Fenix 5X GPS speed spikes

beeing busy running, i didn't test the bike mode yet myself but my guess is that this problem should be reproducible.

1.) do you stop the activitiy manually or have autostop activated (just pause or did you click on "resume later"?)
2.) do you have recording set to 1second?
3.) did you try to reproduce it or are these more or less random spikes?
4.) do the spikes always occur "immediately"/"near" the pause/resume location or random? (not good visible in your examples)

if 4.) holds then the problem seems to me similar like driving through tunnels with older devices: track will go crazy and as a result the speed will follow to go crazy (this is the reason I would always use a cadence/speed sensor for (road)-biking.)

to be more concrete: my guess is that the problem might lie within the smoothing methods which are internally applied to the track-data (especially if 4.) holds!): there is some kind of averaging of the gps-locations going on and maybe (like a moving average for example) the algorithm needs a few seconds of data after hitting the resume button in order to estimate the current location and saving its value again.

at the same time, speed measured from the gps (in m/s), can be easily calculated as the distance between two saved locations (1s recording). a speed of, for example 55kph corresponds to 15m/s, implying that the next measurement point used for calculation (1s recording) was in a 15 meter distance away from the location of the pause.

my guess is that the saved (smoothed(???)) locations at the pause time t and the next moving time t+1 will be significantly different (could also be provoked from bad gps signals, i.e. the watch trying to fetch a good connection again after the pause) while the algorithm to calculate the speed just forgets about the pause - this could create such a spike.

this can clearly be seen in your 179kph spike:
https://connect.garmin.com/modern/activity/1681378649
take a look at the waypoints at which the spike occurs: there is a straight line connecting the points where the speed was measured, while the watch was able to approximate the curve of the road before the break well. .

that beeing said: keep in mind that gc does not show all available and saved data points.

i hoppe to find time to check the behaviour the next couple of days.

how long do you pause approximately or is a rather quick "traffic light - like" stop-start already enough?

>>However, in some activities although I'm at the same place, recorded pause and resume location is some meters far from each other, and this may be caused by GPS satellites changing prosition at that time. Also I sometimes forget to resume and remember after kilometers, but at these times it never did such things.
this would support my thesis: maybe garmin implemented a threshold for "way to unrealistic" distances after resume. who knows.

anyway i am on your side: either your gps is slightly defective or the algorithm should/could be improved :)

do you mind to upload some of the .fit files for me? (not the first one, as you pointed out this might be the most boring and comprehensibly one)


btw: nice rides of yours :)

The first week or so I was getting some crazy max speeds. Recording 130mphs etc! Haven't had it since then though. I do notice minor differences in mph and distance between Strava and the Garmin App. Though I put that down to the different algorithms that the apps use for the data stream.

Do you have those 3D settings, auto climb turned off too? I turned them on, for grins, to see what they did and they caused some odd data to be recorded.

I just have 1 second recording, Gps+Glonass set. Use sys settings off.

Mick,

i hoppe to find time to check the behaviour the next couple of days.

how long do you pause approximately or is a rather quick "traffic light - like" stop-start already enough?

>>However, in some activities although I'm at the same place, recorded pause and resume location is some meters far from each other, and this may be caused by GPS satellites changing prosition at that time. Also I sometimes forget to resume and remember after kilometers, but at these times it never did such things.
this would support my thesis: maybe garmin implemented a threshold for "way to unrealistic" distances after resume. who knows.

anyway i am on your side: either your gps is slightly defective or the algorithm should/could be improved :)

do you mind to upload some of the .fit files for me? (not the first one, as you pointed out this might be the most boring and comprehensibly one)


btw: nice rides of yours :)

The first week or so I was getting some crazy max speeds. Recording 130mphs etc! Haven't had it since then though. I do notice minor differences in mph and distance between Strava and the Garmin App. Though I put that down to the different algorithms that the apps use for the data stream.

Do you have those 3D settings, auto climb turned off too? I turned them on, for grins, to see what they did and they caused some odd data to be recorded.

I just have 1 second recording, Gps+Glonass set. Use sys settings off.

Mick,

There may be some minor differences but these are huge. 3D is on in some activities but no auto climb etc....

i hoppe to find time to check the behaviour the next couple of days.

how long do you pause approximately or is a rather quick "traffic light - like" stop-start already enough?

>>However, in some activities although I'm at the same place, recorded pause and resume location is some meters far from each other, and this may be caused by GPS satellites changing prosition at that time. Also I sometimes forget to resume and remember after kilometers, but at these times it never did such things.
this would support my thesis: maybe garmin implemented a threshold for "way to unrealistic" distances after resume. who knows.

anyway i am on your side: either your gps is slightly defective or the algorithm should/could be improved :)

do you mind to upload some of the .fit files for me? (not the first one, as you pointed out this might be the most boring and comprehensibly one)


btw: nice rides of yours :)

It can be a short pause resume or for some minutes. It is some minutes in the attachment I think.
http://www.f-blog.info/downloads/1803449565.zip

I just found some time to take a look at the .fit file.

First of all, trying to import the file in R via the "fit" library gave an error: "Error in decode_fit_file(filename) : FIT file integrity failed.", while importing other .fit-files work flawlessly with this package.



From the look at it, your GPS seems to be alright: I couldn't finde any abnormal latitude or longitude data and everything seems to be okay-ish besides the recorded speedspike. (even recorded distance don't reflect the speed spike)

The next figure illustrates the trajectory of all the saved datapoints around the SpeedSpike. The location of the SpeedSpike is colored red.

https://abload.de/img/gmaps8usga.jpeg

(Note that you took a rather "steep" curve/corner right before you paused.)

You see, that the track in the neighborhood of the SpeedSpike is quite good. For me it was very surprising to find out that the SpeedSpike is in fact located at the last datapoint which was saved BEFORE you hit the pause button (the pause here took approximately 5 minutes and 30 seconds) - I assumed it right after the pause.
Assuming the speed is saved in real time, this destroys my first guess: the jump to the next location after the pause is "in the future" and won't be used for the calculations. Hence, I can not explain the Spike coming from the difference between the locations before and after the pause anymore.

I had exactly the same problem with Garmin Fenix 5s. I fixed it by changing the below settings.

Speed 3D: must change to inactive

Distance 3D: must change to inactive

Recording: set to 1second

Before:
https://connect.garmin.com/modern/activity/1814677027

Now:
https://connect.garmin.com/modern/activity/1870598897

Regards,
Panagiotis

Except one of two occasions my Fenix 5X GPS runs stable during running activities, it's neither perfect nor a disaster. However, during my biking activities it does strange speed and sometimes position jumps.

For example, during the ride I stop somewhere and pause the activity. Then after a minute or two I resume it and continue. The thing is it displays incredibly high max. speed after this pause/resume and it breaks the high speed reading on Connect too. Strava etc. trims such jumps but Connect doesn't. Also the position it starts recording after resume is some meters away from the road, so it doesn't create a direct route like F3 during same pause/resume.
Also I init biking activity and wait for the green ring. Some location deviation is normal here as it doesn't fixed to all the satellites yet, but it starts the activity with extreme high speeds again.
Also during normal rides I saw some speed spikes, no pause/resume. Speed spikes are short spikes and sometimes reach 170KM/h.

Eg.


I wonder if only my unit GPS is faulty and do you see such conditions on your watches during your rides? I don't want to send it to service but as I don't like its battery performance too, I may have to. But this is my second F5X unit and I know the first one jumped me to a 170 KM/s speed during a ride... So it's maybe not a single unit issue. Mostly they happen GPS only I think but sometimes with Glonass too. So maybe it's not related to that.

Some example activities:
96kph spike at activity start
https://connect.garmin.com/modern/activity/1809479150
Pause/resume spikes:
71kph: https://connect.garmin.com/modern/activity/1807531603
130kph: https://connect.garmin.com/modern/activity/1803449565
55kph spike: https://connect.garmin.com/modern/activity/1797675418
172 kph spike: https://connect.garmin.com/modern/activity/1681378649

Stupid 209kph jump during a short unimportant ride. Strangely Garmin says max. speed is 3,3 kph. :D
https://connect.garmin.com/modern/activity/1799634929

So what about your tracks, do you see such high speed jumps on your rides? Or is it only me?

I had exactly the same problem with Garmin Fenix 5s. I fixed it by changing the below settings.


Speed 3D: must change to inactive

Distance 3D: must change to inactive

Recording: set to 1second

Before:

https://connect.garmin.com/modern/activity/1814677027

Now:

https://connect.garmin.com/modern/activity/1870598897

Regards,
Panagiotis

I have had the exact same problem with runs and it's more exaggerated when doing intervals. It's been happening since I first got the watch. I have been going back and forth with Garmin about this issue and a few others and they have got me to do a hard reset and keep upgrading to the latest beta firmware. I've also given them access to my Garmin account and sent them all of the data from my Garmin folder(when plugged into the computer). I haven't tested since the latest 4.21 but the last time I checked I was still having the same problem.