Hi there I updated to new firmware and since then my drone has turned into a Kamikazi pilot.  When flying over a loch in Scotland it went into a sports mode nose dive. I managed to catch it in time. Sent it into RTH mode then it flew in opposite direction and showed Home point as 500 meters from where I was .
It tried several times to land directly on the water.
Then the next time I flew It lost total contact with the controler and luckily did  RTH. I have had nothing but problems with this update. When I start DJI go 4 it says incompatible firmware. Are you sure you want to instal. Is there anyway to reverse the firmware to an older version. Plz help.

Have you tried setting it to factory defaults and refreshing it again? Then re-calibrate again.

Latest firmware works fine for me

nobody wants to come out and say the digital data that actually flies this creation does corrupt.
it shouldnt, but then it does get corrupt.


so what? make it a habit writing the firmware weekly unless youre certain your craft retains data without EVER getting corrupted?

Could you still fly without issues anytime before?

cutis Posted at 5-10 18:56
Could you still fly without issues anytime before?

Yes never had a problem before

Punchbuggy Posted at 5-10 15:40
Have you tried setting it to factory defaults and refreshing it again? Then re-calibrate again.

Yes did that. Also re-installed  firmware again. still having issues

wow! what does this mean? is it conceivable that discrete electronics are being clobbered to death merely because the prior firmware did not demand undertaking so much tasking that the poor devices just couldn't keep up.
so is suspect firmware overdemanding of electronics capacity?
or, suspect firmware is not overdemanding, and instead underperforming devices releases from factories are starting to surface?
or, the (failed) chips repositing precious firmwares arent good enough retaining integrity without getting smeared?
late entry:  or maybe, the devices actually have run their course of servicing life albeit in warranty period. if you get a good device that withstands, could this be reason why not all mavic air flyers experience misbehaved control? complicating this is uncertainties of individual flyers who augment their devices. such augments could be deviation enough to cause the incidents?

cutis Posted at 5-10 18:06
nobody wants to come out and say the digital data that actually flies this creation does corrupt.
it shouldnt, but then it does get corrupt.

make it a habit writing the firmware weekly unless youre certain your craft retains data without EVER getting corrupted?

If one has to reflash or re-update drones firmware at all, it is not normal, and there is something wrong.
1) Buggy firmware that is overwritting its own code
2) Bad flash memory (dropping or picking bits)
3) Faulty addressing or data steering
4) Sub-components doing I/O (write to memory) to wrong memory locations
5) CPU with separate addressing unit with intermittent problems

Properly implemented, it should not be possible to write to flash memory without special instructions with proper authority level.  Any writing of data should be done to memory seprate from flash memory where firmware is located.

In short, if you have to reflash or re-update drone's firmware to prevent problems as OP has described, send drone in for service.

cutis Posted at 5-12 16:00
wow! what does this mean? is it conceivable that discrete electronics are being clobbered to death merely because the prior firmware did not demand undertaking so much tasking that the poor devices just couldn't keep up.
so is suspect firmware overdemanding of electronics capacity?
or, suspect firmware is not overdemanding, and instead underperforming devices releases from factories are starting to surface?

so is suspect firmware overdemanding of electronics capacity?

Could well be.  New firmware pushing specs of electronics, with some drones built with electronics marginally below specs.  Or New firmware using aspects of electronics that were never tested with previous firmware, resulting in fallout with some versions of drone.  Because those versions of drone used ICs (or sub-assemblies) supplied by different maker, or from a different batch run from same maker.

skullfokker Posted at 5-12 15:16
Yes did that. Also re-installed  firmware again. still having issues

If you can, roll back to previous version.  See if problem continues or goes away.

hedge's pointers elicit notion poor quality assurance quality controls (QAQC) enacted

but entity does not skimp to this end, right?

how to vouch QAQC in fact exists appropriately in this entity?

there is a mixture of flyers who are ALLEGEDLY firmware impacted deleteriously whilst others absolutely not.

it may also be shortfalls balancing workload (QAQC) within this entity, the enterprise clients, professional work clients, the enthusiast cinema photographers flying mavic air.

so, since the trending numbers experiencing frustrates "apparently" coincides to updating, can this mean entity is on top of matters "sufficiently enough", anticipating incurred frustrates, but limiting those numbers in averting the appearance QAQC isn't spread too thinly