Compass calibration

[phantomi]

Hi guys. I want to ask if once you're ready to take off and the FPV don't ask for a compass calibration, it is secure to fly. I'm asking because yesterday I calibrate the compass and for a second flight in the same area the FPV ask for a compass calibration. I reinitiate the fly protocol and once the fpv and the gps were locked it don't ask for the calibration. It is safe to take off in this situation?

 

[Pull_Up]

I don't calibrate every flight. I only calibrate after firmware updates, if I'm flying more than about 150 miles from last take-off or if I've installed or removed any hardware/had the case open. It's easier to do a bad calibration if you calibrate before every flight, in my opinion. However I always do a 30 second hover test at the start of every flight and watch for odd behaviour.

If the aircraft isn't asking for a calibration then it feels everything is ok, obviously. But I'd do that hover test - close in and about 10ft up - and make sure it's a tight hover with no circling or other odd flight behaviour. The fact you are getting recalibrate warnings leads me to suspect you're recalibrating in an area with some ferro-magnetic interference from somewhere. I'd go to a different area, nice and clean, no metal (including underground), no concrete, railings, drain covers etc. When you get a good calibration there personally I'd not bother doing another one unless the aircraft asks for one, or you notice odd flight behaviour during a hover test, or you update firmware or the other things I mentioned above.

Just my opinion, I know there are many who advocate a calibration before every flight. I just feel there's more risk of a bad calibration by doing that, rather than leaving it with a previously good calibration locked in (mainly from bitter experience of flying after calibrating in a stupid place back in the early days!).

 

[phantomi]

I don't calibrate every flight. I only calibrate after firmware updates, if I'm flying more than about 150 miles from last take-off or if I've installed or removed any hardware/had the case open. It's easier to do a bad calibration if you calibrate before every flight, in my opinion. However I always do a 30 second hover test at the start of every flight and watch for odd behaviour.

If the aircraft isn't asking for a calibration then it feels everything is ok, obviously. But I'd do that hover test - close in and about 10ft up - and make sure it's a tight hover with no circling or other odd flight behaviour. The fact you are getting recalibrate warnings leads me to suspect you're recalibrating in an error with some ferro-magnetic interference from somewhere. I'd go to a different area, nice and clean, no metal (including underground), no concrete, railings, drain covers etc. When you get a good calibration there personally I'd not bother doing another one unless the aircraft asks for one, or you notice odd flight behaviour during a hover test, or you update firmware or the other things I mentioned above.

Just my opinion, I know there are many who advocate a calibration before every flight. I just feel there's more risk of a bad calibration by doing that, rather than leaving it with a previously good calibration locked in (mainly from bitter experience of flying after calibrating in a stupid place back in the early days!).

I suspect same thing, i.e. that a solid calibration in an open nature environment is better that re calibrating withing miles of the first one in an unfavorable environment. My option were recalibrating in a concrete floor with metal embeded, so I decided to reinitiate and not calibrate. Thanks!

 

[MattBW]

The manual says if location is more than 30km away from last calibration you should recalibrate so I've always done that. Interesting thread I try to avoid electrical cables etc but hadn't thought about reinforcing bars in concrete. Good points thanks.

 

[N017RW]

I dont think that a set distance is the key.

It is actually how much the declination differs.

My dec. at home in S. Fla is ~-6. I traveled to Western NC where the declination is ~-5. This equated to about 600 miles as the crow flies.

I did not recompensate when I arrived before flying to test compass compensation and observed there was no difference in flight performance. Thus I never compensated in NC.

 

[jimre]

I initially thought declination compensation was a big reason for the compass calibration dance, but it's not. Instead, I think declination compensation is constantly being calculated in the background by the NAZA using GPS data and internal declination lookup tables. The "compass fix" in this latest firmware release apparently allows the NAZA to calculate this more rapidly (almost immediately, vs after 6-8 minutes of flying).

Compass calibration is more about being able to correctly distinguish - in 3 dimensions - the earth's north-pointing magnetic field lines, vs other local sources of magnetic interference. This is why it's especially important to re-calibrate after any equipment changes to the aircraft - since any addition/removal of metal could change the interference patterns. Also why you want to do this outside, away from any ferrous metal objects.

 

[N017RW]

Well I guess there is really two discussions here.

First is why do compass compensation and the second is when to (re)do it.

The first, (i.e. the 'dance') is to compensate for the non-moving magnetic distorting components ON or IN the Phantom.
The so-called hard iron and soft iron distortions. This must be done to compensate or reduce the on-board distortions anytime you add , relocate, or remove hardware from your a/c and because it is a 3 axis magnetometer must be done in both the X and Y axis.

The second is based on how much the current magnetic field, at a macro level, varies from the original location (i.e. the last or previous compensation location) and what effect it may have on compensation results thus possibly requiring re-compensation.

So in my case the declination difference was small but the travel difference was several hundred miles neither requiring me to redo the compass 'dance'.

 

[J.James]

Best way to know when it needs to be done is any time its flying weird and not hovering right. or if when you plug it in to the the naza and the mod values are out of whack. But when its flying great and nothing is wrong its best to not re calibrate it and have it actually take it from dead on to being a lil bit off.

For some reason were I live if I do the compass recalibration outside it usually throws it off and It will not hold still when flying and I get the best calibrations inside my house. I dont know why that is and could be the iron water and gas pipes running all over the place and also old trolly rails under the street. Or even high iron and other mineral content of the ground.

 

[dastagg]

I used to calibrate before every flight and no longer do it. Why? well, I understand that calibrating every time could cause an issue by itself and even DJI recommends not to calibrate every time. I now insure I do a 10' - 10 second hover to insure it is rock steady and if so, I fly and if it moves around, I land and calibrate. This seems to work really well and have only had to calibrate a couple of times since and was mostly due to some ground interference that caused the Phantom to send a warning to calibrate. I moved away from that area and the calibration warning went away and I did the hover test and if OK, I flew with no issues.

 

[N017RW]

Best way to know when it needs to be done is any time its flying weird and not hovering right. or if when you plug it in to the the naza and the mod values are out of whack. But when its flying great and nothing is wrong its best to not re calibrate it and have it actually take it from dead on to being a lil bit off.

For some reason were I live if I do the compass recalibration outside it usually throws it off and It will not hold still when flying and I get the best calibrations inside my house. I dont know why that is and could be the iron water and gas pipes running all over the place and also old trolly rails under the street. Or even high iron and other mineral content of the ground.

That is a good point too.

I did a similar compass experiment when I replaced the DJI Rx/Tx with Futaba about 2 mos. ago.

After the RX swap I did not do any calibration of any kind.
As soon as I took off it began to drift rearward and to the right.

After about 30 sec. of this to verify it wasn't wind, etc. I then landed and performed the compass compensation and I have not done it since.