P4 flyaway

[Jussaguy]

My drone lost signal for the first time ever, completely lost signal, I was 1/2 mile away and through trees, in a noisy (wifi) environment. About 15 seconds later it came back online. I then turned the drone around to come home, when I noticed I had turned it away from me. So, she returned just like it was engineered to, even better because it was now low light, it was dusk when I started, wanted to take advantage of 1mph winds. The fact it re-acquired signal meant the OA wouldn't be needed. I was shaking, thinking the worst, lucky it was only for less then a half a minute. The craft turned around and lowered elevation about 100' and like I said, was well on its way back. I'm starting to trust it a lot more now. Having State Farm covering it helps, but wouldn't want to admit I was flying at night. Also due to it being insured, I've reset my height limit to 400' again, don't want any BS in case of a claim, however I suppose I could just say it's gone but, at 46 not into insurance fraud so I keep it in the up and up.

I hear ya!

The EXACT thing happened to me but I knew it was coming home because that's what it does when it loses connection but the complete disconnect from the aircraft at such a close range is, to put it mildly, disconscerning.


I believe there are updates that need to be coming to either iOS or the Go App itself or maybe the firmware (on either the RC or the craft).

All I know is that when they updated iOS to 9.3.1. and the firmware to 1.6 or something and then 2.7 on the Go App to a completely different looking app and then 2.7.2 to fix some of those problems, there have been issues for me and a contingent of others all over the world. To really have a look go to the DJI forum.

 

[Sim597]

Something nobody brought up, so you've had 2 flights before this on a new update, and both of those were RTH also, so you have 100% of your flights were RTH? And you then were going to fly again from same spot? I mean, seems like a really strong interference or something.
I don't know, maybe it's not, I in fact are going to try and do same flight I lost signal on again, this time a different radio and setup maybe, so maybe not so strange, but I've had maybe 15 perfect flights since updating.

I've heard some say, well, one guy say the Kp index has no effect, but that's in contrast to all those others who say it does affect GPS to the point of failure. I'd love to see a correlation between the days ppl have had fly-aways to the kP index at the time and place these happened, i don't have the software or time, someone could gather data, UAV Forecast is the one that should prove the incident rate, then charge $1.99 for the app if it does have anything to do with our actual hobby.


Sent from my iPhone using PhantomPilots mobile app

 

[thunderbird]

There info about the gimbals have some loose screws, two ppl here have already fixed their gimbals using this method, give it a try and see if any loose screws

Do u have the link about the gimbal loose fix ?
i have mine does make the screaming sound but only when i shake it hard

 

[John Locke]

Do u have the link about the gimbal loose fix ?
i have mine does make the screaming sound but only when i shake it hard

Shaking your drone hard? Sounds like cruelty to drone. Poor thing, it may need a new owner.

 

[John Locke]

False confidence is right I was out yesterday for my P4 and decided to take it into sport mode n-fly 4 foot off the ground at max speed, relying totally on the ground sensors to hold my elevation crashed and about 30 sec, never reached max speed, only lost one prop in crash, but won't totally rely on sensors again

You said "relying totally on the ground sensors". So you're saying you were not watching the craft when you were going 40mph just 4' off the ground. Is that correct?

 

[Sim597]

Do u have the link about the gimbal loose fix ?
i have mine does make the screaming sound but only when i shake it hard

I thinks it's right here under "screaming gimbal" I can look tho


Sent from my iPhone using PhantomPilots mobile app

 

[Sim597]

Do u have the link about the gimbal loose fix ?
i have mine does make the screaming sound but only when i shake it hard

Dji ph4 Gimbal Screaming issue !!!


Dji ph4 Gimbal Screaming issue !!!

Sent from my iPhone using PhantomPilots mobile app

 

[ned]

A few questions, but first, the gory details. P4 with latest firmware / DJI-Go / Apple 6. Two routine flights - both ended with the RTH request. From the same location, 3rd flight -drone became unresponsive and flew away. A return to home was sent. The flight log shows the sent request. The playback says the drone hovered in place at a place that should have been visible by eye. I was walking towards that location when the low battery - RTH message flashed up. According to the flight log, the drone flew directly over my head and landed right behind me at Home. None of the 3 of us watching for it saw it after it flew off. The log shows a second battery low RTH message and the drone flew away from home and into a lake.

The log shows 18-20 GPS, good IMU and magnetometer calibration, but frequent loss of signal - is this common? Is it common for the logs to contain GPS data putting the drone in a place where it clearly was not seen by eye? There are several NaN entries for speed - is this common/normal?

TIA

I calibrated the compass perfectly.It flew off like normal then my P4 lost connection. not sure where it went.

 

[ianwood]

See this video for a possible explanation:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

This video was wrong when he posted it and it's still wrong some 45,000 views later.

 

[Mark The Droner]

How do you figure?

 

[ianwood]

How do you figure?

It's entirely wrong. The whole thing. He has no idea what he's talking about.

 

[normr]

His explanation is plausible, maybe DJI should include GPS coordinates on screen that you could compare with your phone or tablet and if you see a big discrepancy then you can take action before lifting off


Sent from my iPad using PhantomPilots mobile app

 

[Intrepid97]

I always wait for the Satellite to be linked up and give the green go ahead. Also, It wouldn't hurt to perform a test RTH procedure to make sure it has it's bearings.

Really sorry to hear you lost your Phantom. That is a major fear I think most owners have.

 

[ianwood]

His explanation is plausible, maybe DJI should include GPS coordinates on screen that you could compare with your phone or tablet and if you see a big discrepancy then you can take action before lifting off

No, it's not. He has little understanding of how GPS works and how the Phantom uses it. He is wrong.

 

[henick]

He is completely wrong. Phones rarely do that now and when they did it was because it was triangulating a location from cell towers prior to properly acquiring an accurate location from GPS which as everyone knows takes a second to lock. Older gps chips took even longer. It was originally intended to make it faster (or as a fail-safe in a failure such as being indoors) but after many complaints was changed. At the time I was a contractor for a subsidiary of Google API. The issue came up on the maps app when we were trying to put blackberry out of business.

It's actually really the only explanation (that i can think of) that would see the temporary location also show as being relative to the person experiencing the problem, wherever they may be.

The gentleman in the video does say it is his attempt at explaining I don't believe he claims it to be the indefinite reason... so points for that anyway.

 

[sar104]

He is completely wrong. Phones rarely do that now and when they did it was because it was triangulating a location from cell towers prior to properly acquiring an accurate location from GPS which as everyone knows takes a second to lock. Older gps chips took even longer. It was originally intended to make it faster (or as a fail-safe in a failure such as being indoors) but after many complaints was changed. At the time I was a contractor for a subsidiary of Google API. The issue came up on the maps app when we were trying to put blackberry out of business.

It's actually really the only explanation (that i can think of) that would see the temporary location also show as being relative to the person experiencing the problem, wherever they may be.

The gentleman in the video does say it is his attempt at explaining I don't believe he claims it to be the indefinite reason... so points for that anyway.

Agreed - his explanation is almost entirely incorrect, as is his understanding of GPS. The basic hypothesis - that an incorrectly registered home point, sufficiently distant from the actual location to be outside a set distance limit, could lead to the aircraft attempting to get (back) to within the distance limit of that incorrect point, could be reasonable. However, as noted above, Phantoms do not use assisted GPS and, by the time that the FC has locked multiple satellite streams, it is very unlikely that the position error will be more than a few tens of meters, even though the GO app still suggests checking it on the map.

On the subject of phones however, especially those using Google location history services, there still seems to be a tendency for them to report wildly inaccurate positions even after achieving a good satellite lock. I've been analyzing transmitted records from a phone belonging to a missing person (part of our standard search and rescue procedures) and found that the phone would occasionally (a few times per day) report locations that were between a few miles and a few hundred miles in error, sometimes for up to several minutes. That appears to be due to the location service defaulting back from GPS to cell tower ID to save battery life when running in the background and, presumably, referencing an incomplete or inacurate database of towers.