Rookie MA2 owner, decided to fly around the sunset time. Started off good but when it got darker there were warnings about no vision systems because of low light, guess I did not comprehend the significance of it at the time (or what can happen down the road). It was all good until landing. I cancelled auto landing because the AC tried to land about 4-6ft off from the take-off location. I tried to bring it in to land manually at the back porch of my unit. I got it down to knee level and tried to position over a landing pad, which turned out to be a mistake (should have landed immediately in retrospect). I tried to position it perfectly by flying to the side, but it kept going - did not stop when I released the stick. I corrected so it wouldn't hit a glass slider door but it did not stop either and "fortunately" the propellers hit my shorts just above the knee level. The AC fell down onto a weather-beaten wooden decking. It ended up "perfectly" onto all 4 legs, all propellers were folded, it was still powered on and I had the picture on the phone screen. I powered it off. I realized later it must have lost the GPS (no wonder that close to the building) and (with no vision systems) switched to ATTI mode. Can anyone please confirm that this is in fact what happened? Logs are attached. Did I miss the warning (about ATTI mode) in the DJI Fly app or there wasn't one? I was looking at the AC trying to position it.... There are no obvious signs of damage to either propellers or the body. What should I check before flying it again? Thank you for all the help.
MA2 trying to land in the evening - ATTI?
- Thread starter prz3mk0
- Start date
Your flight data looks like this:
DJI Flight Log Viewer | Phantom Help
Upload and view flight logs from your DJI drones
www.phantomhelp.com
The drone did not enter Atti Mode and maintained GPS for the whole flight.
What's the back porch construction?
If it was a reinforced concrete surface that could account for a little wandering.
The logs, if complete and correct flight, should tell the rail.
Always think about why things may be happening.
If your drone tries to land, commanded by you in any fashion, in a place you know is unsafe then intervene. If the place is safe let it land.
I suspect you are right about why the ATTI mode.
If the flight mode had been otherwise the drone can be expected to land within 6-12 “ of TO point.
BTW, is you apt back porch reinforced concrete.
Always think about why things may be happening.
If your drone tries to land, commanded by you in any fashion, in a place you know is unsafe then intervene. If the place is safe let it land.
I suspect you are right about why the ATTI mode.
If the flight mode had been otherwise the drone can be expected to land within 6-12 “ of TO point.
BTW, is you apt back porch reinforced concrete.
Thank you for looking into this
back porch is wooden decking, it was wandering a little on it's own when I got it over the porch (even before I tried to position it perfectly), which never happened before (but this was also first time w/o vision systems), when I did try to position it the AC didn't stop going to the side when I let off the stick, which in my rookie mind suggests ATTI
I am looking in csvview and I see a lot of
[L-GEO] TFR read failed! res :0xffff3072 --what is this?
also
[L-FLYMODE] CTRL rc pause gps_atti --does that mean ATTI?
back porch is wooden decking, it was wandering a little on it's own when I got it over the porch (even before I tried to position it perfectly), which never happened before (but this was also first time w/o vision systems), when I did try to position it the AC didn't stop going to the side when I let off the stick, which in my rookie mind suggests ATTI
I am looking in csvview and I see a lot of
[L-GEO] TFR read failed! res :0xffff3072 --what is this?
also
[L-FLYMODE] CTRL rc pause gps_atti --does that mean ATTI?
Thank you. Where should I look in CsvView for the signals to determine if:
vision was used?
gps was used?
magnetic interference?
I am guessing that low light vision system unavailable (forward, backward, downward?) is not the same as putting it in S mode (forward and backward disabled, but downward is still active). Am I guessing correctly?
Thank you.
Sorry .. I can't help you with CsvView.
I usually refer to the Verbose CSV which you get from a link in the report shown in post #2
But there's something wrong with that, this time and the file is empty.
But as mentioned in #2 ..
Your satellite numbers dropped, but not lower than 8 and GPS health was never lower than 4/5.
The drone did not enter Atti Mode and maintained GPS for the whole flight.
Thank you. I see all the pertinent values would be in Verbose CSV but like you said it is empty. I tried another log and it was empty also.
Strange thing I flew it at night before but never had vision system unavailable warning due to low light (and it was darker and later at night).
I will chalk it up to lack of stability w/o vision system since it had GPS. I will replace the propellers just to be safe.
Thank you for all you help.
I have checked the data of the last 10 seconds of the flight but could not see any obvious uncontrolled movement of the craft.
In the chart below , it can be seen that the roll angle of the craft changed in good correspondence to the aileron input except for a brief moment of 0.5 sec ( circled part ) which could be due to some local air disturbance.
The chart below shows the change in velocity measured by GPS and IMU ( the result of fusing all sensor inputs ). Again, the direction of speed change matches the control input so the craft was generally under control. The craft has been moving at rather low speed until the last 2 seconds during which full left aileron has been applied repetitively and the craft accelerated to the same direction. Some right alieron was applied in the last 0.6 sec but it was too mild and too late to avoid the crash.
In the chart below , it can be seen that the roll angle of the craft changed in good correspondence to the aileron input except for a brief moment of 0.5 sec ( circled part ) which could be due to some local air disturbance.
The chart below shows the change in velocity measured by GPS and IMU ( the result of fusing all sensor inputs ). Again, the direction of speed change matches the control input so the craft was generally under control. The craft has been moving at rather low speed until the last 2 seconds during which full left aileron has been applied repetitively and the craft accelerated to the same direction. Some right alieron was applied in the last 0.6 sec but it was too mild and too late to avoid the crash.
Last edited:
Thank you! This is exactly what I was trying to visualize in the end - compare the AC roll with the input from the sticks. That uncommanded roll (towards the glass slider door) took me by surprise. Also, is it me or did the AC responded slower to the input from the sticks at the end?
The craft responded quickly to stick input as indicated in the chart. The blue line is the roll angle and the red line is the aileron input.
The data indicates that the drone was basically flown into the obstacle under stick command.
Canuk
Well-Known Member
- Joined
- Apr 19, 2018
- Messages
- 852
- Reactions
- 806
- Age
- 59
You should be could to fly as long as you didn't damage any props. Any dents or scratches on the props, change them out. Don't fly with damaged props...it's like sitting on an out of balance washer machine
I did not damage the props but did change them anyway to be safe and since I had 2 more sets from fly more combo. I flew it twice today during daylight for the first time since the crash and it seems fine, hoovers in place and stable. I am keeping an eye on it more instead of watching the screen especially during landing after RTH. I did notice it pitched back in the same direction as it rolled during crash so I took control and landed manually, no problem but VPS was working this time around.
The only new? thing I noticed is that after I press the power button on the battery to power off the AC the camera moves to the right. I do not recall this happening before, although it might have. Is this normal? The camera centers fine after powering on or when pressing FN button twice, it just moves to the right during powering off.
Anytime your landing in low light, your sensors will degrade as they need light to process the data to avoid obstacles. You will have to manually land
There is no indication that you were using the LED landing light. So just to double check, were you using the landing light? For those with night-time experience with the MA2, how effective is the landing light in providing enough illumination when using a landing pad on a high-contrast background?
The LED light came on during landing automatically, I turned it off while bringing it inside the porch and did turn it back on while I got it lower before trying to position it perfectly over the landing pad and before the uncommanded pitch.
I don't believe the landing lights are sufficient for vision sensor. At least it isn't for the M2. It didn't even help much using the gimbal camera to manually get on landing target via FPV.
As stated, without vision sensor assistance providing precision landing and relying totally on GPS, landing can be as far off as 6ft from home/take off point.
As stated, without vision sensor assistance providing precision landing and relying totally on GPS, landing can be as far off as 6ft from home/take off point.
Similar threads
DJI Drone Deals
New Threads
-
Latest version of FlyApp for RC Pro??
- Started by Filmarik
- Replies: 6
-
Is there a 1 click way to delete all photos saved to the smart controller?
- Started by matt@tcdrones
- Replies: 1
-
-
-
Air 3 Trift Suspension Bridge in the Alps: A Decade of Change- Started by GekoCH
- Replies: 0
Forum statistics
