Note to O.P. -- this subject was beat to death about a month ago on this
very forum, so if you're looking for more discussion, it's available.

General consensus, not surprisingly, was that newer devices should be
fine, really old devices may show some failures. Old devices have their
week counter roll over in a little less than 20 years, newer devices
should roll over in about 150 years, if they're taking advantage of the
three additional bits in the week counter.


On 2/24/19 7:22 AM, Soartech wrote:
Just saw this:

Subject: [UKMicrowaves] GPS Frequency standards.. Week Zero alert.
Date: Thu, 7 Feb 2019 18:39:05 -0000
From: Mark GM4ISM



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 1 PPS 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!