Another episode of un-commanded YAW to the left - ruined my timelapse

[Deleted member 94047]

@sar104, @BudWalker; if I may throw another DAT file in to the mix, perhaps it may help. This is from a flight the next morning where a similar but less pronounced yaw movement happened (edit - upon reviewing the pictures I was taking at the time, it looks more or less the same as the previous one - about 30 degrees is my guess). This time, I let it complete and only intervened after the AC settled on its own. I think the uncommanded yaw happened around four minutes (after motor start).

 

[sar104]

@sar104, @BudWalker; if I may throw another DAT file in to the mix, perhaps it may help. This is from a flight the next morning where a similar but less pronounced yaw movement happened (edit - upon reviewing the pictures I was taking at the time, it looks more or less the same as the previous one - about 30 degrees is my guess). This time, I let it complete and only intervened after the AC settled on its own. I think the uncommanded yaw happened around four minutes (after motor start).

That does appear to be a very similar event.



In this case, however, IMU1 had accurately tracked yaw up to that point. The offset between the compass and IMU1 and IMU0 is just the local declination. I have no idea why it is doing that.

 

[Deleted member 94047]

That does appear to be a very similar event.

View attachment 82085

In this case, however, IMU1 had accurately tracked yaw up to that point. The offset between the compass and IMU1 and IMU0 is just the local declination. I have no idea why it is doing that.

Did an IMU switch happen at the end of the reconciliation?

 

[sar104]

Did an IMU switch happen at the end of the reconciliation?

Not that I can see, but I'm less confident than @BudWalker that the active IMU can be identified.

 

[Deleted member 94047]

Ok, thanks. I was just thinking about the discussion you two had regarding this sort of thing happening prior to an IMU switch. I think I will contact DJI, let's see what they have got to say about this.

 

[BudWalker]

That does appear to be a very similar event.

View attachment 82085

In this case, however, IMU1 had accurately tracked yaw up to that point. The offset between the compass and IMU1 and IMU0 is just the local declination. I have no idea why it is doing that.

I have no idea why it is doing that. Me too.

According to the WMM the geoDeclination at the launch site is 2.7° That's shown by the IMU(0) Yaw/magYaw separation until about the time the motors start. It looks to me like the goal of the maneuver at 247 secs is to reconcile the IMU(1) Yaw/magYaw separation - IMU(1):Yaw is held constant while the AC rotates. The result is that both IMU Yaw/magYaw separation converges to about -12.5°.

I looked at a few other MA .DATs that I have laying around. It's not that unusual to see an adjustment being made so that the Yaw/magYaw separation is the same for both IMUs. Here's an example from

Lost drone in a huge rock

Hello Guys, greetings from Brazil! First of all, If this is not the right place to post, my appologies. This is/was my first DJI, I have flown 17h and 223km with it and never felt so powerless as this time. I have taken the time to read 2 complete threads about losing drones in windy...

forum.dji.com

The pilot wasn't asking about any uncommanded Yaw - the incident in that flight occurred much later. But, it looks to me like some uncommanded Yaw occurred. It's difficult to see since there was control stick input at the same time. @Doppler 's .DAT is mostly hovering so the uncommanded Yaw is more pronounced.

Am I ever going to get to the point? Probably not. I'd guess the root cause is that some relatively large errors creep into IMU(1):Yaw. The sensor data seems OK since both gyros are very close.

I'm ready to let go of the idea that it's about getting ready to switch from one IMU to the other.

 

[BudWalker]

@Doppler you might want to consider doing a factory reset. I'm reluctant to make this suggestion since I've never done a factory reset myself. Maybe there are some other members that have done this and could offer advice on this suggestion.

None of the available data in the .DAT point to a HW problem. It could be a FW problem. Have there been previous flights where you would have seen this problem? Since this is only the 17th .DAT it's possible that you just haven't noticed this problem on a relatively new MA.

If it's not HW and not FW then that leaves the possibility that some configuration parameter has been altered in a way that exposes a FW problem. Presumably a factory reset would resolve this.

@sar104

 

[Deleted member 94047]

@Doppler you might want to consider doing a factory reset. I'm reluctant to make this suggestion since I've never done a factory reset myself. Maybe there are some other members that have done this and could offer advice on this suggestion.

None of the available data in the .DAT point to a HW problem. It could be a FW problem. Have there been previous flights where you would have seen this problem? Since this is only the 17th .DAT it's possible that you just haven't noticed this problem on a relatively new MA.

If it's not HW and not FW then that leaves the possibility that some configuration parameter has been altered in a way that exposes a FW problem. Presumably a factory reset would resolve this.

@sar104

Thanks @BudWalker, I will certainly take that as good advice. I don't risk anything by doing that. BTW, I flew three batteries today and there were no yawing issues I could see. But I haven't seen the DAT files yet. Will review them tomorrow.

As for this being ba new MA, it is not that new. The DAT file numbering reset to 00 a while back, after it hit 99 I think. As most MA owners would attest, I have experienced the occasional slight yaw immediately after take off in the past. But never this late into the flight as I did the last few flights. I will do a factory reset if this persists. Thank you again for all the help.

 

[Deleted member 94047]

I went through the three DAT files from yesterday's flights and all seems to be well with the MA. I am guessing these are the parameters I should be looking at (@sar104, @BudWalker could you confirm these please - also, if you want to look at the DATs, I would gladly provide them)


I did an IMU calibration the day before, I am not sure if that is what fixed the issue. Will hold off the factory reset for now.

 

[pj530i]

How do you look at DAT files from a MA?

I can view the txt log from DJI Go 4 with CsvView, but the DAT files I get from the drone itself with DJI Assistant 2 are encrypted or something?

I had an incident a couple days ago where the drone did almost a complete 360 revolution without me touching the sticks. In the same flight there were also a couple of yaw twitches during video recording that ruined a shot. I'd like to see if it was this yaw issue everyone's been having or maybe just a stick calibration issue. I calibrated the sticks after that flight and did a quick flight without any further issues but I was pressed for time so I think that's inconclusive. I've also had several incidents in the past couple months where the drone will yaw on its own, mostly right after takeoff.

Has anyone had any luck getting DJI to acknowledge/fix the problem?

 

[sar104]

How do you look at DAT files from a MA?

I can view the txt log from DJI Go 4 with CsvView, but the DAT files I get from the drone itself with DJI Assistant 2 are encrypted or something?

I had an incident a couple days ago where the drone did almost a complete 360 revolution without me touching the sticks. In the same flight there were also a couple of yaw twitches during video recording that ruined a shot. I'd like to see if it was this yaw issue everyone's been having or maybe just a stick calibration issue. I calibrated the sticks after that flight and did a quick flight without any further issues but I was pressed for time so I think that's inconclusive. I've also had several incidents in the past couple months where the drone will yaw on its own, mostly right after takeoff.

Has anyone had any luck getting DJI to acknowledge/fix the problem?

You won't able to decrypt the DAT files from the aircraft, but the DAT files on the mobile device are readable.

 

[pj530i]

You won't able to decrypt the DAT files from the aircraft, but the DAT files on the mobile device are readable.

ah, found them, thanks.

I'm certain the no-input yaw occurred between 11m and 12m but all the changes in yaw around that time appear to coincide with rudder input. That indicates to me the problem is with the controller. I am 1000% sure I watched the drone rotate in place for 2+ seconds towards the end of the flight while the drone was less than 75' away and my thumbs were no where near the sticks.

There is a smooth, extended change in yaw around 11m 22s, and the position of the drone on the map at that time aligns with my memory of when it happened, but the log indicates there was corresponding rudder input. Attaching log in case anyone wants to take a look..

 

[Deleted member 94047]

I haven't looked at the DAT, but that sounds like a remote calibration issue.