Get more from your Mavic
The largest Mavic community in the world
Join Us Now

Compass error and GPS error while flying

mrwagdog

Well-Known Member
Joined
Apr 17, 2018
Messages
183
Reactions
114
Location
Northern VT, USA
Weird problem yesterday evening and this afternoon flying my MP:

Took off with GPS lock to film a sunset. I flew around for a few minutes. I tilted the gimbal down and then got a compass error, and the mp went into atti mode. It seemed to be flying away - I got very nervous, I may have been giving it forward stick. I was LOS, and was able to get control. GPS mode returned, and a returned after a bit and landed. Here's the log:

DJI Flight Log Viewer - PhantomHelp.com

After landing, I looked at the sensor calibration (IMU and compass) and both were very good. I do have the video of this flight, but I haven't viewed it.

This afternoon, wanting to test out the Mavic, I took off from the exact same spot. I was up in the air a short time and got an error reading something like "working encounters bias large". GO4 said something about IMU needs calibration across the top. I landed quickly, again looked at the sensors and both IMU and compass were good.

I took off and flew again, uneventfully. Here's the log:

DJI Flight Log Viewer - PhantomHelp.com

I'm taking off from my deck, which is the same spot I've flown all my flights from. These are the first time I've encountered any errors. My MP is about a month and a half old. I try to fly daily, sometimes twice. There is a grill, a metal patio table and metal chairs on the deck, but they are at least 8' from where the Mavic is launched, and they've never caused an issue before.

Any thoughts on what's going on?
 

El Dedos

Member
Joined
Feb 28, 2017
Messages
23
Reactions
11
Age
66
Strange I just got an error notice never seen before. Compass Redundancy switch.......don't know what
it means, I'll search around. The Mavic flew fine and it was in an area I always fly.. A metal building with a grill and
mower are 8-9 foot away. Flown 40 + flights here tho.
S. Lewis
 

Canuk

Well-Known Member
Joined
Apr 19, 2018
Messages
922
Reactions
843
Age
56
Strange I just got an error notice never seen before. Compass Redundancy switch.......don't know what
it means, I'll search around. The Mavic flew fine and it was in an area I always fly.. A metal building with a grill and
mower are 8-9 foot away. Flown 40 + flights here tho.
S. Lewis

When you see Compass Redundancy switch it means the drone is switching to its backup compass as the primary one has too much interference to use. The main one is in the rear of the drone. The secondary one is in the front half of the body
 

sar104

Dic mihi solum facta, domina.
Premium Pilot
Joined
May 27, 2017
Messages
13,545
Reactions
17,654
Location
Los Alamos, NM
This is a strange one. Obviously the aircraft switched to ATTI mode due to a large yaw error at around 280 s, but the reason for that is not clear at all. Comparing the actual aircraft location with the IMU-predicted location shows that things start off okay, and then go very wrong:

2018-05-30_[20-17-54]_01.png

After initially heading off WSW, with the GPS (red, 1 s markers) and IMU (green, 1 s markers) data in quite good agreement, at 280 s the IMU velocity data go off the rails and suggest that the aircraft is heading back, first slowly to the NE and then swinging round to the south, while the aircraft actually turns and heads north. For 13 s the aircraft is in ATTI mode and then the IMU velocity data are correct again, although the green IMU track is now offset from the red GPS track due to 13 s of completely incorrect IMU velocities.

Looking more closely at the initial track the period of incorrect IMU data is clearer:

2018-05-30_[20-17-54]_02.png

Looking at the real vs. IMU velocities shows that there is good agreement, as expected, except during the 13 seconds or so when it went badly wrong and switched to ATTI:

2018-05-30_[20-17-54]_03.png

This doesn't look like the typical magnetic distortion problem because the initial track was correct, and it was only incorrect for the brief period of ATTI flight. I think we need to see the mobile device DAT file with the raw magnetometer data to figure out what really happened.
 

fwacone

Member
Joined
Dec 2, 2017
Messages
20
Reactions
3
Age
51
What you describe with your first experience is what I am having on every flight with my drone. Was their any solution?
 

mrwagdog

Well-Known Member
Joined
Apr 17, 2018
Messages
183
Reactions
114
Location
Northern VT, USA
What you describe with your first experience is what I am having on every flight with my drone. Was their any solution?
I moved my takeoff location to out further in my backyard, away from the deck. Haven't had a problem since then.
 

mrwagdog

Well-Known Member
Joined
Apr 17, 2018
Messages
183
Reactions
114
Location
Northern VT, USA
This is a strange one. Obviously the aircraft switched to ATTI mode due to a large yaw error at around 280 s, but the reason for that is not clear at all. Comparing the actual aircraft location with the IMU-predicted location shows that things start off okay, and then go very wrong:

View attachment 39257

After initially heading off WSW, with the GPS (red, 1 s markers) and IMU (green, 1 s markers) data in quite good agreement, at 280 s the IMU velocity data go off the rails and suggest that the aircraft is heading back, first slowly to the NE and then swinging round to the south, while the aircraft actually turns and heads north. For 13 s the aircraft is in ATTI mode and then the IMU velocity data are correct again, although the green IMU track is now offset from the red GPS track due to 13 s of completely incorrect IMU velocities.

Looking more closely at the initial track the period of incorrect IMU data is clearer:

View attachment 39256

Looking at the real vs. IMU velocities shows that there is good agreement, as expected, except during the 13 seconds or so when it went badly wrong and switched to ATTI:

View attachment 39258

This doesn't look like the typical magnetic distortion problem because the initial track was correct, and it was only incorrect for the brief period of ATTI flight. I think we need to see the mobile device DAT file with the raw magnetometer data to figure out what really happened.
I apologize, but for some reason I'm just now seeing this.

As I mentioned in the reply above, I haven't had an error since.

When I get a chance, I would like to find the log files you mentioned to see if this is something I should be concerned with.

Thank you so much for the reply.