Help please and possible cause to catastrophic failure today!

[Gondeep11]

Hello all,

I am here looking for some help/advice regarding the complete loss of my week old Mavic. Here is a little backround about the issue as well as my experience. I am hoping
to gain a better understanding on what possibly went wrong, in addition from many of the posts I have read along with 2 experiences with tech support. I feel I'm going to be in for a long battle.

Working from home today I decided to take the Mavic up for a little lunch flight in-front of my home. All firmware was up to date to the latest 0800 revision (Including Batteries)
This was my 3rd of 4th flight with the mavic, however I'm 50 years old and an accomplished RC Pilot including 3D Helicopter proficency. I give you this info to try to weed out the rookie excuse type posts.

I followed all operational procedures, Controller Power first, Then Drone etc. Compass calibration was done yesterday (Same Location my Driveway)
I proceded to launch via the auto takeoff, Gps Signal green and good to go, connected with 20 Satalites. The Drone proceeds to lift off and keep gaining altitude, once it gets to
about 20ft I notice stick Movements are limited to yaw left and right and no response up/down. The drone proceeds to climb and I decide to excute RTH, ) Flight log Confirms all of this.
The mavic continues it's unresponsive flight to an altitude of 732ft ( Of course warnings are notifying me of the 400ft violation) Nothing I could do... Still shows connected via DGI Go App and Controller. Long story short the drone started to quickly descend at an ever increasing speed from 700+Feet, Long story short the mavic hit my Roof at better than 100mph and exploded into pieces. The flight log for review is at the bottom of Post

I have had the DJI Go app continually crash on my Google Phone / NVIDIA K1 Tablet as well as the new S8+, never in those 3 previous flights have I lost control or dead stick.
Never an major issue as I just re-launched the app and the drone was always in sight so no biggie.

Here is my concern aside from the potentially bad situation hitting a car, neighbor or myself. I don't have Dji Care since I purchased this 2 hours from my international trip last week to film vacation in Cabo.

It seems DJI does not have a good reputation from a Customer experience perspective as well as warranty, can someone shed some light of my flight log so you all can review and give me some insight. I sent it back today (Same Day as crash) and I hope they replace with a new Mavic, this one was exploded!

Any help greatly appreciated...

https://www.phantomhelp.com/LogViewer/JI2GGBA1EUW5F1VS0QJK/#

Regards,

Pete

 

[Egika]

First idea:
This looks like as if landing protection kicked in. The bird detected something that it mistook for the ground and kept trying to rise away from it (in your log VPS altitude stays low all the time!)
When you bring the full stick down flying close to the ground (or what the drone thinks is the ground) it goes into auto landing all the way down.

Did you have something mounted to the Mavic that could block the downward VPS sensors?

 

[Gondeep11]

Thank you for the reply, Nothing at all completely stock... Also Auto take off continued to climb and never stopped at the normal height of 5ft or so.... Just kept climbing with no ability to control?

I'm baffled even in the log you see forced landing and keeps raising altitude

 

[Gondeep11]

Also, I know you cant see stick movements, but on the video log it shows me trying to bring it down for a moment then I just let it go to see if it clears itself up... No avail

 

[Egika]

I only see the VPS altitude staying around 0.2-1ft while the bird actually rose up high.
This faulty reading is the reason for the Mavic going up. And that it goes into auto-land with landing protection and full stick down when close to the ground is common knowledge here in the forum.
Please double check if anything could have mislead the downward sensors!

 

[4wd]

It shows return to home initiated while at about 50 feet, RTH height is set in metres - it could have gone up that high if you assumed it was feet?
If you are flying early or late in the day and Mavic is pointing toward sun it has been known to see that as an obstacle it must ascend to avoid during RTH - this over-rides max altitude settings.
No idea why it eventually crashed down - record ends abruptly at 600 feet which suggests battery disconnect.
600 feet is inadvisabe but well within the Mavic's normal range.

 

[Gondeep11]

Thanks for your insight, it was about 11am if I recall, I believe the bird was pointed to the south.... So based on what you see (and I know you can't predict the future... Do you feel this will be covered under their standard warranty?

 

[Gondeep11]

I only see the VPS altitude staying around 0.2-1ft while the bird actually rose up high.
This faulty reading is the reason for the Mavic going up. And that it goes into auto-land with landing protection and full stick down when close to the ground is common knowledge here in the forum.
Please double check if anything could have mislead the downward sensors!

100 percent sure nothing could have mislead the downward sensors. In addition I wonder why it just fell from the sky? As well as lost the additional flight data?

 

[RaptureUK]

To all the experts out there - is there a way to force back control if it thinks it's auto landing like this ?

Would switching to sport mode and back help in theory ?

Thanks

 

[Egika]

Yes. Sport mode would end the auto landing.
I for myself have disabled the landing protection by default. It also does not add anything I would need anyway...

 

[AlanTheBeast]

Strange that with RTH set to 30 m, it climbs to over 200m.
One curious thing: Looking at the raw data, the VPS height barely changes as the bird goes up. This height is based on the ultrasonic sensors (I'm pretty sure).

I believe that the ultrasonic altimeter failed (or was blocked with dirt) - always telling the bird it was close to the ground - so RTH would follow the special rules. In the graph, right scale is VPS, max height VPS is 1.6 m when the bird was at 98 m height. (below). Normally, when above the height where the sonar can detect, it would read 0. But it almost always has a greater than 0 value - but very low.

Normally, VPS will work to about 10 metres (depending on the surface below) and then show "0" in the data. Will reacquire at about 8 - 10 metres (depending on the surface below).

When you initiated RTH, the distance from the home point was 6.4 metres.
The VPS altitude was 0.2 metres. See the manual extract below.

Would seem the bird used the "below 10 m" rule and began climbing, but never saw the ultrasonic based altitude get above 10 m. So it climbed and climbed ignoring the baro altimeter.

Is it possible that you had dirt in the sonar sensors?

At 58 s, goes into Auto Landing mode and a few seconds later into Forced Landing. At these times, the VPS altitude is 0.2 m. The bird thinks it's close to the ground. Actual is 116 m. It turns heading to be the same as at takeoff (228°).

At 78 s, it goes back to GPS mode. No idea why.

107s, back to RTH. Again, VPS is 0.2 m, and distance from home is 13.7 m. So the same rule as below applies.
Climbs again.

121s: distance from home is 0.1 m (lateral); VPS height is 0.1 m. AutoLanding is initiated.

Baro alt falls, but Vps altitude flirts 0.1 - 0.2 - 0.0 etc.

141.6 s: Data ends. 177.3 m; VPS 0.1 m; descent rate 2.9 m/s.

Root cause? Possibly failed or obstructed ultrasound sensor. Possibly a bug.

Why the data stopped? Could have assumed it was on the ground and shut off the motors and data recording. Not sure.

Left scale: baro height (m); right scale VPS height (m).

Edits: various tidying up.

 

[AlanTheBeast]

BTW: You do know you were in class D airspace as well as less than 5 miles from an airport?

 

[Gondeep11]

alanthebeast

Thank you for taking the time to write such a detailed and informative reply. What you said makes perfect sense! The bird was brand new with only 3 previous short 15 min flights. The chances of dirt or an obstruction to any sensors is highly unlikely, this bird only took off and landed on clean pavement. I believe this was a hardware or software issue. I hope DJI does not try to pull any shenanigans with replacing the bird. In addition, if the do replace I hope it's with a brand new in the box replacement cause anything else will not be acceptable to me. The old bird was demolished and I will not take a referbushed unit for a product that was less than a week old... I hope all goes well with my warranty claim but from some of the posts on google I may be in for a battle... will see what happens!

 

[Gondeep11]

BTW: You do know you were in class D airspace as well as less than 5 miles from an airport?

I did know I was in the class D airspace, since I only planned on taking up the bird 50 ft or so, not sure what protocol should be, am I required to call the tower to inform them of the flight even though it was only supposed to be at an altitude less than 100ft? This is where I need to better educate myself....

 

[Kilrah]

100 percent sure nothing could have mislead the downward sensors.

Apart from them being defective of course, which seems to be the more likely explanation at this point.

 

[AlanTheBeast]

alanthebeast

Thank you for taking the time to write such a detailed and informative reply. What you said makes perfect sense! The bird was brand new with only 3 previous short 15 min flights. The chances of dirt or an obstruction to any sensors is highly unlikely, this bird only took off and landed on clean pavement. I believe this was a hardware or software issue. I hope DJI does not try to pull any shenanigans with replacing the bird. In addition, if the do replace I hope it's with a brand new in the box replacement cause anything else will not be acceptable to me. The old bird was demolished and I will not take a referbushed unit for a product that was less than a week old... I hope all goes well with my warranty claim but from some of the posts on google I may be in for a battle... will see what happens!

I'd say your chances are even or better if you can convince them that the VPS failed and confused the drone.

PRESERVE THE DATA. It shows that the VPS altitude did not behave properly. If you have the data from previous flights I'd like to see that too.

 

[Gondeep11]

I'd say your chances are even or better if you can convince them that the VPS failed and confused the drone.

PRESERVE THE DATA. It shows that the VPS altitude did not behave properly. If you have the data from previous flights I'd like to see that too.

I will have to export the files from my old phone (Just switched phones yesterday) I will look for it and post them here.... So appreciative for your knowledge and help....! Thanks a million!

 

[AlanTheBeast]

I did know I was in the class D airspace, since I only planned on taking up the bird 50 ft or so, not sure what protocol should be, am I required to call the tower to inform them of the flight even though it was only supposed to be at an altitude less than 100ft? This is where I need to better educate myself....

You needed to phone, describe the flight (time, location/area, height) and since it is controlled airspace get their permission.
Airspace Restrictions
Do read the FAA drone site too.

 

[Gondeep11]

I will have to export the files from my old phone (Just switched phones yesterday) I will look for it and post them here.... So appreciative for your knowledge and help....! Thanks a million!

Here are 2 more flights... Let me know your thoughts other than I was breaking the rules on Class D airspace
DJI Phantom & Mavic Flight Log Viewer - PhantomHelp.com
DJI Phantom & Mavic Flight Log Viewer - PhantomHelp.com

 

[AlanTheBeast]

As you can see here, the VPS alt goes to about 12 m, then to 0 (out of range) while the baro goes up. Normal behaviour.
But not sure why the data ends ... w/o landing.
DJIFlightRecord_2017-06-13_[12-49-14]