Well I looked into it some more and I'm now thinking the reason I stated above just isn't going to fly. As best as I can tell all GPS receivers download (and store) a new database of all the satellite's orbits (what I called Ephemeris data above but is properly called Almanac data) whenever they get a lock on any single satellite. A receiver also gets that satellite's precision orbital data (properly called Ephemeris data). Stated another way, each satellite transmits a message containing, among other things, a coarse/"imprecise" database of all the orbits (the Almanac) and a very precise position (Ephemeris) for itself. The Ephemris data is considered stale after 30 seconds and the Almanac after a few months. So if your GPS looses lock but regains the signal before 30 secs, the time to compute a new position is pretty much instantaneous. If you loose lock for more than 30 secs it'll take at least 30 secs (the time it takes for a satellite to send it's complete message) after you regain signal lock to fix your position. If you have turned off your GPS for several months it may take even longer as the receiver's data on which satellites are best to track is out of date. And as I stated before if you move your GPS since it's last lock, it'll take a "long" time for the same reason. All this said, I can't figure out a good reason why time-to-first-fix should have changed
this year vs last year. Perhaps for on turn-on #1 but thereafter it should have been the same (because the Alamanac had been updated). Perhaps something did change in the downlink (though I couldn't find it) or perhaps it's some programming bug in the GPS's firmware. Got me
, I guess so long as it is now fixed it's merely of academic interest.