Another mysterious P3P crash to analyze

Joined
Jan 10, 2016
Messages
218
Reaction score
187
Location
Boston MetroWest
Recently I was flying my P3P on a routine local Litchi mission that I've flown dozens of times before without incident. This time, it didn't come back. Fortunately, I had recently attached a Trackimo 3G GPS tracker to the bird. Within minutes I recovered the drone, which was still powered up and recording video the whole time. Two props were damaged, but everything else just needed a damp cloth to remove debris.

Now I'd like to figure out what happened, so I can learn to avoid this in the future.

The video, shown below, doesn't clearly indicate what caused the craft to lose control and plummet into a forest. The video shows a short downward drop a fraction of a second before the drone began to spiral down. My guess is a bird strike from above and behind. One clue was seen in the end of the video, which I slowed down to show a large bird flying away. We do have lots of crows and hawks in the area, so that's not unreasonable.

I was able to download the .DAT file from the drone. I used DatCon to convert that to an Excel file. Both of these, and the video, are found here. The .DAT file itself is 78MB, whereas the Excel file is 'only' 13.5MB.

Can any gurus here interpret these files to see if something other than a bird strike downed my drone?

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.
 
  • Like
Reactions: Rudder Bug
Can any gurus here interpret these files to see if something other than a bird strike downed my drone?
Did you have signal back to the controller for the incident?
If so I'll have a look but it would be a lot faster and easier to do it from the .txt file from Litchi.
Go to DJI Flight Log Viewer - Phantom Help
Follow the instructions there to upload your flight record from your phone or tablet.
Come back and post a link to the report it provides
 
  • Like
Reactions: DAVPHINS
Did you have signal back to the controller for the incident?
If so I'll have a look but it would be a lot faster and easier to do it from the .txt file from Litchi.
Go to DJI Flight Log Viewer - Phantom Help
Follow the instructions there to upload your flight record from your phone or tablet.
Come back and post a link to the report it provides
No, the bird was out of range of the controller when it dropped. There's a LOT of trees in the neighborhood!
 
No, the bird was out of range of the controller when it dropped. There's a LOT of trees in the neighborhood!
When the data stops at line 10643, the Phantom appears to be in stable flight.
I've gone back 2000 lines in the data and can't see any sudden changes in the pitch, roll & yaw that might indicate a collision.
 
When the data stops at line 10643, the Phantom appears to be in stable flight.
I've gone back 2000 lines in the data and can't see any sudden changes in the pitch, roll & yaw that might indicate a collision.
I've just placed the next sequential DAT and Excel files (FLY380) in the same folder. This file begins with the drone on the ground. There is a 4-minute gap between the two files with no data. However, the camera was recording video continuously!
 
Agreed with @Meta4 as stated above. There is no data in FLY 379 that would indicate a collision or another issue that could have caused the incident. Fly 379 simply ends abruptly @ approx 70 meters in altitude. Fly 380 does not help at all. The aircraft is already on the ground. It doesn't make sense at all that the recorder simply stopped and restarted a new file while on the ground. Below simply shows the data end in FLY 379. The last point recorded is shown in the last image. Maybe @BudWalker can explain the recording end.

FLY 379.png



KML.png
 
  • Like
Reactions: Orvillewrong
Definitely looks like a bird strike. The AC is clearly well above the tree tops so it didn't hit anything, something hit it!
What's odd is how quickly it fell. Did the damage to the props happen in the air? Or from the fall? My guess is at least one prop broke in flight, hence the spiral fall. But the fact that the flight log shows no sudden attitude change is very odd. It almost looks like a CSC motor stop happened. Does anyone know what the flight log would (or wouldn't) record in the event of a CSC motor stop?
 
Agreed with @Meta4 as stated above. There is no data in FLY 379 that would indicate a collision or another issue that could have caused the incident. Fly 379 simply ends abruptly @ approx 70 meters in altitude. Fly 380 does not help at all. The aircraft is already on the ground. It doesn't make sense at all that the recorder simply stopped and restarted a new file while on the ground. Below simply shows the data end in FLY 379. The last point recorded is shown in the last image. Maybe @BudWalker can explain the recording end.

View attachment 105152


View attachment 105153
Don't have anything to add here. A CSC would be recorded in the .DAT
 
Ok here's another question, you can see in the video a sudden change in pitch, we know the max pitch is 30 degrees, if sonething was to cause to AC to pitch more than 30 degrees mid flight would the motors stop? Or would it try to recover? I understand the flight log is unrevealing which is odd in itself. But that aside, what else could cause the AC suddenly drop from what appears to be perfectly stable flight?
 
So powered up my P3P so the gimbal was funtioning to duplicate how the right front leg would appear in view and it appears the AC was struck on the left rear prop. Which props were broken?
 
Two props had small chips and had to be replaced. My guess is that the tumbling started when the drone hit a treetop. The motors were still running after the drone hit the ground, as the original video showed continued thrashing for several seconds.
 
Two props had small chips and had to be replaced. My guess is that the tumbling started when the drone hit a treetop. The motors were still running after the drone hit the ground, as the original video showed continued thrashing for several seconds.
Hmmm very interesting
 
So I just tested my P3P. Did a take off, had my son do a 2 hand grab from the rear on both legs and rotate the aircraft sideways to a 90 degree angle to see if the motors would shut off past 30 degrees, they did not shut off. It continued to try to recover ( and fought pretty hard to do so ). So unless you had a complete hardware malfunction, yes the motors (and props) should have continued to run. And you say all 4 props were still attached with only minor chips? If so then theres no way it should have gone down like that unless something was physically holding onto it and brought it down. The fact that the flight log shows nothing makes me suspect something failed on the main board in the AC.
 
The drone was found lying on the ground INVERTED. It hit the treetops and tumbled to the ground.
 
If so then theres no way it should have gone down like that unless something was physically holding onto it and brought it down. The fact that the flight log shows nothing makes me suspect something failed on the main board in the AC.
You are grasping at straws my friend. First, the motors will not shut off past 30 degrees, there would be no rhyme or reason for them to do so in the first place, by design. Second, there are numerous possibilities for the event occurrence, and without data to back that up, it is wasted breath and speculation. Granted, attempting to explain why the FR stopped and re-started another log file while on the ground is a bit of a mystery, however, that being said, since the FR did re-start another file, it is not likely that this was a main board failure simply due to that event. You basically have 3 primary possibilities. A bird strike being the most likely. The other 2 are a catostrophic motor failure or an ESC failure IMO. But even at that, it is still simple speculation. And as the OP mentioned the tumble was from the trees, pretty obviously.
 
You are grasping at straws my friend. First, the motors will not shut off past 30 degrees, there would be no rhyme or reason for them to do so in the first place, by design. Second, there are numerous possibilities for the event occurrence, and without data to back that up, it is wasted breath and speculation. Granted, attempting to explain why the FR stopped and re-started another log file while on the ground is a bit of a mystery, however, that being said, since the FR did re-start another file, it is not likely that this was a main board failure simply due to that event. You basically have 3 primary possibilities. A bird strike being the most likely. The other 2 are a catostrophic motor failure or an ESC failure IMO. But even at that, it is still simple speculation. And as the OP mentioned the tumble was from the trees, pretty obviously.
You are correct, i am grasping at straws. Its just these facts are intriguing as the P3 is generally very reliable and I would like to know what actually happened here. The reason i tested a motor stop is I've already tested that a motor start is not possible at that angle so i was curious if there was a correlation. It just seems odd that at 70 meters up it could simply fall like that (intact) and not recover on its own. Must have been one major bird strike!!!
 
@3rdof5 .....It doesn't take much..------------------> Bird Strike
Yes i get that. In that video we dont know what the condition of the AC was in after the strike. In this case all 4 props were intact and apparently running, albeit with minor chips, but functional enough to maintain flight. As long as the gps, imu, compass and barometer are functioning, you'd think the AC could recover, even after a solid strike. Weird.
 
Two props had small chips and had to be replaced. My guess is that the tumbling started when the drone hit a treetop. The motors were still running after the drone hit the ground, as the original video showed continued thrashing for several seconds.
How is the AC now? Changed the props and all is well?
 
In that video we dont know what the condition of the AC was in after the strike.
You missed the point. The point being that it doesn't take a large strike to take it down. Also the similarity of the two (visually speaking). Look at the two at the moment of impact. They are practically identical in reaction.
 

Members online

Forum statistics

Threads
143,066
Messages
1,467,354
Members
104,933
Latest member
mactechnic