MM wants to go Backward at takeoff then settles out...

[MiniMan]

I just picked up a second MM. Was testing it indoors in bright light and as soon as takeoff it would go backwards fast and crash. I recalibrated everything (really thought it was IMU) and tested multiple times with a large towel ready to catch it. Same result.

I then decided I would try to correct it by pushing on the forward stick at takeoff. I was able to correct it and low and behold as soon as it achieves it's minimum indoor altitude (with is about 24") it levels out and will hover perfectly.

What could be causing the lurching backward at initial takeoff?

I also believe that if I don't do that initial forward correction it will continue backwards even at a higher altitude. I have not been able to prove that yet because of rain outside.

Thanks in Advance.

 

[TDZHDTV]

Look at the flight data in the app and see if there is a stick input, maybe the stick is damaged or needs calibration

 

[MiniMan]

Will do. But after I do the initial "Correction" I release the sticks and it hovers perfectly....

 

[sar104]

I just picked up a second MM. Was testing it indoors in bright light and as soon as takeoff it would go backwards fast and crash. I recalibrated everything (really thought it was IMU) and tested multiple times with a large towel ready to catch it. Same result.

I then decided I would try to correct it by pushing on the forward stick at takeoff. I was able to correct it and low and behold as soon as it achieves it's minimum indoor altitude (with is about 24") it levels out and will hover perfectly.

What could be causing the lurching backward at initial takeoff?

I also believe that if I don't do that initial forward correction it will continue backwards even at a higher altitude. I have not been able to prove that yet because of rain outside.

Thanks in Advance.

Compass interference at takeoff is the most common cause of that initial behavior, although the later correction is unusual. You will need to post the flight txt logs if you want anything other than wild guesses at this point.

 

[MiniMan]

Let me know if this is a correct file.. Thanks

 

[netsonic]

Yes, this is the correct log. I've uploaded it on phantomhelp log viewer, here is your link: DJI Flight Log Viewer - PhantomHelp.com
You can check stick inputs on the site, in the top right corner, just scroll through the events or timeline.

 

[sar104]

Let me know if this is a correct file.. Thanks

I don't see any anomalous motion in that log - the aircraft responded properly to the stick inputs:

 

[MiniMan]

Thank You. The "Issue" only occurs during the first approximately 2 seconds from the time of takeoff. That is when it starts to go quickly backwards and I give it close to full forward.

So if I am flying indoors is it normal for the flight mode to be P-GPS. I would figure it was flying in ATTI Mode (or Vision) where it is only using the bottom sensors to look at the floor?

 

[MiniMan]

So I was able to take it outside this morning. Full GPS, took it for a flight and a RTH. Worked perfectly. The bottom line is it lurches backward and to the left at takeoff. Which is much different from my other Mini which takes off perfectly straight up.

Indoors seems a bit worse than outdoors, but it may just be the perception of the space.

https://youtu.be/7pcMOA7Kpms

https://youtu.be/2UX5Tp3jKRw

 

[new2mavic]

...I am flying indoors is it normal for the flight mode to be P-GPS. I would figure it was flying in ATTI Mode (or Vision) where it is only using the bottom sensors to look at the floor?

The flight log indicates that the craft was in P-GPS mode with very few satellites so the accuracy was very poor. There must be a lot of windows around. At such low altitude, it should be using the bottom sensors for locating but during the transition from zero altitude to that where the cameras at the belly could focus properly, the craft should be flying in P-GPS mode. As there were insufficient satellites, the calculated GPS location might be drifting around driving the drone to chase after it. I have had the same experience when flying my M2P between tall buildings in the city. The drone was hovering OK before it suddenly flew backward by itself until I stopped it with appropriate stick inputs.

I would suggest you to try again at places with zero GPS signal or very good GPS signals. The symptom will probably go away.
Edit : just realized that you have already done so.

 

[BudWalker]

Mode = P-GPS aka flyCState = GPS_Atti is kind of a misnomer. P-GPS means that the FC is using either GPS or the vision system to determine position - it's not ATTI mode. In this particular flight neither GPS or vision was being used for the 1st 1.25 secs and the M2 was in ATTI mode. Then the vision system started being used and the switch to P-GPS was made. Coincident with this was the ultrasonic height reaching 0.5 meters.


@MiniMan I suspect the problem you're seeing is because of the transition to the vision system in low light conditions. Seems like you've already done a test in better conditions.

@MiniMan I noticed that the .txt didn't begin until 63 secs. How did this happen? Was the M2 powered up and motors once prior to starting the Go App?

 

[tinker]

frist check your props, if even one is damage or nicked it will do that

 

[Slot Tech]

Mine does this also flying indoors with prop cage on. I think it's just the bottom sensors kicking in to get a position hold on the drone. Flys great after the small slide back on take off.

 

[tinker]

you may be right, nenver flown inside yet

 

[MiniMan]

Yes, but my other one does not do this Takes off perfectly straight up?!?!?

 

[lannes]

Yes, but my other one does not do this Takes off perfectly straight up?!?!?

My mini goes straight up, whether indoors or outdoors
Indoors make sure there is as much light as possbile

 

[tinker]

have you added any thing to the unit (light, ect) and once agian check your props, also try it out side with better light

 

[GerdS]

In case the Mavic always will drift to the same direction after takeoff and getting the minimum altitude for the optical flow sensor taking over, you should try an IMU calibration.