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

Little help, please. Mavic Air 1 flips over and dives!!!

landshark23

Active Member
Joined
Dec 28, 2018
Messages
27
Reactions
9
Location
Lacey, WA
Bought my Mavic Air 1 about 2.5 years ago. Had 2 batteries and flew less than 10 times on each with no problems. Then we moved to Seattle and Covid hit, so didn't fly it again for 18 months. When I took it out both batteries appeared totally dead. I bought 2 used batteries on the web, which also were low usage (6 and 15 flights) and also found a service in Portland (hyperflightdrones.com) that recovered my 2 "dead" batteries. They all charge up to full, hold the charge, and hold the charge (one solid, one blinking light) after I run them down to about 30%. Have done this with all of them now about 3 times.

Full Disclosure: I run the Maven control program (not DJI) on my iPhone 12 so I can use VR goggles when I fly (I mostly fly non-DJI FPV freestyle drones). I updated Maven, but no firmware on the drone or controller. I did recalibrate the compass before each flight as Maven required. I also started with new Master Airscrew Stealth props but those broke on the 1st 2 flight crashes so I went back to original props. The 4 flight logs included here are just the ones with issues, all with original props. By the end there are a few scrapes on the ends of a couple of these.

All of my flights since I got the batteries sorted have been hovering and slow cruising at low altitude. The problem is the drone rolls over and spirals down, then recovers if I was high enough or crashes if not. My first 2 flights (testing the batteries) encountered this flip-over after hovering and minimal cruising. It flipped over and crashed from about 6-8 feet high. I was near some power lines and an iron fence and was thinking maybe they caused some interference with the compass or radio. Didn't include those logs cause then I went to our club airfield (large hay field) where I don't have those issues.

I've never looked at any flight logs before this, so if I am not including all the needed info, please just ask. I used my free account on airdata.com. (Bat1 and Bat2 bought on web, Bat3 and Bat4 were my originals that were recovered from too low battery voltage.)

Nov. 17, 2021: Bat1 had 5 flight logs with no indicated problems (short test flights).
Nov. 17, 2021 12:29 -The 6th log starts at 6 min 45 sec and 67% bat. At 11 min. 36 sec. the drone roles over and there is a fast drop in altitude (15'on bar. 6' on sonar). At 11 m. 36.8 s. there is a Low Risk notification "Not Enough Force/ESC Error" repeated 4 times. The altitude readings are confusing, but I think this one may have hit the ground slightly, then recovered and I landed. I was never more than 65' from home, 14.2 mph, 20' high, and radio 100%.

Nov. 17,2021: Bat2 had 1 flight log with no incidents
Nov.17 2021 12:43 - The 2nd log starts at 3 m 11.6s. and 83% battery. At 7 m. 19.5 s. (62% battery) I got the same rollover and drop in altitude (17+' on bar. 6' sonar) was going 16 mph and 68' from home radio 100%. At 7 m. 19.5s. got a Medium Risk notification "Detected side shock / possible collision, aircraft is rolling sharply to the left" (there was no collision with anything and the wind was OK). At 7m. 21.1 s. got the "Not Enough Force/ESC Error" repeated 5 times, and at 7 m. 21.6 s. got Motor is blocked/Not Enough Force/ESC Error 28 times. I think this time it hit the ground and landed upside down.

Nov 17, 2021: Bat3 and Bat4 each flew for 14 min. and had no warnings or incidents. These were primarily to run down the batteries and so were mostly hovering with top speeds of 8 and 10 mph. I took them both back for more testing 6 days later. see below.

Nov. 23, 2021 1:30: Bat3 1st flight. At 5 m. 4.8 s. got Medium Risk notification "Detected side shock /possible collision, aircraft is rolling sharply to the left" (there was no collision). At 5 m. 5.2 s. got Medium Risk notification "Detected Front shock /possible collision, aircraft is pitching sharply forward" (there was no collision). Sonar indicated a 12' drop and Barometer a 15' drop in altitude. I think it crashed. No Motor Blocked or ESC Error notifications. Battery still at 74%. Max speed was 11 mph.

Nov 23, 2021 1:37: Bat3 2nd log starts at 5 m. 14 s. At 9 m. 59.6 s. (battery 47%) got Medium Risk notification "Detected side shock /possible collision, aircraft is rolling sharply to the left" (there was no collision). At 10 m. 0 s. got Medium Risk notification "Detected Front shock /possible collision, aircraft is pitching sharply forward" (there was no collision). Dropped about 20' but was high enough to recover. At 13.5 min. I landed it at 27% battery. Top speed was 12 mph and farthest distance 67'.

Nov. 23, 2021: Bat4 flew with no warnings and no problems. Max speed was 16 mph, max distance was 164 ', max height 33'. Couldn't get it to fail.

Any suggestions or ideas why this thing flips over would be greatly appreciated.
 

Attachments

  • DJIFlightRecord_2021-11-17_12-29-41.txt
    723.6 KB · Views: 2
  • DJIFlightRecord_2021-11-17_12-43-23.txt
    674.4 KB · Views: 1
  • DJIFlightRecord_2021-11-23_13-30-56.txt
    843.7 KB · Views: 2
  • DJIFlightRecord_2021-11-23_13-37-50.txt
    1.2 MB · Views: 4
...Full Disclosure: I run the Maven control program (not DJI)
None of the logs you've provided can be read either by Airdata.com, PhantomHelp.com or be opened in CsvView.

But besides that ... even if they could have been read they most probably hadn't been able to shed any light over this. The reason is that they are created by the Maven app ... all 3:rd party apps usually creates a mobile device log with far less content compared to the DJI GO4 app & furthermore they don't create that second mobile device .DAT log which contains all raw sensor & motor data ... which I strongly suspect we will need in your case.

I advise you to install the latest GO4 app and use that until you can log another flight with these hiccup's

If you can't find the latest DJI apps in the app stores (believe it's mostly an issue if you use android), you need to go to the DJI site, download it there & side load it to your mobile device. Go here for the GO4 app --> DJI GO 4 - Download Center - DJI
 
  • Like
Reactions: landshark23
None of the logs you've provided can be read either by Airdata.com, PhantomHelp.com or be opened in CsvView.

But besides that ... even if they could have been read they most probably hadn't been able to shed any light over this. The reason is that they are created by the Maven app ... all 3:rd party apps usually creates a mobile device log with far less content compared to the DJI GO4 app & furthermore they don't create that second mobile device .DAT log which contains all raw sensor & motor data ... which I strongly suspect we will need in your case.

I advise you to install the latest GO4 app and use that until you can log another flight with these hiccup's

If you can't find the latest DJI apps in the app stores (believe it's mostly an issue if you use android), you need to go to the DJI site, download it there & side load it to your mobile device. Go here for the GO4 app --> DJI GO 4 - Download Center - DJI
Thanks for your response. I will get the GO4 app tomorrow (1:00 am here now) and redo my testing as soon as it stops raining. This is Seattle so that may take a few days.

I didn’t understand your comment that the logs couldn’t be read by airdata.com. I uploaded them from Maven and airdata read them fine. Then I uploaded them again from Maven to my pc and attached to the post. If those are not in the correct format, maybe I can give you access to my airdata account or send them to you from the account? If they don’t have sufficient data anyway, it’s ok. Just offering the idea in case weather delays getting the GO4 logs.
 
Yeah ... share a link to Airdata directly to one of the flights where you had this problem & I can confirm if a GO4 app created DAT log is needed (which I strongly suspect).
 
Bought my Mavic Air 1 about 2.5 years ago. Had 2 batteries and flew less than 10 times on each with no problems. Then we moved to Seattle and Covid hit, so didn't fly it again for 18 months. When I took it out both batteries appeared totally dead. I bought 2 used batteries on the web, which also were low usage (6 and 15 flights) and also found a service in Portland (hyperflightdrones.com) that recovered my 2 "dead" batteries. They all charge up to full, hold the charge, and hold the charge (one solid, one blinking light) after I run them down to about 30%. Have done this with all of them now about 3 times.

Full Disclosure: I run the Maven control program (not DJI) on my iPhone 12 so I can use VR goggles when I fly (I mostly fly non-DJI FPV freestyle drones). I updated Maven, but no firmware on the drone or controller. I did recalibrate the compass before each flight as Maven required. I also started with new Master Airscrew Stealth props but those broke on the 1st 2 flight crashes so I went back to original props. The 4 flight logs included here are just the ones with issues, all with original props. By the end there are a few scrapes on the ends of a couple of these.

All of my flights since I got the batteries sorted have been hovering and slow cruising at low altitude. The problem is the drone rolls over and spirals down, then recovers if I was high enough or crashes if not. My first 2 flights (testing the batteries) encountered this flip-over after hovering and minimal cruising. It flipped over and crashed from about 6-8 feet high. I was near some power lines and an iron fence and was thinking maybe they caused some interference with the compass or radio. Didn't include those logs cause then I went to our club airfield (large hay field) where I don't have those issues.

I've never looked at any flight logs before this, so if I am not including all the needed info, please just ask. I used my free account on airdata.com. (Bat1 and Bat2 bought on web, Bat3 and Bat4 were my originals that were recovered from too low battery voltage.)

Nov. 17, 2021: Bat1 had 5 flight logs with no indicated problems (short test flights).
Nov. 17, 2021 12:29 -The 6th log starts at 6 min 45 sec and 67% bat. At 11 min. 36 sec. the drone roles over and there is a fast drop in altitude (15'on bar. 6' on sonar). At 11 m. 36.8 s. there is a Low Risk notification "Not Enough Force/ESC Error" repeated 4 times. The altitude readings are confusing, but I think this one may have hit the ground slightly, then recovered and I landed. I was never more than 65' from home, 14.2 mph, 20' high, and radio 100%.

Nov. 17,2021: Bat2 had 1 flight log with no incidents
Nov.17 2021 12:43 - The 2nd log starts at 3 m 11.6s. and 83% battery. At 7 m. 19.5 s. (62% battery) I got the same rollover and drop in altitude (17+' on bar. 6' sonar) was going 16 mph and 68' from home radio 100%. At 7 m. 19.5s. got a Medium Risk notification "Detected side shock / possible collision, aircraft is rolling sharply to the left" (there was no collision with anything and the wind was OK). At 7m. 21.1 s. got the "Not Enough Force/ESC Error" repeated 5 times, and at 7 m. 21.6 s. got Motor is blocked/Not Enough Force/ESC Error 28 times. I think this time it hit the ground and landed upside down.

Nov 17, 2021: Bat3 and Bat4 each flew for 14 min. and had no warnings or incidents. These were primarily to run down the batteries and so were mostly hovering with top speeds of 8 and 10 mph. I took them both back for more testing 6 days later. see below.

Nov. 23, 2021 1:30: Bat3 1st flight. At 5 m. 4.8 s. got Medium Risk notification "Detected side shock /possible collision, aircraft is rolling sharply to the left" (there was no collision). At 5 m. 5.2 s. got Medium Risk notification "Detected Front shock /possible collision, aircraft is pitching sharply forward" (there was no collision). Sonar indicated a 12' drop and Barometer a 15' drop in altitude. I think it crashed. No Motor Blocked or ESC Error notifications. Battery still at 74%. Max speed was 11 mph.

Nov 23, 2021 1:37: Bat3 2nd log starts at 5 m. 14 s. At 9 m. 59.6 s. (battery 47%) got Medium Risk notification "Detected side shock /possible collision, aircraft is rolling sharply to the left" (there was no collision). At 10 m. 0 s. got Medium Risk notification "Detected Front shock /possible collision, aircraft is pitching sharply forward" (there was no collision). Dropped about 20' but was high enough to recover. At 13.5 min. I landed it at 27% battery. Top speed was 12 mph and farthest distance 67'.

Nov. 23, 2021: Bat4 flew with no warnings and no problems. Max speed was 16 mph, max distance was 164 ', max height 33'. Couldn't get it to fail.

Any suggestions or ideas why this thing flips over would be greatly appreciated.
Just as a WAG from previous postings here on the forum - any time a drone like a Mavic flips - it tends to be a motor issue; as Mavics typically do not flip any other time. The "death spiral" sounds a lot like the motor issue as well, as reported by others on the forum with such issues. You can search the forums for those issues and see what pops up as well.

@slup or one of the other file guru's will be able to tell you more when they have files that can be read for certain.

Just as a test, crank the drone up and watch the props / motors to see if one "seems" or "is" spinning differently than the others. May be too hard to tell.

Sad, that you've had these issues and let Covid stop you from flying. I never stopped and it was typically better flying as far fewer people were out.
 
Checked the log from post #7 ... that shows your MA1 suddenly pitch up & roll to the left at 599sec into the flight, with the sticks in mainly neutral position (value 1024 in below chart). This strongly indicate that the thrust is lost in the rear left corner ... this state recovers just 4sec later.

1638296261406.png

Also saw this in the message stream ...

Message stream.jpg

This can indicate that a mobile device .DAT log ending with FLY067.DAT actually is written.

Go here :

iOS:
With iOS devices you need to access the app files via computer, either using iTunes or with a file system browser such as iExplorer. The DAT logs are in Apps » DJI GO 4 » FlightRecords » MCDatFlightRecords.

Android:
With Android devices the file system should mount when plugged into a Windows machine, or via various software options on a Mac. The DAT logs are in DJI GO 4 » FlightRecords » MCDatFlightRecords.

If you find a DAT log there that ends with FLY067.DAT ... retrieve it & attach it directly in a new post here in this thread. If no file is found it's back to install the GO4 app & use that until it happens again ...
 
  • Like
Reactions: 4006448
"If you find a DAT log there that ends with FLY067.DAT ... retrieve it & attach it directly in a new post here in this thread. If no file is found it's back to install the GO4 app & use that until it happens again ..."

I got a confirmation this am from the Maven web site - Maven does not create DAT logs. So, I have loaded the latest DJI Go 4 app on my phone, and will get out and retest today if the weather holds.
 
Ok , was able to fly yesterday on the DJI Go 4 app. and have all the logs. Flew 4 batteries and created 8 .txt flight records. The problem occurred on battery 2 and 4 (not on 1 or 3, which I flew in Sport mode but still could not get it to flip). So problem has occurred now on all 4 batteries.

The 3 DJI .txt logs with errors are in order attached::
1st 12 min. of bat2 (crashed upside down), last 2 min. of bat2 (flipped over but recovered and landed), 1st 6 min of bat4 (crashed upside down).

For the .dat files, I tried to associate the 10 .dat files to the 8 .txt files. I even loaded them to airdata.com, but the error notifications are not repeated, and neither the file names nor the clock times they were stored make it clear to me. My best guess based on last modified times are the 1st 7 .dat are for Bat1 (I was updating FW and doing a couple takeoff/landings to test the DJI pgm). 15-56-21 is for Bat2, 16-19-14 is Bat3, and 16-38-29 is Bat4.

If there is some better identifier I can use to corrrelate I'll be glad to try. I include all 10 here with my apologies. It may be harder to correlate without all 8 .txt files, so I'm glad to correlate them if someone can give me the way.

Thanks again for the help.
 

Attachments

  • 2021-12-01_15-55-03_FLY075.DAT
    74.9 KB · Views: 1
  • 2021-12-01_15-55-30_FLY076.DAT
    155.8 KB · Views: 1
  • 2021-12-01_15-56-21_FLY077.DAT
    6.5 MB · Views: 1
  • 2021-12-01_16-19-14_FLY078.DAT
    5.3 MB · Views: 1
  • 2021-12-01_16-38-29_FLY079.DAT
    5.9 MB · Views: 2
  • 2021-12-01_15-52-47_FLY074.DAT
    712.8 KB · Views: 1
  • 2021-12-01_15-34-53_FLY073.DAT
    5.4 MB · Views: 1
  • 2021-12-01_15-34-09_FLY072.DAT
    78.2 KB · Views: 1
  • 2021-12-01_15-31-29_FLY071.DAT
    854.5 KB · Views: 1
  • 2021-12-01_14-25-57_FLY070.DAT
    3.2 MB · Views: 1
  • DJIFlightRecord_2021-12-01_[16-39-24].txt
    628.6 KB · Views: 1
  • DJIFlightRecord_2021-12-01_[16-11-02].txt
    213.4 KB · Views: 1
  • DJIFlightRecord_2021-12-01_[15-57-30].txt
    1,000.9 KB · Views: 1
Looked into DJIFlightRecord_2021-12-01_[16-11-02].txt where this problem showed up... and the corresponding DAT log FLY077.DAT.

In below chart, where the incident starts up at 923,7sec, you have just released the forward flight stick (black) ... & your craft starts to pitch up for braking (green) & the heading speed (light blue) starts to go down. But just there all goes haywire as the craft also fast rolls to the left (blue) ... & the movement mainly turns your craft's belly away from the ground so the VPS sensors actually lose the height reading momentarily (magenta).

(Click on the charts below to make them larger)
1638524483103.png

And as I suspected, your craft loses thrust in the rear left corner. In below chart you again have the magenta VPS height as a reference only ... but just when thing's goes bad the flight controller commands 100% (blue) on the rear left corner to counteract the uncommanded pitch & roll. But looking on the RPM for that motor (red) doesn't show max rev. at all ... instead a choppy increase not reaching max at all.
1638524867872.png

So we clearly have a loss of thrust in the rear left corner ... all craft movements starting off this incident confirms that. The loss of thrust, can be due to 2 thing's ... either a prop failure or a motor that doesn't perform due to some reason.
We see that in this case the motor doesn't do what the flight controller commands ... it doesn't go full rev. to counteract the movement... so we have a bad performing motor.

The reason for the bad performing motor can be several ... but the easiest to check is if it's blocked in some way. If the motor is commanded to increase the RPM but can't turn properly that will create an increase in amp draw. And that is exactly what we see here ... the amp draw spikes way higher than the other 3 motors.

Your rear left motor is bad ... either it's a seizing bearing, or you have a ding in the bell making it come on contact with the magnets, a bent motor shaft, again making the magnets come in contact with the bell... or you have something lose inside the motor, jamming it intermittently.

1638525281726.png
 
  • Like
Reactions: BudWalker
slup,

Wow, that's a very impressive analysis! I'll take the motor apart and see if anything falls out (not expected). If it was any other drone than a DJI, I might try and replace the motor (I build my own freestyle drones), but I'm not going to crack this thing apart. I've never really flown this drone other than a few test flights, so I won't miss it. Looks like it's headed for sale for parts and the batteries.

Thank you so much for your time and analyzing the motor problem. I could never have figured it out.
 
Lycus Tech Mavic Air 3 Case

DJI Drone Deals

New Threads

Forum statistics

Threads
131,095
Messages
1,559,761
Members
160,078
Latest member
svdroneshots