disaster...crash what on earth happened? any ideas?

IMG_1943.PNG
 
Were you standing under anything? I flew mine from under s bridge and havoc ensued. It reassigned the home point randomly and kicked into ATTI mode. I had to manually fly back and land in 18 mph wind without hitting anything, or flipping on landing. Had all of the usual bystanders back away as I brought it in. White knuckle...


Sent from my iPad using PhantomPilots
 
Your wind data comes from HD, which as I understand their website, is derived from algorithms which factor motor performance, stick inputs and GPS data to infer a wind speed. If, as the OP reported, there was GPS error, then I'd be cautious about the wind speed and direction from HD...
Yes.
 
Yes indeed, I ignore the wind data, I have had too many spurious results.
I have found the wind data from healthy drones to correlate very well with actual observed conditions.

I made no suggestion about the accuracy/reliability of the depicted readings on this occasion. Simply an observation that strong gusts were recorded in HD.
 
I have found the wind data from healthy drones to correlate very well with actual observed conditions.

I made no suggestion about the accuracy/reliability of the depicted readings on this occasion. Simply an observation that strong gusts were recorded in HD.
OK, sorry if I mis interpretated, but I have no confidence in the HD data on wind.
 
OK, sorry if I mis interpretated, but I have no confidence in the HD data on wind.
No problem at all.... these wind readings are at the end of the flight right where the crash occurred. It's probably erroneous data, agreed. The most likely way for HD to derive wind data would be to compare flight controller commanded pitch angle to GPS reported velocity and compass heading. data obtained during the crash would be unreliable. The problem with Hd is there does not seem to be enough data captured (sample frequency) to be helpful in most cases.
 
There might be something to this regarding wind. If the wind was higher than the craft can overcome, it would get pushed around a lot. Looking at the logs a bit deeper the craft was seriously banking hard and the motors were ramped up the last 15 seconds of the flight. And if I'm reading this correctly the craft was angled AWAY from the tower
No problem at all.... these wind readings are at the end of the flight right where the crash occurred. It's probably erroneous data, agreed. The most likely way for HD to derive wind data would be to compare flight controller commanded pitch angle to GPS reported velocity and compass heading. data obtained during the crash would be unreliable. The problem with Hd is there does not seem to be enough data captured (sample frequency) to be helpful in most cases.
The angle of the craft before the crash would be INTO the wind for the wind readings from healthydrones to be correct. This is not the case according to the data I see.
 
There might be something to this regarding wind. If the wind was higher than the craft can overcome, it would get pushed around a lot. Looking at the logs a bit deeper the craft was seriously banking hard and the motors were ramped up the last 15 seconds of the flight. And if I'm reading this correctly the craft was angled AWAY from the tower

The angle of the craft before the crash would be INTO the wind for the wind readings from healthydrones to be correct. This is not the case according to the data I see.
Take a look at thrustAngle. It's computed from motor speeds. It can be viewed in either a SigPlayer or the GeoPlayer. In this incident it shows the AC being propelled towards the tower.
upload_2017-2-7_16-2-23.png

The yellow line is the direction the AC is being propelled by the motors. The green line is the AC heading.
 
There might be something to this regarding wind. If the wind was higher than the craft can overcome, it would get pushed around a lot. Looking at the logs a bit deeper the craft was seriously banking hard and the motors were ramped up the last 15 seconds of the flight. And if I'm reading this correctly the craft was angled AWAY from the tower

The angle of the craft before the crash would be INTO the wind for the wind readings from healthydrones to be correct. This is not the case according to the data I see.
I wouldn't like to be flying there today...


IMG_0288.PNG
 
Could it just be as simple as turbulence, caused by the tower? I'd imagine a 22 mph gust would cause quite a bit of friction on the surface of the tower and could make the air unstable. Just a guess, though.

p3s- fishing
p4 Black- Inspections
 
I would be interested to know if compass error and consecutive GPS disable can be caused by out of range readings, here banking too much possibly caused by wind (with airplanes the autopilot can disengage when it looses airspeed indicators).

Sent from my iPhone using PhantomPilots
 
Last edited:
Guys....

Im new to this, but I will 99% guarantee it wasn't down to wind, it was a calm day and there weren't any notable gusts. If there had been, I would not have flown it.
The drone was never out of range at all.....the signal strength was at 100% throughout and GPS had at least 15 satellites!

On a different note, DJI online chat told me to email europe tech support with various details so that I could arrange to get the drone returned for investigation.

I emailed them on Monday afternoon, twice. Ive had an automated response but nothing else.

How long do they take to contact you with details of what to do next? Its 3 days now and nothing...
 
Guys....

Im new to this, but I will 99% guarantee it wasn't down to wind, it was a calm day and there weren't any notable gusts. If there had been, I would not have flown it.
The drone was never out of range at all.....the signal strength was at 100% throughout and GPS had at least 15 satellites!

On a different note, DJI online chat told me to email europe tech support with various details so that I could arrange to get the drone returned for investigation.

I emailed them on Monday afternoon, twice. Ive had an automated response but nothing else.

How long do they take to contact you with details of what to do next? Its 3 days now and nothing...

We are in U.S., but have had significant delays in communication and responses to email... two things I can suggest:

- use the chat feature on their support site to stay in touch with them frequently (squeaky wheel gets the grease)

- if you haven't already, suggest posting on DJI's forum with similar information as this thread, include your CAS number and one of the forum mods may be able to help speed things up... not sure that's feasible since they've just celebrated Chinese New Year and probably have a backlog.

We initiated contact with DJI on an incident which happened on Christmas Day and do not yet have our P4 back... requires a great dose of patience dealing with them.
 
Guys....

Im new to this, but I will 99% guarantee it wasn't down to wind, it was a calm day and there weren't any notable gusts. If there had been, I would not have flown it.
The drone was never out of range at all.....the signal strength was at 100% throughout and GPS had at least 15 satellites!

Honestly, I don't think for a second that you had gusts over 20m/s. In weather like that you would have very little to no control over the drone. As far as I know HD is calculating wind from the drone speed and from the power required to achieve that. So something has caused your Phantom to receive bad data. It doesn't help if you have GPS "lock" but the data is corrupted. Also there is good chance your IMU's were effected. Everything points towards that tower. Close proximity to magnetic source will wreck havoc on your instruments. It's worth mentioning that with good data drones like Phantom 4 don't just go flying on their own. In general, I would go as far as to suggest that fly-aways are either cause of lack of GPS data or wrong home point. GPS is not bullet proof. When it fails many tend to panic and are incapable of flying their drone back in ATTI mode especially if it's windy.

In professional shoots I have seen high end drones loose their bearing when close to large structures. That is to be expected. One capable of flying manual mode will have no problems. If you want to avoid that there are two options. Either don't fly close to structures or invest in a high end system with fault tolerant differential (ground based unit) GPS. Phantom is awesome little drone but one can't expect too much from a system costing just 1.5K. Differential GPS system alone costs around 4K.




Sent from my iPhone using PhantomPilots
 
Honestly, I don't think for a second that you had gusts over 20m/s. In weather like that you would have very little to no control over the drone. As far as I know HD is calculating wind from the drone speed and from the power required to achieve that. So something has caused your Phantom to receive bad data. It doesn't help if you have GPS "lock" but the data is corrupted. Also there is good chance your IMU's were effected. Everything points towards that tower. Close proximity to magnetic source will wreck havoc on your instruments. It's worth mentioning that with good data drones like Phantom 4 don't just go flying on their own. In general, I would go as far as to suggest that fly-aways are either cause of lack of GPS data or wrong home point. GPS is not bullet proof. When it fails many tend to panic and are incapable of flying their drone back in ATTI mode especially if it's windy.

In professional shoots I have seen high end drones loose their bearing when close to large structures. That is to be expected. One capable of flying manual mode will have no problems. If you want to avoid that there are two options. Either don't fly close to structures or invest in a high end system with fault tolerant differential (ground based unit) GPS. Phantom is awesome little drone but one can't expect too much from a system costing just 1.5K. Differential GPS system alone costs around 4K.




Sent from my iPhone using PhantomPilots

I had a reading come back from HD with wind speeds at 300mph! :)


Sent from my iPad using PhantomPilots
 
  • Like
Reactions: WBN3

Recent Posts

Members online

Forum statistics

Threads
143,066
Messages
1,467,352
Members
104,933
Latest member
mactechnic