Hi guys,
I'm hoping someone with more experience could take a quick look at the .bin file attached and provide some input on what is causing this quad to drift quickly off when it is placed in loiter mode.
This is running a modified version of 3.2.1 on a pixhawk with an external (mast mounted) 3DR gps/compass. The config is almost identical to several others I've built which have loiter great, so I'm baffled here.
Thanks
Brad
Replies
As an update to this problem (because I hate to read posts like this where the answer isn't posted), relocating all power wires to get them more than 5 inches away from the external compass has resolved the issue.
Looking at the logs more I see large changes in magx, magy and magz with throttle input so I suspect this is the source of the problem. Also the mission planner auto analysis tool gives this:
Test: Balance/Twist = GOOD -
Test: Brownout = GOOD -
Test: Compass = FAIL - Large change in mag_field (80.56%)
Max mag field length (630.25) > recommended (550.00)
Test: Dupe Log Data = GOOD -
Test: Empty = GOOD -
Test: Event/Failsafe = GOOD -
Test: GPS = WARN - Min satellites: 5, Max HDop: 3.23
Test: IMU Mismatch = GOOD - (Mismatch: 0.12, WARN: 0.75, FAIL: 1.50)
Test: Parameters = GOOD -
Test: PM = FAIL - 10 slow loop lines found, max 9.73% on line 18523
Test: Pitch/Roll = GOOD -
Test: Thrust = GOOD -
Test: VCC = GOOD -
Tomorrow I will reroute some wires away form the external compass and compass wiring, do a compasmot calibration and see if that improves things.
Can someone tell me which compass is the internal and which is the external in the dataflash log files? There is a MAG and MAG2 and I'm not sure which is which.
http://diydrones.com/forum/topic/listForTag?tag=loiter
http://www.google.pl/?gfe_rd=cr&ei=F9r1VezbHuGv8we0o6TYCA#q=loi...
Please read abovet looks like loiter mode software intelligence is affected badly.
I can review loiter mode algorithmics since I develop SelfEgo AI-based fuzzy logic control algorithmics and control systems.
What is obvious to the developer is not clear to the autopilot run by the drone.
In theory GPS receiver should be placed as close as possible to the center of the mass gravity. Otherwise, 3D geolocation error should be calculated real-time ( 3D geolocation of the center of the mass gravity is calculated).
Loiter mode looks to be a term coined by DIYdrones.
There is nothing what can be done since developers of autopilot are not part of the discussions.