P4P Flyaway

[TopView Imaging]

Title says it all. Someone on Reddit suggested I head over here and post logs for the pros to take a look at. I'm not sure how to do that. Any help would certainly be appreciated. I've clicked on the export CSV button in the DJI Go 4 app but do not know where to go to get the file.

Short description. Flew that day twice before with no problems. Two different locations. Second location, landed and swapped batteries. Put it back up and it disconnected and just left the area about 2 minutes into the flight. The feed on the screen froze and then showed an "aircraft disconnected" message. I've seen this flash onscreen very briefly, a few times previously, but the connection always returned before the auto RTH feature engaged. The last message to pop up on the screen was a warning about high winds and to fly carefully. It did NOT tell me to land ASAP.

The aircraft was NOT in ATTI mode.

 

[Fly Dawg]

You can start at the link below. Follow the instructions then post a link back here to that.

DJI Flight Log Viewer

 

[Fly Dawg]

By the way, true "Fly Away's" are extremely rare. There are normally underlying issues that lead to that, and generally speaking the higher percentage is Pilot error. We can look at the data once you get it posted.

 

[TopView Imaging]

I must be making too many changes to my post too quickly. It's giving me a hard time. Below is the link to the phantom help with my uploaded log
DJI Flight Log Viewer - PhantomHelp.com

 

[Fly Dawg]

Taking a quick look at this data, I see no warnings ( Other than the High Wind ) the log abruptly ends at 1:19....in a tree clustered area according to the GE data. Right now I would suggest a sudden power loss. Have you looked in that area? Give me few minites and will post a GE profile.

 

[Fly Dawg]

In the area below with the red indicator ( That was the high wind warning )..................As unfortunate as it is there have been numerous power loss's recently on the P4's. Primarily due to the battery loosening in flight. I can't be certain from this particular data set but that is what it appears to be.

 

[TopView Imaging]

In the area below with the red indicator ( That was the high wind warning )..................As unfortunate as it is there have been numerous power loss's recently on the P4's. Primarily due to the battery loosening in flight. I can't be certain from this particular data set but that is what it appears to be.View attachment 99389

I was well clear of those trees at 170’ altitude. Also there wasn’t even a hint of voltage dropping. We searched the area as thoroughly as we could. Granted it is fairly heavily wooded in that particular spot.
Going with your assumption, do you think DJI will take any responsibility? Do I have any chance of warranty possibility?
I’m just struggling to wrap my head around this. I’ve taken good care of the equipment and flown diligently and safely. I don’t understand how this could happen.

 

[Fly Dawg]

I understand your altitude was fine, as was all the other data. What you have to remember is once there is a sudden power loss there is no more data transmission. You would not see a voltage drop in the data. As far as DJI goes, I can't say yes or no, but from your data the "incident" is inconclusive at best. The P4's also record a .dat file on the device. You might try retrieving that and see if it provides any more detail. The primary data will be on the aircraft itself, if you can locate it. If this were a true "Fly Away" you would have seen other evidence in the data to indicate such. You were only roughly 800 ft out so I would not think signal loss is a factor. If that were the case RTH or Hover would have initiated, assuming your failsafe was set correctly. I will refer this @sar104 but I wouldn't be surprised if were the same conclusion.

 

[Fly Dawg]

Here is a plot of your battery. Notice the data just simply ends.

 

[sar104]

Here is a plot of your battery. Notice the data just simply ends.

View attachment 99390

That does look like sudden power failure - possibly the dreaded battery disconnect. The aircraft DAT file would be the only way to confirm power failure since the telemetry has neither fast enough sample rate nor low enough latency for the voltage drop to show up in the txt files.

I did get around to doing a simple test to look directly at voltage collapse rates when the battery disconnects. Even the aircraft DAT files only record at 50 Hz, which only gives a few data points during the voltage drop. I used a P2, since it is much easier to get to the internal connections to measure voltage directly, and sampled at 3 MHz while pulling out the battery with the motors running.

The results were a little surprising given the DAT file data for battery disconnect, where the voltage drops to around 7 V (the minimum required to support the FC) in around 80 ms.

In this case the voltage drops by around half, reproducibly, in about 4 ms, so much faster, with an interesting shoulder in the data at around 3.5 V (FC shutdown?). It's possible that the P2 circuits are substantially different, but that's not a very satisfactory hypothesis. If anyone wants to take apart a P3/P4/Mavic to do a similar test it would be interesting to compare. I'm not very enthusiastic about dismantling any of mine.

 

[TopView Imaging]

Guys,
Thank you so much for your help on this. I really appreciate you taking the time to look into it and provide input. What's the story with the "dreaded battery disconnect"? Sounds like something that happens more than it should? Is there anything one can do to prevent this? I always made sure to seat the battery as solidly as possible. Is this a general defect or poor design? Is DJI doing anything about it?

 

[sar104]

Guys,
Thank you so much for your help on this. I really appreciate you taking the time to look into it and provide input. What's the story with the "dreaded battery disconnect"? Sounds like something that happens more than it should? Is there anything one can do to prevent this? I always made sure to seat the battery as solidly as possible. Is this a general defect or poor design? Is DJI doing anything about it?

There has been a recent noticeable increase in reports of sudden power failure on P4Ps - just like yours. The DAT files from those that were retrieved all suggest a battery disconnect. DJI also added a battery latch sensor to the v2, which may be just a coincidence, or may indicate a propensity for the batteries to appear latched when they are not.

 

[TopView Imaging]

Wow...that is quite the "coincidence"...adding the battery latch sensor. Well, at this point, with the hull being MIA, I guess I'm probably screwed out of $1500.

 

[sar104]

Wow...that is quite the "coincidence"...adding the battery latch sensor. Well, at this point, with the hull being MIA, I guess I'm probably screwed out of $1500.

It shouldn't be that hard to find - it will have fallen pretty much straight down at the last recorded location, and I'd expect it to be within 10 m or so of there:

45.814526, -92.919875

 

[Fly Dawg]

Agreed with @sar104, In this image the extended bottom path line is approximately 50ft from the last reported position. Taking in to account that your forward travel at that point was 25.3mph, the AC would have fallen at an angular path, so is should be just a bit forward of the actual last location, but not much. This is assuming very light to no wind.

 

[sar104]

Agreed with @sar104, In this image the extended bottom path line is approximately 50ft from the last reported position. Taking in to account that your forward travel at that point was 25.3mph, the AC would have fallen at an angular path, so is should be just a bit forward of the actual last location, but not much. This is assuming very light to no wind.

View attachment 99408

Good point - I had overlooked how fast it was going when it shut off.

 

[TopView Imaging]

Found it

 

[sar104]

Found it

Excellent. Coordinates?

 

[Paul2660]

Was the battery still installed?

Glad you found it.

Paul C

 

[PhantomWetSuits]

Wow...that is quite the "coincidence"...adding the battery latch sensor. Well, at this point, with the hull being MIA, I guess I'm probably screwed out of $1500.

The New Version 2.0 does solve this problem with the switch , But we have had to post this video to many times because of that battery disconnecting on the older versions of the Phantom 4 series..

We ordred a 2.0 to see if the battery still has the loose wiggle. Has anyone checked that.

Most likely you did not see this video.
Approved Vendor Phantomrain.org