Scary moment!.. Almost a goner

[Noël]

Today I've had my true 1st really scary moment!

I’m on holiday in Austria at the moment, and today I was following my son up the T-bar lift with my P2V.
No problems, and at the top of the run (about 300 meters away) I rotated the P2V slowly to the right while it was hovering in GPS mode.
Having done this I wanted to get in front of my son to film him coming towards me.

This is when the **** hit the fan!

Whatever input I gave on the RC the P2V went (looking at it) to the left!
As the P2V was 300 meters away, and in front of a mountain covered with snow, it was also hard to see it visually.
(I got really lucky at 12 seconds into the video..)

Facts:
- Before takeoff the P2V had GPS lock.
- The compass was calibrated one day before at the same place.
- At the moment of the trouble the P2V was flying in GPS mode.
- IOC had no effect what so ever!
I am not sure if the P2V had a course lock as it was to sunny to see the green rapid flashes before takeoff.

The only way I could get the P2V back was by going into Atti mode!

Small part of the video.
http://youtu.be/i8uu_6TVSnA
At 1:03 I managed to get back control by going into Atti mode.

 

[pault]

No idea what caused your problem but a good reason to have it in NAZAM mode. It sounds like without ATTI you would have lost her. As it was you were lucky not to hit any of those trees !

 

[pawnmeister]

Please if you would, enlighten those of us contemplating NAZA mode, how it helped you get your Vision back to you. Its this type of story that is pushing me to make the jump from Vision mode after about twenty flights. Thanks in advance for any info.

 

[iDrone]

The theory is that for whatever reason when the GPS & compass are in conflict with eachother the autopilot may try to "correct" is position by flying off to the new position where it thinks its supposed to be. If you've set Vision to NAZA mode and assigned Vision's switches for GPS and ATTI, you can disconnect the GPS from the autopilot and regain control of the aircraft by switching to ATTI mode.

iDrone

 

[Pull_Up]

Glad it was ok. To see if we can get to the bottom of it, can I just clarify, as there's a bit of a contradiction in what you said:

Facts:
- Before takeoff the P2V had GPS lock...
<snip>
I am not sure if the P2V had a course lock as it was to sunny to see the green rapid flashes before takeoff.

Are you meaning by GPS lock you had single flashing green on the leds (more than 6 sats acquired?). It's just that with mine the rapid flashing of home location locked comes pretty quickly after (sometimes just before) the last single red flash clears. Lack of a home location lock would certainly explain why IOC Home lock didn't work. IOC Course Lock doesn't require 6 the same 6 sat minimum as home lock.

Could it be possible you locked in a new home location and/or a new course lock orientation by toggling between the two modes when trying to get her back? It was certainly odd behaviour in the video, but not like a "flyaway" situation... almost like a "stick to the spot" one!

 

[themosttoys]

Glad it was ok. To see if we can get to the bottom of it, can I just clarify, as there's a bit of a contradiction in what you said:

Facts:
- Before takeoff the P2V had GPS lock...
<snip>
I am not sure if the P2V had a course lock as it was to sunny to see the green rapid flashes before takeoff.

Are you meaning by GPS lock you had single flashing green on the leds (more than 6 sats acquired?). It's just that with mine the rapid flashing of home location locked comes pretty quickly after (sometimes just before) the last single red flash clears. Lack of a home location lock would certainly explain why IOC Home lock didn't work. IOC Course Lock doesn't require 6 the same 6 sat minimum as home lock.

Could it be possible you locked in a new home location and/or a new course lock orientation by toggling between the two modes when trying to get her back? It was certainly odd behaviour in the video, but not like a "flyaway" situation... almost like a "stick to the spot" one!

+1 That is exactly what it looks like.

Rule of thumb to live by "do not panic". Also, there is a reason DJI put all those warnings when switching from Vision mode (where in any situation, just let go of the sticks) to NAZA mode where you must keep track of S1 & S2 positions and be 100% aware of what they are capable of (Ex: resetting home.) Just like anything, experience leads to calm decision making. Chesley Sullenberger is the best example of calm under pressure.

 

[Noël]

I'm sure that it had more than 6 sats acquired (it was 8 @ takeoff).
I did not see the rapid flashing of home location locked as it was to hard to see on the snow.

Could it be possible you locked in a new home location and/or a new course lock orientation by toggling between the two modes when trying to get her back?

No definitely not. I only tried the IOC home switch once and after that the Atti.

I think that the problem was GPS loss.
I did some testing today (same location) and it took a real long time to get more than 6 satellites.
When I got the lock I did not get more than 7!
The location is surrounded by mountains. Not sure if this could cause this.

 

[Pull_Up]

When GPS signal is lost for 3 seconds or more then atti mode will kick in. Would that tally with the initial movement towards the trees, was it "with" the wind? It doesn't explain why you couldn't actually control it after that, though... But it does look likely to be GPS related somehow.

 

[LandYachtMedia]

And lets not discount that if you were flying very close to very hard rock a local magnetic anomaly could be present.

 

[Noël]

When GPS signal is lost for 3 seconds or more then atti mode will kick in. Would that tally with the initial movement towards the trees

I forgot about the Atti mode kick in after 3 seconds.
The "auto mode Atti kick in" defiantly did not correspond with the event and control movements I gave it.
So in this case we can forget the GPS loss.

was it "with" the wind? It doesn't explain why you couldn't actually control it after that, though... But it does look likely to be GPS related somehow.

No wind that day, and if any it was the other way.

And lets not discount that if you were flying very close to very hard rock a local magnetic anomaly could be present.

I do not think that I was close enough for this kind of effect.

I did check the lift at the top today, but no big motor's or anything else that could cause a magnetic anomaly.
All commands I gave the P2V did correspond till the time after I turned the p2V 180 degrees around (it's not in this video).

I've just had a look at the original video again, and at takeoff the camera of the P2V was facing the same direction as the way the P2V was going when in trouble.

 

[pault]

'Local' magnetic anomalies can extend over a large distance - several miles. The whole mountain could contain seams of ferrous rock and this could cause conflict with the GPS maybe.

 

[Noël]

Well in that case it could have caused it, but I personally do not think so.
I will ask the locals tomorrow if any "magnetic anomalies" are know here.

Anyway.. New little longer video here: http://youtu.be/l1jBHBo5vnY

@ +/- 2:22 I did flick it back in GPS mode again, and all was okay again.

 

[Byrdimaging]

I agree with the fact that you should have nazam enabled so you can manually override GPS in a case such as this. I have, on humorous occasions been geocaching with my phone and all of a sudden I am 100 yards always from where I was two seconds ago. During this "GPS drift" I had 8 sats at all times and never lost GPS. I imagine this would cause a problem with a phantom's autopilot.

Another point is with 8 sats, assuming they are evenly distributed around the sky. If you get close to a mountain and block half of them you now have 4 SATs. The other problem with flying close to a mountain is iron ore. Your compass may have been getting interference.

One last point look at any handheld GPS or your phone using a GPS app not maps. Get 5 SATs and look at your accuracy. Then wait for 8 SATs and watch your accuracy get a lot better. What I do is wait for my P2V to get GPS set home point, then I wait for 8 SATs and reset my home point with S1. This way I know home was set with higher accuracy.

 

[Noël]

Apparently the resort I'm in is a big wellness resort where a lot of believers go to in the summer to get cured.
Maybe the "vibes" here caused the P2V to go haywire?

Anyway.
I have done some measuring with my phone and a magnetic field detector and did measure a bit more magnetic activity at top of the run.
Bottom was 43uT. Top was 75uT.
Not sure if this was casing the trouble.