Unexplained Control Loss

[Hawgguy]

One minute after takeoff, hovering at around 150 feet deciding which way to go. Put into sport mode. As soon as I pushed the stick forward began pitching and spinning - possibly inverted. Crashed in the driveway 100 feet from where I took off. 3 props and gimble busted. Front leg dinged. Already started the refresh process.

Somehow I'd feel better if I were doing something stupid/reckless - at least there would be an explanation.

Wind was mild to moderate by west Texas standards. Perhaps a gust combined with full forward in sport mode? Really unsettling.

Already ordered a new Mavic. The refreshed one will go to my 17 year old son..... it was his drone anyway.

 

[Cyberpower678]

How about a flight log? Can you upload your?

 

[Hawgguy]

Here is the log from Litchi

 

[BD0G]

There have been multiple reports on the forum of a similar phenomena. This behavior has largely been attributed to noise being produced by one (perhaps 2) of the Mavics motors during Sport Mode. The High RPM's of the motor producing EMI that causes a compass shift from one to the other. Then when the compass data is compared between the two on board by the Mavic Pro it becomes "confused". This confusion leads to loss of control.


DIY rear compass cabling fix

Here is a thread that describes how a potential DIY solution to this issue was created. Since you are using a Refresh to have the Mavic replaced then this is not something you would need to do if the new replacement unit had DJI's new fix (twisting the cabling that leads to the rear motors) already performed.

 

[Cyberpower678]

There have been multiple reports on the forum of a similar phenomena. This behavior has largely been attributed to noise being produced by one (perhaps 2) of the Mavics motors during Sport Mode. The High RPM's of the motor producing EMI that causes a compass shift from one to the other. Then when the compass data is compared between the two on board by the Mavic Pro it becomes "confused". This confusion leads to loss of control.


DIY rear compass cabling fix

Here is a thread that describes how a potential DIY solution to this issue was created. Since you are using a Refresh to have the Mavic replaced then this is not something you would need to do if the new replacement unit had DJI's new fix (twisting the cabling that leads to the rear motors) already performed.

If you ask me, it looks like the drone was run into the ground by the pilot. The throttle was pulled back all the way until the drone hit the ground. I'll analyze more later. I'm currently occupied.

 

[Hawgguy]

I appreciatge any insights as I don't know how to analyze the log. I don't *THINK* it was pilot error because I was not really doing anything but starting to go forward shortly after takeoff, and I don't think I could have induced all of the pitching & rolling if I tried. It was a fairly violent maneuver. I was fully caffeinated, so I can't blame it on that. That being said, if it is PE, I'll own it, but I need to be convinced.

 

[Cyberpower678]

I appreciatge any insights as I don't know how to analyze the log. I don't *THINK* it was pilot error because I was not really doing anything but starting to go forward shortly after takeoff, and I don't think I could have induced all of the pitching & rolling if I tried. It was a fairly violent maneuver. I was fully caffeinated, so I can't blame it on that. That being said, if it is PE, I'll own it, but I need to be convinced.

I'm not accusing you of such yet, I just quickly glanced at the flight data. I need to do a thorough analysis before I call it.

 

[Hawgguy]

Here is the Healthy Drones link

Not too useful as the flight was so short - insufficient control data to analyze.

Looking at the compass turn rates - sustained 100-150 degrees per .1 s seems kinda high.

 

[Cyberpower678]

Here is the Healthy Drones link

Not too useful as the flight was so short - insufficient control data to analyze.

Your compass was highly erratic, and the power draw of the drone was above 20A. That is way too high. I'm guessing the compass readings were from the violent movements as you claim, but it's worth mentioning.

 

[Cyberpower678]

I looked at a lot of data here, and the drone lost control of itself. It moved fast, and moved around. Analyzing your stick movements, the movement of the drone was not consistent with the stick movement. It was highly erratic and was unable to obey the pilot's commands. Unless the pilot hit something when moving, I cannot explain the behavior of the drone, and this is something that should be covered by warranty.

 

[BD0G]

Battery had 39 charge cycles on it. So this Mavic has been flown a number of times without issue it appears. Has it been flown in Sport Mode flat out before without becoming erratic?

 

[Cyberpower678]

If you look at the RC stick controls as compared to the drone's behavior itself, they're inconsistent.

 

[Hawgguy]

Battery had 39 charge cycles on it. So this Mavic has been flown a number of times without issue it appears. Has it been flown in Sport Mode flat out before without becoming erratic?

Most definately.

 

[Hawgguy]

Message and data sent to DJI

 

[Hawgguy]

Blue line is Rudder, Green line is Yaw

 

[SkyRider]

Did a motor fail or a prop come off? Could that explain it?

 

[Hawgguy]

Did a motor fail or a prop come off? Could that explain it?

Flight data does not indicate any problems with the motors. All props were attached (3 broken) when recovered.

 

[Cyberpower678]

Flight data does not indicate any problems with the motors. All props were attached (3 broken) when recovered.

Indeed the motors were operating, albeit very aggressively. The current draw was extreme. Normal usage is around 8A to 10A, and the logs show that this thing spiked all the way up to 20A.

 

[Blades]

Here is the log from Litchi

What is the build date on it. The compass and motor wire issue was only in the first orders I believe when the were trying to get out the orders.

 

[Blades]

There have been multiple reports on the forum of a similar phenomena. This behavior has largely been attributed to noise being produced by one (perhaps 2) of the Mavics motors during Sport Mode. The High RPM's of the motor producing EMI that causes a compass shift from one to the other. Then when the compass data is compared between the two on board by the Mavic Pro it becomes "confused". This confusion leads to loss of control.


DIY rear compass cabling fix

Here is a thread that describes how a potential DIY solution to this issue was created. Since you are using a Refresh to have the Mavic replaced then this is not something you would need to do if the new replacement unit had DJI's new fix (twisting the cabling that leads to the rear motors) already performed.

This problem was only on the early production when they were trying to rush them all out the door I thought? Whats the build date on the MP?