Phantom 3 Professional - Crashed into Lake

[Tenly]

I lost my Phantom 3 Professional today in the strangest way.

I took off in P mode and was flying around over the lake at 30-40m up - following a friend who was riding a Sea Doo. Out of the blue, I received a GPS warning briefly. The warning went away before I could take any action, but I started hearing back - and the. I got the message again and flipped into "A" mode. About 10 seconds later, I got the connection lost message and had lost site of it. I assume it dove straight down into the water but I have no idea why. I was definitely switched over into "A" mode, and in a safe area out over the lake a good 5 or 10 seconds before I lost connection.

By the time I got on the Sea Doo and over to the last location that had been updated on the map, there was no sign of it. (Do these things float at all? Even briefly?)

Replaying my flight log, the very last entry on it shows an altitude of 20m and a vertical speed of -10m/s. I haven't been able to pull the log and upload to HealthyDrones yet since I'm not at home.

I need a root cause analysis of this thing. If I did something to cause this, that's fine. Is there any kind of defect that we might find that would make this DJI's fault and compel them to replace it?

If it's not my fault and if DJI won't fix it - I don't know what to do going forward. I'm not going to buy a replacement if this is just a scenario that can happen randomly without warning.... But if I clearly did something wrong and can learn from it, I'll order another one immediately.

I should be able to send the link to the HealthyDrones report before midnight. Until then - are there any guess as to what MIGHT have gone wrong???

 

[Procharged97]

**** that sucks man, sorry to hear that

 

[DroneDestination]

If it's not my fault and if DJI won't fix it

If your logs prove that it indeed wasn't pilot error, or something in the surroundings, (i.e. bird strike) DJI should replace it. Otherwise a MicroCenter $600-700 P3P looks pretty good from where you're standing......

 

[Reed L]

So it wasn't a fly away, it was a drop in the lake... Sorry to hear it, sounds like a motor failure at the least. Uploading the logs will tell it all

 

[Tenly]

Thanks....I see that somebody else has the exact same issue yesterday. They described the problem better than I did.

I didn't see any birds in the area - but how would they be able to tell from the logs?

When it first happened - I was suspicious that someone might have tagged me with a drone zapper... I've never seen one of those in use. What happens to the drone if targeted by one of them?

 

[Reed L]

Thanks....I see that somebody else has the exact same issue yesterday. They described the problem better than I did.

I didn't see any birds in the area - but how would they be able to tell from the logs?

When it first happened - I was suspicious that someone might have tagged me with a drone zapper... I've never seen one of those in use. What happens to the drone if targeted by one of them?

They can tell by the differences in the changes of rpm, speed and more. The log spells it out.

 

[Tenly]

Here's a link to the log...
HealthyDrones.com - Innovative flight data analysis that matters

I've opened a ticket with DJI Suport.

 

[DroneDestination]

Looks like the bird started RTH and never made it. I think it hit something, but what?

 

[Tenly]

Looks like the bird started RTH and never made it. I think it hit something, but what?

There was nothing around for it to hit - unless it was a bird that was too small for me to see - but if that was the case, I would have expected to see a dead bird floating in the water near where I lost contact. I searched the area on a Sea Doo for a good 30-45minutes looking for any kind of debris and couldn't find anything. There was a light chop on the lake so I was unable to see down into the water even a little bit.

I also called this a flyaway instead of a water crash because I didn't see it hit the water and the DJI Go logs don't show it hitting 0 altitude before they stop. I assume that the Phantom ended up underwater but I can't prove it - nor can I pinpoint the exact location that it entered the water. Looking at the last trajectory reported, even if it stabilized and didn't dive bomb into the water, it was heading out into open water and would have eventually run out of juice before hitting land.

In any case - a flyaway seemed more accurate since it definitely took on a life of its own and flew "away" from where I expected it to hover.

In the logs, it also looks like it tried to activate the Vision Positioning System at about 40m altitude. I assume that would be when I flipped the switch to A

The RTH has kind of a last ditch panic effort to make something happen after all else had failed. I would say that it can be ignored as any part of root cause analysis since the problem was already in full swing by the time I hit the button.

I guess what I'm trying to determine is why things went bad initially. Bad GPS sensor/Compass? Prop spin off? Motor failure? Bird strike? Targeted by a drone zapper?

I'm also confused by what happened (or didn't happen) when i toggled to A mode. I expected that to take the wonky GPS completely out of play and cause the drone to just hover at a constant altitude, drifting with the wind - but that's not what happened. It continued to fly itself.

I don't know if it's also worth calling out that DJI Go reported "Safe To Fly (GPS)" as I took off with 11 Satellites locked. The HealthyDrones report shows GPS is in red at that time. Also - when I received the "GPS Signal Error", the flight record playback still shows that I was connected to 12 satellites...

I don't know. Still waiting to hear back from DJI support but honestly, based on my past interactions with them - I'm not expecting them to offer much in the way of help. I'll definitely update this thread with their response.

 

[BudWalker]

Here's a link to the log.....

Can you upload your .txt to https://www.phantomhelp.com/LogViewer/Upload/ The analysis provided there is more informative than HealthyDrones, especially for this kind of incident. That site will also provide a link that you can post here that will others will have access to the analysis and the log.

 

[Tenly]

Can you upload your .txt to https://www.phantomhelp.com/LogViewer/Upload/ The analysis provided there is more informative than HealthyDrones, especially for this kind of incident. That site will also provide a link that you can post here that will others will have access to the analysis and the log.

Sure. As requested....
Phantom Log Viewer - PhantomHelp.com

 

[henick]

Interesting one. It definitely tried to RTH after a GPS error but then had compass and a grab bag of other problems. Interesting to find out wth happened...

 

[BudWalker]

Sure. As requested....
Phantom Log Viewer - PhantomHelp.com

I had assumed that HD wouldn't be adequate. In this case it does tell the story, although indirectly. Very likely your P3 suffered a broken or lost prop. In HD if you go to the Sensors page and select the Compass tab you'll see that HD is saying you had multiple compass errors at the end of the flight. HD arrives at this conclusion by assuming that a high rotation rate can't possibly be happening and it must be a faulty compass incorrectly saying the AC is rotating at a high rate. But the AC is spinning due to the rotational torque caused by the loss of one prop.

 

[Tenly]

Interesting one. It definitely tried to RTH after a GPS error but then had compass and a grab bag of other problems. Interesting to find out wth happened...

Yeah. It also looks like it recovered from the initial GPS error and stabilized for a few seconds before encountering the error again.

Based on what's in the log, what are your guesses as to the level of help that DJI will be willing to provide?

 

[Tenly]

I had assumed that HD wouldn't be adequate. In this case it does tell the story, although indirectly. Very likely your P3 suffered a broken or lost prop. In HD if you go to the Sensors page and select the Compass tab you'll see that HD is saying you had multiple compass errors at the end of the flight. HD arrives at this conclusion by assuming that a high rotation rate can't possibly be happening and it must be a faulty compass incorrectly saying the AC is rotating at a high rate. But the AC is spinning due to the rotational torque caused by the loss of one prop.

Thanks for taking the time to have a look. Is your determination 100% proven out by the information in the logs, or is it an educated guess? Are you 90% certain that's what happened or like 75%?

It's very unlikely that the props would have spun off. Making sure they are tight is part of my pre flight routine - however a breakage in flight is possible I suppose? I could have hit a big June bug or something. Would that typically be a hard enough collision to snap a prop?

So - moving on to confirming your diagnosis - if I were able to recover the Phantom, we'd see if it had a missing or broken propeller - but it's underneath somewhere between 4-14 feet of water. Does anyone have any experience searching for one of these underwater? Would it show up on sonar or a fish finder if I were able to borrow one?

 

[BudWalker]

Thanks for taking the time to have a look. Is your determination 100% proven out by the information in the logs, or is it an educated guess? Are you 90% certain that's what happened or like 75%?

It's very unlikely that the props would have spun off. Making sure they are tight is part of my pre flight routine - however a breakage in flight is possible I suppose? I could have hit a big June bug or something. Would that typically be a hard enough collision to snap a prop?

So - moving on to confirming your diagnosis - if I were able to recover the Phantom, we'd see if it had a missing or broken propeller - but it's underneath somewhere between 4-14 feet of water. Does anyone have any experience searching for one of these underwater? Would it show up on sonar or a fish finder if I were able to borrow one?

Unless the AC is recovered it's an educated guess. I really should know better but I'm willing to go 95% certain. Every time we've looked at a broken/lost prop incident via the .DAT the AC always spins. Every time we've also looked at HD it's shown a compass error coincident with the prop issue.The fact that it's the last thing that happens while it's falling is especially compelling.

 

[BudWalker]

@Tenly I'm going to amend my assessment and say instead that there was a propulsion issue with one of the motors. This is quite often a broken/lost prop. But, it can also be an ESC issue with a motor.

 

[Tenly]

@Tenly I'm going to amend my assessment and say instead that there was a propulsion issue with one of the motors. This is quite often a broken/lost prop. But, it can also be an ESC issue with a motor.

LOL! I'm not going to test you on it or use it against you if you're wrong. Like I said - I appreciate you taking the time to look and was just wondering what your degree of certainty was...

Since then I've actually found that I have the video from the flight in my DJI Go "videoCache" folder. Looking at the video, the first anomaly that occurs is a steep dip and turn to the right, followed by an equally steep dip and turn to the left. In the very last captured frame, it looks like the prop is still attached but not spinning. It's hard to tell if it's just a high shutter speed that stopped the propeller in the frame - or if it really was stopped. Earlier in the video, there is definitely visible motion blur. In any case - here are the frame grabs. Let me know if these provide any new or different insights.

This one is from the instant before the problem occurred...

This shows the pitch of the aircraft at the original onset of the problem...

This is one from somewhere between the initial problem onset and the end of the broadcast...

This one is the final captured frame. Not sure why I didn't get a frame or two after, or as it was entering the water - but well - this is all I got...,

 

[bobomet]

I too think a prop flew off or broke. I can't see the A/C spinning out in the way it appears to have for any other reason. Did you include in your preflight a "tight check", making sure your props were given a snug tightening turn? Did you look your props over for signs of cracks or wear?
If it didn't hit a bird, then a prop failed.

 

[WetDog]

Thanks for taking the time to have a look. Is your determination 100% proven out by the information in the logs, or is it an educated guess? Are you 90% certain that's what happened or like 75%?

It's very unlikely that the props would have spun off. Making sure they are tight is part of my pre flight routine - however a breakage in flight is possible I suppose? I could have hit a big June bug or something. Would that typically be a hard enough collision to snap a prop?

So - moving on to confirming your diagnosis - if I were able to recover the Phantom, we'd see if it had a missing or broken propeller - but it's underneath somewhere between 4-14 feet of water. Does anyone have any experience searching for one of these underwater? Would it show up on sonar or a fish finder if I were able to borrow one?

If you knew someone with a portable 3D sonar like bass fisherman, etc use you might well find it, especially if you have a reasonable notion of where the bird is. They have resolution in the order of inches at those depths.

Maybe somebody needs to market a sonar pinger for Phantoms that is water actuated - like real planes have.