Notes: Running latest software (22.40,GPS 2.70). Problem is repeatable and validated with a number of time sources. Bluetooth connection to the phone was off during all the testing. I do not know how long the issue has been there or if related to a recent update. Although I have not yet done a master reset I have done numerous re-starts and soft resets (holding down button for 20+secs). I have tested with numerous activity selections and with all three combinations of GPS source. This is NOT a general time issue - the watch is keeping time just fine.
I track my runs during orienteering events of which we have just started back and I have done two recently. The GPX export is used, in conjunction with the electronic punching and results to analyze routes etc. Two weeks ago I had an issue with the analysis as it appeared the times in the GPX file where all shifted forward by about 20 seconds so did not match up with the data in the results from the punch system. I confirmed that the issue was only showing with my GPX and did not appear to be a result of a sync problem with the punching software system and controls. After seeing the same issue with the second event I had a closer look. The repeatable issue is shown in the linked video (note that the video jumps forward in areas where I cut out sections not needed. I have used multiple reference time sources but the issue is so obvious when watching that this is almost irrelevant. Despite saying this, it would easily go unnoticed with normal use, especially as it would be rare to compare against another time source.
The Video: Basically, starts showing that the Fenix watch page is the same time as the reference watch (previously synced to GPS and validated against other time sources within a second). On the Fenix, select (not start) and activity that includes the time of day on the data screen (note that this only shows the minutes). As soon as the initial GPS signal is found the minute on the data screen jumps up to the next minute so the reference time is 8:51:50 and the 5x activity shows 8:52. Starting the activity at 8:52:00 on the reference watch and run it for a minute or so - note that the 5x will continue to show the jump to the next minute earlier than the reference. Stop and save the activity and then return to the watch face and the time again matches the reference time.
Related or not?? Included in the end of the video I have shown what is happening when syncing the time to GPS - the time syncs and is then ahead of the reference time than about 20 seconds later it adjusts back to the correct time.
Downloaded the GPX file and checked the time stamps (shown in the video) - the first one is at 08:52:24 so 24 seconds after starting the activity at 08:52:00 (or close). After the initial time this occurred I analyzed the track locations matching location to the control on the ground and in this case the shift throughout the GPX file was about 18 seconds throughout.
Related or not?? Included in the end of the video I have shown what is happening when syncing the time to GPS - the time syncs but jumps ahead of the reference time by about 22 seconds, then about 8 seconds later it adjusts back to the correct time - jumps backwards from 08:54:39 to 08:54:20 (correct time).
If it wasn't for the need to sync up time stamps between the watch and another source I would likely never have seen this. I cannot think of any reasonable reason for this given it is consistent, easily tested and impacts both the watch and the resulting GPX files uploaded to Connect. I cannot see how it can be a direct GPS issue so somehow the watch is screwing up the time stamps from the GPS chip?
I'm assuming I will be forced to do a master reset soon but am curious if anyone else can reproduce this or is it limited to an issue with my watch (or can think of some other thing I can do to test further).
