Mavic RTH point moving without the Mavic even running.

[TowerGuy100]

I have noticed that the home point seems to move when I test the RTH, and the Mavic doesn't really hold it's heading while flying straight. So today I noticed that the home point is moving while the Mavic is just sitting on the ground without even the motors running.

Fresh IMU and compass calibrations have no affect and show normal. What else can I try?

Thanks for any ideas.

 

[Bunny Too]

I have noticed that the home point seems to move when I test the RTH, and the Mavic doesn't really hold it's heading while flying straight. So today I noticed that the home point is moving while the Mavic is just sitting on the ground without even the motors running.

Fresh IMU and compass calibrations have no affect and show normal. What else can I try?

Thanks for any ideas.

The GPS's in the Mavic and your device will show a little discrepancy, that happens with mine too. It has not affected my flying.

 

[TowerGuy100]

The GPS's in the Mavic and your device will show a little discrepancy, that happens with mine too. It has not affected my flying.

That certainly seems possible, but when I try to land with RTH, it misses the home point by many feet. That was my first hint something was up. If I let it sit longer, I have seen it go to 15' all while sitting there not moving.

 

[Bunny Too]

That certainly seems possible, but when I try to land with RTH, it misses the home point by many feet. That was my first hint something was up. If I let it sit longer, I have seen it go to 15' all while sitting there not moving.

Return to home has a variable of a couple feet (the error rate in the gps) Try "precision landing" and it should return within a few inches (mine does.)

 

[Qoncussion]

It wouldn't hurt to recalibrate your compass and IMU...

 

[TowerGuy100]

It wouldn't hurt to recalibrate your compass and IMU...

As I said in the original post, fresh calibrations have no affect.

 

[BudWalker]

.... So today I noticed that the home point is moving while the Mavic is just sitting on the ground without even the motors running.
....

This is normal if it's just a few (< 10) feet. It will set the HP just as soon as it has a few satellites and then reset it when it has enough for GPS mode.

 

[TowerGuy100]

This is normal if it's just a few (< 10) feet. It will set the HP just as soon as it has a few satellites and then reset it when enough for GPS mode.

This is with 16 satellites, and after sitting, unblown, for 20 minutes. It just keeps moving. I reset the home point, and then it shows that I am moving. Non-stop. This cannot possibly be normal.

 

[BudWalker]

This is with 16 satellites, and after sitting, unblown, for 20 minutes. It just keeps moving. I reset the home point, and then it shows that I am moving. Non-stop. This cannot possibly be normal.

You're right. That's not normal. Does the gpsHealth bar display on the controller show 4 or 5 bars?

If so, then can you retrieve the .DAT from the tablet (retrieving the .DAT from the Mavic is likely to be difficult or impossible). Look here to see how to do that.

 

[Qoncussion]

As I said in the original post, fresh calibrations have no affect.

Sorry, my bad.

 

[TowerGuy100]

You're right. That's not normal. Does the gpsHealth bar display on the controller show 4 or 5 bars?
View attachment 30408
If so, then can you retrieve the .DAT from the tablet (retrieving the .DAT from the Mavic is likely to be difficult or impossible). Look here to see how to do that.

I got the .DAT file. I believe this is the one that had the drifting while sitting.

 

[TowerGuy100]

I got the .DAT file. I believe this is the one that had the drifting while sitting.

Crap, I just uploaded it to the viewer and that isn't it. However, it was the most recent flight. The drifting was occurring while sitting on the ground after this flight and I didn't fly for there test. Maybe no flight record as i didn't fly?

 

[BudWalker]

I got the .DAT file. I believe this is the one that had the drifting while sitting.

That's the .txt not the .DAT we need. In fact, there won't be a .txt since there was no flight. Follow the instructions I referred to.

How to retrieve a V3.DAT from the tablet

 

[ErikDF]

If you take off from a landing pad or anny distinct pattern the probability of exact landing will increase. I have mad my own landing pad from a white photography reflector and black tape

 

[BossBob]

As a demonstration of how gps position wanders I suggest that Android users have a look at an app called GPS Averaging. Even with 20 satellites it take my phone 60 gps measurements to achieve an accuracy of +/- 2.0 m. You’ll only get the really accurate RTH using the visual positioning system.

 

[Cut.Aussie]

That’s well within the normal tolerance for civilian GPS accuracy

 

[TowerGuy100]

That's the .txt not the .DAT we need. In fact, there won't be a .txt since there was no flight. Follow the instructions I referred to.

How to retrieve a V3.DAT from the tablet

I am sorry, I didn't catch the last step. Here is the most recent DAT file. Thanks again for the help.

Oops, can't upload DAT files here. Here is a link to it: Dropbox - 2018-02-03_15-27-15_FLY045.DAT

 

[BudWalker]

I am sorry, I didn't catch the last step. Here is the most recent DAT file. Thanks again for the help.

Oops, can't upload DAT files here. Here is a link to it: Dropbox - 2018-02-03_15-27-15_FLY045.DAT

FLY045 isn't the .DAT you wanted since it contains 2 flights. Those flights started at 2018-2-3 20:27:15 GMT and look like this


The .txt started at 2018/02/03 20:42:30.001

Anyway, I looked some at FLY045 to see if it exhibited the behavior you described after power on and before launch. Specifically I looked at the eventLog stream to see how the home point was being set. Seemed to be the normal initial setting followed by an updated setting shortly after the motors started.

But, I'm not sure I understand what the behavior is that you're concerned about. The Mavic is turned on and allowed to sit on the ground for 20 mins? When you say the HP is moving does the Go App announce that's it been set/moved several times? Or, do you just see the location moving around in the Go App map display? If the latter and less than 10 feet then that's normal I think.

There were a couple of comments about GPS accuracy. The GPS data is made more accurate and faster by fusing it with IMU data. I suspect this process doesn't start until the AC is launched. That's why the AC will appear to move around on the ground some but can then hover with much smaller lateral excursions.

 

[TowerGuy100]

I will make a fresh one and try again. The app does not say anything. I don't normally let it sit, but I was just watching the distance indicator before a takeoff and noticed it moving around. In the video I posted, I reset the homepoint and you can see that it immediately indicates movement. I let it sit to see how far it would go. I have done this several times now and it has gone as far as 30'.

After a 4000' foot max distance flight yesterday, as I manually landed in the same spot I took off from, I noticed that the indicated height was negative 6' before I even hit the ground. I have only had this since Christmas, but I don't recall ever seeing that before either. Even hovering at 10' in no wind the Mavic climbs and descends a couple of feet and this change shows in the altitude display.

It just feels like the gps is not working 100% despite having 14+ satellites each time. Despite fresh IMU and compass calibrations, it also doesn't seem to fly perfectly in the direction the drone is pointing. It yaws slightly left and right during flight. I figure maybe all of this is related.

I used DJI Assistant app to calibrate the sensors just for the hell of it, with no change. I just want to get a feel for if this is normal or if I should contact DJI for support options. I don't have DJI coverage beyond normal warranty.

 

[Helo]

I have noticed that the home point seems to move when I test the RTH, and the Mavic doesn't really hold it's heading while flying straight. So today I noticed that the home point is moving while the Mavic is just sitting on the ground without even the motors running.

Fresh IMU and compass calibrations have no affect and show normal. What else can I try?

Thanks for any ideas.

Many people get hung up on the number of satellites the gps is receiving. In truth all you need is three to provide a location (four for altitude). Time, very accurate time (thus the need for atomic clocks) is needed to insure that when a measurement is taken from the three satellites it is at the exact same time. This creates a triangulation that can now calculate your location.

The fact that all of the GPS Satellites are in constant motion means that the triangle on which the accuracy depends is constantly changing.

The ideal configuration for the three satellites being used is 20° above the horizon and spaced 120° apart, although this rarely happens and if so only for a short period of time. Many hand held GPS receivers will show you the position of all the satellites the receiver is collecting data from and on a separate page it will show the strength of the satellites (it will also give the # of the individual satellites). Some units will give you estimated accuracy in feet.

This is where the movement you are seeing comes in. As the satellites orbit the earth the shape and size of that triangle is constantly changing and further from the ideal spacing mentioned above the less accurate the calculation.

This is a layman's explanation of GPS positioning and it much more complicated than this as I'm sure Alan and others will point out.

Bottom line if it bothers you don't zoom in so close and you won't even notice it. If it still bothers you get a new hobbie because all the calibration in the world on your Mavic (or any other GPS device) is not going to change it.