Commits on Jan 12, 2012
  1. Initial version of a quaternion-based AHRS and an ahrs demo.

    Non-functional, unbuildable, but useful calculations are there for use.
    committed Jan 12, 2012
Commits on Jul 30, 2011
  1. Adjust throttle when tilted.

    committed Jul 30, 2011
Commits on Jul 2, 2011
Commits on Jul 1, 2011
  1. Do the attitude compensation based on the same current state.

    This fixes a possible problem if the intgrated angle values overflow
    after compensation using the magentometer vector and before the
    accelerometer based compensation.
    committed Jul 1, 2011
Commits on Jun 29, 2011
  1. Fix the usage of current rotation rate.

    This makes the flight actually almost stable.
    committed Jun 29, 2011
  2. Add a simple protocol for enabling/disabling options in-flight.

    The heading-hold mode is untested at this point, will probably follow
    up with a commit or two to fix the signs and values.
    committed Jun 29, 2011
Commits on Jun 28, 2011
  1. Basic yaw rate reporting.

    committed Jun 28, 2011
Commits on Jun 27, 2011
  1. Handle the case when all 6 receiver channels are connected to arduino…

    … pins.
    This works much better because we can also read the gyro.sw. and the CH5
    potentiometer states and use them for something, and we can more
    accurately read the cyclic stick values, independently of the HOV.PIT.
    pot, of the Idle Up mode and of the throttle state.
    Unfortuntely it uses up an additional two cables and two input pins, one
    of them being the arduino LED output.  Or is it possible to do this with
    only every second channel connected? (so three pins only?)
    committed Jun 27, 2011
  2. Add trivial yaw control.

    committed Jun 27, 2011
  3. Better comment.

    committed Jun 27, 2011
  4. Add a really simple motor control loop.

    Do not use this for any real flying yet.  There's no feedback from
    motors taken into account at this point, there's no out-of-range
    detection and there's no yaw control.  The pitch and roll control is
    prone to getting stuck in endless oscillation.  The AHRS is not yet
    ideal either.
    committed Jun 27, 2011
Commits on Jun 26, 2011
  1. Remove one bit of resolution of time period in gyro integration.

    This should make overflows less likely, may need another bit of resolution
    less at one point.
    committed Jun 26, 2011
Commits on Jun 25, 2011
  1. Add a minimal README.

    committed Jun 25, 2011
  2. Add a minimal license note.

    committed Jun 25, 2011
  3. Break the trig maths and IMU sensor code out to separate files.

    Also average accelerometer readings over a 0.3s time for initial test.
    committed Jun 25, 2011
  4. Initial version of a simple integer-based AHRS.

    At this point do not rely on the heading, do not make quick movements
    (still suffers from some overflows) and optimally do not fly with it
    at all.
    committed Jun 25, 2011
Commits on Jun 20, 2011
  1. Include an I2C driver.

    committed Jun 20, 2011
  2. Add some likelies/unlikelies.

    committed Jun 20, 2011
Commits on Jun 19, 2011