Three phantom 2+ Vision dive-aways

Joined
Dec 4, 2014
Messages
4
Reaction score
0
Hi forum. I'm an advertising photographer and have been using a Phantom (unsuccessfully, as you'll see) mostly for sharing and learning, and a DJI s1000 with 5d3 for work. I feel I've got to share my story so hopefully someone can shed some light on what's happened with my P2V+ experiences. As the topic states, I've had three separate Phantom 2 Vision pluses each fly themselves, each time into a left-diving crash.

The first one was preordered directly from DJI when they first were announced. I had months of successful flights, had updated firmware a few times along the way, and setup in NAZA mode. Its demise I chocked up as my own fault, as the bird had been in the water (due to my own stupidity) some weeks earlier although I had seemingly revive it and it had flown numerous times successfully in the meantime. I initially and I suppose still do attribute this crash to a possible delayed effect from that submersion. Or a user error involving GPS setting. Anyway, here's how the fatal flight occurred. I was flying in a heavily treed area near a lake cottage and was therefore unable to get GPS signal. My intention was to fly it out of the trees and over the water then allow a friend to fly after it picked up GPS. I had done the compass calibration dance. Took off and within 30 seconds the app requested compass calibration. So I flew it back and recalibrated. Took off again. Seemed okay so I flew it out of the trees and over the water. My friend was watching the phone and could see it pick up satellites on the app. I watched the phantom blink green and immediately dive to the left at full speed and about a 30 degree falling angle straight into the lake. Tried unsuccessfully to find it. I blamed myself also for having flown it in GPS mode when it didn't have satellites. I figure I should been flying in ATTI while without signal and then more carefully switched to GPS. Nevertheless, it's almost like it behaved as an iPhone does sometimes when it first tries to locate itself on maps and the blue dot jumps a mile or two as it picks up satellites. It decided it wanted to be a mile to the left and a hundred feet lower and nothing was going to stop it. Splash and lost forever.

So I kicked myself and bought another. Again many successful flights. Until I was flying for a client at a motorsports event. Had food flights with it that day. Took off to shoot a quick video of an area of the event's vendor area. Relevant or not, it so happens that where I took off was somewhat near the GoPro display trailer, and also near some relatively small residential style overhead power lines. I had done the calibration dance nearby that day, so I didn't start by calibrating, but a few seconds after takeoff the app requested compass calibration. So I landed and calibrated. Took off again and while flying the controls seemed imperfect. I had been flying in ATTI and it was pretty calm but it was behaving like I had underestimated a swirling wind. So I decided to get out of the somewhat congested area I was in and gained altitude to get above the trees, power lines, etc, and switch to GPS to see if that would work better. As soon as I switched up to GPS, I had the same dive to the left as I described in the last crash. To the left and slightly away, 30 degree 25 mph powered dive straight into trees a few hundred yards away. On this one I had time to try different stick positions to no effect, and after that tried switching back to ATTI with no change. Couldn't believe it. People were nearby and I had been flying over crowds earlier, and in fact I had to fly the s1000 over people later that afternoon. And of course couldn't use the phantom for the rest of the shoot. I was mad. I had bought this one on Amazon, so I exchanged it for a new one citing faulty product as a reason. Was the crash due to wifi interference from GoPros or a small power line? Maybe, but why only immediately when I switched to GPS?

Amazon sent me a new one so I humbly and cautiously proceeded. Same story - updated and implemented NAZA mode. Maybe 50-100 successful flights over a month's time. Then on a fateful mission, this one in the desert (sand dunes actually, as far away from 2.4 or 5.8 interference or power lines as you can get), I had the same left-diving uncontrollable flyaway. Because I was on the top of a sand dune I had time to try nearly everything in the four or five seconds as it dove a few hundred yards away. Even turned off the transmitter and the whine of the propellers never changed a bit. It was on a mission. Conditions were low winds and early morning, I had actually flown it in nearly the same location the day before in much windier conditions. Dive to the left. Third one.

I searched the internet for similar stories, talked to the far-smarter-than-I-am fella who built my big bird, and a dealer who has sold hundreds of phantoms and can't find anyone to relate to my recurring dilemma. After some deliberation I returned this one to Amazon for a refund and purchased a Phantom 2 and had it set up for GoPro. Figured I had to try something different or I'd be even more mad if I didn't change anything and it happened again. I'm about a dozen flights into this one and all seems fine, for now.

Would sure appreciate the insight anyone can give. Best regards, JL
 
Out of curiosity are you powering off and powering the phantom back up after your compass calibration to let the IMU do its initialization?

Also, are you powering it up on a fairly level open grassy (non-ferrous) area? Ie: not powering up and then setting it on the ground, and not placing it on cement (with rebar) hidden inside.
 
Although it would seem unlikely, suffering three simmilar crashes, I'm leaning towards magnetic particles stuck to the magnets in your front left motor. I've had this happen once, and they come down at a forward and left sloping dive.
I found my bird and pulled everything apart trying to figure out what happened. That's when I saw some tiny black (sand?) bits in the left front motor. Again, not sure if something like that happened to you, but if you take off or land in loose dirt or sand, it's certainly possible.
 
Doesn't make sense that the same exact thing would happen with multiple drones. Does your vehicle (maybe parked nearby) have it's own wifi hot spot that could somehow (I don't know how) be interfering with the signal? I know some of the new cars now are their own wifi hot spot.
 
Looking for a common factor , I would wonder about a corrupted app. I think before I would try again, I would remove the app completely and reload it. Three drones all misbehaving in a similar manner would make me look for the common thread.
 
Amazon replaced your crashed Phantom after 50-100 flights? Multiple times? That is some stellar return policy. I'm glad I bought mine from Amazon.
 
Wow 3 separate drones all the same issues. To me this tells me the issue is with something your doing wrong (obviously not on purpose) or something about the area your flying in. If this was a defect with the Phantom we would all have the issues but for it to repeatedly only happen to you sounds super odd. Sorry I don't have the answer, I know how painful it is to lose a $1,200+ drone.
 
MapMaker53 said:
Doesn't make sense that the same exact thing would happen with multiple drones. Does your vehicle (maybe parked nearby) have it's own wifi hot spot that could somehow (I don't know how) be interfering with the signal? I know some of the new cars now are their own wifi hot spot.
Thought about the wifi issue with the gopro people (second event). I would lean towards control (5.8 ghz) or a common mechanical/user problem.
 
With all 3 displaying the same pattern I would have to question whether you are going through the compass calibration properly or if you are placing your visions next to a subwoofer in your trunk.
 
Seems to be the common hitch in the stories... "took off and immediately asked for compass calibration again... calibrated and took off"

Why is it asking to calibrated so often? what are your storage situations?

Subwoofer?
Too close to the S1000 beefy magnet motors?

If you go to fly in an area... and you calibrate first... and then it immediately asks you to calibrate again after takeoff ... there is something wrong... it's time to stop and pack up... get into Phantom assistant and see if compass values are consistent or if it needs degaussing.


If you are storing your compass to close to a heavy magnetic source it will start to share magnetism or become charged magnetically.... when you move it away from that heavy magnetic source it will begin to lose that charge over time... so an hour later the compass will show different value ... and then 30 minutes later different value again... the Phantom can't adjust for that so it asks for additional calibration... then awhile later... "boom"
 
Hi guys. Thanks so much for the replies. I understand we may never know, but it felt good to vent and share. I think I've run through almost all of the theories laid out above myself, and pretty much reach a dead end each time. The only one that caught me off guard is the question of whether I power off after doing calibration. I haven't been doing that and hadn't seen it suggested before. Is that true?

Other than that, in regards to the storage questions it's kept in a waterproof hard plastic case with laser cut foam, I forget the brand. Two of the three flights would have taken off near large off road vehicles. But many many more other successful flights would have as well. In each of these instances the s1000 was not nearby. Each of the instances were in completely different states, so no common location problem. I'm curious what the app could have done to cause this, but it's a possibility. Same iPhone and app in all instances.

In my mind it seems it has to have been a problem with GPS somehow. The common factor is that it either just acquired GPS, or I had just turned on GPS mode. Not sure what I might have done wrong, but that's the only thing that makes sense to me. All this talk about compass interference - would a sudden unresponsive dive really be what happens with a slow changing compass reading? Keep in mind that it flew okay up until it suddenly dove.

Another thing I could mention is that even though I had kept them in the padded case and tried to baby as much as possible, each of these was subjected to probably more than average vibration over time, due to my work with boats and off road vehicles. I had a theory going that maybe the gps/compass module was coming disconnected or something. But except for the first one which was lost forever, the other two both flew again. They were both in bad shape and the camera/gimbal ruined, etc. But we unbent the plastic arms and patched them up enough to fly just to test and see if a motor was seized or if something had come disconnected, etc. And both of those flew okay after their crashes. Go figure.

Thanks again for your time and help. JL
 
The only thing that makes sense to me is messing up your compass by storing near a magnetic source.

Nothing else makes sense.
 
happydays said:
lsteigerwalt said:
Check out this post viewtopic.php?f=27&t=30912
I'm willing to bet this was your issue every time.

On three separate phantoms?
Yes Connor, the common thread here though is the vibration that the birds are suffering. I can't say why they worked for a long time beforehand (unless it took them that long to be affected by the vibration).
 
Fplvert said:
Although it would seem unlikely, suffering three simmilar crashes, I'm leaning towards magnetic particles stuck to the magnets in your front left motor. I've had this happen once, and they come down at a forward and left sloping dive.
I found my bird and pulled everything apart trying to figure out what happened. That's when I saw some tiny black (sand?) bits in the left front motor. Again, not sure if something like that happened to you, but if you take off or land in loose dirt or sand, it's certainly possible.

I've suffered this issue. Thankfully with a £17 quad-copter, lol. It yawed violently, constantly. Basically any throttle would make it go up and yaw left until it crashed. When I opened it up, something was stuck to a motor. Something metal. I took it off and it stopped spinning.
 
Hi JL,

I've had exactly the same issue with my P2V+ v2.0. The drone started to behave a bit strange when I was on holiday, drifting sideways unexpectedly a couple of times, but I could land it safely every time. Back home it really went wrong: I took off from an empty stretch of grass land, and almost immediately the drone started to dive away just as you described- straight into a creek! After some searching I was able to recover the drone and removed the battery as quickly as possible, and let the drone dry for some time (but I never tried to see if it still worked).

In the meanwhile I contacted the Dutch reseller. They asked me to describe what happened, so I wrote a lengthy description of what went wrong. I was quite sure that it wasn't pilot error: I did the compass calibration, waited for all the lights to go green before take-off, etc. Eventually they asked to send in the drone for repair. I got the bird back a couple of weeks later, they told me that the flight computer was replaced, free of charge, but I didn't get a replacement battery though. Anyway, I went for a testflight, which went well initially, but after about 10 minutes of flight, the drone started to tilt again which resulted in an uncontrollable dive-away, until it hit the ground badly, ripping off the camera.

So I contacted the reseller a second time, and I was asked to return the Phantom again. After a while they contacted me, saying that they would ship a new P2V+ to me, a V3.0 this time. My contact person told me that they had an argument with the DJI agent / repair agency, from what I understand they just wanted to repair the drone again and ship it back a second time, but the reseller insisted on sending me a new one. Good service!

I still have no clue about what might have caused the dive-aways. On both occasions I had good flight conditions and there was no sign of GPS problems. My theory is that it had to do with the compass. It was definately not a 'regular' fly away, I was quite close to the quadcopter during the two crashes and there was no strong wind whatsoever. The drone did somehow still react to the RC but there was no controlled flight possible, given the tilting and diving behavior.

I'm still a bit reluctant to bring out my new V3.0 P2V+ but of course I will fly again, probably tomorrow. My lesson learnt is that these drones will never be completely reliable so I will be extra careful when flying. I hope that other people will use their brains too and not fly close to people, roads, et cetera. When it goes wrong a crash can happen almost instantly and you can never predict how the drone will land once it goes berserk.

Yours,
Wiebe
 
Wiebe, thanks for the comment. Sounds exactly the same as what I experienced. As I said in my initial post, I migrated to a non-vision. Haven't flown it too much yet, but so far so good.

I read the reports above with interest (about the GPS connection issue). But in my case after two of the crashes the heli picked up satellites and and flew again, without ever taking off the cover. Seems that rules out a broken connection issue.

Best regards,
JL
 
I know these forums attract people world wide. I would love to know what the percentage is of phantoms are having these kind of problems. I've had a Phantom 1 for a couple of years and have never had any problems with it. That's the reason I decided to get a P2v+. Now I'm starting to get scared to fly the thing as I don't want to see $1300 fly into the sunset
 

Recent Posts

Members online

No members online now.

Forum statistics

Threads
143,090
Messages
1,467,565
Members
104,974
Latest member
shimuafeni fredrik