Compass calibration over a big pile of metallic trash :S

[Cyril]

Hi guys,

I'm flying my P3P with the last firmware, done 100+ flights so far without problem. Recently I started flying from an offshore rig (basically thousands of tons of steel in the middle of the ocean).

I did a good compass calibration onshore, then I flew to the rig (50 miles flight). There I did 8 flights offshore with no issue, taking off from the top of my hard shell backpack (because if I put the P3 straight on the helideck I had a compass error, obviously), and landing directly on the floor easily.

Flight #9 : take off OK, P-GPS, the drone starts drifting badly after 3 seconds, getting worse and worse when I tried to counter the drifting (lot of pitch, speed, loss of altitude). The VPS is off, not because I had any problem with it, but just in case I tried to get higher to avoid possible magnetic interference, but it went even worse. No compass error to be mentioned, by the way. I managed to land it (who know how ?) with a few bounces, but I really thought I would lose it (overboard or crash on the helideck).

I presume it's a problem of compass, but I don't understand why it started mis-behaving at the 9th flight. So, question 1 : would a switch to ATTI mode have helped me to land it smoothly ? And question 2 : do you think I can fly it again (here on the rig) if I do a new compass calibration (on this big pile of steel) ?

Thanks

 

[RoyVa]

I would stop and save the bird. Way to much interference. Your compass is getting out of whack from all that interference. It's just going to get worse. Only way to correct it would be getting into a rubber raft away from all that steel, calibrate and fly but it will happen again. Compass will get disoriented. Sorry but the only way to fly from there would be a cheap o quad without GPS and only manual... Make sure she floats ... Not gonna be easy!

 

[Steve Moore]

The compass should have any bearing on drift (pun intended). That's more a GPS or gyro confusion, especially with that extent of drift. The direction the bird was facing wasn't squirrely was it?

 

[Steve Moore]

Actually... look at the flight log data. Specifically at the time it started and WAS drifting. Were the GPS coords changing? Airspeed and distance changing? I'm betting there was no movement indicated by airspeed, but possibly by GPS. There are a lot of factors out there.. big *** chunk of metal, over water, and a pretty strong satellite/microwave/high freq transmitter very close. Does it fly without problem back on land? I wouldnt hesitate to try it again, but I probably wouldnt push it much till I nailed down the cause.

 

[Cyril]

I have been replaying the flight and the bird was spinning (slowly), drifting & losing altitude at the same time, as if unable to get balanced. The GPS signal is always strong here (18 satellites as soon as it is turned on) so I don't think it's a GPS problem. The wind was light that day, about 5 knots. Gyro or compass then ? Well, I think I'll stop flying from here until I get back to shore for a proper calibration.

About Gyro, is it calibrated by doing an IMU calibration ?

Thanks.

 

[Bratinov]

I once tried to take off in a demolished building, lots of metal and concrete... keept getting gyro errors and IMU calibration requests.
0 issues when I changed location.
Did a IMU calibration after that, just to be safe.