This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

barometer drift

Hi, with latests fw barometer readings ( and so elevation) are ok only after a factory reset, for a few of days, then slowly drifts, resulting in elevation inaccuracy.
I have to factory reset the watch once a week !
Is there a way to recalibrate pressure sensor without factory reset the watch ?
  • https://support.garmin.com/en-US/?faq=FmE1UxYdn56rYdmThGbO06&searchType=noProduct

    In addition to that it sometimes recalibrates during activities with GPS at unpredictable moments.
  • Hi, with latests fw barometer readings ( and so elevation) are ok only after a factory reset, for a few of days, then slowly drifts, resulting in elevation inaccuracy.
    I have to factory reset the watch once a week !
    Is there a way to recalibrate pressure sensor without factory reset the watch ?


    Definitely, VA3 barometer is a piece of junk. As you said, it callibrates (sometimes), and after some wearing, elevation starts driftting.
    I have notice it get worst after taking shower.
  • Someone pointed out that the VA3 Music has baro-altimeter holes in different location than regular VA3.
    Anyone have one?
    I am curious if VA3 music has same issues.
  • There are many complaints about drifting on this forum, but you should be aware of the fact that drifting can be perfectly normal for a barometric altimeter. Simply because the atmospheric pressure changes with weatherconditions from day to day. Barometric altimeters need calibration on a daily basis or even more frequent.

    Also, this is not aviation grade equipment, so do not expect too much of its accuracy.

    I do not experience any significant drifting, and that may be due to the fact we are having steady weather conditions here for quite some time now.

    The real problem with the va3 is its calibration mechanism: very little control over when it calibrates, and it only calibrates against GPS altitude data which in many cases is highly inaccurate.
  • Today, I went for a short LOOP 3k walk at lunchtime, for about 30 minutes (Walk default activity on the watch).
    The weather didn't change during the 30 minutes walk, hence most of the discrepancies in the Elevation must be due to the VA3 elevation bug rather than any weather changes...

    I am on FW 5.2, GPS+Glonass and Smart recording.
    The elevation went crazy once again ... see below.

    I started and ended the walk at the exact same physical point (ie same elevation known as 54m on OS map).
    The watch started at 49m (ok!) and ended, on the same spot, at 77m (not OK)!
    in 30 minutes, I had a drift of + 28m !!! Sure this is not aviation grade altimeters but most competitor watches manage to have an accuracy of +/- 5m (not +/- 28m!).

    Also, I noted that the elevation was recorded in steps rather than a smooth climb... I walk with a clear sky above me (open street/path, no trees, no high rise buildings etc, most houses being bungalows (no floor) and well away from the walkway). In a few places (cliffs), the elevation jumps up and down... does each jump = a GPS Elevation sampling???
    In real, the climb is smooth and steady gradian. On the watch, I walked flat, then climbed a cliff, walked flat again, climb another cliff etc... (I had no climbing gear with me!!!).

    Also, note at about 34:26m the elevation jumped from 53m to 77m! Again, I don't recall climbing a cliff face of 24m. !
    (re-calibration / GPS Fix???)
    If this jump didn't happen, the watch would have had the endpoint at 53m (instead if 54m in real!)

    This has been going for quite some time now, I wonder if Garmin will ever fix that defect?

    I will escalate this again with Support (this watch is already a replacement watch received 2 weeks ago for the same issue (elevation going crazy!).
  • That's definately not ok! And indeed weather doesn't change in such extreme ways....
    If my theory is right those jumps would indeed be some sort of calibration attempts, and replacing the watch isn't going to help.

    I think Garmin should rethink this elevation thing competely:
    • No calibrations during an activity, That way at least climbs and descents will be recorded properly
    • Consider an option to always auto-calibrate to zero at the start of each recording. That way only climbs and descents are recorded, not the actual altitude. I think for many activities that will be sufficient.
    • Reinstate the possibility for manual calibration to a known height for those who care about the actual altitude.
    Let's just hope the engineers at Gamin will do something about this soon...

  • My gess is Garmin's engineers read very few this forum.
    Many users that report these issues to "official" support (isn't this the official Garmin's Forum??) receive the feedback that they are not aware of any elevation issue.
    And Brazilian official support is even worst: they have now knowlegde about the existence of this forum!
    So, I very disapointed with Garmin smartwatches. This is my first and my last of this brand. And when it definitely brokes, it goes directly to the trash. And all my friends ask me about this watch, I say only terrible things about it!
  • My gess is Garmin's engineers read very few this forum.

    I'm just a newbie, but indeed i haven't encountered any Garmin officials inhere sofar.... Maybe it helps just a little bit if we all report this to the support center.

    I happen to be software engineer, and i find myself investigating the possibility to create an app that does things better.... The watch is not perfect, but i like a lot of it. And it's altimeter can do a lot better then this.

  • Yes, this is generally a user-to-user forum, and you need to use the support contacts to report things to Garmin. But to be honest, drift over time is kind of a normal thing with baro sensors, and that's why re-calibration is needed It also involves things like the temperature from the internal sensor. From the doc on rawAmbientPressure and meanSeaLevelPressure:

    The raw ambient pressure in Pascals (Pa).

    This returns ambient (local) barometric pressure as measured by the internal pressure sensor. The data is the temperature compensated information read directly from the internal sensor.

    The mean sea level barometric pressure in Pascals (Pa).

    This returns barometric pressure calibrated to sea level. Since pressure varies dues to several factors, a GPS-based altitude must first be obtained, then the ambient (local) pressure is measured by the pressure sensor before conversion to a calibrated barometric pressure value.



    The GPS for MSL is what happens when GPS is first started in an activity an d a 3d GPS fix occurs (the wait 30 seconds thing)

    As far as doing an app, that would be with Connect IQ. But there, you can see what's going on, but not change things. There are already a number of apps in the store that give you info on pressure - rawAmbientPressure, ambientPressure, and meanSelLevelPressure, and with getPressureHistory, a graph of the MSL pressure over time.

    Here's a widget that gives you the Altiture and MSL pressure (with graphs) and a compass. myABC Here I'm just reporting what the firmware is saying.
  • Indeed it won't be possible to change the built-in altimeter, but from what i have seen sofar it is possible to create a datafield that calculates and records climbs and descends based on that rawAmbientPressure value. It could simply calibrate itself to zero at initialize, and that would at least provide an accurate climb-descend graph without the crazy jumps that many ppl (including myself) are experiencing. I do not believe those jumps are caused by the pressure sensor itself.

    It won't provide the actual altitude that way, but in many cases that's not important at all. I just want to see how much i have climbed and descended during my bike rides, the actual altitude above sealevel is not really relevant.

    I'll probably be playing around with this idea the coming weeks :)