Shartzilla said:
Dadcat said:
The problem only exists in GPS Mode, Course Lock and Home Lock where the bird is trying to fly a GPS course. It doesn't hook or anything in ATTI so no point to testing it. 7
Course lock does not use GPS at all. It relies solely on the compass to determine orientation.
Page 20 of the Naza Manual.
http://download.dji-innovations.com/dow ... .26_en.pdf
Agreed, however what we are seeing boils down to a difference in the location of North between the compass and the GPS.
If you have ever done a map reading course or bush walked before the days of the GPS system, think back to what you would do.
1 Obtain map of the area.
2 Find GR of your start location on map.
3 Find GR of your destination on map.
4 Calculate distance and bearing from start to destination.
5 Subtract ( in Australia ) magnetic declination using date of map and declination data on map to get magnetic ( compass ) bearing
6 Set prismatic compass to bearing.
7 Sight object in distance on bearing.
8 Walk to object, counting paces.
9 At destination?
10 No, goto 7
11 rest, have a cuppa.
The phantom firmware in course lock will be flying along the compass bearing ( angle between magnetic north and the direction the front is facing ) however as it's learning the local magnetic declination it's north position will be changing - it's comparing the compass bearing and the GPS bearing and adjusting the compass bearing so both agree, effectively making the compass point to true north.
Because it has locked in a bearing before it has calculated the local declination it will fly off course until course lock is reset after the local declination is learnt.