Compass Error on attempting yo take off at canal lock
- Thread starter Skylark
- Start date
I took off from a metal boat a couple of weeks ago. Before doing that, I checked with some knowledgeable folks including a DJI insider on if I could do that and what I should do to prevent a fly away.
The recommendation was to not calibrate the P3 on the ship. I was told by the DJI insider that calibrating 150 miles away on land was sufficient.
When I started the P3, I held it above my head to put some distance between the metal deck and the P3. I then placed it on a wooden table where I launched it.
To make sure everything was ok, after launching I let it hover about 10 feet above the deck for a minute to make sure it was stable and then flew over the water.
I had two flights on two separate days with no problems. Here are a couple of shots from those flights.
The recommendation was to not calibrate the P3 on the ship. I was told by the DJI insider that calibrating 150 miles away on land was sufficient.
When I started the P3, I held it above my head to put some distance between the metal deck and the P3. I then placed it on a wooden table where I launched it.
To make sure everything was ok, after launching I let it hover about 10 feet above the deck for a minute to make sure it was stable and then flew over the water.
I had two flights on two separate days with no problems. Here are a couple of shots from those flights.
Last edited:
It won't, because the error induced by taking off from metal is now the base point for all of the location calculations. When it is away from the metal and that interference is not longer effecting the calculations, the compass may start reporting that it is a long way from where it thinks it was at take off, or traveling in an incorrect orientation to actual orientation. I believe this is one of the primary causes of "fly-a-ways"
A compass can't report that, only gps can. All a compass has the ability to do is report which way north is.
The compass directional info, calibrated with speed info, must match where the gps thinks the P3 is. It is a checks and balances system. (consumer gps is not accurate enough on its own +/-6-15ft) The compass has a whole lot more to do with location than where north is. Thats why compass errors are so critical to pay attention too. That's why taking off on metal and then moving away does not solve your compass error issues, it makes them exponentially worse. Now the checks and balances get checked, but they are not balanced. The FC tries to balance, post haste, and then you get the speed of light fly-a-way. Thats also why the FC now goes automatically goes into atti mode with major compass errors. It was a safety feature not available to the Naza Phantoms or custom builds. Thats also why it is recommended to switch manually into atti with compass errors. You often dont want it switching back to gps and trying to reconcile compass errors on its own.
I just can't work out why, when the P3P gets all this incorrect data, it just decides, rather then staying happily where it is, to zoom off somewhere else.
Has it been programmed to do this?
Surely it could have programming that told it to ... 'stay put and obey RC signals' ... when it gets scrambled data. Of all the possible options, I think that the very worst option is ... "fly quickly away from operator until crash".
Hey Andy nice location. You could use the P3 to spot fish!
Was the boat mainly steel, fiberglass or aluminum? Or a combo?
GPS or Atti?
Cheers.
Because it doesn't know what still is.
The scrambelled data might make it think it's drifting 30mph to the left (due to wind) so it will attempt to fly 30mph to the right in order to "stay put"
ianwood
Taco Wrangler
Close your eyes, spin around 30 times and then try to walk in a straight line. That's what happens when the compass is all jammed up. Bad data is bad data. Don't take off from or near metal!
I had big compass problems on a p3 the other day taking off from both the bridge of a large cargo ship, and unsuccessfully from the dock about 20m from it. It initially didn't display any compass warnings before takeoff from the bridge, but once in the air and about 30m away it did, so I switched to ATTI expecting it to fly fine without using compass, but it still influenced it. There was an ever increasing backwards drift, and the gimbal started rolling all over the place with movement so it got brought back in pretty quick. Not the best thing out over the sea. Got it job done by flying from 200m back which had no issues, even when flying within 30m of the ship.
It was a steel hulled ship and I was using GPS.
Look closely at the stern of the ship and you'll see that there are shark cages being deployed into the water. So fish spotting for sure
ianwood
Taco Wrangler
This is a good example. The gimbal uses compass data to determine orientation. The only option when metal is nearby is to get the Phantom as far away from it as possible before turning it on and then keeping it away from metal at all times while it is on.
Ian, I have really appreciated your input throughout this thread. Could I trouble you for your expert opinion as to how the video I posted was shot with the Inspire 1? How did take off and how was it flying rock solid amongst all that ferrous material? This simply puzzles me.
ianwood
Taco Wrangler
It's not impossible to stay in control in a tricky environment. But most likely they avoided magnetic interference when calibrating and when taking off. That is when the FC defines parameters it will use for the whole flight.
You can encounter it in the air but if you avoid it in these two situations, you can most likely get away from wherever you're encountering it in the air.
You can encounter it in the air but if you avoid it in these two situations, you can most likely get away from wherever you're encountering it in the air.
The other consideration is for a professional company an inspire 1 trashed for a shoot may be an acceptable loss so they were happy to shoot in a riskier environment. It's not as clean cut as it will work or it won't. But the risk of something going wrong increases as the environment becomes less ideal.
If you're in the same position happy flying! Unfortunately I'm not in the position where I can treat my P3 as a consumable so I only fly in low risk environments.
The full video (not posted in this thread) shows multiple take offs from the concrete flooring (rebar?) with flight paths amongst metal beams. I shall try to get some more info from the pilot himself on this shoot, but he may not reply how he did it! But many thanks for your input.
"Just because" doesn't work here. Please answer the question:
How?
The compass has nothing to do with the location - just which way the drone is pointing. The GPS reports the latitude-longitude position to the controller. The compass reports the heading. There is no sensor on the Phantom to determine speed, that is calculated from subsequent positions from the GPS.
Bad compass data will not cause a flyaway. "It happened to me" is not a technical description of how bad compass data can cause a flyaway. It's what we in the engineering world call a coincidence. The controller goes into ATTI mode when there is a significant compass error not to prevent a flyaway, but because without reliable heading information from the compass, your Phantom won't know which way to point when it performs a "return to Home".
Similar threads
