2022 GPS rollovers: Furuno, Airmar, JRC, and who else?

While I certainly knew that it was not June 8 2002 when I took this photo last Saturday, the Sunrise/Sunset times shown above are also quite wrong, and it’s hard to tell what other calculations get screwed up when a networked device like this Maretron DSM starts receiving a system date that is 1,024 weeks old. Yes, the problem seems like the well-publicized April 1999 GPS week number rollover, except that it just happened a few weeks ago on the attached Airmar PB200, and lots of older gear from Furuno, JRC and probably other brands are experiencing a similarly delayed and under-publicized rollover in 2022. The date changes will undoubtedly mess up some boaters in the months ahead, so let’s make ourselves aware.

Furuno announced a 1/2/2022 GPS week number rollover date affecting many older devices
Furuno announced a 1/2/2022 GPS week number rollover date affecting many older devices

The root cause of GPS week number rollover is a well-known programming limitation in the early GPS satellite transmissions, and it’s happened twice so far — at midnight August 21, 1999, and again on April 6, 2019. When Ben Stein well explained the latter rollover, organizations like the U.S. Coast Guard were already warning boaters about it and many marine electronics companies offered firmware fixes.

But few of us realized then that many devices got through the 1999 rollover without problems only because their GPS receiver chipsets had been jiggered to push the 1,024-week date change forward a couple of years or more. In fact, Furuno’s long-posted GPS rollover notice lists equipment that rolled over on 12/20/2020 or 6/20/21 or earlier this month, and JRC announced rollovers that will happen on the 15th of May.



So what we have here is a series of rolling rollovers that are apparently only being publicized by the manufacturers, if at all. For instance, I only powered up the 13-year-old PB200 WeatherStation because Kees Verruijt encountered the rollover in his test lab over the holidays. It turned out that Airmar is aware of the problem but they say it can’t be fixed, and they further note that rollover is not an issue for the GPS chipsets in the many WeatherStation models that superseded the PB200 many years ago.

Moreover, an unfixable GPS date is not necessarily an insurmountable problem, and I think I’ve found a solid way to use the PB200’s remaining faculties (as discussed below). Then again, the wrong system date can cause major problems…

What will a GPS rollover mean on your boat?

When I searched for “rollover” on FurunoUSA’s Forum, it was no surprise that some tropical cruisers experienced issues right after the Jan. 2 rollover, including a Class B AIS that would no longer transmit. Forum moderator “Johnny Electron” thoroughly explained the update options, and the FA50 procedure looks fairly DIY reasonable largely because only an Ethernet cable and PC is needed. (By contrast, updating a GP33 requires a modified NMEA 0183 cable and I’d mind Furuno’s “Not recommended for end-users” advice.)

Satellite communications devices can also be crippled if they’re getting an invalid date from their own GPS receiver or one they are networked to. But other rollover effects can be quite subtle. For instance, the automatic online updating of a device that Kees was testing failed after the PB200 because the connection’s website certificate wants to see a valid date.



A standalone GPS with the date 19.6 years wrong — like my antique Garmin 45 — will probably calculate your boat’s correct position, speed, and true course (watch out for magnetic variation errors). But I was happy to use Airmar’s WeatherCaster PC software along with an Actisense NGT-1 bridge to turn off all of the PB200’s GPS related PGNs.

So now networked devices can still see the PB200’s valid Wind, Heading, Temp and Baro output, while Date, Position, etc are provided by another GPS receiver. And let’s acknowledge that position fixing technology has evolved a good deal since the rollover-prone GPS chipsets were manufactured, with ultra-accurate multi-constellation and multi-band GNNS the latest.

I’m sure that there are other ways to work with unfixable misdated GPS gear that can still perform some functions well, and I hope that readers will share them. But the bigger story here may be what happens in the spring when more older but trusted marine electronics don’t work right anymore, perhaps in subtle ways. So please spread the word — it’s far better to fix these issues before the problems start — and please report on any recent or anticipated rollovers not mentioned above.



Ben Ellison

Ben Ellison

Panbo editor, publisher & chief bottlewasher from 4/2005 until 8/2018, and now pleased to have Ben Stein as a very able publisher, webmaster, and editing colleague. Please don't regard him as an "expert"; he's getting quite old and thinks that "fadiddling fumble-putz" is a more accurate description.

24 Responses

  1. Great research, Ben! We knew something was up, but I had no idea it was so widespread. One wonders what other “easter eggs” are hidden away in our smart devices these days 🙂 .

  2. Speaking of Magnetic Variation, I found a really neat map page at: https://www.ncei.noaa.gov/maps/historical_declination/ which shows the changing magnetosphere over time. As for the difference between 2002 and 2022, the further North you are, the bigger the difference!

  3. Fred Murphy says:

    Great information Ben!

  4. Howard says:

    I have a Raymarine RS125 GPS providing data over Seatalk to my E120 MFD. Per a Raymarine document the 125 will have an issue in 2022. Document in this forum thread: https://forum.raymarine.com/showthread.php?tid=7453
    However, in another forum thread the date for the RS125 is stated as April 16, 2025. https://forum.raymarine.com/showthread.php?tid=7275&page=2
    I am not concerned about date error, I use my iPad to get tide/current data. However, I use magnetic heading and course to match my compass, the true/magnetic variation will be incorrect on the autopilot and MFD once the date error occurs.

    • Ben Ellison Ben Ellison says:

      Thanks, Howard. That’s quite a list, showing many Ray rollovers from this year through 2035:

      https://forum.raymarine.com/attachment.php?aid=2852

      I suspect that 2022 is the year for your RS125, because it’s on the official list, but I don’t know for sure.

    • Howard says:

      I just checked a magnetic variation calculator and learned that in the past 20 years the declination has only changed by 1.25 degrees in my area (Rhode Island). I can’t read a compass that accurately so I will not be concerned with the declination error after the RS125 rolls over. My E120 and Autopilot still work well, I can still update the Navionics chart in the E120, and I use my iPad running iNavX for routing and autopilot control. (INavX NMEA autopilot sentences out via WiFi to boats network, iKommunicate converts to wired 0183 to drive Raymarine S3G AP computer) The E120 then gets next waypoint and course info to plot on chart and RADAR displays via Seatalk.

      • Jim McGowan says:

        Our navigation aids team was taking another look at GPS rollover, and determined that the Raystar 125 will experience that in March of 2023. That buys you a bit more time, Howard, and through the 2022 boating season. Hopefully knowing a more firm date will help you in any future decision to upgrade to a newer GPS receiver. If you have any question, please don’t hesitate to contact me.

        Jim McGowan
        Raymarine Electronics
        [email protected]

        • Howard says:

          Jim,
          Thank you for the updated information on my Raystar 125, this is good to know and helps with future planning.
          Howard

  5. Tricia Evangelista says:

    Thanks for writing about this. On our way south from Stuart to Marathon FL on 1/14/22, I suspected that our AIS was not transmitting – Furuno FA50. We were receiving fine. Had not run the boat since we were home for the holidays so something happened between November and January. Spent most of the night trying to figure out what was going on. Sure enough, a software update was required to correct for the GPS week rollover, something we had never heard of. Fortunately, our mechanic and electronics guru sent us an email he had just received from his electronics supplier advising of the problem. Downloaded the software and installed via Ethernet cable to the Furuno hub. Not sure if we would have figured out the problem without that well timed email heads up.

    • Ben Ellison Ben Ellison says:

      Thanks, Tricia, and tip of a Panbo cap to your guru. I think that a lesson for all is to get acquainted with whatever support services your electronic manufacturer provides, like forums and online service notices, and then to check them from time to time for issues involving your devices. And/or follow your example by cultivating a good guru 😉

  6. Fred Murphy says:

    Is it possible that having a unit off during the rollover period would allow the unit to work its way through it and restart without the problem

    • Tricia Evangelista says:

      Our Furuno FA50 (and all other Nav equipment) was off from 11/15/21-1/13/22. The rollover occurred on 1/2/22. The FA50 filed to transmit when fired up on 1/14/22. Installing the software update corrected the problem.

    • Fred, the problem is routed in how the GPS software determines the date. The only date info contained in the GPS data stream is a “week number”, which. because of the number of bits involved, “rolls over” back to zero every 1024 weeks – about 19 years. If the unit/software is new, then the date can be interpolated to allow for this rollover – but the maximum possible lifespan would be those 19 years unless the software is updated.
      So unless the GPS data stream is modified to include some sort of “epoch number” (which might cause even more problems elsewhere!), GPS receivers will ALWAYS have this issue unless they are updated at least every 19 years. Manufacturers are not going to update their software every week, of course, so whatever version you have will not start the week your unit was made, but at some week in the past – sometimes many years ago! Modifying the software has costs, too, so manufacturers have to decide if it is worth the effort to continue modifying the software for ancient receivers – and not every unit as the ability to even have its firmware updated without major disassembly and intervention.

  7. Fred Murphy says:

    Is it possible that having a unit off through rollover period would allow the unit to reboot after rollover period and com through this event
    It seems that AIS units are susceptible to this problem. My boat has Garmin GPS that pushes position info into the network. Is my unit susceptible to failure?

    • Ben Ellison Ben Ellison says:

      Fred, when a GPS device experiences a rollover, it usually takes a reboot to get the correct position again. But the reboot does not correct the date problem. If your Garmin GPS came to market less than 10 years ago — very rough guess — it probably won’t roll over soon.

  8. Fred Murphy says:

    Thanks Hartley and Ben. Great knowledge! I guess worst case scenario is buy a new GPS. Anybody know a good new simple one that will put out NMEA 0183 data. I guess I could run on NMEA 2000 with additional wire.

    • The RayMarine RS-150 I have is rated to 2033, while the older RS130 goes to 2027. Both are NMEA2000 output, however. The RS125 ends in 2022, which is why they are all over eBay right now 🙁 . I’m not sure I would trust the cheap Chinese NMEA-0183 GPS’ that are all over eBay and Amazon unless they tell you directly when they will fail (and they will ALL fail at some point!).
      BTW, the reason a reboot works for location (but not date) is that it forces the GPS receiver to reload the “almanac” which is how it knows where the satellites are “right now” (and therefore where it is – “right now”!). The date, however, is extracted from that week number and only a software update will re-align it to reality.

  9. Donald Joyce says:

    Sigh,
    It would seem the clever people that bring us all this fantastic stuff could have firmware that anticipates something like this and handles it appropriately without needing to be updated etc….ie if week 1023 occurs the firmware would note the current date, project the next week date and all would be good when week 0 comes around. Since this isn’t the case, what is it that is prevents such a simple remedy? I feel clueless.

    • Donald Joyce says:

      I’ve thought about this a bit further and realize that the organization behind GPS, namely the US Defense Department, would be in serious trouble if they had to reset each GPS they utilize. Clearly they have implemented a hands off solution.

  10. Systems which display or depend on tide can be affected by the rollover too. Tide calcs are based on time/date/position.

    Bottom mapping systems like Olex take clock from the GPS and will thus calculated tide wrong and record bottom wrong if fed with a rolled-over GPS system.
    Other systems I work with like TimeZero get clock from the computer OS and are not affected.

    I’m pretty good with mad-scientists type wiring and could not get a GP-33 to update software.

    • Wow, Jason – I hadn’t thought of that, but I am sure that the plotter needs the date to look up tide & current info for display. We use the plotter info for tides & currents a LOT!

Join the conversation

Your email address will not be published.