Elevation problem

Former Member over 5 years ago

I did read a lot of discussion here:

elevation-consistently-inaccurate-reading-too-low

Looks that I am not alone who has problems with elevation. Problem which happen in my case is ONLY during activity. I also think that there are several different problems.

I did check pressure and elevation graph in watch during the day (holding ABC and scroll), values were OK. I did compare pressure with my weatherstation and very close national weatherstation. I did stay at home during all day so altitude was same but elevation in watch change a little (+-5m).

I did start mtb ride at 5:30pm and for 1h ride I did make cca 17km. I did make several small up and down hills. Before I did stop activity I did see that elevation is not correct because I did start and stop almost at same place.

Here are graphs just after I came home:

you can see that elevation at the end is higher. I did start at 228m and stop at 264m.

pressure graph, I was SHOCKED, how pressure can change so much, cca from 1016 to 1022. I did find out that it happen during descent.

I did check my local weather station:

I also check national weather station, and there is also nothing strange. I did put marks to be able compare with watch 6 hours graph. I think, there is something wrong with sensor or with firmware of Instinct.

I did export activity to GPX and then enable elevation correction in web and export that too. Here is comparison

Blue is original GPX with elevation based on barometer. Green is GPX which has elevation correction enabled.

The significant change start at cca 0.55h = 33min of ride during normal descent.

In other threads someone mentioned sweat can make some problem. And there is also thread where rain is mentioned. Do you think that this can be problem in my case? It is 4 hours after I finish my activity now. Watch still showing wrong values of pressure and elevation. If it is problem of sweat or water then how long it takes to dry? or is it going to correct itself?

What I also know is that instinct is slow in elevation reading.

I can see that in fast descent. Another day I went slower on downhill and also slow down before next uphill. Results was quite OK. Of course, barometric altimeter will not catch all small hills.

Settings:

- altimeter auto cal = off

- barometer watch mode = auto , Tell the true, I don't understand this settings clearly.

thanks for any comments

Top Replies

All Replies

0 Former Member over 5 years ago

I did keep watch on hand during night. Elevation profile is flat at 277m. Pressure profile is almost same as profile at local weather station BUT with wrong values, showing last value 1022 but actual pressure is 1015.71.

Going to calibrate barometer:

#1 entering correct elevaion 216m

#2 do I know sea level pressure. I am not sure now, if I can enter pressure from my local weather station.

choosing YES and watch is showing 1014.77 so changing that to 1015.71 {value from local weather station and very close to national one}

question is: where watch take that value of pressure during calibration. I have seen that before, watch estimate almost correct value compare with old one which was totally wrong.

0 Former Member over 5 years ago

just find this thread design-flaw-sensor-opening-in-the-wrong-place-what-now

0 Former Member over 5 years ago

altimeter auto cal = off

- barometer watch mode = auto , Tell the true, I don't understand this settings clearly.

This may help.

support.garmin.com/.../

0 Former Member over 5 years ago in reply to Former Member

OK, it is more clear now.

I did think that it is good known problem and solution is OK. Sigma computers using barometric altimeter already for quite long time. So I trust AUTO mode.

Altimeter mode: Calculating altitudes using pressure has a direct solution. The problem is that the pressure changes over time. So if your ride is completely FLAT, the watch calculates altitude differences due to pressure changes. If your ride is short, then this is not a big deal, but if the ride lasts 5 hours, the deviation can be quite big.

Barometer mode: watch will not calculate elevation

OK, I will try altimeter mode for next rides.

Question is if I can calibrate altimeter (watch mode = altimeter) during activity. For example all day outdoor hiking and time to time I will be on places where I know exact altitude, if watch is showing wrong one, can I calibrate without stopping activity?

0 Former Member over 5 years ago in reply to Former Member

if watch is showing wrong one, can I calibrate without stopping activity?

I've just tried to do so and yes, I was able to manually calibrate during an activity. But, this was with my 945. I don't own an Instinct, but I suspect it will be the same. I started an activity, went in to settings, then sensors.

0 Former Member over 5 years ago

UPDATE #1

I was thinking that I will make some test: ride without wearing watch, ride with watch on right hand, ride with diff settings of barometer mode. So I could find out what can be problem. I decide to search in my previous ride and find problematic ones which has same part which I will ride again. I will compare elevation profile saved during ride using barometer with profile which is result of elevation correction on web.

Ride 108:

First 5km is very good. Then looks that sensors is kind slow. Maybe sweat.

Ride 115:

Again, first 5km quite OK then start to be shift. I did calibrate altimeter 3 times during ride.

First 16km of ride 108 and 115 is same.

Ride 120:

Looks that altimeter was shifted almost from start.

0 Former Member over 5 years ago

UPDATE #2

First test: ride without wearing watch. This test should show if sweat or close barometer hole cause problem. But I don't have heart strap so I will lose HR data. Solution could be If I could cover barometer hole. So I did draw gadget and then make 3D print.

you can download freecad drawing and STL for 3D printer : /cfs-file/__key/communityserver-discussions-components-files/173/gadget.zip

I did taped that gadget to watch and go for ride.

Ride 122:

I would say a very nice result. As you can see for the first 12 km, the watch nicely copied the terrain. Between 12 and 14 km it is not perfect, but I think that the height profile from the correction is also not perfect. Anyway between 10 and 12 km I am pretty satisfied because if you compare with previous rides (in previous comment) you can see that there always was big diff.

Ride 123: (same settings and with gadget)

I would say that first 9km is not so bad. After that looks that sensor start to be slow. I cam home and it tooks some time to analyse results.

so what happen?

Here is same chart but time base

Problem start at about 38minutes of ride, cca 14min before end of activity. Here is pressure chart in watch, photo is taken cca 30 minutes after activity end.

I did check my local weather station and national one and there was not anything strange. Weather was stable.

Conclusion:

I think that we are facing two problems. First is mentioned in thread about wrong design, looks that sweat in barometer hole can make problems. Second is that FW is not correctly recognize pressure pattern. I have barometer watch mode = auto for all tests.

anyway, I will keep using gadget with same settings. Maybe later I will try barometer watch mode = altimeter.

0 dav.dc over 5 years ago in reply to Former Member

That is really intersting, thanks a lot for sharing your tests.

I was also thinking about making some protection but I didn't know how to do it nicely. I will maybe try your solution if I can have access to a 3D printer

0 kHanDaichi over 5 years ago in reply to Former Member

Well done!
Thank you for all the tests, as well as for the 3D protection.
The design problem is quite obvious.
Garmin should hire you!

Bravo!
Merci pour tous les tests, ainsi que pour la protection réalisée en 3D.
Le problème de conception est assez flagrant.
Garmin devrait vous embaucher!

0 Former Member over 5 years ago

UPDATE #3

Ride 124:

I did hope that everything will be OK. But during the ride I already knew that it will be wrong. Anyway, here is chart.

I wanted to go same track as on ride 120 but I did skip one hill. So I did modify ride 120 for comparison.

RED = ride 124 but with elevation corection

GREEN = ride 124 using gadget

BLUE = ride 120 without gadget

NOTE: ignore part between cca 3.7km and 5.3km, Blue ride was over hill which I did skip on green.

Another interesting comparison is betwen ride 123 and this one = 124.

Rride 123 after 3.1km is same sa ride 124 after 8.9km. And check uphill, it is almost same.

EDIT: forgot to mentioned, there was normal weather, nothing strange and here is pressure graph cca 1h 50min after activity