P4P v2 Fly away today

[sar104]

Above, I thought you were saying Litchi creates a regular txt log file (in addition to the CSV log file everyone is familiar with).

That's correct, it does, at least under iOS. The txt log is located in Litchi » Documents » SDK_logs » FlightRecord. I don't have Litchi on my Android device so I can't confirm that it exists or where it can be found, but it seems unlikely that the Android version wouldn't also create one.

 

[sar104]

Apparently DJI dont cover it for warranty if its using a third party app?

That's not correct. However, the lack of appropriate log data may be an issue, depending on the app.

 

[Meta4]

Apparently DJI dont cover it for warranty if its using a third party app?

Although this is often repeated in forums, it's not correct.
The app you fly with is not important.
The two important factors are:
1. The cause of the issue was a DJI issue
2. You are able to provide evidence of (1)

If you have lost the drone, DJI won't use a Litchi flight record and you are out of luck.
But if you have the drone, DJI can use the internal dat file from the drone to investigate the incident.

 

[msinger]

Apparently DJI dont cover it for warranty if its using a third party app?

If you lose your drone, you'll have nothing but the flight logs to submit to DJI for review. If you don't have DJI flight logs (from DJI GO), they likely won't want to look at them. No drone... no flights logs... no warranty assistance.

Just flying with a 3rd party app won't void your warranty though.

 

[jbain1]

You might want to send it in to be repaired. They will give you a quote - it might be significantly less than a new drone. I sent one in that I thought was not repairable - quoted $550 and I got a new P4P v2

 

[Born to Fly (Remotely)]

This is very interesting.

I have had this experience, more than once. I was using Go, not Litchi. What is in common, is the drone suddenly ignoring the remote, gaining altitude and flying off to wherever the mood took it. I am trying very hard to find out why. In each case I was able to recover the Phantom, barely, but as much by luck as anything. In no case did GTH work. Changing to ATTI mode gave me some sort of intermittent control, even though I was line of sight, (stretching it when the drone had flown as high and far away as it chose to) but, by the smallest of margins, landed it somewhere recoverable. A perfectly normal flight, everything going smoothly, and then this sudden mad flight. At first, no responsiveness to the remote, and then, sometimes responding, which enabled me to point it the right way, suggest it go "forwards" and "down". After that it was all luck.

I have toyed with many theories as to the cause. I don't know. I'm pleased to find I am not alone. It is very disconcerting as I don't know when it will happen. The worry isn't just losing the drone, but if the battery dies before my little random bits of control point it somewhere safe, the damage done could be anything. It doesn't happen often, it happens rarely. But more than once to me. But I don't want it to crash into a, well, anywhere bad.

site guy, did you ever get to the bottom of this? Did your logs get analysed by someone, and did that lead to a solution, or even an explanation?

Would be good to share with all. You said you still had plenty of battery, I can tell you it is much scarier when the battery is low.

I can just imagine standing in front of the FAA and the Judge and saying "it wasn't me, the drone just had a mind of it's own", after those fast spinning blades come crashing down in a playground.....

Really keen to learn how to not have this experience again.

 

[Meta4]

I am trying very hard to find out why.
...
I have toyed with many theories as to the cause.
...
Really keen to learn how to not have this experience again.

No point wondering and guessing.
The first thing to do is to look at the recorded flight data to find out what actually happened.
Go to DJI Flight Log Viewer | Phantom Help
Follow the instructions there to upload your flight record from your phone or tablet.
That will give you a detailed report on the flight data.
Come back and post a link to the report it gives you.

It would help to give a description of what happened and what your launch spot was like.

 

[braden77]

Had a very similar experience with my P4pro V2 6 months ago. The drone was fine one moment and then suddenly lost the link, flew out of view over a tree line and disappeared. I was freaking out with no idea what to do or why it happened. Was able to recover it two days later, downloaded the data file. Apparently, it was in RTH mode (set to do so if disconnected) on its way back with 11 gps satellites locked on when it just shutdown and fell out of the sky from 300ft with 80% battery. Filed a past warranty claim with DJI, shipped everything in for “repair”and prayed. A week later they notified me that it would be completely covered under warranty and then 7 days later I had a brand new P4p v2 delivered to my doorstep. And only explanation they gave was that the “motors shut off while in mid-flight”. After reading some online horror stories about warranty claims I wasn’t sure how this was gonna turn out, but BIG shout out to DJI for standing behind the product and rushing out a replacement!

 

[Weston Ney]

First thing to try if your drone starts to fly away is immediately switch to atti mode. If that doesn't work pull your left stick down and to the right corner and push and hold return to home. The propellers will stop and the drone will crash, but it will keep it from flying away and hurting property or people. Best to have it crash right there in front of you, and who knows, you could even have a chance of catching it! A friend of mine caught his Phantom 3 as it was crashing once and saved it from further damage.

 

[RodPad]

@Born to Fly (Remotely), @siteguy,

"site guy, did you ever get to the bottom of this? Did your logs get analysed by someone, and did that lead to a solution, or even an explanation?

Would be good to share with all."

Yes, that would be great!


Rod

 

[Born to Fly (Remotely)]

Thanks for the help. I have downloaded the flight records, into excel. Wow, a lot of data.
I departed from a wide open beach with good visibility, directly out to the ocean. Sand hills behind me. I can see where it went from a nice consistent 30.2m on the way back to me, and began climbing. No error messages. . On the way up, it was reading P-GPS until at 178m it changed to Go Home, although I hit Return to Home from the outset. At 215feet the message RTH Preascending, was recorded. At 186feet I went to ATTI mode. I briefly accidentally had it on sport mode, used to the P3 controller I suppose. That was for nearly a minute. No response. At 253 feet I see an entry: warnings:Mobile device CPU fully loaded. Related performance will be affected.

Then at 518feet the message changes to Go Home. Then "GPS position no match". At 560 feet the log says "RTH Cruise, followed by lots of "Cruise" messages, and "Braking to avoid obstacle". It didn't but climbed to 675feet. It is 948m away and not coming closer. There is an error in the log, "Cache Full".

In ATTI mode it starts to inch toward me and descend. Message days "In ATTI mode, change to GPS for safety". 495 feet and 650m away message changed to auto land. Then 1 message that says "Command Error". None of this is over the water, it had flown over my head and into the housing area behind me. 387 feet and 655 metres message changes to Forced Land. Critically low battery warning. It landed 616 metres away, but somehow had managed to find it's way back to the beach, skimmed the Sandhills and the battery died about 2 metres short of the water. A miracle.

So I can see that my iPad is a little too old, small, short on memory perhaps. At it's furthermost point there were a couple of messages telling me to adjust the antenna for better reception. I don't know what obstacle it thought it was avoiding up there. Maybe the bait dropper. Seems unlikely as it hadn't tried to avoid the bait dropper at any other time. The sun was up when this flight started. By the time it got back, the sun had set. I have read that sunset can confuse the sensors.

Many Thanks

 

[Born to Fly (Remotely)]

178 feet, not metres, sorry

 

[Meta4]

Thanks for the help. I have downloaded the flight records, into excel. Wow, a lot of data.

There are some good eyes here.
If you post the txt file, you could get some experienced analysis.

 

[RodPad]

This is very interesting.

I have had this experience, more than once. I was using Go, not Litchi. What is in common, is the drone suddenly ignoring the remote, gaining altitude and flying off to wherever the mood took it. I am trying very hard to find out why. In each case I was able to recover the Phantom, barely, but as much by luck as anything. In no case did GTH work. Changing to ATTI mode gave me some sort of intermittent control, even though I was line of sight, (stretching it when the drone had flown as high and far away as it chose to) but, by the smallest of margins, landed it somewhere recoverable. A perfectly normal flight, everything going smoothly, and then this sudden mad flight. At first, no responsiveness to the remote, and then, sometimes responding, which enabled me to point it the right way, suggest it go "forwards" and "down". After that it was all luck.

I have toyed with many theories as to the cause. I don't know. I'm pleased to find I am not alone. It is very disconcerting as I don't know when it will happen. The worry isn't just losing the drone, but if the battery dies before my little random bits of control point it somewhere safe, the damage done could be anything. It doesn't happen often, it happens rarely. But more than once to me. But I don't want it to crash into a, well, anywhere bad.

site guy, did you ever get to the bottom of this? Did your logs get analysed by someone, and did that lead to a solution, or even an explanation?

Would be good to share with all. You said you still had plenty of battery, I can tell you it is much scarier when the battery is low.

I can just imagine standing in front of the FAA and the Judge and saying "it wasn't me, the drone just had a mind of it's own", after those fast spinning blades come crashing down in a playground.....

Really keen to learn how to not have this experience again.

You stated "site guy, did you ever get to the bottom of this? Did your logs get analysed by someone, and did that lead to a solution, or even an explanation? Would be good to share with all.

What are you sharing ?
(F)acts or (F)iction

Rod

 

[Born to Fly (Remotely)]

Hi Rod,

Sorry,you have lost me if that was directed at me. This is what happened and I just excruciatingly picked the data out of the flight records, which to be honest, if not for this forum, I didn't know I could do. Fact or Fiction??

 

[Born to Fly (Remotely)]

txt file

 

[Meta4]

txt file

Your flight data looks like this: DJI Flight Log Viewer - PhantomHelp.com

the drone suddenly ignoring the remote, gaining altitude and flying off to wherever the mood took it. I am trying very hard to find out why.

It looks like you had no problem with left/right control and the drone responded normally to left/right joystick inputs.
It also went up when you pushed the left stick forwards.
Your description is a bit vague but I'm guessing you were talking about the drone:

• getting to the home point at 11:59 but not descending
• not descending when you pulled the left stick down as happened at 13:33 and other times after that.

The reason for that shows quite clearly in the recorded data.
You had a problem with the downward facing sensors that indicated to the drone that it was almost on the ground.
The VPS has a range of up to 10 metres but even though the drone was much higher, VPS is showing very small numbers when it should have shown nothing.
This would have been visible on your flying screen the whole time.

This might have been due to a gimbal guard or something else blocking the view of the downward facing sensors.
You could have solved the issue in flight by disabling the VPS sensors.

 

[sar104]

Your flight data looks like this: DJI Flight Log Viewer - PhantomHelp.com

It looks like you had no problem with left/right control and the drone responded normally to left/right joystick inputs.
It also went up when you pushed the left stick forwards.
Your description is a bit vague but I'm guessing you were talking about the drone:

• getting to the home point at 11:59 but not descending
• not descending when you pulled the left stick down as happened at 13:33 and other times after that.

The reason for that shows quite clearly in the recorded data.
You had a problem with the downward facing sensors that indicated to the drone that it was almost on the ground.
The VPS has a range of up to 10 metres but even though the drone was much higher, VPS is showing very small numbers when it should have shown nothing.
This would have been visible on your flying screen the whole time.

This might have been due to a gimbal guard or something else blocking the view of the downward facing sensors.
You could have solved the issue in flight by disabling the VPS sensors.

Agreed - there was a major problem with the VPS height reading and it appears that the sensor was obstructed.

There is no point in trying to troubleshoot any further until that is fixed. IMU looks mostly consistent:

 

[Born to Fly (Remotely)]

OK Thank you folks. You do have experience at reading these logs!
So, this takes me back to why the downward looking sensors thought it was almost on the ground? Why did it start climbing when, while it was out to sea, all I did, as usual, is hit RTH?

Your saying there was an obstruction to the downward looking sensors. This makes sense to me. Lets say they "saw" the bait dropper. In flight I guess it would think it was too close to the ground and so in P-GPS mode starts climbing? Trying to RTH would have the same problem then?

When it was flying out of VLOS into places I am not allowed to go, and the battery was dying, I was quite anxious. I could well imagine I could have made some joystick errors, as I did when putting it into Sports mode briefly. But still should have been able to bring it home in ATTI mode.

What did in the end enable it to get back to the beach, even if nowhere near the home point? I do recall being relieved at one point that I could see it was pointing towards to ocean, although I wasn't able to then just fly it back normally. At least if I could get it moving, then it crashed, or I crashed it, it wouldn't be on a populated area.

I saw a few messages about obstacle avoidance. But not many. Do the downward looking sensors create those messages, or did a bird fly by?

I am now growing more suspicious that the bait dropper was what the downward sensors picked up on. It was mounted under the camera, where I thought the angle of view of the sensors wouldn't see it (else they'd see the camera first), and after perhaps 20 flights configured like this, the drone behaved perfectly.

Thank you for your fast analysis. I appreciate that you have explained what went wrong.

What I want to do is to NOT have to turn off the security features of the drone just to be able to use a bait dropper. That seems unreasonable to me. I don't think the downward looking sensors have a sufficiently wide angle view to see something under the camera, reading the specs. And it has worked over and over. But I am still suspicious that is the cause.

By the way, you will not be offending any bait dropper manufacturers if you reply. I made it myself. It is lightweight, has no electronics, no metal parts, doesn't rely on line tension to release and is small enough to, in theory, avoid the sensors while not interfering with the camera. Well, all up until this particular flight anyway......

Thanks Again

 

[Born to Fly (Remotely)]

Sorry, while you are being so indulgent, would you say my iPad is underpowered, or am I misinterpreting that? Can fly these without any ipad/iphone/android device anyway, so doesn't feel like the iPad can cause problems, yet the error messages about cache full, and the other one about the mobile device, seems point to my ipad.......