DJI A2 X8 yaw issue

[Dagger One]

Hi guys.

I was hoping to hear your feedback on this issue I experienced recently with the A2..

I was flying with a new gimbal when I noticed that it made an extra rudder input when I released the rudder stick back to center position. It is really minor and it still flies well but still it is pretty uncomfortable to have any extra movements than your own.

I just got a little clip of it when it did this for the first time and you can see it here:

It is hard to notice this on a video but I am yawing slowly to the right and then center the rudder stick but then it continues the yaw with a bit more aggressive yaw to right just a few degrees.

So I updated the firmware of the A2 because I still had the 2.4 and also calibrated the IMU and the GPS and made another test flight today. The issue is still there.. It still makes thous extra movement but it seems to only do it when I yaw to the right.

Here is the testflight from today (notice the sound the motors make at the end of yaw to the right):

What I think might be wrong is motor alignment. But still it is really close to be perfect. And I'm not sure how sensitive the A2 is to motor alignment.

But please let me know what you think!

All thoughts are greatly appreciated

 

[Ti@goo]

I have a Naza V2 and that happens to me also. Wether I YAW to the right or left it also gives an little input to one of the sides..

 

[maxwelltub]

Sounds silly but do a motor test on you bench and make sure all your motor inputs are correct.

 

[Dagger One]

I have a Naza V2 and that happens to me also. Wether I YAW to the right or left it also gives an little input to one of the sides..

Good to know that it does not only happen to me.. You are flying X8 as well? I did take all the props off and tried to see if it would make this extra input while on the ground without the props and it does not. So I really doubt that this is because of the motors or ESCs.
I don't feel comfortable flying it until I find the cause of this issue. I sent DJI a support request about this but highly doubt that they will ever answer...

 

[Dagger One]

Sounds silly but do a motor test on you bench and make sure all your motor inputs are correct.

I somehow think that this is more likely to be a issue for the flight controller than motors/ESCs. But how would you make sure that the motor inputs are correct?

 

[maxwelltub]

Watched the video again, this is totally normal behavior. Also motor alignment is more of an issue on tubular arms and with older flight controllers. That noise is the motors speeding up to stop the rotation. If you think about air as a fluid picture your aircraft as a boat. if you paddle in a circle you have to back paddle to stop that movement. X8s in my opinion fly better with using the two stacked quads method. DJI has all of the counter clockwise props on top and all clockwise on bottom. If you put motor 1 on top, 2 on bottom, 3 on top for bottom and so on it will fly better, and the flight controller doesn't know the difference as long as 1-4 are counter and 5-8 are clock. But this reaction to end of yaw is normal.

 

[maxwelltub]

I somehow think that this is more likely to be a issue for the flight controller than motors/ESCs. But how would you make sure that the motor inputs are correct?

there should be a motor test button in the software. Can't recall what page but its bottom right. I have mixed motor inputs before and it flew a little funny.

 

[Old Man]

X8s in my opinion fly better with using the two stacked quads method. DJI has all of the counter clockwise props on top and all clockwise on bottom. If you put motor 1 on top, 2 on bottom, 3 on top for bottom and so on it will fly better, and the flight controller doesn't know the difference as long as 1-4 are counter and 5-8 are clock.

But this reaction to end of yaw is normal.

Dead on the money. Flight responses with an X-8 are much better as described above. The DJI method is not as stable nor does it provide the best control response.

Extreme fine tuning of PID's will help minimize the yaw issue but a small amount of yaw will likely remain. The larger in mass the aircraft the more pronounced this can be. Large propellers with motors and ESC's that are being under driven in voltage will also be slower in counter control response. The rate at which a MR is yawed also contributes to the condition. Something else to consider is checking the transmitter trims and channel centering in the transmitter.

 

[Dagger One]

Thank you guys for the feedback, I really appreciate it

The thing is that when I yaw to the right it seems like it wants to stop the yaw motion as you described but in the wrong direction.. If you look at the shorter video again you can see that the movement from 10 seconds in is just her own. I am yawing slowly to the right and then finish the yaw when it makes this quick extra yaw to the right. I also noticed as the "Old Man" said that the yaw movement was bigger with the gimbal than when it only flew with its batteries in the second flight. This yaw issue never really happened when yawing to the left.
Also one thing which I haven't yet mentioned. When I flew with the gimbal and first experienced this it got stuck in a very slow yaw motion which I had to counter with a slight rudder input. That is something which had happened to me before when I flew a fairly large quad with a superx flight controller and the alignment was not correct.
I will do the motor test as you recommended and another testflight to see if I get better results or better understanding of the issue.

Are you guys sure that this might not be because of motor alignment? It is as you mentioned an square tube arms so if the motors would be off in alignment it would be really little.

The transmitter is working as it should, no trims or anything. Also I have been flying another multi-rotor with the same transmitter without any problems

Thanks again

 

[Old Man]

Motor alignment is not hard to check, but can be time consuming. Probably the greatest drawback of round booms is the ease with which they permit misalignment. It's worth a look but likely less contributory to the issue than you think.

In the video it was difficult for me to discern between the counter yaw you described and what might be thumb input. That's kind of important because if you inadvertently input any opposite yaw input the aircraft will demonstrate that if the PID's are too low. Something to consider is a bit more tuning to see if much of the counter yaw can be dialed out. Also for consideration is when the aircraft is flown in different configurations the gain settings will work better for one than another. Gear down and gear up has an effect on tuning. This is an area where the fixed gear crowd like those on Cinestars have an advantage because the gear is always in one place but rotates with the gimbal so there's no change in mass position to impact tuning. Tuned without a gimbal and you have an aircraft that will perform best without a gimbal. Adding a low slung gimbal changes mass distribution, which in turn impacts FC tuning. The FC will need to be a bit more aggressive to deal with the additional mass and pendulum effect of the gimbal. That rule of a body in motion tends to remain in motion unless another force is applied to counter the motion is really visible with multirotors.

Check the motor alignment for sure but perhaps a little more time and effort with the gains might be in order. There's also the condition where all the tops spin one way and the bottoms another. Your call with what you choose to do with that but no matter what you are doing, only do one thing at a time to avoid polluting the results of a change with items that make defining the solution impossible. BTW, Nice build!

 

[SleepyC]

Have this problem as well. I'm going to do the 2 stacked quads (counter rotating) method. It's pretty easy in the custom motor mixer, just reverse every other motor on top and bottom, and make sure each arm is counter rotating.

 

[Dagger One]

Thank you very much Old Man for the detailed reply

I know it is hard to see on the video what is my input and what is her own. So I made another video which should clarify it:

The gains are actually set pretty high, if I did get them any higher it would begin to oscillate. I might try the two quad setup but I'm still a bit skeptical about that.

 

[Dagger One]

Have this problem as well. I'm going to do the 2 stacked quads (counter rotating) method. It's pretty easy in the custom motor mixer, just reverse every other motor on top and bottom, and make sure each arm is counter rotating.

Please let me know how it goes!

 

[maxwelltub]

If you have an osd check to input on the motors to make sure they are all giving even inputs. Also try in Atti mode. I dont think its motor alignment because it looks like those booms are not round. Not sure what your experience with big multirorotrs is but this may just be normal behavior. Also when you say gains have you tried adjusting just the yaw gain, that shouldn't give you oscillation even if its cranked up. you can also try lowering it see if that makes any difference.

 

[Dagger One]

Thanks for the reply

My experience with big multirotors like this one is a bit limited.. I tried to get the OSD data but wasn't able to do so. I find the software to be really complex and I never could read anything from it. I tried to find some instructions on the web but couldn't really find any.. I only wish it was as simple as the Super x..
I will play a little with the gains.

 

[maxwelltub]

Ya the software sucks, and I've tried search and posting on this forum for anyone who can explain the best ways to use the software but with no luck. If you can get the data open find the motor input graph. If for some reason you have uneven torch it may be fighting hard to yaw one way then the other way. To me it looks normal but I can feel it so don't want to miss lead you either.

 

[Dagger One]

Yeah I can't understand why they haven't updated or just made a new software. But I will try again and hopefully get something out of it. Thanks again!

 

[maxwelltub]

because most of their users are prosumer who wouldn't use it.

 

[Old Man]

My guess has one of the yaw PID's a little too aggressive.