Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
AW5 LTE GPS accuracy problem
- Thread starter rbart
- Start date
- Sort by reaction score
Ok, strange. I was using WorkoutDoors also.
To nail the issue more down: I have a aluminium AW 5 LTE and am using the original sport band. Crown was up, LTE on and iPhone and Wifi were in range when I had very bad results. iPhone left behind at home.
Forgot: watchOS 6.01.
My experience FWIW - I've done several outdoor runs with my series5 (no phone) using the Apple Workout app.
I've seen two anomalies - one route hiccup where the GPS clearly pulled location points ~200 meters off the track briefly then corrected itself. Other appeared to be the GPS took about 3 minutes to start recording location information, but did fine once it started.
I've also had a couple weird gps tracks when on my fishing kayak using the outdoor-cycle workout since it's a pedal kayak -- however I had my phone with me so the watch GPS wouldn't have been involved.
I've seen two anomalies - one route hiccup where the GPS clearly pulled location points ~200 meters off the track briefly then corrected itself. Other appeared to be the GPS took about 3 minutes to start recording location information, but did fine once it started.
I've also had a couple weird gps tracks when on my fishing kayak using the outdoor-cycle workout since it's a pedal kayak -- however I had my phone with me so the watch GPS wouldn't have been involved.
I did a test just right now. Switched off BT on my phone, Wifi on my AW off and LTE on AW on.
Waited until I had a blue bar in Workouts with 4 satellites, started a walk for 1km and the result was mixed:
![Bildschirmfoto 2019-10-07 um 19.32.33.png]( "Bildschirmfoto 2019-10-07 um 19.32.33.png")
![Bildschirmfoto 2019-10-07 um 19.34.32.png]( "Bildschirmfoto 2019-10-07 um 19.34.32.png")
From a distance perspective, I measured 1.025 with GPSies and the resulting track is 1.04km which is within tolerance.
What makes me nervous are the very strange hiccpus the watch takes from time to time.
Will repeat this with LTE switched off.
Waited until I had a blue bar in Workouts with 4 satellites, started a walk for 1km and the result was mixed:
From a distance perspective, I measured 1.025 with GPSies and the resulting track is 1.04km which is within tolerance.
What makes me nervous are the very strange hiccpus the watch takes from time to time.
Will repeat this with LTE switched off.
I have submitted a feedback to Apple (https://www.apple.com/feedback/) but I don't know if they read it.
I have contacted support for my first S5 watch, they have just sent me links to the documentation, how to start a workout, how to enhance accuracy and told me my watch is fine
Nothing interesting
If LTE corrupts GPS tracks, I will return my watch and take a GPS only one.
I have contacted support for my first S5 watch, they have just sent me links to the documentation, how to start a workout, how to enhance accuracy and told me my watch is fine
Nothing interesting
If LTE corrupts GPS tracks, I will return my watch and take a GPS only one.
My bad, I got confused here. The magnet interferes with the compass and not GPS indeed.
Did my second test today. Same procedure (disconnected from iPhone, pre checking of GPS reception) but now with LTE off.
Track looks better, but not perfect. (Yellow is with LTE, blue is LTE off):
![Bildschirmfoto 2019-10-08 um 18.00.41.png]( "Bildschirmfoto 2019-10-08 um 18.00.41.png")
I will raise a support ticket with Apple, found something under MySupport. Would be happy if you folks with problems could also raise a ticket so we get this thing resolved. Thanks!
Edit: colleague in German forum gave me these links:
Apple Bug Report für Customer: https://www.apple.com/feedback/
Apple Bug Report für Developer: https://bugreport.apple.com/web/
Also in chat with Apple support.
Track looks better, but not perfect. (Yellow is with LTE, blue is LTE off):
I will raise a support ticket with Apple, found something under MySupport. Would be happy if you folks with problems could also raise a ticket so we get this thing resolved. Thanks!
Edit: colleague in German forum gave me these links:
Apple Bug Report für Customer: https://www.apple.com/feedback/
Apple Bug Report für Developer: https://bugreport.apple.com/web/
Also in chat with Apple support.
Last edited:
Your test is very interesting and looks like my own tests ...
It's always difficult to be sure to reproduce a run in the same condition, but it seems that LTE localization is used in case of poor GPS reception ...
And it's not accurate at all.
I have 2-3 days to decide if I keep my cellular watch or return it to get a wifi/GPS version
It's always difficult to be sure to reproduce a run in the same condition, but it seems that LTE localization is used in case of poor GPS reception ...
And it's not accurate at all.
I have 2-3 days to decide if I keep my cellular watch or return it to get a wifi/GPS version
It shouldn't be that way really. But if it is that could be the reason why I'm not suffering poor GPS track with LTE on. I have done all my runs on very good LTE coverage.
I have the notion that calibration improved. But you are probably right with the LTE thing. Where I walked and had the hiccups, the LTE reception is not good. Not sure about the GPS reception though, have to check.
I would wait with the exchange until the next updates. You can always switch off LTE if it disturbs too much.
To the testers.. have you tried using a 3P app, like WorkOutDoors that will let you confirm GPS signal strength before starting your activity? I have read that once you are moving GPS acquisition can struggle more. I did a walk recently, about 4 miles and there was one little hiccup but the track looked really good. I had my phone with me but turned off BT.. but it may have turned itself back on.
I will try again today with just the watch, one way with LTE and one way without. WIll be shorter tha 4 miles.. don’t have that much time.
I will try again today with just the watch, one way with LTE and one way without. WIll be shorter tha 4 miles.. don’t have that much time.
Two samples where we can see big glitches:
On the firt one, it shoult be on the blue line, on the second one, the top track must be near the yellow mark
It was at the middle of my run, but in some areas where LTE was probably poor and GPS also.
But I never has such glitches without LTE
On the firt one, it shoult be on the blue line, on the second one, the top track must be near the yellow mark
It was at the middle of my run, but in some areas where LTE was probably poor and GPS also.
But I never has such glitches without LTE
Attachments
Just did a quick 20 minute walk. No phone. 10 minutes out with LTE on and 10 minutes back with LTE off. Both tracks look good IMO. I couldn’t pick a winner. THere is a slight shift in the sidewalk in one part of the route and the track accounted for it well IMO.
I have made some tests during my last night run ...
... and it's difficult to have clear point of view.
I have made one half with LTE ON and one half with LTE OFF.
The trcking look similar in both cases, not perfect but not so bad.
I think LTE coverage was quite good in this area. I have logged with WOD, I don't know if I can analyze more by exportin GPS accuracy data into a GPX ?
... and it's difficult to have clear point of view.
I have made one half with LTE ON and one half with LTE OFF.
The trcking look similar in both cases, not perfect but not so bad.
I think LTE coverage was quite good in this area. I have logged with WOD, I don't know if I can analyze more by exportin GPS accuracy data into a GPX ?
you could try with mygpsfiles.com
Maybe first export FIT file and split it into halfs with fitfiletools.com. This is my spontaneous idea, there are probably a lot more efficient ways.
With mygpsfiles you can perfectly compare half 1 and 2.
which is a little ray of hope.
I have the theory that streaming via LTE has an impact. Could be a CPU problem. But very vague, need a test. Too bad I am injured that limits my testing capabilities. :-(