[time-nuts] GPS PRN 32

Martin Burnicki martin.burnicki at burnicki.net
Thu Jan 28 03:08:52 EST 2016


Magnus Danielson wrote:
> Charles,
> 
> Considering how the UTC parameters (IS-GPS-200H, Table 20-IX and section
> 20.3.3.5.2.4) got upset for some of the SVNs, the correction from GPS
> time (as corrected for that PRN) into UTC got a shift of almost 13,7 us.
> As a GPS receiver receives information, solves for position and time
> (fixed timing receivers naturally only for time), correct into GPS time
> and then into UTC you end up with having to select the information from
> one of the GPS satellites for translating the GPS time into UTC time.
> Depending on the state of that GPS receiver, it may or may not select
> this bad info, and it may change selection at any time. That's why we
> have observed receivers of the same brand and same FW fail at different
> times, but starting having problems at the same trigger time. While
> there can be receivers that have some form of protection from this
> particular problem, I guess many don't, and in this case, the specifics
> of the receiver FW and the actual state of the receiver may have cause
> the full range of heavy to no impact. I would be careful to make to much
> judgment of various receivers because of this. It is clear that it's a
> serous hit to at least most of them.

Absolute correct, IMO.

> I remember the impacts from PRN 31 and PRN 32. The re-occuring GPS WN
> wrapping issue is another. This instance is rather unique in it being
> related to the ground infrastructure seems to have provided bad upload.
> 
> There is an intense amount of work behind the scenes. Some patience will
> be needed before even a minimalistic statement can be found.

Now there's a statement available. See my other post.

Martin



More information about the time-nuts mailing list