Not stable when hovering...help

Joined
Feb 15, 2014
Messages
20
Reaction score
0
I recently had issues with compass calibration and reset the compass with the magnet as per the DJI video.

I was flying in my back yard , it is surrounded by walls, the phantom doesnt want to hover, sometimes circling or drifting.

I also did a control re-calibration through the software and it passed.

Any thoughts / please help?

[youtube]http://www.youtube.com/watch?v=oeO3lqL7Kmc[/youtube]
 
Scratch what I originally wrote, just watched the vid properly again....

Where did you do the compass recalibration? If you did it in a nice open area (as you should) but this walled area has different ferro-magnetic properties you may need to recalibrate the compass. If it was done in this area, you may need to find somewhere more open to do a compass dance.
 
Do you think this is a compass issue?

Also, IF the compass calibration passes, then I dont need to check values in the software?

I did do compass calib in the same back yard :(
 
Your compass lights can turn green after the compass dance but this isn't in the ideal range.
What are the nuimbers in the assistant software?

Majke sure you are nowhere near any concrete or metal while doing the compass calibration outside.

Also do an advance IMU claibration, and provide a screenshot on the vlaues.
 
jedi88 said:
Do you think this is a compass issue?

Also, IF the compass calibration passes, then I dont need to check values in the software?

I did do compass calib in the same back yard :(

Unless they are out of range the values aren't important - what can cause the drifting is the compass thinking north is there because of something affecting it's magnetic field when calibrated (say, rebar in concrete, or underground metal pipes/drains) versus what the GPS thinks. Go and fly it in a nice clear area and see how it hovers there. If it's good, you know it's "stuff" in the yard causing the issue, if it's still off do a compass calibration in the nice clear area and it should then hopefully perform better.
 
How many sat's are showing in the app?

IMHO your video shows a fairly stable hover (depending on the number of sat's locked in.) I don't think the P2V moved more than 3 feet.

Besides GPS / Compass and props (you are using the stock DJI props and they are completely undamaged?), another possibility (considering the environment in the video) not yet mentioned is that the P2V could be being effected by it's own prop wash. Take it out into a more open space and climb to 10 meters or more. With DJI props (rotors), good compass calibration, 10 or 12 sat's locked, at least 20 to 30 feet of altitude and low wind, you will see a good hover. Also remember that it is flying. :)

Happy flying
 
jedi88 said:
Mav456 said:
get new props

...
I had 6 sat's locked ...

6 is the bare minimum it needs and if they were in a bad constellation at the time it is the most likely cause for intability.

In my experience the compass is much overrated in exlanations of GPS mode stability.

Doing the compass dance every time before flight seems very unnecessary to me. I very much beleive in doing calibrations of any kind only if there is a measurable indication that they are necessary. This is not limited to the P2V; it is general experience from all kinds of instrumentation I have operated.

I have done the compass dance exactly once so far, immediately after buying the P2V. As long as it shows sufficient stability in GPS mode I will not do it again. I have flown half way around the world from the original location and the P2V still is stable in GPS mode. Locations I have flown at include: Malibu CA US, San Diego CA US, Lake Havasu AZ US, Atlanta GA US, Roermond NL, Düsseldorf DE. The magnetic variation in these places is widely different too, from 13.2E via 4W to 0. No re-calbration necessary so far.

I find quality of GPS recepion to have a noticable impact on stability, and of course strong wind. I will not fly in significant wind with less than 8 sats.

This is my experience, northern hemisphere only so far. Yours may vary.
 
if you look at the tree's it looks like its a bit windy may be that as got something to do with I know mine as moved like that when its a bit windy
 
I've noticed that the stability of my P2V's hover is directly proportional to how many satellites it sees. With 6 (the minimum) it will drift several feet, especially if there's much wind. When it finds the 7th satellite, the stability improves noticeably. By 8 or more, it becomes virtually rock solid when there's little wind.

Your video looked like a reasonable 6 satellite hover to me.
 
Visioneer said:
I've noticed that the stability of my P2V's hover is directly proportional to how many satellites it sees. With 6 (the minimum) it will drift several feet, especially if there's much wind. When it finds the 7th satellite, the stability improves noticeably. By 8 or more, it becomes virtually rock solid when there's little wind.

Your video looked like a reasonable 6 satellite hover to me.

I would say thats right when I have had 12 satellites it does not move much in hover even in the wind
 
Visioneer said:
I've noticed that the stability of my P2V's hover is directly proportional to how many satellites it sees. With 6 (the minimum) it will drift several feet, especially if there's much wind. When it finds the 7th satellite, the stability improves noticeably. By 8 or more, it becomes virtually rock solid when there's little wind.

Your video looked like a reasonable 6 satellite hover to me.

I have the same experiences. if circumstances like wind and visibility permit it, i try to gain some altitude to increase the number of satellites the receiver can lock on. i do so maintaining some forward speed for additional stability.

i do not fly far away unless i have at least 9 sats locked.

speculation: flying a couple of straight lines seems to improve gps accuracy. i have no idea if this is real or imagined, if it is due to the gps receiver getting a better fix due to more sats or better lock, the compas being calibrated with bearings obtained by gps, ...
it appears to me that we have a lot of speculation going round, because accuracy is hard to really measure and because we do not know how gps mode works in detail.

having said that, visioneer's observation about the number of satellites is one that i fully share and i have heared quite some people relate.
 
Sorry for late response...

So when I take off, the phantom seems to drift still... Here is a screen shot of the compass.

I think it looks pretty out?
 

Attachments

  • compass.png
    compass.png
    51.1 KB · Views: 961
Does anyone KNOW how the Vision's innards (algorithms) operate? I would have thought maintaining a position would be a function of GPS (and maybe altitude) info. Wouldn't a compass fix merely provide heading info? ... which couldn't tell it where it was relative to a given location, just what direction it was facing relative to where it was facing then.
 
Visioneer said:
Does anyone KNOW how the Vision's innards (algorithms) operate? I would have thought maintaining a position would be a function of GPS (and maybe altitude) info. Wouldn't a compass fix merely provide heading info? ... which couldn't tell it where it was relative to a given location, just what direction it was facing relative to where it was facing then.

that is exactly what i *suspect*. however i do not *know* it because i have no knowledge of the algortihms the naza uses. so far my suspicion/theory is consistent with my observation of the p2v behavior. it would be useful to hear other people's observations.

observations:

calibrated only once so far and operated normally in locations with widely varying magnetic variations from 13e to 4w. conclude that calibration is not really necessary due to variances in compass deviation.

got 'needs calibration' messages whenever starting up near large chunks of metal such as inside car boot, on table with massive metal foot. conclude that re-calibration is necesary when metal/magnetic environment changes, e.g. when mounting metal hardware on the p2v.

hover stability increases with number of satellites reported. conclusion: hover instability caused by poor/varying gps fix.
 
I've had issues with circling a bit when hovering lately but after much investigation I think for me it boils down to limping along on props that need replaced.
 

Recent Posts

Members online

No members online now.

Forum statistics

Threads
143,066
Messages
1,467,352
Members
104,933
Latest member
mactechnic