GPS Epoch 6th April 2019 - Failed Equipment list, fixes and updates etc

[Towger]

As most of you are aware the next major GPS Epoch is on the 6 of April 2019 and is a GPS Week Roll Over.

A basic description can be found on the www.gps.gov by Edward Powers of  USNO: https://www.gps.gov/cgsic/meetings/2017/powers.pdf


Possible Failure
-----------------

HP-Z3805A: GPS Time & Freq Receiver

    Web: https://www.keysight.com/en/pd-Z3805A%3Aepsg%3Apro-pn-Z3805A/gps-time-freq-rcvr?cc=US&lc=eng

    GPS Module: Symmetricom ?

    Firmware: 1.01

    Comment:
    This was listed on ebay during the week, seller says he is off loading it before it fails on the 6 April.
    https://www.ebay.com/itm/HP-Z3805A-10811-DOCXO-16-channel-GPS-Frequency-Time-Receiver-10-MHz-1PPS-/132995407934?_trksid=p2047675.m43663.l10137&nordt=true&rt=nc&orig_cvip=true

[EOC_Jason]

The Motorola VP modules have flash memory that stores the date/year, so even if it power cycles it should be okay. Some of the other older modules like the UT+ have already started to have rollover issues in their firmware, but a workaround is to manually set the date BEFORE it tracks any GPS satellites and it continues on its happy way....

Someone got a good deal on the HP unit!

[kilobyte]

Thanks Towger for the information about the GPS Epoch issue.

The Symmetricom UCCM GPSDO are using a Furuno GT-8031F receiver module witch have a rollover on 22 September 2022 according to the following Document.

https://furuno.ent.box.com/s/fva29wqbcioqvd6mqxn5rt976dkaxudj

There is also a workaround described by setting a time with the ZDA command.

[Towger]

Thanks EOC_Jason and kilobyte

The Furuno GT-8031x (epoch 2022/09/18) is found in a number different ebay 'e-waste' UCCM GPSDOs.  Some of the 'boxed' versions on eBay have no display/microcontroller with just 10Mhz and 1PPS outputs.  It is not clear if they will still operate/lock properly without the date being set.  The others will probably be fine with a firmware update to set the date on boot up.  I am assuming they just need any date in the next epoch to initialize, rather then the current date?

[Theboel]

I have some 58503A and I am worried for 6th april, if any one has any idea how to solve this problem I really appreciate
 

[texaspyro]

Virtually no GPS receivers will be affected by the week rollover in April.  They usually set the start of the 1024 week cycle from the date the firmware was created, not the start of the GPS epoch.   Also, GPS receivers don't stop working when the epoch rolls over... they just start sending dates 1024 weeks in the past.  (there are a couple of very rare devices out there that do go titsup).

If you are using Lady Heather, it has automatic GPS rollover correction and will correct for bogus dates.

[awallin]

some makes/models listed here with issues or no issues:
https://kb.unavco.org/kb/article/preparing-gps-gnss-receivers-and-hardware-for-the-april-6-2019-gps-week-number-rollover-wnro-867.html

[EEVblog]

Virtually no GPS receivers will be affected by the week rollover in April.  They usually set the start of the 1024 week cycle from the date the firmware was created, not the start of the GPS epoch.   Also, GPS receivers don't stop working when the epoch rolls over... they just start sending dates 1024 weeks in the past.  (there are a couple of very rare devices out there that do go titsup).

I've heard that during the last GPS rollover in 1999 you could just reset the GPS and leave it for a while until it gets the full almanac again. Not really any different for leaving the batteries out for a few months.

[EEVblog]

I do wonder if this will affect the 1pps output on my CSIRO SRS standard...

[peteb2]

Any reports by now of reaction in GPS anywhere around the world for the April 6th WNRO issue? .... in my country we’ve also come off Summertime for April 7 @ 2am meaning everyone gets their extra hour back. a couple colleges at my place of employment had done some maintenance to the TV Station’s reference standard which is locked from GPS. The manufacturer of the equipment didn’t believe ‘WNRO’ would totally compromise the technology. It’s now 3.30am, i’m the duty On Call engineer & no one has called from the station, so all must be OK.

For my personal car however, onboard GPS & entertainment centre (android base, contracted Tomtom GPS) had a mandated update for WNRO or the GPS would degrade and the car’s clock would be always wrong... The 20 min upgrade with engine running (!) went fine but the hands free mobile phone accessory is now unsupported... bricked permanently. Worse... the Dealer has no idea how to fix it. Not good as it was a handy feature...

[texaspyro]

I recently got in a Navsync CW25-TIM receiver and added support for it to Lady Heather.   Along the way, I found out that it (and a lot of Navsync devices) will be affected by the rollover... in a major way.

When the GPS week is 0, it stops tracking sats for the week.   These are the only receivers that I know of that actually stop working when rollover hits.  Conner-Winfield offers firmware updates for affected devices.  The updates are in .rar format that requires winrar v5.xx to extract.

[Theboel]

nothing happen with my 58503A except LH report RO (roll over),
the 1PPS and 10Mhz output working as usual and time lab show nothing changed

[texaspyro]

I recently got in a Navsync CW25-TIM receiver and added support for it to Lady Heather.   Along the way, I found out that it (and a lot of Navsync devices) will be affected by the rollover... in a major way.

When the GPS week is 0, it stops tracking sats for the week.   These are the only receivers that I know of that actually stop working when rollover hits.  Conner-Winfield offers firmware updates for affected devices.  The updates are in .rar format that requires winrar v5.xx to extract.

Yep, right on schedule, the CW25 crapped its pants.   I upgraded the firmware, but no joy... still not tracking sats after 3 hours.
 

[LapTop006]

I do wonder if this will affect the 1pps output on my CSIRO SRS standard...

Shouldn't.

My TrueTime XL-DC actually rolled over at some point in the last year (it lives back in Melbourne, so I only see it a few times a year), everything was still solid, just the day & hour was wrong.

[texaspyro]

After 10 hours the CW25-TIM with the new firmware finally started tracking satellites again...

[remois]

  Hi,

  Could you care to elaborate on the ZDA command?

  Thank you

[apollomaker]

Oh, hell. Well, my new BG7TBL unit got nailed. Everything was working beautifully then  . Satts failing to track to Lady Heather. What a mess. If anyone gets a fix or info please post or send me a PM please.

I will do the same. Lady Heather is showing the automatic "ro" correction but no joy. Will let it run for some time and see what happens.

Thanks --- Bob

[apollomaker]

Hello -- have been observing my BG7TBL for some time and it is thermally stabilizing but it is not being disciplined as it was before the Rollover issue. Not as tight by about 1 to 2 orders of magnitude -- like a standard thermally stabilized OCXO. I have sent an information request to the source I purchased it from for any information about a fix/patch??? Will see what I get back.

[texaspyro]

Oops... airplanes grounded by rollover:

https://arstechnica.com/information-technology/2019/04/gps-rollover-apparently-cause-of-multiple-flight-delays-groundings/

[apollomaker]

Wow, yes, when you think about all the aviation, marine, automotive systems that are based on GPS, it is scary. Well, I hope that all producers of these systems get some patches/fixes out. No word back yet for the BG7TBL unit I am trying to run here.

[janoc]

Wow, yes, when you think about all the aviation, marine, automotive systems that are based on GPS, it is scary. Well, I hope that all producers of these systems get some patches/fixes out. No word back yet for the BG7TBL unit I am trying to run here.

They aren't "based" on GPS, there are always multiple navigation systems available to fall back on. So even if the GPS crapped out mid-flight for whatever reason, there is always the usual VOR/DME, NDB radio navigation, the planes have IMUs, there is radar coverage and ATC then there are compasses & charts (yes, even A380 still has a good old magnetic compass) and even celestial navigation. E.g. when the Russians were jamming GPS over the Baltic sea recently, multiple planes reported the interference and inoperative GPS - and that was about all that it has caused for aviation.

The reported issues grounding/delaying flights were more by abundance of caution (why to take a risk and fly with an inoperative system when you don't have to, even though there is redundancy) than anything really serious.

Marine navigation systems are similar, there is always one or more things to fall back on.

The only exception is automotive, but there the worst thing that could happen is that someone's satnav or fleet tracking stops working.

Also many new navigation receivers work with multiple systems - e.g. GPS, Glonass and Galileo.

[apollomaker]

Wonder of wonders. Well, after some time, my BG7TBL unit began disciplining again. Looking at Lady Heather, the week count shows 1026. And Lady Heather is reading the unit correctly (for the most part) which it would not do after the Epoch rollover. I had "written off" the unit since the supplier has provided no information.

Any ideas about what exactly might have happened?? The unit is now working as it did before the rollover. 

[Ohm_My]

Is the week number supposed to actually reset to a value after the rollover event? I can't tell whether or not my GPSDO handled the event successfully or not.

It shows the correct date / time and all appears to be ok except for the week number. Is it supposed to reset itself to some value as a result of the rollover event or is my thinking not quite correct?

[texaspyro]

Is the week number supposed to actually reset to a value after the rollover event? I can't tell whether or not my GPSDO handled the event successfully or not.

It shows the correct date / time and all appears to be ok except for the week number. Is it supposed to reset itself to some value as a result of the rollover event or is my thinking not quite correct?

Receivers that don't handle the week number rollover properly will usually start the week over at 0.  Those that have some WNRO mitigation will add a multiple of 1024 weeks to the week number sent by the satellites.