|Subject:||[Paparazzi-devel] ArduIMU update|
|Date:||Thu, 03 Mar 2011 20:59:14 +0100|
|User-agent:||Mozilla/5.0 (X11; U; Linux i686; en-US; rv:18.104.22.168) Gecko/20110223 Lightning/1.0b2 Thunderbird/3.1.8|
Hi,have committed some ArduIMU changes to https://github.com/martinmm/paparazzi for review.
It is based on ArduIMU v1.7 and fixes the main issue which was printing serial data while in I2C interrupt, also includes Steves fixes. Additional features are running at 60Hz, having the possibility to use the Paparazzi ROLL_RATE loop and transferring the gps data as it is received rather than at a constant, non synchronized rate. The v1.7 uses gps 3D speed for centrifugal correction whereas the v1.8.2 uses 2D speed. Premerlanis paper is a little vague on that but I would think that 3D should be more precise when climbing/descending. Maybe 2D is used as simple gps receivers don't have 3D speed?
An example airframe file is at mm/fixed-wing/fw_ins_arduimu.xml It controlled a Funjet in 55m circles climbing 6m/s nicely.
Next step is to move the calculation over to the LPC/STM using Christophes code. The Arduino will then only do AD converting and averaging (at a higher and more steady rate).
The Wiki says that the I2C queue length should be increased and the I2C clock rate reduced. Both does not make sense to me, sounds a little magic. What issues were solved with that?
|[Prev in Thread]||Current Thread||[Next in Thread]|