Replies

  • There's two things that an airspeed gives you:

    1. it scales the control surface deflection based on real data, not based on a poor guess. This means that the roll and pitch controllers become actually tunable. At the moment, this is sort of irrelevant because the attitude solution needs work and the roll and pitch controllers are poor, but we're going to have better attitude and hopefully better controllers soon.

    2. it allows the plane to use pitch to control the airspeed, thus reacting correctly to a stall situation (it will automatically try to pitch down to recover airspeed)

    There are a couple of disadvantages:

    1. It adds a level of tuning, and at the moment we don't have a way to isolate the airspeed PID from the throttle PID and tune it (we badly need a FBW mode for this)

    2. If you set a speed that is faster than your plane can achieve, it will pitch down until it hits the pitch limit, attempting to achieve the speed right up until it hits the ground. Just make sure your max airspeed is set reasonably

    Mounting the pitot tube on a plane with a tractor prop is easy - put it on the wing. If your plane is foam (most are) you can just dig out a spot for the sensor and then glue the pitot on in front of it. It should be about 30cm away from the prop laterally.

    There's also a chance that we'll be able to ascertain true airspeed from GPS and attitude, soon, so there is also the potential to improve flight without an airspeed sensor very significantly.

  • It's the first third of the wing from the wing root that is where nothing is mounted to the wing that you dont want to be affected by the propeller, I think mounting it any closer than that will cause problems. (possibly make the pitot tube stick out in front of the propeller, if its too close it will still be affected, you could probably tell by just holding the aircraft and doing the runup if it is causing an error in the measurement) If you measure a safe distance from the wing root to use, then use this same measurement vertically like you said on the vertical stabilizer, or anywhere above or below. It is the "cone" from the propeller that gets bigger as you go aft that you are trying to avoid, it is the same distance vertically as it is horizontally, and back at the tail it could be larger or more than the "first third" is up towards where the wing is normally located. full scale aircraft have nothing mounted in this area of the wing that would be affected by airflow or pressures, like the pitot tube and stall warning.

  • Developer

    Hi Mike,

    ArduPlane flys very well without an airspeed sensor. In fact, tuning ArduPlane with an airspeed sensor enabled is often more difficult than without one.

    One common misconception is that some people assume that ArduPlane uses the GPS ground speed as an input to throttle control. It doesn't do that. The throttle control is based on the nav_pitch. The code adds more throttle when the plane is pitched up, and less throttle when pitched down. The nav_pitch is in turn based on the altitude error.

    The key parameter is to set TRIM_THROTTLE correctly for your airframe, setting it to a level that gives you a comfortable cruise speed in level flight. Work out what that level is while flying in FBWA mode, then set it in TRIM_THROTTLE and the plane should fly very well without an airspeed sensor.

    The only time when an airspeed sensor is really important is for automatic landing, when the plane will be flying close to the stall speed.

    Cheers, Tridge

    • Hi Tridge,

      I was wondering how does APM plane show me an airspeed (not groundspeed) if I don't have an airspeed sensor?

      I posted this in a new thread: http://diydrones.com/forum/topics/how-does-arduplane-calculate-airs... 

      My glider tends to stall around 25kph when loaded, or 30kph plus depending on the angle of bank (which is expected of course).

      I read your post at https://groups.google.com/forum/#!topic/drones-discuss/zdaEfa_EfBY and basically I would also love a mode where I can plug in my stall speed (I know what it is from test flights), APM Plane would calculate a predicted stall speed depending on the bank angle (that's known math) and then add a generous fudge factor to allow for airspeed errors which are as you point out more of an issue at lower speeds (percentage wise).

      I started a new thread for that:

      http://diydrones.com/forum/topics/suggestions-for-stall-avoidance-m...

      Best,

      Marc

This reply was deleted.

Activity