P4 started to fly off on its own

Joined
Jun 28, 2016
Messages
203
Reaction score
65
Age
66
Location
Alexandria, Virginia
Website
fishcrow.com
My P4 did fine on the last flight a few months ago. I started a flight yesterday. Instead of hovering in place, it was drifting all over the place. No wind, it had plenty of satellites (the controller was green and GPS), and the controller was set to P mode. Then it suddenly started to climb on its own. I felt like it was resisting my efforts to bring it down. I managed to bring it in for a hard landing on a driveway (no apparent damage). Is there something that can be reset? Do I dare to try another flight? Does it need to be sent in to be checked out and possibly repaired?
 
My P4 did fine on the last flight a few months ago. I started a flight yesterday. Instead of hovering in place, it was drifting all over the place. No wind, it had plenty of satellites (the controller was green and GPS), and the controller was set to P mode. Then it suddenly started to climb on its own. I felt like it was resisting my efforts to bring it down. I managed to bring it in for a hard landing on a driveway (no apparent damage). Is there something that can be reset? Do I dare to try another flight? Does it need to be sent in to be checked out and possibly repaired?
The obvious thing would be to look into what caused the incident.
Post your .txt file and we'll see if it sheds any light on it.
 
Thanks a lot! I was using Litchi (not a programmed flight). Where should the txt file be located? Sorry this is new to me. I did a search here for txt files. Found lots of posts, but didn't find where this file is located.
 
Thanks a lot! I was using Litchi (not a programmed flight). Where should the txt file be located? Sorry this is new to me. I did a search here for txt files. Found lots of posts, but didn't find where this file is located.
Litchi logs aren't as easy to read so they usually take a bit longer to interpret.
For Litchi, the file required is a CSV file.
Instructions for finding it are here:
 
Thanks so much for the help! It turns out that I had done this before years ago when using the DJI app. There are two CSV files from yesterday. I tried to upload them here, but I wasn't able to select them when clicking the attach files button below. Do I need to do something with them first?
 
Thanks so much for the help! It turns out that I had done this before years ago when using the DJI app. There are two CSV files from yesterday. I tried to upload them here, but I wasn't able to select them when clicking the attach files button below. Do I need to do something with them first?
If you upload them to Phantomhelp, you can post the link to the Phantomhelp summary file for each.
Or you could post them to dropbox or similar and post a link.

One more thing, it might help to describe the launch spot.
 
Any chance you have a gimble guard installed underneath? Often the cause of a climb without input, Blocked sensors.
 
I would suggest a Sticks calibration but if it’s that far off it should be beeping like crazy. Hope answer in the logs.
 
As can be seen in the images at the links that I provided, it was in the front yard of a residence. No metal objects nearby. I have done hundreds of flights without ever having anything like these two flights.
I'm working on it but Litchi logs are horrible to try to interpret and there are two of them.
What was the launch surface and did you power on the drone at the launch spot?
 
It was drifting all over the place during the first flight.
The speed is very slow (which makes it even harder to sort out in the data) but since the drone dropped to atti mode despite having plenty of sats locked, and the course it took it looks like a small yaw error issue.
You were lucky with this one.
Yaw errors often end up badly.
 
It was initially behaving the same way during the second flight, but then it started to climb on its own. The log says there were no satellites, but the controller was green/GPS.
The second flight was another yaw error situation (possibly caused by the same launch point ?)
Again because of the slow speed and short flight, it's hard to see much.
The climb is only a metre or so with no corresponding joystick input.
Maybe also related to the yaw error - I can't be sure.
 
The launch surface was the front yard of a house where I have done several previous flights without incident.

A possible clue: When I opened Litchi, there was a delay while something updated. I didn't ask it to update Litchi, but that is what seemed to happen. Do I dare try another flight? Maybe with the DJI app? If so, I will get out in the middle of a big field.

Yet another possible clue: Two months ago, I miscalculated in planning a Litchi flight in an area with variable terrain. The P4 crashed into some vegetation. It was fortunate that a clump of small branches made a soft catch, and there didn't seem to be any damage (other than the tip of a prop broken off). I did several flights after that incident with no problems. So it doesn't seem to be related to the problem that occurred yesterday.
 
The launch surface was the front yard of a house where I have done several previous flights without incident.

A possible clue: When I opened Litchi, there was a delay while something updated. I didn't ask it to update Litchi, but that is what seemed to happen. Do I dare try another flight? Maybe with the DJI app? If so, I will get out in the middle of a big field.
Updating Litchi won't cause issues like that.
Yaw errors are caused by the gyro sensors being initialised incorrectly so they are not aligned to show proper compass directions.
It happens most often when the compass is deflected by ferrous material nearby during startup.

A simple test would be to setup the drone to fly and note the compass direction it's facing on the ground and compare with the compass direction the drone icon in the radar display or map window is facing.
They should be the same (eg east and east etc).
If there's a difference, that's a big clue that there's a compass problem.
If you do some more testing, it would be much, much easier to analyse flight data if you use DJI Go 4.
 
We finally got some decent weather today. So I did another test. I set the P4 on the ground away from anything metal. As shown in the attached photo, I used a tether in case it decided to fly off again. This time, it behaved like my drones have behaved in hundreds of previous flights. No idea what caused the problem and hope it won't happen again. One valuable lesson came out of the incident. I sometimes launch the drone from a ship or boat. I realize this can be risky due to metal on deck, but it's something that I need to do at times. As a precaution in the future, I will attach a tether at startup when on the water. Unlike in the attached photo, I will rig the tether so that it can be released from the ground.

tether.jpg
 
We finally got some decent weather today. So I did another test. I set the P4 on the ground away from anything metal. As shown in the attached photo, I used a tether in case it decided to fly off again. This time, it behaved like my drones have behaved in hundreds of previous flights. No idea what caused the problem and hope it won't happen again. One valuable lesson came out of the incident. I sometimes launch the drone from a ship or boat. I realize this can be risky due to metal on deck, but it's something that I need to do at times. As a precaution in the future, I will attach a tether at startup when on the water. Unlike in the attached photo, I will rig the tether so that it can be released from the ground.

View attachment 120964
Well that's a good one,you need to get trust back,,,good idea to prevent flyaway but also risky ,,are you doing auto take off or manual, myself would check imu and compass is ok,yip stick calibration to make sure stick inputs are corect
 
Well that's a good one,you need to get trust back,,,good idea to prevent flyaway but also risky ,,are you doing auto take off or manual, myself would check imu and compass is ok,yip stick calibration to make sure stick inputs are corect
My main concern about a possible flyaway is to avoid damaging someone's property, causing an accident, etc. I Idid a manual start, but it's relatively strong line that wouldn't be likely to get caught up in the props. I intend to run all of the checks for sure. Thanks for the inputs!
 
My main concern about a possible flyaway is to avoid damaging someone's property, causing an accident, etc. I Idid a manual start, but it's relatively strong line that wouldn't be likely to get caught up in the props. I intend to run all of the checks for sure. Thanks for the inputs!
Now this flyaway thought is on all pilots minds and it happen to anyone,anytime,anyplace,,fortunately we can do checks and precautions to help avoid this but dont matter how much precautions we take it can happen for some dumb reason,,agree on property damage, I had p4 free fall onto my own house roof after battery error shut down in flight and smashed the concrete tiles on roof,again lucky it was mine because when happen I was trying to get back overhead ,I cant afford fix someone's stuff so I avoid flying around like I used to,,,I also use me drone for fishing and I carry braided line out and dont matter if strong string or line if something does occur the drone will tilt and tangle alright,,I been close and lucky at times,I just mention auto take off cause it will rise to height and stop so make enough line to allow for that,,as mention above try flight with go and if abnormal its easier to read what's happening in the data,,,do your checks,give it some tests,blow those cobwebs outside and best of flights
 

Recent Posts

Members online

Forum statistics

Threads
143,094
Messages
1,467,602
Members
104,980
Latest member
ozmtl