I have a friend who had a crash today with his P4 Pro. He had to get up quick on a job so he took off before it had locked onto enough satellites to give the green bar and home point set. He is in P mode at takeoff btw. But he claims he does this quite often and has never had a problem. But today as soon as it got maybe 35 ft in the air and hooked up to enough satellites it just takes off left right into a tree and fell to its death. Any ideas on what may have happened in this case? I have never taken off until I got the home point recorded message for the app. I'm just wondering how you would know if your IMU was even all booted up and ready to work right if you don't wait for the green ready bar!
Dangers of taking off before GPS Satellites have all hooked up?
- Thread starter KevMo Photog
- Start date
You should suggest to your friend upload the flight log to phantomhelp log viewer and share the link here- there may be a few with a much better idea what might have occurred then.
GadgetGuy
Premium Pilot
- Joined
- Jun 18, 2015
- Messages
- 6,881
- Reaction score
- 2,199
Sounds like a compass error from lauching off a metal surface. The wild left hook is legion among compass errors of the past. I've seen it, and covered it, like Allstate!
Meta4
Premium Pilot
- Joined
- Jan 8, 2014
- Messages
- 15,303
- Reaction score
- 9,247
launching before getting GPS shouldn't create any problems.
Normally, you'd expect to pick up sats soon after and the Phantom will record a home point wherever it is when that happens.
The yellow bar that you get before you have enough GPS sats tells you.
GG's suggestion above is a likely explanation for the incident.
If not that, was there a wind that could blow the Phantom in atti into the tree?
I did. And were gonna try to do that tomorrow night. He is in Salt Lake City Utah. I'm guessing a compass error as well.
Thx Meta 4. Wind def was not cause. He also is telling me that he did not get any compass warnings. That's a concern if it was caused by compass error. We'll try to load the data tomorrow. It pretty much totaled his bird. Fell to the pavement! Ouch!launching before getting GPS shouldn't create any problems.
Normally, you'd expect to pick up sats soon after and the Phantom will record a home point wherever it is when that happens.
The yellow bar that you get before you have enough GPS sats tells you.
GG's suggestion above is a likely explanation for the incident.
If not that, was there a wind that could blow the Phantom in atti into the tree?
A tip that could help before fly is to quickly check if craft is correctly located in the map including heading.
Overkill on my opinion as magnetic North will be the same, until you travel to another country or so. And difference would be tiny anyway. Most recent DJI manuals are not recommending anymore to calibrate every time, if I'm correct.
Meta4
Premium Pilot
- Joined
- Jan 8, 2014
- Messages
- 15,303
- Reaction score
- 9,247
That's completely unnecessary and does nothing to make your flight any safer.
In normal situations there is no need to ever recalibrate your Phantom's compass - even a new Phantom straight out of the box.
Hey meta4 and msinger! Here are my buddy's Dat. files. We have started the bird up and everything seems to be okay besides the gimbal drift and the cracks on the case. But would really appreciate any help in figuring out what happened on this flight yesterday. Again, he got no compass warnings at any time! Once he got it in the air 10-15 ft, it took off straight towards the tree. Also, he is a very experienced flyer and has been doing so for over 5 years. This is the first time a "crash" and even error has happened for him. Thank you guys!
Meta4
Premium Pilot
- Joined
- Jan 8, 2014
- Messages
- 15,303
- Reaction score
- 9,247
msinger
Approved Vendor
- Joined
- Oct 30, 2014
- Messages
- 31,961
- Reaction score
- 10,787
- Location
- Harrisburg, PA (US)
- Website
- phantomhelp.com
msinger
Approved Vendor
- Joined
- Oct 30, 2014
- Messages
- 31,961
- Reaction score
- 10,787
- Location
- Harrisburg, PA (US)
- Website
- phantomhelp.com
Thank you Phantom Help! Please see the attached txt. file from the flight that day.
Meta4
Premium Pilot
- Joined
- Jan 8, 2014
- Messages
- 15,303
- Reaction score
- 9,247
Here's what it looks like: DJI Flight Log Viewer - PhantomHelp.com
There's not much data to work with, the flight was all over in 12 seconds and no home point or GPS location was ever recorded so everything was in atti mode.
The Phantom was launched with the battery showing 77% which can cause bad things to happen - but the flight didn't last long enough for this to have been a factor.
The pilot gave full left rudder from 5.8 secs to 6.3, and 7.3-7.5.
Further strong left rudder was applied from 9.1-10 secs, together with full left aileron and full elevator from 9.4-9.9 which would have caused the Phantom to move quickly to one side.
The pitch, roll and yaw data seems to match the joystick inputs.
The compass error message at 12 seconds is after the crash and the Phantom is on the ground.
We can't tell where the flight was or anything about the local environment.
I would have to suspect that the joystick input was what caused the Phantom to crash into the nearby trees.
There's not much data to work with, the flight was all over in 12 seconds and no home point or GPS location was ever recorded so everything was in atti mode.
The Phantom was launched with the battery showing 77% which can cause bad things to happen - but the flight didn't last long enough for this to have been a factor.
The pilot gave full left rudder from 5.8 secs to 6.3, and 7.3-7.5.
Further strong left rudder was applied from 9.1-10 secs, together with full left aileron and full elevator from 9.4-9.9 which would have caused the Phantom to move quickly to one side.
The pitch, roll and yaw data seems to match the joystick inputs.
The compass error message at 12 seconds is after the crash and the Phantom is on the ground.
We can't tell where the flight was or anything about the local environment.
I would have to suspect that the joystick input was what caused the Phantom to crash into the nearby trees.
Meta4
Premium Pilot
- Joined
- Jan 8, 2014
- Messages
- 15,303
- Reaction score
- 9,247
Here's the relevant part of the data:
Yellow = hard joystick movement
Blue is horizontal velocity which seems to match the application of full elevator and near-full aileron.
No other horizontal velocity shows in the data.
Applying full stick in atti mode close to obstacles can lead to undesired consequences.
This one is quite the puzzler. There actually was a switch from ATTI to GPS+ATTI at time 3.89 but gpsHealth was 0 throughout and the number of satellites never got above 7. There is more about why the PhantomHelp converter doesn't show the switch to GPS+ATTI below.
From your description it sounded like the the kind of erratic flight that would occur when the AC was in GPS+ATTI but shouldn't have been.
Since there was not enough sats the limited amount of GPS data didn't make it's way to the .txt. If there were GPS coords in the .txt it might have been possible to determine if Yaw was correct at launch, i.e. was this a launch-from-geomagnetically-distorted-site incident. The .txt shows that the heading was -4.1 degrees prior to launch. Do you recall if the P4Pro was pointing north at launch?
The reason that the PhantomHelp converter doesn't show the switch to GPS+ATTI is that it uses additional data to determine ATTI vs GPS+ATTI. Specifically, the gpsIsUsed signal is used. If this value is FALSE then the flyCState is reported as ATTI even though it may actually have the value GPS+ATTI. Makes sense to me but maybe there is a subtlety that will yield the reason for this incident.
@KevMo Photog would it be possible to provide the .DAT for this flight. Look here to see how to do that. It'll be large so the .DAT will need to be Dropboxed or GoogleDrived first and then provide a link here.
@msinger
Similar threads
