Tweet
Ciao a tutti,
non so a quanti possa interessare questa notizia ma la ritengo interessante per chi utilizza il GPS. Potrebbe essere un problema simile all'atteso del millennium.
Da quello che ho capito, qualche apparato GPS utilizzato potrebbe cambiare data ma fornire l'impulso di 1pps corretto
Mi dispiace che le informazioni siano tutte in inglese, le ho ricevute oggi dal gruppo tecnico della rsgb. Spero qualche volontario metta in luce, in italiano, il probabile problema e che cosa potrebbe succedere.
73
Gian
I7SWX
2a.
Possible GPS problem 6th April
From: Peter Martinez
Date: Sun, 24 Feb 2019 11:00:00 CET
My attention has been drawn to the appended post by Mark GM4ISM. I think my own GPS (Trimble Palisade) will be OK. It already gives me the wrong date - probably because it missed the last week-number rollover in August 1999. After April 6th it might stay 19 years ago or it might jump to 39 years ago, but the time and the PPS should be OK.
73
Peter G3PLX
-------- Forwarded Message --------
Subject: [UKMicrowaves] GPS Frequency standards.. Week Zero alert.
Date: Thu, 7 Feb 2019 18:39:05 -0000
From: Mark GM4ISM via Groups.Io <gm4ism=blueyonder.co.uk@groups.io>
Reply-To: UKMicrowaves@groups.io
To: UKMicrowaves@groups.io
For those of us reliant on GPS based frequency standards and location
systems, there is an upcoming event that may upset some GPS engines.
GPS week numbers only go up to 1024 and we hit another 'week zero' event
on 6th April this year.
This can and does cause some older GPS engines to become rather upset.
Some completely loose lock, some may still produce 1PPS etc but output
the wrong date and time
Many of these recover at week 1 but that is not guaranteed.
Some testing has been done by a colleague on a number of GPS based
systems using a locally generated GPS signal set that simulates the week
zero signals.
A number of older GPS engines have failed this test and the frequency
standards associated with them have been shown to go into holdover and
start to drift....
Amongst the ones likely to fall over are the Rockwell chipset widely
used by amateurs. The exact model with 10KHz (Jupiter) out has not been
confirmed but pretty much all of the related models, used in a number of
GPS disciplined oscillators, fell over.
The Jupiter in my old RUH system already needs a poke of serial data at
start-up, I believe to correct a previous timestamp rollover. I guess it
may need a different one soon.
I am aware that if all else fails, UBlox GPS engines have been
successfully used in the RUH design standard in place of the Jupiter, as
some Ublox engines can be programmed to output the required 10KHz
instead of 1PPS
All (slightly older ) UBlox modules that have been tested were
unaffected so it is likely that newer ones will be OK too,
Some older Rapco GPS standards fail the Week Zero test which makes them
think they are the best part of twenty years younger. Their recovery
into week 1 is not certain.
I am aware that at least one model of these has an internal firmware
week counter which will rollover in July if I recall, further
complicating matters
I also have one of these and its useful life is now in question.
The Trimble Palisades out there are an unknown quantity.
I will try to get one tested. I use one of these with a VE2ZAZ system
and so does GB3CSB
Recovery in week 1 has not been tested in my world, a full week in
holdover is not acceptable so units are being replaced.
So the upshot is that for the duration of week 0, your GPS freq
standard, if based round an older GPS engine, may fail to lock and
provide an accurate 10MHz
I have a Rubidium standalone, which I fear I may need because none of my
references are modern.
You may see people flogging off some professional GPS frequency
standards. Some may be real bargains if there is a 'fix', other may
become simply a source of a reasonable 10MHz OCXO Caveat emptor!
Happy Binary Birthday GPS!
Mark GM4ISM
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
2b.
Re: Possible GPS problem 6th April
From: David Barber
Date: Sun, 24 Feb 2019 11:44:33 CET
Just out of curiosity I wonder why this / these issues are out there?
Did the clever people who designed the hardware and wrote the firmware for
these GPS modules 'forget' this would happen? Was there a cost saving in
doing it this way?
Or were these modules never expected to be in service for a period beyond
which the issue would be encountered?
David Barber
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
2c.
Re: Possible GPS problem 6th April
From: Peter Martinez
Date: Sun, 24 Feb 2019 12:24:56 CET
David said:
Did the clever people who designed the hardware and wrote the firmware for
these GPS modules 'forget' this would happen?
I am guessing that this problem is just poorly-written software. The Millenium Bug was a similar problem, where programmers chose to record years with only the last two digits (so it should have been called the Century Bug!) so at year 2000 these programs registered the year as 1900. In this case the GPS satellites only have 10 bits to count the passing weeks and then return to zero, which it will do midnight 6/7th April. GPS receiver designers should have been able to predict this and avoid it, but evidently some of them will malfunction and some may never work properly thereafter.
73
Peter G3PLX
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
2d.
Re: Possible GPS problem 6th April
From: Dave Sergeant G3YMC
Date: Sun, 24 Feb 2019 13:27:30 CET
There seems to be quite a lot of confusion on this matter, and even
with far newer satnaves. Tom Tom have a page about it on their website
with a box to type your satnav serial number to check. My Start 20
comes up as needing a firmware update but when I check it online there
is none available. It seems it was patched in a software update early
last year which it had installed already during the normal updates.
Together with as reported on the radio last week them sending emails to
customers with a link to click that looked just like a spam. I imagine
most current satnavs will continue to work just fine, but clearly we as
amateurs who use them in different modes may find otherwise.
73 Dave G3YMC
On 24 Feb 2019 at 9:53, Peter Martinez via Groups.Io wrote:
My attention has been drawn to the appended post by Mark GM4ISM. I
think my own GPS (Trimble Palisade) will be OK. It already gives me the
wrong date - probably because it missed the last week-number rollover in
August 1999. After April 6th it might stay 19 years ago or it might jump
to 39 years ago, but the time and the PPS should be OK.
http://davesergeant.com
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
2e.
Re: Possible GPS problem 6th April
From: Neil Smith G4DBN
Date: Sun, 24 Feb 2019 14:20:53 CET
Tim N3QE has a Z3801 which is going through it's second rollover and will be into its third 1024-week epoch.
For the technical detail about rollover, search the Time-nuts forum for "Rollover" (it is indexed on google). Loads of intricate stuff on there about the issues around rollovers.
Good history at http://www.leapsecond.com/notes/gpswnro.htm
whenre Joe Gwinn discusses (amongst other things) the use of the drifting offset between GPS time and UTC as an identifier for which 1024-week GPOS epoch you are in.
https://www.febo.com/pipermail/time-...ne/067840.html
https://lists.febo.com/mailman/listi...lists.febo.com
Beware of getting assimilated into the world of time-nuttery, you'll start losing sleep because your 1Hz offset at 10GHz is an appalling lapse of precision.
Neil G4DBN
PS, check out volt-nuts as well. As if you don't have enough to worry about with time and frequency precision
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
2f.
Re: Possible GPS problem 6th April
From: mike
Date: Sun, 24 Feb 2019 19:43:55 CET
Almost certainly the latter.
ISTR the Y2K bug got everyone up in a lather too.
Lets wait and see what your GPS engine does..As for mine, I really dont care
so long as the 1pps stays accurate...
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
3a.
non so a quanti possa interessare questa notizia ma la ritengo interessante per chi utilizza il GPS. Potrebbe essere un problema simile all'atteso del millennium.
Da quello che ho capito, qualche apparato GPS utilizzato potrebbe cambiare data ma fornire l'impulso di 1pps corretto
Mi dispiace che le informazioni siano tutte in inglese, le ho ricevute oggi dal gruppo tecnico della rsgb. Spero qualche volontario metta in luce, in italiano, il probabile problema e che cosa potrebbe succedere.
73
Gian
I7SWX
2a.
Possible GPS problem 6th April
From: Peter Martinez
Date: Sun, 24 Feb 2019 11:00:00 CET
My attention has been drawn to the appended post by Mark GM4ISM. I think my own GPS (Trimble Palisade) will be OK. It already gives me the wrong date - probably because it missed the last week-number rollover in August 1999. After April 6th it might stay 19 years ago or it might jump to 39 years ago, but the time and the PPS should be OK.
73
Peter G3PLX
-------- Forwarded Message --------
Subject: [UKMicrowaves] GPS Frequency standards.. Week Zero alert.
Date: Thu, 7 Feb 2019 18:39:05 -0000
From: Mark GM4ISM via Groups.Io <gm4ism=blueyonder.co.uk@groups.io>
Reply-To: UKMicrowaves@groups.io
To: UKMicrowaves@groups.io
For those of us reliant on GPS based frequency standards and location
systems, there is an upcoming event that may upset some GPS engines.
GPS week numbers only go up to 1024 and we hit another 'week zero' event
on 6th April this year.
This can and does cause some older GPS engines to become rather upset.
Some completely loose lock, some may still produce 1PPS etc but output
the wrong date and time
Many of these recover at week 1 but that is not guaranteed.
Some testing has been done by a colleague on a number of GPS based
systems using a locally generated GPS signal set that simulates the week
zero signals.
A number of older GPS engines have failed this test and the frequency
standards associated with them have been shown to go into holdover and
start to drift....
Amongst the ones likely to fall over are the Rockwell chipset widely
used by amateurs. The exact model with 10KHz (Jupiter) out has not been
confirmed but pretty much all of the related models, used in a number of
GPS disciplined oscillators, fell over.
The Jupiter in my old RUH system already needs a poke of serial data at
start-up, I believe to correct a previous timestamp rollover. I guess it
may need a different one soon.
I am aware that if all else fails, UBlox GPS engines have been
successfully used in the RUH design standard in place of the Jupiter, as
some Ublox engines can be programmed to output the required 10KHz
instead of 1PPS
All (slightly older ) UBlox modules that have been tested were
unaffected so it is likely that newer ones will be OK too,
Some older Rapco GPS standards fail the Week Zero test which makes them
think they are the best part of twenty years younger. Their recovery
into week 1 is not certain.
I am aware that at least one model of these has an internal firmware
week counter which will rollover in July if I recall, further
complicating matters
I also have one of these and its useful life is now in question.
The Trimble Palisades out there are an unknown quantity.
I will try to get one tested. I use one of these with a VE2ZAZ system
and so does GB3CSB
Recovery in week 1 has not been tested in my world, a full week in
holdover is not acceptable so units are being replaced.
So the upshot is that for the duration of week 0, your GPS freq
standard, if based round an older GPS engine, may fail to lock and
provide an accurate 10MHz
I have a Rubidium standalone, which I fear I may need because none of my
references are modern.
You may see people flogging off some professional GPS frequency
standards. Some may be real bargains if there is a 'fix', other may
become simply a source of a reasonable 10MHz OCXO Caveat emptor!
Happy Binary Birthday GPS!
Mark GM4ISM
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
2b.
Re: Possible GPS problem 6th April
From: David Barber
Date: Sun, 24 Feb 2019 11:44:33 CET
Just out of curiosity I wonder why this / these issues are out there?
Did the clever people who designed the hardware and wrote the firmware for
these GPS modules 'forget' this would happen? Was there a cost saving in
doing it this way?
Or were these modules never expected to be in service for a period beyond
which the issue would be encountered?
David Barber
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
2c.
Re: Possible GPS problem 6th April
From: Peter Martinez
Date: Sun, 24 Feb 2019 12:24:56 CET
David said:
Did the clever people who designed the hardware and wrote the firmware for
these GPS modules 'forget' this would happen?
I am guessing that this problem is just poorly-written software. The Millenium Bug was a similar problem, where programmers chose to record years with only the last two digits (so it should have been called the Century Bug!) so at year 2000 these programs registered the year as 1900. In this case the GPS satellites only have 10 bits to count the passing weeks and then return to zero, which it will do midnight 6/7th April. GPS receiver designers should have been able to predict this and avoid it, but evidently some of them will malfunction and some may never work properly thereafter.
73
Peter G3PLX
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
2d.
Re: Possible GPS problem 6th April
From: Dave Sergeant G3YMC
Date: Sun, 24 Feb 2019 13:27:30 CET
There seems to be quite a lot of confusion on this matter, and even
with far newer satnaves. Tom Tom have a page about it on their website
with a box to type your satnav serial number to check. My Start 20
comes up as needing a firmware update but when I check it online there
is none available. It seems it was patched in a software update early
last year which it had installed already during the normal updates.
Together with as reported on the radio last week them sending emails to
customers with a link to click that looked just like a spam. I imagine
most current satnavs will continue to work just fine, but clearly we as
amateurs who use them in different modes may find otherwise.
73 Dave G3YMC
On 24 Feb 2019 at 9:53, Peter Martinez via Groups.Io wrote:
My attention has been drawn to the appended post by Mark GM4ISM. I
think my own GPS (Trimble Palisade) will be OK. It already gives me the
wrong date - probably because it missed the last week-number rollover in
August 1999. After April 6th it might stay 19 years ago or it might jump
to 39 years ago, but the time and the PPS should be OK.
http://davesergeant.com
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
2e.
Re: Possible GPS problem 6th April
From: Neil Smith G4DBN
Date: Sun, 24 Feb 2019 14:20:53 CET
Tim N3QE has a Z3801 which is going through it's second rollover and will be into its third 1024-week epoch.
For the technical detail about rollover, search the Time-nuts forum for "Rollover" (it is indexed on google). Loads of intricate stuff on there about the issues around rollovers.
Good history at http://www.leapsecond.com/notes/gpswnro.htm
whenre Joe Gwinn discusses (amongst other things) the use of the drifting offset between GPS time and UTC as an identifier for which 1024-week GPOS epoch you are in.
https://www.febo.com/pipermail/time-...ne/067840.html
https://lists.febo.com/mailman/listi...lists.febo.com
Beware of getting assimilated into the world of time-nuttery, you'll start losing sleep because your 1Hz offset at 10GHz is an appalling lapse of precision.
Neil G4DBN
PS, check out volt-nuts as well. As if you don't have enough to worry about with time and frequency precision
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
2f.
Re: Possible GPS problem 6th April
From: mike
Date: Sun, 24 Feb 2019 19:43:55 CET
Almost certainly the latter.
ISTR the Y2K bug got everyone up in a lather too.
Lets wait and see what your GPS engine does..As for mine, I really dont care
so long as the 1pps stays accurate...
View/Reply Online | Reply To Group | Reply To Sender | Mute Topic | Top ^ | New Topic
3a.