time, the GPS's internal clock may begin to drift back toward the GPS standard time.

    :)
  • jonnop 101 Points
    Right now, GPS clocks are 15 or 16 seconds AHEAD of UTC.:)
    I was wondering why I kept arriving a few seconds early. Now I know. :o
  • SkipD 91 Points
    Boulder Colorado facility transmits time information a few times a day; these clocks can pick it up usually within 6 hours or so.
    The clock at Boulder CO feeds the WWV transmitters which are just north of Fort Collins CO and they are transmitting time data 24/7 every day of the year.
  • SergZak 341 Points
    I have a couple of those atomic time-sync clocks. They auto-sync via the transmitter every 6 hours or when you pull, then replace the battery.
  • I have a couple of those atomic time-sync clocks. They auto-sync via the transmitter every 6 hours or when you pull, then replace the battery.
    Yup... Most of the ones I've had operate like this when first started.

    Receiver is activated CONTINUOUSLY.
    Receiver receives its first signal some time later (hours, perhaps)
    Receiver sets its clock and then shuts off the radio.
    Radio wakes up a minute or so ahead of the next broadcast, and shuts down after receiving it.
    If no received signal after 24 hours, restart the process.
  • mikes 0 Points
    GPS time is locked to UTC(USNO), the time at the US Naval Observatory.

    There are several levels of error. UTC(USNO) differs from UTC, and the amount can only be calculated after the fact. This error is on the order of 10 ns. Here's the recent offset:
    image
    Then, GPS differs from UTC(USNO) by up to 1 us, but usually very much less, often no more than 25 ns. Recent offsets can be viewed here: http://www.nist.gov/pml/div688/grp40/gpsarchive.cfm

    Then, there's the inaccuracy of the GPS receiver. A GPS timing receiver, which has a fixed and known location, can solve the equation for time within about 10 ns. A moving GPS, which needs to solve for both time and location, won't do nearly as well, maybe a couple of hundred ns.

    So, a good timing GPS will typically be within about 50 ns of UTC. An ordinary consumer GPS should be within 500 ns.

    All of this is much more accurate that the time received by radio clocks, which suffer from variable propagation delays on the order of 300 us, due to atmospheric conditions (GPS can and does correct for this).

    As someone mentioned, there is an integer offset between GPS time and UTC. That difference is sent as part of the GPS navigation message, which repeats every 30 seconds. A consumer GPS would be expected to use that to display UTC (offset to local time zone). There's not much use for uncorrected GPS time.

    So, the answer to the question is that your GPS keeps what most people would consider very accurate time, typically within 1 millionth of a second (1 us) of UTC, although it's possible for it to be off a bit more than that. Many GPS receivers don't offer any way to communicate time of that accuracy to external devices (i.e. your Garmin GPS probably can't get time that accurately from the actual GPS receiver inside of it). Time is often passed via serial messages, and that may introduce much larger errors, on the order of 10-20 thousandths of a second. Overall, your Garmin should know the time to a small fraction of a second.

    The time displayed depends on the device, there may be software delays which are more than any of the errors mentioned above. But for any reasonably made device, the time displayed should be accurate to the second.
  • Thanks mikes. That is a very detailed explanation.
    It is within a second, so that is good enough for me.

    ns and us are more than I can even comprehend.
  • gatorguy 328 Points
    Hmm. . . So within nanoseconds :D

    Wouldn't each nanosecond of timing error introduce up to 1 foot of position error? Believe I'd read that at an Air Force site.
  • kenp 92 Points
    As someone mentioned, there is an integer offset between GPS time and UTC. That difference is sent as part of the GPS navigation message, which repeats every 30 seconds. A consumer GPS would be expected to use that to display UTC (offset to local time zone). There's not much use for uncorrected GPS time.
    Are there still consumer GPSs where you can select whether GPS time or UTC is shown? I seem to temember that being a choice some time ago.

    Back in the 90s I was involved in a couple of amusing incidents during field measurement projects when one system was synched to UTC and another was synched to GPS.

    Ken
  • Tim 1497 Points
    Hmm. . . So within nanoseconds :D
    I guess I'm just looking at this from an experiential rather than theoretical standpoint. I've got a number of GPS devices on my desk with a 3D fix, Garmin, Magellan, DeLorme, and TomTom. Those with a "sync time" function, I've pressed it. I'm looking at the clock on each and they are not in sync with each other. Most are within two seconds of each other, one is about eight seconds out.

    So regardless of GPS time, UTC time, the time the GPS chipset is calculating... whatever, the displayed time isn't within nanoseconds as they differ from each other by a noticeable margin. The original poster was asking if it was accurate to within a couple of seconds and I'd say that is true. But if the displayed time was accurate to within a few nanoseconds then I shouldn't see any difference in time between devices that have a 3D fix and time synced.
  • gatorguy 328 Points
    I think we're both correct Tim, depending on how you consider the accuracy. I just don't get many opportunities get little digs in. You're always so darn right. :lol:
  • Of course, it all goes out the window when you come up on a big hour-long traffic jam caused by some idiot that failed to properly secure the port-a-potties on the trailer he was hauling. Then, the nano-seconds that my nuvi's destination ETA may be off are the least of my worries. :lol:
  • mikes 0 Points
    Hmm. . . So within nanoseconds :D
    Well, 25 ms (which is a reasonable estimate of how accurately the GPS software in a Garmin can determine real time) is 25,000,000 ns, so yes.
    Wouldn't each nanosecond of timing error introduce up to 1 foot of position error? Believe I'd read that at an Air Force site.
    That's based on the fact that RF travels about 1 ft / ns (within a couple of percent).

    But GPS can determine position without any reference to real time. As I mentioned there are errors between UTC and UTC(USNO), then again between UTC(USNO) and GPS time. Neither of those errors affect GPS position accuracy.
  • gatorguy 328 Points
    Thanks for the detailed explanation!

    I misunderstand completely now. :)
  • luisr 0 Points
    I read m ost of this topic and checked my GPS to see if it had seconds display. I also happened to be in front of the computer and had my cell phone at hand. It was 5:59 PM and waited for 6:00 PM to arrive. All three changed within two seconds. That's pretty darn good for most purposes. My cell phone network is pretty accurate and the computer is set to adjust itself periodically as per default Windows settings.
Sign In or Register to comment.
↑ Top