Hi David, and a Happy New Year,
I take it you are still waiting for a reply to your last post, (though Ash tells me he has sent you a PM)?
I normally leave track analysis to Lee as he can analyse track files using software much quicker and more efficiently than I can. (I have to do it manually - by physically looking through the files in Wordpad on my Windows laptop - which is much less efficient and FAR more time consuming.) I am surprised that Lee hasn't responded yet, but I do know that he has been very busy with development work - as well as his 'day job' so guess he has probably just missed the post or hasn't had time to get round to looking at the files yet.
In the absence of a response from Lee, I have nonetheless taken a look at all the files you uploaded to Dropbox over the past three days and found the following.
I started by examining your first track file from 19th November - which starts with bootup (at Henstridge) at 140556. Everything starts up as expected, and I can see regular GPS reports ($GPGGA and $GPRMC files) at one second intervals right through the log until the file ends suddenly at 143557, presumably with your Rosetta shutting down several minutes before your reported GPS dropout over Lymington at 1441. Running the file on the PilotAware / Aircrew Playback Tool shows that your Rosetta must have shut down just as you were about to cross the A338 well before you reached Lymington.
The second track file, which is a continuation of the above flight, starts at 144245 as you headed out over the Solent, with a fresh reboot of your Rosetta AFTER your reported 1441 dropout. This file shows a completely normal reboot with GPS position reports at 1 second intervals all the way through 'til the file stops after landing at Sandown - so in this case completely normal.
The first set of logs from 25th November shows pretty much the same thing - normal startup at 0945 and regular GPS reports at 1 second intervals throughout the track (for two and a quarter hours) until the log stops suddenly just after a routine set of GPS reports at 120019. This again indicates that the GPS was working normally throughout the entire flight up to the point where the Rosetta shut down at least a minute and a half before your reported dropout at Thelwall. Unfortunately this track won't replay on the PilotAware / Aircrew Track Replay Tool, which makes it a bit more difficult to say exactly where the Rosetta stopped working.
The second track file for 25th November starts at 13:42 - presumably as you were getting ready to Depart Barton for your return flight. Analysis of this log again shows regular GPS reports at 1 second intervals for 90 minutes - right up to 151249, shortly after which the track log again stops - again in my opinion indicating a shutdown of Rosetta, rather than a GPS failure. This track file also unfortunately won't run on the Replay Tool.
The final track file starts with a system restart at 151814, and after restart, again shows regular GPS reports throughout the log until the system shuts down at 1538.
In short, all of these files show solid reporting of GPS positions through fairly long flight periods, with in the case of the first file on 19th November and the First and Second Files on 25th November a sudden shutdown of the Rosetta unit rather than a simple GPS failure, so we need to look elsewhere than the GPS for the reason.
My primary suggestion would be to fully investigate the power feed to Rosetta. I am aware that (like myself) you are powering your Rosetta from a Charge4, so that should be pretty much bombproof, though it's worth checking that you aren't overloading it with other devices. I would concentrate my attention on the power cable and connections between the Charge4 and Rosetta. Are you using the supplied power cable, or have you fitted a longer one - and if so is it of the recommended gauge? Check for a solid connection between the microUSB plug and socket at the Rosetta end. Do the 'wiggle test' while observing the red 'Power' LED (which is inside the case on the antenna end of Rosetta under the 869MHz antenna socket). The Power LED must show solid constant Red - If it blinks or goes out at all, you have a bad connection. Also look for Voltage Warnings or Errors on the PAW Home Screen, though I can't see any of those reported in the track logs.
Another less likely possibility - is there any chance of the unit having been subjected to significant overheating? - for example is it mounted in close proximity to the cabin heating system? Again, I can't see any evidence of this in the log, but could have missed it.
Please let us know if you find anything and we can take things from there.
Best Regards
Peter