Hello everyone,

I purchased myself a Mavic Pro a couple days ago.  Everything about this drone has impressed me apart from one thing.  During flights at higher altitudes, the mavic begins to drift (enough so that i have to make a controller input to keep it on course), and the aircraft yaws very slightly.  This issue only occurs when flying in a particular direction and at higher altitudes.  When hovering, even at high altitudes, it is rock solid.  

I have gone thorough all calibration processes, (controller, compass, IMU), yet the issue still seems to presist.  I have considered that fact that this behavior might be caused by wind, although I would think that the GPS would keep it steadier than it is.  My previous Dji product, the phantom 3 pro, never had a single problem like this.


Any help would be greatly appreciated,

Thanks!

You sound like an experience pilot - therefore don't take the below personally - just pointing out the obvious just in case:

- did you calibrate the compass for sure without any outside input? that is - pockets clear, no big metal buckle on the belt, phone in airplane mode and far away from the Mavic (once you start), no suspected metal in the ground (like sprinklers etc)?

- did you try to rotate the propellers? that is - place them on the opposite arm compared to before? not a very high probability but still could be a funky propeller there.

Hope this helps a bit or if it doesn't you'll solve it soon enough. Happy flying. Cheers!

You could use the HD Sync app and then analyze the data to see what the problem is.  It should helps explain the drift problem.

I like the rotate prop idea by rnrnrn.  However, IMHO the likely cause is winds aloft or compass error.  Depending on the amount of drift, you may be able to notice a difference, on the map display, between the direction the AC is actually moving  and the direction of the AC arrow - while drifting.  If the drone is drifting (while moving forward) in the direction of the Arrow,  it's likely a compass problem.

Griffith Posted at 2017-6-1 06:14
I like the rotate prop idea by rnrnrn.  However, IMHO the likely cause is winds aloft or compass error.  Depending on the amount of drift, you may be able to notice a difference, on the map display, between the direction the AC is actually moving  and the direction of the AC arrow - while drifting.  If the drone is drifting (while moving forward) in the direction of the Arrow,  it's likely a compass problem.

When viewing the map, the arrow is in fact facing the correct direction of the aircraft, but you can clearly see it drifiting.  

The issue also happens when in RTH.  It will fly to and land in the correct spot in which it took off, but it makes some very strange maneuvers to get to the location.  Almost as if it tries to fly sideways.  Another interesting thing i have noticed is that it doesnt behave like this in every flight, which is why I want to disregard the idea that the props are the problem.

rnrnrn Posted at 2017-6-1 00:27
You sound like an experience pilot - therefore don't take the below personally - just pointing out the obvious just in case:

- did you calibrate the compass for sure without any outside input? that is - pockets clear, no big metal buckle on the belt, phone in airplane mode and far away from the Mavic (once you start), no suspected metal in the ground (like sprinklers etc)?

I dont believe the calibration is the issue.  I have calibrated this aircraft dozens of times, away from any objects that could interfere with it, and getting the same result in flight.

I have yet to rotate the props, and definitely will when I get the chance, but I do not believe this is the issue.  The aircraft does not behave like this every flight.   

Thanks for the tips!

Have you eliminated winds-aloft as a possibility?

DefconX3 Posted at 2017-6-1 07:29
When viewing the map, the arrow is in fact facing the correct direction of the aircraft, but you can clearly see it drifiting.  

The issue also happens when in RTH.  It will fly to and land in the correct spot in which it took off, but it makes some very strange maneuvers to get to the location.  Almost as if it tries to fly sideways.  Another interesting thing i have noticed is that it doesnt behave like this in every flight, which is why I want to disregard the idea that the props are the problem.

I think Griffith might be right I would check compass IMU can also cause drift but it wouldn't be intermittent you would see it constantly. Unless it's the wind.

You can do IMU test.
Raise AC to 8 feet
Hover for twenty seconds
Fly fwd 2 feet
Fly back 2 feet
Left 2 feet
Right 2 feet
Up 2 feet
Down 2 feet
Yaw left Yaw right


Each time returning to hovering point, if you Aircraft returns to good horizontal each time you have a good IMU
This is also good exercise to check compass and that you have good gps, do this before every flight and you will have more confidence in your set up.
Good luck

hallmark007 Posted at 2017-6-1 09:17
I think Griffith might be right I would check compass IMU can also cause drift but it wouldn't be intermittent you would see it constantly. Unless it's the wind.

You can do IMU test.

Went through the IMU checking process you described, everything seemed flawless.  Drone was rock solid, without a sign of issue.  

However, I looked through some of my flight data and noticed something off.  During a RTH, the path recorded was perfectly straight, even though the drone on video looked like it was drifiting.  I found out today that my drone, rather than drifting, is actually flying at an angle.  In other words, the angle in which the drone is flying is not in line with where the camera is facing.  This gave the illusion that it was drifting off course.  

This finding leaves me to suspect that wind is at fault for this behavior.  From what I'm assuming, the drone is being forced to yaw.  The drone compensates for this by adjusting the motors RPM's which keeps it flying straight, but not inline with where the camera is facing.  But if my speculations are true, why doesn't the gimbal react to this and move inline with the drones direction?

Another slight issue that I noticed.  When making maneuvers such as turning a full 360 degrees, it almost seems like the camera is temporarily disoriented.  The camera began to very slowly pan to the right for about 30 seconds, until finally stopping in the correct direction.     

DefconX3 Posted at 2017-6-1 14:50
Went through the IMU checking process you described, everything seemed flawless.  Drone was rock solid, without a sign of issue.  

However, I looked through some of my flight data and noticed something off.  During a RTH, the path recorded was perfectly straight, even though the drone on video looked like it was drifiting.  I found out today that my drone, rather than drifting, is actually flying at an angle.  In other words, the angle in which the drone is flying is not in line with where the camera is facing.  This gave the illusion that it was drifting off course.  

Well I'm not to sure have you tried gimbal calibration, camera and aircraft should be on the same heading, unless it's very windy AIRCRAFT shouldn't be veering.
360 should give you perfect circle might be glitchy in sports mode.

If you've done all the calibrations and this continues you might be forced to send for repair

hallmark007 Posted at 2017-6-1 15:14
Well I'm not to sure have you tried gimbal calibration, camera and aircraft should be on the same heading, unless it's very windy AIRCRAFT shouldn't be veering.
360 should give you perfect circle might be glitchy in sports mode.

I've been messing around with this issue for the last couple of days.  It does seem like the severity of the issue changes based on location.  Today may have been the worst have seen it.  After taking off and hovering, the gimbal began to turn to the right, even though the aircraft was not moving.  It got to about 45 degrees off the aircrafts forward position, then stopped.  I have tried calibrating the gimbal several times  to no avail.  

Im trying to rule out every possible problem before sending the drone back to DJI for repair.  The last thing I can think of is the possibility of magnetic interference.  Typically I feel that the gimbal behaves worse based on the surrounding interference.  The only part that stumps me is the fact that my Phantom 3 has behaved perfectly in all of the areas in which I flew my Mavic.  I feel that I would have had at least had some issues with the Phantom 3 if interference were to completely screw up the Mavic's camera system.

DefconX3 Posted at 2017-6-5 21:16
I've been messing around with this issue for the last couple of days.  It does seem like the severity of the issue changes based on location.  Today may have been the worst have seen it.  After taking off and hovering, the gimbal began to turn to the right, even though the aircraft was not moving.  It got to about 45 degrees off the aircrafts forward position, then stopped.  I have tried calibrating the gimbal several times  to no avail.  

Im trying to rule out every possible problem before sending the drone back to DJI for repair.  The last thing I can think of is the possibility of magnetic interference.  Typically I feel that the gimbal behaves worse based on the surrounding interference.  The only part that stumps me is the fact that my Phantom 3 has behaved perfectly in all of the areas in which I flew my Mavic.  I feel that I would have had at least had some issues with the Phantom 3 if interference were to completely screw up the Mavic's camera system.

Got some big subwoofers in the trunk with your Mavic?

Seriously, make sure when you transport it you are not driving around with it near a speaker in the car.

It really does sound like a compass calibration error. If it think North is a few degrees off it would exhibit this kind of behavior. Also local magnetic declination can have an effect. And that effect can change depending on where you fly.

Google magnetic declination and the name of your town and see if there is some mineral anomaly in the soil or something.

Ron in Utah Posted at 2017-6-5 21:46
Got some big subwoofers in the trunk with your Mavic?

Seriously, make sure when you transport it you are not driving around with it near a speaker in the car.

Heh..  No big subwoofers..  YET!

I want to blame it on Magnetic Interference all so badly.  However, as stated, my P3P never had an issue with this sort of thing.  

DefconX3 Posted at 2017-6-5 21:16
I've been messing around with this issue for the last couple of days.  It does seem like the severity of the issue changes based on location.  Today may have been the worst have seen it.  After taking off and hovering, the gimbal began to turn to the right, even though the aircraft was not moving.  It got to about 45 degrees off the aircrafts forward position, then stopped.  I have tried calibrating the gimbal several times  to no avail.  

Im trying to rule out every possible problem before sending the drone back to DJI for repair.  The last thing I can think of is the possibility of magnetic interference.  Typically I feel that the gimbal behaves worse based on the surrounding interference.  The only part that stumps me is the fact that my Phantom 3 has behaved perfectly in all of the areas in which I flew my Mavic.  I feel that I would have had at least had some issues with the Phantom 3 if interference were to completely screw up the Mavic's camera system.

As a last resort before you send in try a cold IMU calibration followed by gimbal calibration make sure you do it on a perfectly level surface,
Another thing you should check is your compass heading, you can do this when aircraft is started outside using small red arrow on the map in bottom left hand corner, make sure the point of the arrow and the heading of the aircraft are pointing in the same direction.

DefconX3 Posted at 2017-6-5 21:16
I've been messing around with this issue for the last couple of days.  It does seem like the severity of the issue changes based on location.  Today may have been the worst have seen it.  After taking off and hovering, the gimbal began to turn to the right, even though the aircraft was not moving.  It got to about 45 degrees off the aircrafts forward position, then stopped.  I have tried calibrating the gimbal several times  to no avail.  

Im trying to rule out every possible problem before sending the drone back to DJI for repair.  The last thing I can think of is the possibility of magnetic interference.  Typically I feel that the gimbal behaves worse based on the surrounding interference.  The only part that stumps me is the fact that my Phantom 3 has behaved perfectly in all of the areas in which I flew my Mavic.  I feel that I would have had at least had some issues with the Phantom 3 if interference were to completely screw up the Mavic's camera system.

The DAT file contains a lot of information that might be helpful in diagnosing this. Pick a flight where the problem was most pronounced, retrieve the DAT file using these instructions, then upload it to DropBox and provide a link to the file and I'll take a look at it.

DefconX3 Posted at 2017-6-1 14:50
Went through the IMU checking process you described, everything seemed flawless.  Drone was rock solid, without a sign of issue.  

However, I looked through some of my flight data and noticed something off.  During a RTH, the path recorded was perfectly straight, even though the drone on video looked like it was drifiting.  I found out today that my drone, rather than drifting, is actually flying at an angle.  In other words, the angle in which the drone is flying is not in line with where the camera is facing.  This gave the illusion that it was drifting off course.  

"But if my speculations are true, why doesn't the gimbal react to this and move inline with the drones direction?"


The gimbal does not move the camera in the direction you are wanting to go.  It should be facing forward of the drone, which is what you seem to be describing.  So far, this all sounds like winds aloft.  By the way, what altitude were you flying?

DefconX3 Posted at 2017-6-5 21:16
I've been messing around with this issue for the last couple of days.  It does seem like the severity of the issue changes based on location.  Today may have been the worst have seen it.  After taking off and hovering, the gimbal began to turn to the right, even though the aircraft was not moving.  It got to about 45 degrees off the aircrafts forward position, then stopped.  I have tried calibrating the gimbal several times  to no avail.  

Im trying to rule out every possible problem before sending the drone back to DJI for repair.  The last thing I can think of is the possibility of magnetic interference.  Typically I feel that the gimbal behaves worse based on the surrounding interference.  The only part that stumps me is the fact that my Phantom 3 has behaved perfectly in all of the areas in which I flew my Mavic.  I feel that I would have had at least had some issues with the Phantom 3 if interference were to completely screw up the Mavic's camera system.

" It got to about 45 degrees off the aircrafts forward position, then stopped.  I have tried calibrating the gimbal several times  to no avail.  "

45 degrees is beyond what it should be rotating.  This is sounding more like a repair trip.

AG0N-Gary Posted at 2017-6-6 10:51
"But if my speculations are true, why doesn't the gimbal react to this and move inline with the drones direction?"

Wind could have been a factor.  I was flying at 400 Ft, but I am now starting to see it react strangely just 10ft off the ground.  The camera turned about 45 degrees off the drones forward position when it was simply hovering.  

AG0N-Gary Posted at 2017-6-6 10:50
" It got to about 45 degrees off the aircrafts forward position, then stopped.  I have tried calibrating the gimbal several times  to no avail.  "

45 degrees is beyond what it should be rotating.  This is sounding more like a repair trip.

I re calibrated the gimbal once again.  This time, it actually seemed to work quite well.  The camera is no longer acting strangely, and is behaving as expected.  

On this test flight, it was a considerably windy day.  The drone hovered perfectly in the wind, but over the span of moving forward by around 100ft, the drone had drifted about 5-10 feet off course.  While i'm pretty sure this was caused by the wind, I want to make sure that my Mavic is behaving as it is supposed to.

Have you had similar behavior with your mavic when flying in higher wind velocities?