You must Sign-in or Register to post messages in the Hobby Squawk community
Registration is FREE and only takes a few moments

Register now

Announcement

Collapse
No announcement yet.

GPS vs Airspeed test on a T-33

Collapse
X
Collapse
First Prev Next Last
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • GPS vs Airspeed test on a T-33

    When I started logging telemetry data, I noticed that GPS was all over the place. Sometimes it seemed right and sometimes it was way off. GPS measures ground speed using only the horizontal plane so if you are diving straight down, GPS would read 0 even if you are going in at 100mph. Also, when inverted or at steep bank angles, readings were much lower than actual. Seems that GPS is only accurate straight and level.

    I did a comparison video of GPS sensor vs an actual air speed sensor that uses a pitot tube. Check out the video.



    Jose Soriano

  • #2
    Don't forget how inaccurate GPS really is. People seem to think somehow that GPS can determine your position at 1cm accuracy....far from that. What we see on the screen of our sat nav systems in a car is merely the correlation of the GPS data to a road map.
    For instance, when the GPS says to take an exit on any highway, and you continue straight on, you can observe how long it takes before the system notices that you're not on the exit. The system has this circle of uncertainty, which can go as large as a 15m radius. As long as this circle around your actual position is intersecting with the planned route, the sat nav system will assume that you are following the planned route.
    Ground speed is used in aviation to calculate ETA and the wind vector, it's pretty useless for other purposes.

    Comment


    • #3
      Originally posted by HangarQueen View Post
      Don't forget how inaccurate GPS really is. People seem to think somehow that GPS can determine your position at 1cm accuracy....far from that. What we see on the screen of our sat nav systems in a car is merely the correlation of the GPS data to a road map.
      For instance, when the GPS says to take an exit on any highway, and you continue straight on, you can observe how long it takes before the system notices that you're not on the exit. The system has this circle of uncertainty, which can go as large as a 15m radius. As long as this circle around your actual position is intersecting with the planned route, the sat nav system will assume that you are following the planned route.
      Ground speed is used in aviation to calculate ETA and the wind vector, it's pretty useless for other purposes.
      That is a function of the proprietary algorithm the car GPS uses to figure out if you are on the road, not a function of the GPS sensor's absolute accuracy. And its speed calculation is differential, so signal propagation errors will cancel over the short term. A better example is how well does the GPS speed correlate with the car's speedometer. Every car GPS I've ever had is within 2mph of the speedometer and spot-on when compared with radar. Ditto for the GPS in my (full size) plane when compared with ATC's radar.

      Comment


      • #4
        Originally posted by Amahoser View Post
        When I started logging telemetry data, I noticed that GPS was all over the place. Sometimes it seemed right and sometimes it was way off. GPS measures ground speed using only the horizontal plane so if you are diving straight down, GPS would read 0 even if you are going in at 100mph. Also, when inverted or at steep bank angles, readings were much lower than actual. Seems that GPS is only accurate straight and level.

        I did a comparison video of GPS sensor vs an actual air speed sensor that uses a pitot tube. Check out the video.



        Jose Soriano
        You might try relocating the GPS antenna.

        Comment


        • #5
          The car GPS algorithm show us its confidence in the accuracy with this "circle of confidence". It takes a given deviation from the planned route before the algorithm has enough confidence to display the correct position.
          Even in aviation they use tricks of the trade like differential GPS to counter the uncertainty of the absolute position data.
          GPS groundspeed calculation gets rid of most of that problem indeed by differentiating, that is true.
          But my point remains that ground speed is of relatively low importance in aviation, except for what I mentioned above.
          I only use air data sensors with pitot tubes for my models, not perfectly accurate either, but it has clear advantages over GPS sensors:
          - most importantly: they measure airspeed, which is the data that we actually need
          - cheaper than GPS sensors in most cases
          - consume less power than a GPS sensor
          I'm a fan of openXsensor on an Arduino board. In my gliders, I use it to make an airspeed compensated vario sensor.
          I have never been interested in ground speed in RC, really. Only for navigation purposes in 1:1 aircraft. Or in my car 🙂

          Comment

          Working...
          X