Problems with Altimeter

TeaRacer said:

That's why I think if the Solar is so inaccurate, the altimeter is defective.

That cannot be TeaRacer.  You have already been told the problem is with sweat. How dare you argue against "hundreds of similar controlled experiments"? (Try not to laugh after that quote). Stop making ridiculous suggestions to anyone who's asking for assistance and conform! :) 

Because thousands of users do not have these problems. I had a problem with the old Intinct, exactly as you write, at the same time I had a new one and there were no problems. Understand that it can't be sweat, because I had both run and I took turns. What more can I say? The old Instinct was thrown by the humidity, sweat, rain hundreds of meters. Solar decides a maximum of meters by sweat, rain and humidity. Do you understand me already? I'm not making fun of anyone, I'm just saying my experience, so please don't attack me.

How do you explain to me that the old Instinct didn't point exactly on my hand and the same route with the new Solar was okay? It was at the same time, the same conditions, the same SW, but different watches (chipset, etc.). Thank you

You have already been told the problem is with sweat. How dare you argue against "hundreds of similar controlled experiments"?

Looks like you misunderstood the experiment I have shown. The sweat is causing the steep drops of elevation shown in the previous screenshots of @henzim.  The same drops shown on the elevation profile from my liquid experiment.

The steady elevation shift is rather caused by the change of temperature and/or by the changes of the atmospheric pressure (though I would not completely exclude the build up of sweat either). I am not sure how it is with Instinct Solar, but with the standard Instinct, a change of just a few of °C results in the difference of tens of meters. It could be that the firmware of Instinct Solar is better thermally calibrated, but perhaps also not.

You can do another simple experiment -  start a dummy activity, let the watch sitting on a desk (same elevation, stable temperature), wait till the internal temperature and the elevation stabilize, and then expose the watch to some temperature changes (warming up, cooling down). My watch shows deviations to +/- 50m in such experiment. Note: you have to do the experiment preferably outdoors or with an open window, to exclude the influence of air pressure fluctuations in a closed room.

Also look up the weather data for your location for exactly the time of your activity, and find out what the atm. pressure was at the moment of the start of the activity, and at the end. A millibar of pressure increase will result in the drop of elevation of roughly two meters

TeaRacer said:

That's why I think if the Solar is so inaccurate, the altimeter is defective.

If the altimeter were defective, it would not record the elevation correctly when worn in the bag, as henzim wrote. The deviations clearly come from external factors like the temperature, wind gusts, the sweat, and/or the wrong position/fit of the watch on the wrist, otherwise it would not show diametrically different profiles when protected from the external factors vs. when worn on the wrist.

So how is it possible that the old Instict was showing me large variations, on the same wrist, at the same time with the same firmware on the same route? And not Solar? Therefore, it is clear that Solar has been improved and if it shows such large deviations as the old Instinct, it is probably a defective piece. This will be decided by an authorized service center. But such large variations in watches that have military standards, and reputable brands, are large fluctuations. There are many reviews of Solar on the Internet, and I have not read anywhere that he has problems with altitude. Today I went in extreme conditions (snow, big wind, then sun) the usual route, where the old Instict always showed how many times it describes large deviations. Solar is + - in the order of meters accurate.

Yes, it is possible. An obstacle in the sensor may not be very visible. It would be ideal to try another piece of the same watch, after agreement with the support of garmin. Because I understand that with such a large deviation, the user is dissatisfied.

TeaRacer said:

, so please don't attack me.

I don't think you understood my post. I completely agree with what you're saying and my sarcasm wasn't directed towards you. 

TeaRacer said:

Yes, it is possible.

BTW, try doing the same controlled experiment with your Solar, as I posted previously: sit motionless at the desk, with the Instinct firmly attached to your wrist. Start a Walk activity (never mind the GPS lock). Wait 5 minutes to see whether the elevation is stable. Then pour some luke-warm or body-warm water on your wrist near the pressure sensor opening (better have a towel or a recipient under the arm ready). You can make some light wrist movements to simulate the running. Wait 2 more minutes, and stop & save the activity. Then post the elevation profile it recorded. Would be interesting to see whether there is really a difference here between the Solar and non-Solar versions.

So I'm sorry, English is not my language, so it's probably a language barrier.