Loss of control after 1st takeoff of the day.

[Beamer]

@sar104 , thank you. I can't say that I understand what all that means, but it tells me that I'm not just crazy. I will look into the IMU cal procedure when I get the time.

Can you think of any reason why, with an out of cal IMU, I wouldn't see this issue across every flight? Seems odd that a cal issue would show itself intermittently.

 

[paulatkin73]

so, what sar104 meant - calibrate both IMU and compass, repeat flights, carefully, 2-3 times and then look (or post here) those .dat files to see if IMU signals still going to differ between IMU0 and IMU1. most likely they will not - as it was due to the lack of a calibration - but if they will - that means a factory defect and you will need to open a warranty case.

 

[BudWalker]

This is probably not a compass issue - it's definitely an IMU problem. IMU0 and IMU1 (active) disagree on virtually everything. The rate gyros, in particular, are completely different, resulting in completely different solutions for the aircraft attitude. It's pretty much impossible to tell what is correct.

View attachment 77288

I would start with an IMU calibration, and then another test flight.

@BudWalker - any thoughts on this?

Although IMU(1) is the active IMU it seems that IMU(0) has correct data.

The light green trace shows the IMU(0) path and the darker teal green shows the IMU(1) path. @Beamer 's description matches the IMU(0) path.

Hard to say, but it seems possible that if IMU(0) had been active there would not have been an incident. Still thinking about this...

 

[paulatkin73]

Hard to say, but it seems possible that if IMU(0) had been active there would not have been an incident. Still thinking about this...

it is peculiar - but in the majority of posted logs, on most of those issues it seems it was the IMU1 that was either producing wrong data or stopped producing data.
where is IMU1 physically located on the board?

 

[sar104]

Although IMU(1) is the active IMU it seems that IMU(0) has correct data.
View attachment 77291
The light green trace shows the IMU(0) path and the darker teal green shows the IMU(1) path. @Beamer 's description matches the IMU(0) path.

Hard to say, but it seems possible that if IMU(0) had been active there would not have been an incident. Still thinking about this...

Maybe - but even IMU0 shows significant disagreement with the magnetometers:



That's what left me wondering if either could be trusted.

 

[Cymruflyer]

Sounds like he has the MAX8 version of the Mavic 2 PRO...

 

[sar104]

Sounds like he has the MAX8 version of the Mavic 2 PRO...

Ouch.

 

[Cymruflyer]

You guys just blow my mind with your knowledge of this stuff, I sit here in awe!

 

[Beamer]

Ouch.

...???...

Oh THAT Max 8. That's messed up, lol.

 

[BudWalker]

Maybe - but even IMU0 shows significant disagreement with the magnetometers:

View attachment 77298

That's what left me wondering if either could be trusted.

The Yaw/magYaw separation was much worse for IMU(0).

IMU(0) Yaw/magYaw separation was OK until about 12 secs. But, by comparing the .txt Vel[N,E, D] velocities with the .DAT velocities it's clear that IMU(1) was the active from the start. It didn't switch to IMU(1) at 12 secs because of the IMU(0) Yaw/magYaw separation.

 

[sar104]

The Yaw/magYaw separation was much worse for IMU(0).
View attachment 77302
IMU(0) Yaw/magYaw separation was OK until about 12 secs. But, by comparing the .txt Vel[N,E, D] velocities with the .DAT velocities it's clear that IMU(1) was the active from the start. It didn't switch to IMU(1) at 12 secs because of the IMU(0) Yaw/magYaw separation.

Right - but if IMU0 were operating normally then it should have continued to agree with the compass.

 

[yankfanp]

I'm hoping this is pilot error, but maybe it is faulty hardware/software?

I have been intensely surprised and crudely awakened to the realities of drone piloting on more than one occasion now when my Mavic 2 Pro attempted to fly off (twice now), shortly after the first takeoff of the day. Rebooting the drone stopped the issue at the time.

The first time, the drone barely returned, having little control with the joysticks maxed out. If I released the sticks (to zero positions) it continued to attempt to rise and fly forward.

The second time was nearly the same, but with even less control and I ended up doing a slow ditch, barely controlled, crash into the sand.

I never tried to let the drone continue it's flight and then RTH, so I'm not sure if that would have even helped (just an afterthought).

Is it possible that the controller was in a non-zero position (slightly pressed) at boot to cause this?

I doubt it, but maybe compass calibration could cause this?

Should I send it in to have it looked at? It happened maybe 1 in 10 flights, but now I fear it happening every time I boot up at a new location.

If anyone has ideas to consider which might be the culprits, please let me know. Obviously I'd like to avoid this in the future.

I wish I could really help you out with something technically wise but all I can offer is that I have been flying my M2P since September of 2018 and after tons of flights, I have never had this happen. In conclusion, I would be requesting some definitive investigation by DJI in some manner.

 

[Cymruflyer]

Never had a problem with my 2 Pro either. Glad I have never experienced that. I always wait until all GPS satelites are up and the announcement that the home point has been updated, before ever taking off.

 

[Bill combs]

I'm hoping this is pilot error, but maybe it is faulty hardware/software?

I have been intensely surprised and crudely awakened to the realities of drone piloting on more than one occasion now when my Mavic 2 Pro attempted to fly off (twice now), shortly after the first takeoff of the day. Rebooting the drone stopped the issue at the time.

The first time, the drone barely returned, having little control with the joysticks maxed out. If I released the sticks (to zero positions) it continued to attempt to rise and fly forward.

The second time was nearly the same, but with even less control and I ended up doing a slow ditch, barely controlled, crash into the sand.

I never tried to let the drone continue it's flight and then RTH, so I'm not sure if that would have even helped (just an afterthought).

Is it possible that the controller was in a non-zero position (slightly pressed) at boot to cause this?

I doubt it, but maybe compass calibration could cause this?

Should I send it in to have it looked at? It happened maybe 1 in 10 flights, but now I fear it happening every time I boot up at a new location.

If anyone has ideas to consider which might be the culprits, please let me know. Obviously I'd like to avoid this in the future.

have you tried to recalibrate the r.c.?

 

[sar104]

have you tried to recalibrate the r.c.?

Just to avoid any potential confusion, the RC is not the problem.

 

[Beamer]

I submitted a ticket with DJI for now. I suppose we'll see what happens. As of now, they plan to look over the Flight Records in question. I asked for this prior to returning the drone. Unfortunately, it takes them up to 3-5 days to process everything.

I will let ya know how it goes and what they come up with.

Thanks to everyone for their input, I really appreciate it. I tend to second guess myself on things like this, so it's good to see that there's a bit of a consensus that I'm not just imagining things.

(I sure wish I understood the different sensor data that is included in the flight records, I'm sure I'll gain a better understanding of everything with time.)

Cheers!

 

[sar104]

I submitted a ticket with DJI for now. I suppose we'll see what happens. As of now, they plan to look over the Flight Records in question. I asked for this prior to returning the drone. Unfortunately, it takes them up to 3-5 days to process everything.

I will let ya know how it goes and what they come up with.

Thanks to everyone for their input, I really appreciate it. I tend to second guess myself on things like this, so it's good to see that there's a bit of a consensus that I'm not just imagining things.

(I sure wish I understood the different sensor data that is included in the flight records, I'm sure I'll gain a better understanding of everything with time.)

Cheers!

I would expect that they are going to determine hardware failure, and offer to replace the aircraft.

 

[Beamer]

That would be a plus, because its kinda nerve-racking every time I first turn it on now. Hopefully that won't last.

 

[Beamer]

Well, I got a response from DJI, and as promised, I said I'd share. (I wrote them a lengthy response...)

The aircraft was piloted under GPS mode after taking off, but the GPS signal was relatively weak(only 3bar 8 sats available), thus the aircraft was unable to hover in place perfectly.

According to the analysis, the incident was not caused by any product malfunction.

We can still proceed with the repair, however, you will cover the repair charges.

Your understanding and cooperation will be highly appreciated.
Thanks a lot for choosing DJI! Our goal is to provide you the best customer service you only deserve.

What do you suppose the 'repair' is they are offering that I pay for?

 

[sar104]

Well, I got a response from DJI, and as promised, I said I'd share. (I wrote them a lengthy response...)



What do you suppose the 'repair' is they are offering that I pay for?

That's not the right answer, and their analysis was clearly incomplete. Did you send them the DAT file? If not, then you need to do that and, either way, point them to the large disagreement between the two IMUs.