This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

GPS Accuracy

Former Member
Former Member
So it begins.

I will have mine Fenix 5 on Saturday and will start doing comparisons to an Ambit 3 Peak. I don't have an F3 to directly compare to as of now.

Anyone have an F3 and F5 to compare?
  • I agree you shouldn't be waiting 5 minutes, but one full minute rather than haring off the moment the green ring appears might be sensible. Green ring means 'enough satellites for a 3D lock', which is only 4, and a few more seconds could give you 6-8, better initial position and less chance of the lock going berserk if you lose sight of one single satellite.

    My F3 gives a better lock at 60 seconds than green ring alone in a place where my 310xt regularly took 5 minutes to lock at all, because of the number of metal-framed buildings around it.


    I think dcrainmaker may have suggested just wait until pace shows zero before setting off? Certainly on my 735 after I get a lock it shows pace for some more seconds before dropping to zero.
  • On the F3 and now F5, I've generally found that the pace and alititude numbers settle within a few seconds of the green ring appearing (as long as I don't move around obviously). I don't think I've ever had to wait 5 minutes for a good GPS fix, but it can take noticeably longer if I'm forced to start an activity from inbetween tall buildings or under a thick tree canopy (not ideal, but sometimes you've got no choice).
  • Thanks webvan ;) And hoping to break the 1:25 barrier on the HM, ran 1:27:26 last year so should be possible... Gonna be tough though!

    Took a better look at the routes, the 'towards' ones is taken by the 735xt. It starts on the proper side of the road and only drifts off around the water (the only place where the F5 performs better). Shapewise the tracks are actually quite close I guess, but the 735xt is more accurate wrt the route I actually ran.

    The 'back' one is closer, both not great but hard to tell which would be the better one. Because I was unable to start/stop the watches at exactly the same time it's hard to comment on the 4sec pace difference.

    1k autolaps were indeed not in sync between the 2 watches, but again hard to say which one was most correct.

    Btw, just walked the dog again and this time set the F5 to GPS only, utter crap still (might be even worse then GPS+GLONASS) so that was no success. Might try smart-recording next time (but think I used that once before, without luck. Also, it kinda sucks that you can set GPS vs GPS+GLONASS activity specific but not smart vs 1sec recording). Atm I actually seriously think about taking them both during my HM, would be a very nice test...
  • I think dcrainmaker may have suggested just wait until pace shows zero before setting off? Certainly on my 735 after I get a lock it shows pace for some more seconds before dropping to zero.


    Shows zero and stays there for ten seconds or so, would be my take. Once it's done that, you can be sure it's got enough satellites for a stable lock, and it hasn't locked on to a reflected signal accidentally (which would cause a slow drift in position). But, broadly, yeah, somewhere between GREEN RING TIME TO RUN and 5 MINUTES MINIMUM :)

    When you're getting initial lock, still as possible is good, so satellite signals aren't interrupted . When you're moving, as fast as possible is good, as it helps with reflection rejection.
  • Shows zero and stays there for ten seconds or so, would be my take. Once it's done that, you can be sure it's got enough satellites for a stable lock, and it hasn't locked on to a reflected signal accidentally (which would cause a slow drift in position). But, broadly, yeah, somewhere between GREEN RING TIME TO RUN and 5 MINUTES MINIMUM :)

    When you're getting initial lock, still as possible is good, so satellite signals aren't interrupted . When you're moving, as fast as possible is good, as it helps with reflection rejection.


    I have a 620 (I'm upgrading to the 5X, which just arrived) and I've always found it best to stand still until the pace shows 0:00 and then do a quarter or half turn. At that point the pace will start changing again before quickly going back to 0:00. Then it's got a good lock.

    I was kind of hoping my 5X would be better than the 620 in that regard, but I'll have to try it out both ways to see how it fares.
  • Took a better look at the routes, the 'towards' ones is taken by the 735xt. It starts on the proper side of the road and only drifts off around the water (the only place where the F5 performs better). Shapewise the tracks are actually quite close I guess, but the 735xt is more accurate wrt the route I actually ran.

    The 'back' one is closer, both not great but hard to tell which would be the better one. Because I was unable to start/stop the watches at exactly the same time it's hard to comment on the 4sec pace difference.


    When I took a look at the routes on GC, the F5 route actually looked better to me because it seemed to cut through fewer buildings. But you know the exact location of your run better; I was just guessing from the track.

    Btw, just walked the dog again and this time set the F5 to GPS only, utter crap still (might be even worse then GPS+GLONASS) so that was no success. Might try smart-recording next time (but think I used that once before, without luck. Also, it kinda sucks that you can set GPS vs GPS+GLONASS activity specific but not smart vs 1sec recording). Atm I actually seriously think about taking them both during my HM, would be a very nice test...


    Can you post the dog walk? I'm just curious what such a bad track look likes?
  • Former Member
    0 Former Member over 8 years ago
    Replaced the watch

    Hi,

    I got my F5 replaced yesterday, due to a potentially faulty unit. At first, I thought the GPS of the replacement watch performed better . Here is a walk that I took with my first watch.

    https://connect.garmin.com/modern/activity/1648371856

    Then here is that same walk with the replacement watch. In reverse order

    https://connect.garmin.com/modern/activity/1648863105

    As you can see there is a considerable improvement.

    The next day which is today, I went for a morning run. Here are the results. The GPS track is all over the place. Same issue that I was having with my first watch.

    https://connect.garmin.com/modern/activity/1649593859

    Any help or pointers would be greatly appreciated. All of these walks/runs have been taken with GPS + GLONASS and also with 1s second reading. The walk with the replacement watch was done with the original firmware that came on the watch.
  • When I took a look at the routes on GC, the F5 route actually looked better to me because it seemed to cut through fewer buildings. But you know the exact location of your run better; I was just guessing from the track.



    Can you post the dog walk? I'm just curious what such a bad track look likes?


    Sure :)

    F5:
    https://connect.garmin.com/modern/activity/1650104018

    735:
    https://connect.garmin.com/modern/activity/1650105756
  • Sure :)

    F5:
    https://connect.garmin.com/modern/activity/1650104018

    735:
    https://connect.garmin.com/modern/activity/1650105756


    Yikes. I mean both have lots of wiggles and wanders, but that F5 track is awful. It looks like it never had your location locked at the beginning because you start in a building and promptly walk through a different building :eek:
  • Yikes. I mean both have lots of wiggles and wanders, but that F5 track is awful. It looks like it never had your location locked at the beginning because you start in a building and promptly walk through a different building :eek:


    Yeah, the F5 is with green gps and 'press start' msg though (which is about 2 mins after the gps going green in this case) so should be enough to get atleast something. And it's not as if we're surrounded with high buildings or thick leaves here, so meh. Not impressed to say the least :/