Mirek6
Second Officer
Flight distance : 609724 ft
Canada
Offline
|
James,
I am assuming that your second record is the one which shows fly-away.
First 40 seconds uneventful. Than interesting things start to happen:
39.1s at 29 m wind from East picks up.
47.5s at 45 m wind from EEN – quite strong (I see it through Spark’s roll and pitch)
53.7s at 58 m wind strengthens to about 30 km/ hour from EEN.
Spark has trouble keeping up position and starts drifting to SSE. It is being pushed to south by southerly vector of the wind yet it is overcompensating to the East (irregular gusts from easterly direction which are suddenly dropping at about 56 seconds).
This is all happening while commanded position stays the same (no input from sticks in any direction but up).
Since Spark should stay put but found itself in a wrong physical position, it throws yaw error at 58 sec.
By pure inertia – since wind eased for couple of seconds, Spark lurches to the North, passes its commanded position while trying to break.
Within 2 seconds it follows up with compass error which shows that Spark’s systems are confused. It knows that commanded position is different than physical position, it knows that it is hunting and is in a wrong spot but it does not know which system failed (compass, IMU, or GPS).
To protect integrity of flight, it drops to ATTI. This is correct reaction.
Since its pitch is still to the North, it continues this way for another 5 seconds while you instruct it to gain altitude until it reaches 76 metres. At the same time wind from the East starts to blow harder again pushing it to the West.
At 1m 4s you push throttle down instructing Spark to lower its altitude – it obeys.
You do nothing to change its direction of drift.
You keep your stick down for less than 4 seconds. Spark dropped to 70 metres and is still drifting away.
You lose signal several seconds later while Spark is just 61 metres from you.
This pattern shows issues with Spark’s firmware.
Yes – wind was gusty and quite strong at the altitude you were flying. However, this does not justify total confusion of the Spark’s sensors and firmware. Spark should have detected the wind and should have warned you – it did not. Instead, its systems got confused by the wind and caused it to drop to ATTI.
To add insult to injury, 61 metres is nothing – you should have strong signal. Why did the connection drop? Did you go around the corner of your house and was not in LOS?
Give this case and log to DJI. According to me it is Spark’s firmware issue which calls for warranty replacement.
Mirek
|
|