Tale of woe

Flight log will show something though eh!

1 Like

Indeed the biggest problem is what was actually happening, the phone went dark and video was lost, however when I went out of the fly part of the app, the front page said connected go fly, on going back in there was no picture and it said no connection.Foolishly I spent too much time trying to reestablish the connection to the phone instead of flying the aircraft.
The find my drone tool said that the drone was exactly at the end of the yellow line, but it wasn’t., and yes the outer ring is the 690 point.
Yes, the controller was connected so I was getting telemetry giving height, speed and distance away but trying to work out which way to go was a headache, all you really have is the distance away counter, but if it is counting down that doesn’t mean that it is coming directly towards you, only in your general direction.
But the really confusing thing is why the RTH took it further away. On set up the map showed exactly where I was, the compass was facing the right way and the initial part of the flight followed the path shown on the map, so why didn’t it come back?
It was a mavic pro and the flight log shows exactly what you see on the map, indeed it was taken by a guy called sar104 from mavic pilots forum and he is a genius at reading the logs, unfortunately you can’t read what ain’t there!

1 Like

Looks like you did not double check that the home point was actually the place you were standing when you took off…
:scream:

With the MP you don’t have to double-check anything. It tells you, as it takes off: “… the home point has been set …”

Didn’t calibrate compass at the new location before launching …

1 Like

Personally, I’m far less sure that is the cause.

Over the last 5 years of drone ownership I’ve driven to Italy, and around Italy, and back, twice … once with my Phantom 2, and more recently for 3 months with my MP …. and around the UK with both.

I recalibrated the compass on the P2 just once, when I got it, because it said you should.
As yet, I’ve never recalibrated the compass on the MP even once.

The change in magnetic declination in Europe is too little to make much difference wherever you are.

On drone/DJI/phantom/MP forums, over those 5 years, I’ve read a good many reports of flyaways and come to my own conclusions …
a) If it flies OK the first time … don’t fiddle with it
b) the more times it’s recalibrated, the more chance there is that one of the calibrations goes wrong, and you’re worse off than not having done one.
c) I’d look at that chart if I’ve travelled much further, and decide whether to bother with a recalibration.

Indeed, the compass accuracy is needed whilst in flight. If you recalibrate where there is some very localised warp in the magnetic field on the ground, once it’s flying and away from this anomaly you’ve basically just built in an error.

Interestingly, of all the people I’ve known that virtually never do a compass calibration, none have had a flyaway, and of the ones that I have read about that had a flyaway, most are reported with " … I did the usual recalibrations before the flight … ".

Personal opinion, and I’m sure there’ll be exceptions and those that want to disagree, but … if you understand what it is you’re recalibrating, why/when it “might” be necessary, and the limited impact one that is slightly out of calibration can have on a flight … you’d probably reach a similar conclusion to myself.

I realise that this doesn’t bring back your MP and, should you follow them, my thoughts mean you’re no closer to understanding what DID happen.
Believe me … I’d really like to know the real reason, if it can be found an proven, even if it means I have to completely change my thoughts on the compass recalibration issue. :wink:

2 Likes

Personally I never calibrated Mavic until prompted to do so.

Think that’s the advice from DJI these days, calibrate only when asked to do so.

Too easy to blame the pilot but fact is the tech is not fool proof and errors are going to occur.

It’s just a shame that there is no flight logs and recourse with DJI.

1 Like

Hmmm - yeah … and no.
If the cause is a local anomaly … I’d rather move 10-20m and see if it still asks me to.
As I say above, a local anomaly can have the opposite effect to that required.

1 Like

Yeah but have you ever tried to do a bad calibration Dave?
It dosent let you, pure and simple. The only option is to move the Mavic.

I did some testing while back. If you introduce a magnetic field wether it ferrous material, rocks or even a magnet, calibration will always fail unless the magnetic field is moving on same axis as the magnetometer.

But your right, get a compass error, move and if that fails calibrate.

Errrr - I’ve only ever done one! :stuck_out_tongue:

There’s all sorts of folklorey type stuff around this. “Don’t wear your iWatch”, etc., which I somewhat take with a large amount of salt. I’ve yet to see conclusive proof.

“You can’t take off from something large and steel - it screws the compass calibration” … you’ve seen my manhole cover take-off vid. But then a concrete (I assume reinforced) trig-point and it throws a wobbly! (The top-plate is brass - so non-ferrous). But a recalibration wouldn’t have helped. It just didn’t like being on a trig point.

I know people have been unable to take off from the decks of cruise ships … but there’s also plenty of vids out there that show it happens, too.

As to whether you CAN do a bad calibration … I’m sure it’s possible … even if most times it will prevent you from doing so. And I also think the current generation of DJI drones are better than the early ones.
What is more certain is, that you’re less likely to get a bad calibration if you don’t do them. :wink:

Edit: I was actually mid writing a post as to what happens if your compass is incorrectly calibrated as a post here. I think I’ll use it to start a new thread to discuss the subject. I’ll make it more detailed, in that case.

1 Like

The watch thing is sort of true.

I need to calibrate a lot as fly with and without a tracker.
Wearing my watch will result in a failed calibration 9 times out of ten.

1 Like

But why were you recalibrating in the first place?
Mine has never asked to be recalibrated in 18 months.

Or does wearing an iWatch cause it to request a recalibration? … which is what I understood and, whilst I don’t have one, my friend and I cable-tied his to my MP and it booted up, and flew, without any issue.

No I’m calibrating because I’m using or not using the GPS tracker. I try to leave it on all the time so it’s not an issue but often forget it.

Its not an apple watch, just a normal stainless watch. Watches are only a problem if your doing a calibration, that’s why I said sort of true. :stuck_out_tongue:

1 Like

Ah - OK - I’d read iwatch specific scare stories.
That early morning vid of a sailing boat motoring along off the south coast … the iWatch was on the MP then.
My friend was going frantic as it flew off nearly 1km over the sea after I’d said I’d just try a hover after take-off … and then … wheeeeeeeeeeeeeeee
LOL!

1 Like

Yes I did check that the compass and map matched, and it has only been calibrated when it said to do so on initial purchase.
The flight was meant to be a quick downhill trip to film some canal and river mill scenes and back. I was aware that behind me but nowhere near my intended course were a double row of pylons, however when I tried to fly it back and the counter went down and then up I had passed my position whilst still 200m away and then inadvertantly crossed over the power lines, so when I tried RTH it might be possible that the double row affected the compass which is why it never came home, Who knows!

1 Like

I’ve flown close to power cables without suffering any ill effect.
What was the RTH height set to? Sufficient to clear the cables?
… or, were you high enough to clear them when you lost control? … since it would stay at that height for RTH, not descend to the set height.

It was high enough to go over the top, but I can think of no other reason for it not to come home.

1 Like

How old was the Mavic mate?

Have you contacted DJI? They are sometimes not helpfull when the flight record ends abruptly but have seen a few cases where it’s been replaced under warranty.

OK - let’s add some meat to this compass discussion. (It’s taken me an hour to locate a video I took over 4 years ago.)

Back then, when people were having this same compass calibration discussion after each flyaway (very frequently), I carried out a bit of an experiment with my P2.

The compass on the P2 is (should be) mounted on one of the landing-gear legs … and when I changed the rather fragile P2 legs for some more robust carbon fibre ones (that, importantly, also provided better ground clearance for the H33D gimbal), I had to relocate the compass.

Everyone was saying I HAD to recalibrate it! So, this got me interested to find out what would happen if I didn’t. And it flew without an issue.

I became more intrigued … so I undid the cable-tie used to hold the compass onto the bracket on the leg, so that it was dangling totally free …. AND out of the alignment it held when on the previous, and new, legs … and wobbling about in the downdraft.
Again, it flew … without an issue … and hovered as accurately as it ever did. (Video below.)

I tried RTH, and it flew straight to me over a distance of about 400m (the size of the field - RTH height set to 5m - so if it went off at a tangent I’d see which bushes/trees it went into).

Whilst I don’t have video of the RTH exercise (I wanted to hold the controller just in case!), I do have video of it hovering, below, on quite a windy spring day.

The dangling compass is marked in this frame-grab, and as it should be fixed in the other two photos that I’ve just taken.

image

IMG_20180613_200950

IMG_20180613_201002

Video:

Note, also, powerlines in the background that I’ve flown VERY close to - specifically to find out if it affected the P2’s stability or compass.

1 Like

Only thing worth mentioning is the P2 having a single magnetometer and Mavic having dual.

It’s not true redundancy as that would need 3 but allows mavic to detect errors by comparing the 2 against IMU calculated heading.

If you add anything to Mavic it will normally give compass error as it affects one compass more than other. You might also see it yaw without input as it tries to resolve issue itself.

I’ve found the odd time I’ve removed tracker and forgot to calibrate it lets me take off. I can fly in a straight line but the first time I yaw it goes nuts as the compass does not match what IMU has calculated. It then quickly drops into ATTI mode.

It’s a good system to be honest and works quite well.

The general consensus is that the dual compass has cut down on many of the compass errors with previous models.