For people taking off from the car hood

[BaNZ]

I landed and took off from my car hood a few times. Most of the time it was fine but I do get a lot of compass recalibration and interference.

However it started drifting sideways after my last take off. I've stopped doing that after I search through the forums. So do becareful if you're planning to take off or land on the hood.

 

[msinger]

Great tip!

Another common problem area is concrete sidewalks (most of which contain rebar). You really want to make sure the Mavic is not near any type of magnetic metal objects when powering it on and/or taking off.

 

[Splats]

How far away from metal object do you need to be? A few metres, 10m, 30m?

 

[msinger]

3+ meters is usually safe enough.

 

[Jake 61]

Tried that once and will never do it again.

 

[Jameslavis]

Wouldn’t put that on my baby. Far different amount of 0s between drone and truck.

 

[macoman]

I rather land the Mavic on my hand than in concrete or metal platform. Too much issue with the compass.

 

[4wd]

If you do 'need' to do it, the compass error is kind of fake and will clear if you hover a few feet up.
The worst thing you can do is recalibrate because it says error, while still in the poor location.
The app shows you which way the Mavic thinks it is pointing, it's wise to routinely check that too before zooming off.

 

[Whitwellian]

About a year ago I intended to take a flight with a Phantom 4 on the coast at Hemsby in Norfolk (UK) but got a "compass error" warning. Tried a recalibrate with no success, so decided not to fly

Recalibration back where I was staying was not necessary but I did it just for completion.

Later that holiday I took another walk down to the sand dunes where I'd had problems and realised that the concrete blocks, no doubt once used as part of the WW2 defence system and now just used as part of the system to prevent erosion of the sand dunes that protect the village were heavily reinforced with iron bars.

Since then I've had another (safe) flight from a few yards away with a Mavic Pro.

Just remember to take the presence of any ironwork into consideration before flying (including vehicles)

 

[Huddy]

My car bonnet slopes way too much to use as a take-off point, but the car roof would be OK. I do wonder if the interference might be caused more by the engine rather than the bonnet. Either way, it’s probably a better option for landing rather than for taking off.

I’ve had the same problem taking off from the seaside pavement outside my house. The drone doesn’t like it, and I assume there are metal reinforcements underneath. Either that, or I’m picking up interference from the adjacent electric tramway.

It just goes to show how many forces are at work around us that we’d know nothing about if we weren’t using a drone.

 

[Lake_Flyer]

I take off from the bonnet of my old Landrover Defender frequently, when the ground is not suitable. Landing on the bonnet with an I1 is not really an option. I sometimes have a piece of board on the roof rack which allows me to take off and land safely, even while standing on it myself. The Defender body is all aluminium and doesn't interfere with the compass, as I have tested many times.

 

[BudWalker]

If you do 'need' to do it, the compass error is kind of fake and will clear if you hover a few feet up.
.......

This isn't such a good idea. A compass error in this situation is likely caused by Yaw being initialized incorrectly because the mag data was incorrect. After launch the mag data will become correct but there will be a delay (5 secs to 120 secs) before Yaw is corrected. During that time the FC will want to correct any drifting using that compromised Yaw. This has been the cause of many launch time fly aways.

 

[Travisimo]

This isn't such a good idea. A compass error in this situation is likely caused by Yaw being initialized incorrectly because the mag data was incorrect. After launch the mag data will become correct but there will be a delay (5 secs to 120 secs) before Yaw is corrected. During that time the FC will want to correct any drifting using that compromised Yaw. This has been the cause of many launch time fly aways.

And a striking reason why having an actual ATTI mode on all of DJI's drones should be standard. If the FC began to compensate for drift, a pilot could simply switch to ATTI mode and compensate visually.

 

[BudWalker]

I'm guessing that 80-90% of what follows is true.

In the good old days the P3 would detect when there existed the possibility of a compass error. It would then switch to ATTI and all was well. Except that, over time, the number of false positives kept increasing in an attempt to make sure that legitimate compass errors were detected. Seemed like a day didn't go by without there being a false positive mid-flight compass error and the masses left wondering why.

Then, with the introduction of the Mavic Pro there was an "improvement". The Mavic would attempt to reconcile a compass error by slowly rotating with Yaw held constant until that constant Yaw value agreed with magYaw (compass derived Yaw). When that occurred the Yaw value would resume being updated. Didn't need to provide a way for the pilot to switch to ATTI. Worked great, but only when it worked. If the Yaw/magYaw separation was too large to begin with erratic flight would result. Presumably the FC would not attempt to do the normal position corrections during this procedure. But the flyCState is still GPS+ATTI (how's that for confusing?).

I'm not sure when the next "improvement" was made (it could have been coincident with the improvement above). Knowing that it can't use Yaw and GPS derived coords the Mavic will use vision data to hover while it's trying to reconcile a magYaw/Yaw separation with gpsUsed being false and visionUsed being true.

Then, the P3 was given the same "improvement"

 

[OldBlueHeron]

I take off from the bonnet of my old Landrover Defender frequently, when the ground is not suitable. Landing on the bonnet with an I1 is not really an option. I sometimes have a piece of board on the roof rack which allows me to take off and land safely, even while standing on it myself. The Defender body is all aluminium and doesn't interfere with the compass, as I have tested many times.

View attachment 35178

Like that Defender mate.

 

[AnzacJack]

It would be good if a material existed that could make a landing mat to shield magnetic interference

 

[mnoutdoors]

I take off from a aluminum patio table works great

I have also used my Jeep hardtop ( plastic )

 

[ddesens]

I have noticed that my air seems super sensitive to metal objects at take off. This is compared to my p3's and p4.

 

[Endeavor]

I think the very first thing I learned was never try to take off from the bed of my truck as I get immediate compass errors. I have not had any problems with concrete as most sidewalks do not have mesh in them unless they are on private commercial property. I hand catch and use a landing pad for take offs to keep dirt away from the lens. In a year I have calibrated the compass exactly once and as long as I stay away from cars on take off never any compass errors or erratic flights.

 

[Rotorcrafty]

I think the very first thing I learned was never try to take off from the bed of my truck as I get immediate compass errors. I have not had any problems with concrete as most sidewalks do not have mesh in them unless they are on private commercial property. I hand catch and use a landing pad for take offs to keep dirt away from the lens. In a year I have calibrated the compass exactly once and as long as I stay away from cars on take off never any compass errors or erratic flights.

I take off from the bonnet of my old Landrover Defender frequently, when the ground is not suitable. Landing on the bonnet with an I1 is not really an option. I sometimes have a piece of board on the roof rack which allows me to take off and land safely, even while standing on it myself. The Defender body is all aluminium and doesn't interfere with the compass, as I have tested many times.

View attachment 35178

When I try to take off from the bonnet of my Defender it gives magnetic interference error. The bonnet is aluminium but just below it is a huge chunk of cast iron called the engine. I can take off and land on the roof at the rear of the vehicle without problems.
There is a material that shields magnetic fields called "mumetal". It was used in old CRT instruments to shield the CRT from magnetic fields that might interfere with the electron beam deflection. Search mumetal. There is a wiki on it.

Eric