Bug with GPS elevation and sensor pressure

Hi there,

While testing a quick app I made to get familiar with the framework, I noticed a bug with data provided. I think this bug can be corrected easily at the next update.


I made a video to explain the below:
www.youtube.com/watch

* BUG 1 : No GPS elevation *

When getting Position.altitude (which should provide GPS elevation), I instead get altitude from the altimeter.

Example:
My actual altitude is 420m (and the ambient pressure is 962mb)
Barometer setting to read 420m is about 1018mb (this is the pressure at sea level)
I set the altimeter at 1000mb, I now read 330m on the altimeter (all this is normal).

The GPS reading also gives 330m instead of 420m! This is not normal

All sources currently provide the same data:
Position.altitude = from altimeter (this is not normal, it should instead be the GPS elevation)
Sensor.altitude = from altimeter (this is normal)
Activity.altitude = from altimeter (this should be according to user preference, either GPS or Altimeter)


* BUG 2 : Ambient pressure and Baro setting *

Sensor.pressure returns [altimeter setting + pressure variation since setting] instead of [ambient pressure].

In the example above, Sensor.pressure returns 1000.xx mb instead of 962 mb.

This is not a large issue because assuming that Sensor.pressure returns the barometer setting, one can calculate the actual ambient pressure. But it's a big calculation that's already been calculated elsewhere in the watch, so it would be best to just provide it to the app.

Best way to solve this problem:
1/ Make Sensor.pressure return the ambient pressure
2/ Add a data point "Sensor.barosetting" and provide here the barometer setting
3/ even better: allow Sensor.barosetting to be read/write so user can update it through the app (e.g. accept "std" setting with just one click instead of going in the menus and typing 1013).


Thanks for reading!

Garmin, is there any way these two bugs be corrected at the next update?
  • I understand you concerns, though this is getting beyond the scope of Connect IQ. It's now involving sensors, positioning, device firmware, and Garmin Connect. :) I'm not able to provide answers myself, but I will start contacting representatives from other teams to see what I can find out. I appreciate the effort you've put into this!
  • Former Member
    Former Member over 9 years ago
    Brandon,
    I think I just locked onto something about the GPS as I toy with Sport Tracks 3.1 settings and I posted about it in the THIRD post (by the end) of my Formatting F3 thread. Look for LATE FINDING in the third post.

    If you have any grip, any at all, with Garmin Engineers and Developers, I beg you to use it because we can't go on with the watch acting up this way.
  • Apologies for not getting back to you sooner on this. I contacted the device team in late August, and they've been discussing your issue in light of changes they have planned that should improve the accuracy issues you've been seeing. I wish I could provide more detail, but in all honesty I don't know any more than that.

    The wait-and-see approach probably isn't the kind of answer you'd prefer, and it's not generally the answer I like to give, but I trust that the device teams have things in hand. When it comes to sensors in particular (e.g. GPS, barometer, magnetometer, etc.) we have some really smart people that have been working with positioning and sensors for a long time. Not that they always get things perfect or don't make mistakes. :)

    It hasn't been part of Garmin culture to make use of the forums in the past (it's regarded primarily as a resource for customers to help other customers), but I've leveraged them heavily for Connect IQ. Feedback from folks like you has been invaluable! Thanks for the information you've provided on this issue.
  • Former Member
    Former Member over 9 years ago
    It's ok Brandon, thanks for coming back anyway.

    Last week I had a tip about a sub-patch in the works and a few days ago I had it confirmed. Meanwhile, the user Runswithbeers has been theorizing lack of WAAS signature in the GPS compartment. If this is true (and I think only the senior team could confirm), although it could be only a part of the problem, its configuration must be in a non-accessible part of the software or in a part that can't be tampered without the SDK.

    Could you please look into the possibility of building an IQ that shows the "constellation"?
    Even though I don't think this is the problem (and if it was, there have to be specific technical reasons behind the choice not to activate it at this time), a screen such as this would help us crossing out an eventual glitch in the WAAS.

    There's something fishy about thatgreen circle.
    I noticed something during yesterday's test: the presence of a single standing person looking at my F3 while I was waiting for the GPS lock turned the signal from GOOD to POOR and then RED in a loop.

    The GPS accuracy APP is much too simplified, whereas, the constellation might be helpful in these cases to narrow down to possibilities that have been discarded so far.