Suunto 9 Highly inaccurate speed readings

[surfboomerang]

I already created a support request for the highly inaccurate speed reading of the Suunto 9 vs my previous Ambit 3 peak. I’m just curious if someone else noticed the same behaviour…

This is the support request I filed:

===================================================================

A couple of months ago I replaced my Ambit3 peak with the Suunto 9 Baro. I noticed that the GPS recordings of the Suunto 9 are highly inaccurate. When I export the activity as GPX and analyze it, I noticed that the calculated speed is constantly spiking up and down with about 10-20 KM/H. This is also noticeable on the screen of the watch.

The battery setting during all the activities is set to “Performance” with GLONASS on so the GPS is set to “Best”.

I included 3 fragments as attachment to show the spiking. Attachment 1 and 2 show the spiking of the Suunto 9 watch and Attachment 3 shows my previous Ambit 3 recordings.
All the activities are exported as GPX from Strava.

What I already did to try to solve it:

• Soft reboot the watch
• Update the watch to the latest firmware (2.5.18)
• Enabled/disabled GLONASS

Can you tell me if this is a known issue that can be solved in future firmware updates or that this issue is hardware related?

===================================================================

And these are the attachments I included to demonstrate the difference:

Suunto 9:

Ambit 3 peak

[Dimitrios Kanellopoulos]

Can you try to upload those reading in .fit file on quantified-self.io ? I would love to see those graphs fro the fit files

[surfboomerang]

I exported the FIT files from sports-tracker and uploaded them to the Quantified self site.
Here are parts of the same activities as my previous post:

Suunto 9

Ambit 3 peak

So for running (suunto 9 capture 2 and Ambit 3 capture) it looks quite ok now. For windsurfing however (suunto 9 capture 1) the spikes are still there.

Another thing I noticed when I analyzed the files on the Quantified Self page was my heart rate graph from Suunto 9 and Ambit 3 peak. In both activities I used the same HR belt but look at the graphs below…

Ambit 3 peak

Suunto 9

[Dimitrios Kanellopoulos]

the HR is not filteredd in that tool. The more spikes the worse the belt is.

About windsurfing I am not sure I have never tested that high speed sport.

[surfboomerang]

The Ambit activity was synced from Movescount to sports-tracker. The Suunto 9 moves were synced with the Suunto App…Is it possible that the data of the Ambit 3 was already filtered by Movescount?

I will do another run today and treat the HR belt connectors with gel to see if that will improve the readings.

[Dimitrios Kanellopoulos]

@surfboomerang yes it can be.

I would just not trust any GPX exports from ST.

And in general GPX

[surfboomerang]

I know, but how am I suppose to export a fit file from Suunto app at this moment?

[Dimitrios Kanellopoulos]

@surfboomerang good one.

there was somewhere one hacky thing posted. I need to dig that up.

If you have strava sync you can via web-> Go to the activity and append in the url /export_original that will give you the .fit file that was send to strava

[surfboomerang]

Thanks, I will try that this evening.

[surfboomerang]

@Dimitrios-Kanellopoulos I exported the same activities as before, but this time from Strava with the method you mentioned. Worked like a charm!
After that I analyzed it in the Quantified Self site, but unfortunately the graphs for windsurfing are identical to the one I exported from Sports-tracker. So the spikes and drops are still there.

I also did a new run today with the heart rate belt and some contact gel, but I still see the spikes in the heart rate graph. The Suunto App shows the HR graph ok (so it filters out the spikes)

[Brad_Olwin]

@surfboomerang said in Suunto 9 Highly inaccurate speed readings:

@Dimitrios-Kanellopoulos I exported the same activities as before, but this time from Strava with the method you mentioned. Worked like a charm!
After that I analyzed it in the Quantified Self site, but unfortunately the graphs for windsurfing are identical to the one I exported from Sports-tracker. So the spikes and drops are still there.

I also did a new run today with the heart rate belt and some contact gel, but I still see the spikes in the heart rate graph. The Suunto App shows the HR graph ok (so it filters out the spikes)

For HR, it is likely your belt is bad, not the sensor. When I routinely see HR spikes and I have cleaned the belt, it is time for a new one. Also, spikes can be worse in the winter if you wear fabrics that transmit a lot of static electricity.

[Dimitrios Kanellopoulos]

@surfboomerang I need to add filtering at that tool then

[Dimitrios Kanellopoulos]

@surfboomerang btw joke as these spikes are natural to happen with others as well.

Now back to the windsurf issue. Could you for just my sake use a swim profile or standuppaddling? Some water profile that is designed for on water GPS. I suspect that is the issue. SUP will do I think to make the gps more relaxed on reflections.

S9 uses different GPS profiles for each type of movement. FYI

[surfboomerang]

Always worth a try.

Although I doubt that it will work because I’ve seen this same behaviour in a car on cruise control. See https://forum.suunto.com/post/16076. I used sports profile “Other” there. (or at least I think it is translated to “Other”)

I will try with the SUP profile for windsurfing next time. Hoping for some good wind in the near future. And of course some free time

[GingerBeardMan]

I find, when running, if I do fartleks, intervals or other workouts that include quick changes of pace my suunto 9 will log me as having achieve super human speeds at the beginning of the interval before settling down to something more realistic.

[surfboomerang]

@Dimitrios-Kanellopoulos Because the wind forecast is not very good in the near future, I decided to test the profiles by car to see if the SUP or WIndsurf profile does make a change.

I started an activity with a SUP profile and drove on the highway with the car on cruise control (about 110 km/h). Below is the resulting graph:

I also did the test with the Windsurfing profile (cruise control at 100 km/h):

The windsurf profile looks pretty clear this time, but I doubt that this is because of the profile I think I was lucky with such a clean graph this time. As you can see on my first graph it can still spike really up and down.

What really grind my gears is that the spikes are almost always identical in height and in a lot of the cases has the same drop value as the spike a second before.

An example to clarify:
If I drive an average of 110km/h and I see a spike of 120km/h, in a lot of cases it follows the next second with a drop to 100km/h. So it averages again to 110km/h, but the recorded max speed is 120km/h

It is so common that the spikes are almost identical in height and drops that I can’t really imagine that this is a GPS reception issue. More like a GPS recording issue. Like the watch is recording too fast ahead and than slowing down to compensate…

If you look at the line below, I see that the trackpoints are perfectly in line with the driving direction, but they do not have even distance between them some are closer of further apart, although I was driving at a constant speed.

Is there a possibility to see the number of satellites or reception quality from the recordings?

[Dimitrios Kanellopoulos]

@surfboomerang keep in mind that the spikes are vise versa. Aka you stopped.

[Dimitrios Kanellopoulos]

@surfboomerang I am working on QS v2 to show those differently and support more data btw

[surfboomerang]

Little status update:
Suunto support requested to return the watch for investigation.

Lets see if they can find something…

[surfboomerang]

Watch is replaced by Suunto support. Just got my new one.
My first run looks promising. No real spikes like I had before. I’ll test high speed sports soon to see if my issue is really resolved.