DJI Mavic, Air and Mini Drones
Friendly, Helpful & Knowledgeable Community
Join Us Now

Concerns over firmware

ipcctv

Member
Joined
Apr 5, 2018
Messages
9
Reactions
2
Location
Cheshire, UK
I bought my AIR a while back and was immediately spooked by the comments that seemed to suggest the firmware was flakey. I did the dilligent thing and watched lots of videos on how not to lose the drone. I've had the drone out just 3 times and flown for all of 20 mins. During this time I've been prompted to re-calibrate the compass twice and the IMU once. Yesterday before this 3rd flight ever I upgraded the firmware and re-watched the videos. I was sensibly reminded that the firmware can blow away all kinds of settings. This in itself sounds iffy.

What shocked me was the descriptions of the last firmware:

  • Improved firmware data safety.
  • Fixed an issue where GPS location changes inexplicably, and when RTH is abnormal due to compass interference.
  • Optimized vision arithmetic to improve flight performance at both high altitudes or over unclear ground patterns.
  • Improved camera stability.
  • Improved safety for ActiveTrack and SmartCapture.

Questions and observations.

1) What does data safety mean? GDPR? Firmware being corrupted or what?
2) How can (DJI) claim to fix something you find inexplicable?
3) Is the compass interference real or imagined sometimes by your firmware?
4) Any thoughts/suggestions on my compass problems

I do wonder if my prompts to have my compass re-calibrated twice from 3 flights and the stories of fly aways can sometimes be duff firmware and with release comments like this, I'd not be surprised.

On the plus side, I still have my drone and on the minus side, I got a great picture of my 'egg in the nest'

Tom.
 
Hi Tom,

I agree with most of your analysis here... relating to the "inexplicable" thing, well, I just credit that sort of nonsense to very bad use of Google translate.

I fly my Air from the same spot every time while at home (the roof of my building), and I am prompted to calibrate compass erratically, sometimes I am, sometimes I am not. I do a pre-flight check to see the compass interference in Advanced Settings -> Sensors State -> Compass, and never take off with anything other than green. There is a lot of iron in the building rebar and other buildings and construction cranes around, so I steer clear of those when I calibrate. Also, I calibrate by taking the aircraft above my head as far as my arms reach out, and doing the little 360 dances... (my daughter has great laughs whenever I do this, and I am sure anyone looking thinks I am a complete clown).

I am worried by the FW not being totally tested. It seems to me DJI skips the testing phase and goes straight to release, thinking it might be cheaper to release the FW and THEN handle the issues. Since that makes us ALL testers for DJI, I usually take it fairly slow and try to gauge the AC's capabilities and issues before pushing the limits (given that it has been over 3 months since the last release, my Mavic Air is, in essence, a limited product until these things are resolved).

Right now, there are MANY issues with 0.400, including, but not limited to, motor current errors (when both sticks are pushed up to decrease altitude and go forward at the same time, at least in my case, which is to say on almost every approach to home), yaw and pitch issues, the AC changing direction rapidly while hovering (seen this 3 or 4 times on mine), getting a warning that you are limited to Beginner Mode while flying, low connection range, the remote charging the device in Android, etc. Also, DJI seems to be spending most of its time on their new launch, instead of working the issues with the MA.

My main concern with the compass is that it WILL disable GPS mode if it fails mid-flight, so I am constantly reminding myself that I may need to fly the drone back by hand and WILL need to have a good connection to it (be within range) because if it loses GPS and I lose connection, its bye bye Air. To be honest it is a little uncomfortable to fly under these conditions, since if the compass fails at long range, and the wind picks up the AC while flying away from the HP, it could easily be a flyaway due to lost connection. The hope of re-connection is almost nil when the compass is actually onboard the AC, and it is unlikely to "pick up" GPS again, from what I've read.

In any case, these issues are not being addressed expediently by DJI, so fly with caution.

E.
 
Hi Tom,

I agree with most of your analysis here... relating to the "inexplicable" thing, well, I just credit that sort of nonsense to very bad use of Google translate.

I fly my Air from the same spot every time while at home (the roof of my building), and I am prompted to calibrate compass erratically, sometimes I am, sometimes I am not. I do a pre-flight check to see the compass interference in Advanced Settings -> Sensors State -> Compass, and never take off with anything other than green. There is a lot of iron in the building rebar and other buildings and construction cranes around, so I steer clear of those when I calibrate. Also, I calibrate by taking the aircraft above my head as far as my arms reach out, and doing the little 360 dances... (my daughter has great laughs whenever I do this, and I am sure anyone looking thinks I am a complete clown).

I am worried by the FW not being totally tested. It seems to me DJI skips the testing phase and goes straight to release, thinking it might be cheaper to release the FW and THEN handle the issues. Since that makes us ALL testers for DJI, I usually take it fairly slow and try to gauge the AC's capabilities and issues before pushing the limits (given that it has been over 3 months since the last release, my Mavic Air is, in essence, a limited product until these things are resolved).

Right now, there are MANY issues with 0.400, including, but not limited to, motor current errors (when both sticks are pushed up to decrease altitude and go forward at the same time, at least in my case, which is to say on almost every approach to home), yaw and pitch issues, the AC changing direction rapidly while hovering (seen this 3 or 4 times on mine), getting a warning that you are limited to Beginner Mode while flying, low connection range, the remote charging the device in Android, etc. Also, DJI seems to be spending most of its time on their new launch, instead of working the issues with the MA.

My main concern with the compass is that it WILL disable GPS mode if it fails mid-flight, so I am constantly reminding myself that I may need to fly the drone back by hand and WILL need to have a good connection to it (be within range) because if it loses GPS and I lose connection, its bye bye Air. To be honest it is a little uncomfortable to fly under these conditions, since if the compass fails at long range, and the wind picks up the AC while flying away from the HP, it could easily be a flyaway due to lost connection. The hope of re-connection is almost nil when the compass is actually onboard the AC, and it is unlikely to "pick up" GPS again, from what I've read.

In any case, these issues are not being addressed expediently by DJI, so fly with caution.

E.
 
Thanks for your reply. It makes pretty grim reading and has added a few concerns. I was flying in fairly mild wind the other day and it was wobbling a fair bit when I let go of the RC. I got my drone with the hopes of it being a relaxing hobby, but as you say, it seems like it makes one nervous. You would think that compass, GPS and RTH would be the most tested bits of all and rock solid.

I used to do testing of hardware, software and firmware in a previous job and this screams of a lack of control and or skilled coders. If I'm unluck enough to lose mine, I'd never get another. Data leakage to me (my slant on what that means) is definitely bad programming.
 
  • Like
Reactions: Drgnfli
...Also, DJI seems to be spending most of its time on their new launch, instead of working the issues with the MA.
E.

Personally, I think that the delayed release of the new product is directly related to the current firmware issues some folks have experienced (including myself). I'm also inclined to believe some of the problems may be exacerbated due to bad components on some of the aircraft. Something's rotten in Shenzhen, and the lack of any FW updates in nearly 3 months indicates they still haven't quite figured it out.

That being said, I was finally able to get a replacement MA that appears immune, and it is a pure joy to fly. It's what I expected it to be from the get-go. I actually haven't touched my P4P since. So there is hope.
 
Personally, I think that the delayed release of the new product is directly related to the current firmware issues some folks have experienced (including myself). I'm also inclined to believe some of the problems may be exacerbated due to bad components on some of the aircraft. Something's rotten in Shenzhen, and the lack of any FW updates in nearly 3 months indicates they still haven't quite figured it out.

That being said, I was finally able to get a replacement MA that appears immune, and it is a pure joy to fly. It's what I expected it to be from the get-go. I actually haven't touched my P4P since. So there is hope.
 
That feels like there are plenty of us with flakey hardware then and the firmware iterations could be a weak and failing attempt to work around them. If proven, there should be an intrinsic right to return kit for a replacement, but I fear current law would leave people stuffed with the burden of proof being gone along with the MA that flew into the sunset. :-(
 
That feels like there are plenty of us with flakey hardware then and the firmware iterations could be a weak and failing attempt to work around them. If proven, there should be an intrinsic right to return kit for a replacement, but I fear current law would leave people stuffed with the burden of proof being gone along with the MA that flew into the sunset. :-(
Again, it's just conjecture on my part, based on the experience I had. My original MA worked fine until the update, and when I sent it in they claimed it had a bad main board. If so, why did it work fine until the .0400 update? Anyway, I'm guessing that if they do have a bad batch, they'll never admit to it. My only suggestion, if you're unwilling to wait for an update, would be to open a repair ticket and send it in for repair (likely replacement) until you get one that's immune. I'd suggest you do it before the warranty expires.
 
Thanks for your reply. It makes pretty grim reading and has added a few concerns. I was flying in fairly mild wind the other day and it was wobbling a fair bit when I let go of the RC. I got my drone with the hopes of it being a relaxing hobby, but as you say, it seems like it makes one nervous. You would think that compass, GPS and RTH would be the most tested bits of all and rock solid.

I used to do testing of hardware, software and firmware in a previous job and this screams of a lack of control and or skilled coders. If I'm unluck enough to lose mine, I'd never get another. Data leakage to me (my slant on what that means) is definitely bad programming.
Being completely honest. That "butterflies in your stomach" feeling does get better with experience and a little time. Nowadays I can fly my MA to within 1,500 meters with pretty good signal, and mostly disregard the motor current error warning or alter my approach aggressiveness, so as to avoid it. Also, I cannot lie and say I have had a single instance of a bad RTH or being kicked to ATTI mode, but I did get one RTH where the drone attempted to head back by way of the course it had flown over, instead of directly home. This was my mistake, as I hit the "smart RTH" at 30% battery, and had my low battery setting to 25%. That 5% would not have been enough to fly straight home if the drone flew back over its initial path.

My main concern here is that DJI knows full well of the issues, there are countless posts in their forum, and they have taken over 3 months to sort the issues. It seems to me, like I said, that we are unwilling beta-testers for them, and the best ones they can get, because no other company would expect their testers to spend 1,200 on one of their products to do the testing for them, and risk their money in the process.

I love the MA, and I am enjoying it, I just wish the software was more "solid." The way I lost the fear of losing it was thinking in economic terms. If I can get 43 hours of enjoyable flight time out of the MA, then I would not mind losing it after hour 44, since it would've cost the same to get 43 movies or any sort of $14 of entertainment for 2 hours, hehe.

It'll get better on your end, I cannot say if it will improve on DJI's end.

E.
 
That feels like there are plenty of us with flakey hardware then and the firmware iterations could be a weak and failing attempt to work around them. If proven, there should be an intrinsic right to return kit for a replacement, but I fear current law would leave people stuffed with the burden of proof being gone along with the MA that flew into the sunset. :-(
 
Yeh I might just do that. It galls me that mine in pristine condition could be swapped for a refurb when I only recently paid £800 for it. Not on really. I wish the firmware was open source and then we'd know where the problem was. I'm mindful to ask for a new replacement and see what their response is. An official statement from the company would be nice. Hint if DJI are monitoring this.
 
You can't find a mass produced, complicated electronic product today that doesn't have firmware/software issues of varying degrees of severity. It's just part of ownership. The reality of it is that millions of people are flying DJI's drones and most people likely do not have issues. There will always be defective devices, horror stories, etc. and you can't let that ruin it for you. Searching the internet will also provide a grossly disproportionate picture of reliability, because happy people generally do not post such. There is always a chance you will lose or crash your drone, and if you stay with the hobby long enough, it's probably a "when" rather than "if". I have flown my MA dozens of times without even the slightest hiccup on the new firmware. As for data security, unless you're an engineer on DJI's data security team, anything you read is likely going to be speculation. Many of us have smartphones, smartwatches, smart speakers, internet connected cars, a PC, etc. etc. and I can only imagine how much information is being collected.

My advise to you is just get out there and enjoy yourself while taking all precautions necessary for responsible flying - if it's still too worrying or stressful for you, maybe it's time to reevaluate. If you think something is wrong with your drone, send it to DJI. DJI's customer service may not always be the best, but we all knew that going in.

Also your compass needing re-calibrated probably means it passed by something with interference, it's not unusual (even a rock can do this). Just do what the app tells you and chances are you will be fine.

I also personally don't agree with the beta tester comments - there are countless products many times (even orders of magnitude) more expensive than a Mavic Air that have major issues, recalls, service advisories, questionable firmware updates, etc. A product not being absolutely flawless for it's entire life cycle does not make customers bets testers, in my opinion.
 
The "data safety" thing in release notes usually refers to trying to prevent people from gaining priviledged access to their aircraft and tweaking it or circumventing some of DJI's nonsensical restrictions, or do things like get detailed aircraft logs that can be used to challenge "wrong" analysis of warranty cases.
 
Last edited:
You make some good points in terms of trying to enjoy etc and I will take that on board. I also thank you for your slant on data security. If it means that, then thats much more acceptable, but that wouldn't need a f/w update as it's collected and sent by the phone isnt it? I've not done this yet. Thats just app software if I'm right.

I dont quite agree on the level of sophistication etc as an excuse for buggy code given the nature of the product. Rapid release methodology may be good on some things, but not drones. It shouldn't be even getting so many releases unless they are feature upgrades. This one admits weird things (inexplicable) were going on. The conclusion is that this release may be just be a best guess at whats going wrong. That would make us beta testers.

My problem is that with defective navigation chips, the result isn't like your PC, smart phone, watch etc. Your PC doesn't fly away and take its logs with it. Also, I've not had a BIOS upgrade every few months on any system I've owned. I have had flakey Sat Nav and Smartwatch firmware from Garmin, but they fessed up. I entirely accept people have done odd things and blamed it on the MA only to be put right by experts reviewing logs. I hope I don't end up being one of them.

Do people really manage to get into the drone to patch flight logs? I would have thought very few people could do this.

I'm going to keep mine in a tin box for a bit and see if it gets better, but on balance, I think its a pup. I didn't know DJI customer service was bad by the way else I'd have bought something else. I should have researched better but was taken in by the hype. In a twist of the proverbial knife, it was the RTH feature amongst others that I wanted. Next sunny day I'll be out though and I hope this all settles down.

regards Tom.
 
  • Like
Reactions: Drgnfli
Hey Tom,

Completely agree, if my Android smartphone has a FW issue, it is unlikely to land on someone's head, unless I throw a tantrum and throw it out of the balcony. The intrinsic safety issues of drones would make you think that the manufacturer would be more inclined to fully test FW and HW. When you get on a Boeing or Airbus, you're not expecting the same level of commitment that you do from your TV or Washing Machine's manufacturer, but much higher. With drones, the very simple truth is that if you have a flyaway or crash, you may completely lose your investment. DJI's own policies do not allow replacements without recovering the drone (Care Protect) and seldom do replace missing drones unless they can prove 100% that there was not one bit of pilot error involved.

Another thing to consider, despite DJI slowness to release patches and lack of due diligence with testing, is the fact that DJI IS the best drone manufacturer out there... I do not own any other drones, I am thinking of buying a Beebop 2 just to test it out and have a wider picture of the market, but DJI is the most solid option, so I recognize this and I am not simply bashing them out of pure joy, but simply because in the case of the MA's FW, the deserve a little bashing.

Drones are very hackable, so yeah, I understand DJI trying to fix backdoors. Last I heard you could remove the MA's height limit. I bet DJI understands that if a drone manages to down an airplane because of some idiot pilot, they are the ones who will untimately take the economic flak and fallout from the incident.

E.
 
I bought my AIR a while back and was immediately spooked by the comments that seemed to suggest the firmware was flakey. I did the dilligent thing and watched lots of videos on how not to lose the drone. I've had the drone out just 3 times and flown for all of 20 mins. During this time I've been prompted to re-calibrate the compass twice and the IMU once. Yesterday before this 3rd flight ever I upgraded the firmware and re-watched the videos. I was sensibly reminded that the firmware can blow away all kinds of settings. This in itself sounds iffy.

What shocked me was the descriptions of the last firmware:

  • Improved firmware data safety.
  • Fixed an issue where GPS location changes inexplicably, and when RTH is abnormal due to compass interference.
  • Optimized vision arithmetic to improve flight performance at both high altitudes or over unclear ground patterns.
  • Improved camera stability.
  • Improved safety for ActiveTrack and SmartCapture.
Questions and observations.

1) What does data safety mean? GDPR? Firmware being corrupted or what?
2) How can (DJI) claim to fix something you find inexplicable?
3) Is the compass interference real or imagined sometimes by your firmware?
4) Any thoughts/suggestions on my compass problems

I do wonder if my prompts to have my compass re-calibrated twice from 3 flights and the stories of fly aways can sometimes be duff firmware and with release comments like this, I'd not be surprised.

On the plus side, I still have my drone and on the minus side, I got a great picture of my 'egg in the nest'

Tom.

Use this combination of FW and software. It is field tested (roughly 10 hours of flight time) and is rock solid. DO NOT allow updates.


App: 4.0.8
Aircraft: 01.03.0700
Remote Controller: 01.03.0700
Flight Database: 00.00.01.04
Basic App Fly Safe Database: 01.00.01.05
Precise App Fly Safe Database: 01.00.01.05

Good luck.

D
 
Is it possible to point newbies like myself to instructions as to how to downgrade FW/software (and also where to find it)? Many thanks.
 
Lycus Tech Mavic Air 3 Case

DJI Drone Deals

New Threads

Forum statistics

Threads
130,588
Messages
1,554,134
Members
159,591
Latest member
Albrecht0803