Mavic Pro - acting strange

[AnotherMavicPilot]

Hi there,

I was out flying my drone yesterday for the first time in a few months and experienced some odd behavior, I would like someone to try and make sense of the logs if possible.

A brief description of what happened:

I flew my drone to the top of a local hill to get some shots of a new mountain bike track that had been recently made. A few minutes into the flight I got an obstacle avoidance warning, despite being high enough to avoid any obstacles.

So I flew back home to my position to take a closer look at the drone. I hovered the drone a couple of feet in front of me and rotated it around just to confirm there was nothing obstructing the front sensors.

At this point, my drone disconnected from both the tablet and controller, even though it was only a few feet in front of me. I held down the left stick to try and get the drone to land, but it ignored my inputs. Eventually, it reconnected and I was able to land my drone.

This experience left me feeling uncertain about what was happening and so I didn't try to fly it again, I'm now concerned it may happen again and so I'm reluctant to attempt another flight.

A few things to note:

I changed the date on my tablet to the 13/06/19 in order to get the maps working within the Dji App (I've never had problems with this in the past).

I'm using the Titan Atlas antenna upgrade, antenna battery was charged.

I've experienced disconnects previously on just the tablet only, I've put this down to a possible faulty USB cable, I was using that same cable.


I'm hoping someone with more experience can help make sense of the logs and hopefully give me a better idea of what went wrong, so I can avoid this in the future.

Many thanks.

Link to my log on Google Drive:

FLY327.DAT

drive.google.com

 

[sar104]

Did you intend to post the flight log?

 

[AnotherMavicPilot]

FLY327.DAT

 

[AnotherMavicPilot]

Did you intend to post the flight log?

Yeah tried to upload my log, must be too big, so I've included a Google Drive link, thanks.

 

[new2mavic]

From 498 ~ 525 seconds after the starting of the motors, the radio signal strength went to zero. I think this is the uplink signal as the downlink was working OK. The receiving and recording of telemetry data such as motor speed and VPS height were not interrupted.

You mentioned that you have applied throttle down control during the black-out period but this does not show up in the chart. May be the throttle signal here represents that received by the drone, not that transmitted by the RC. If the uplink is broken, the drone not being able to receive the control signals is expected.



May be there is some intermittent problem with the antenna upgrade ?

Regarding the obstacle avoidance message, I have seen it a few times when flying my M2P 200 m above water but I have never seen it again since that flight. May be it was due to some insects flying across the drone ?

 

[sar104]

The uplink failed. The reason isn't clear but I'd certainly be wary of non-OEM "upgrades".

498.024 : 33935 [L-RC]rc_connect:1=>0
498.024 : 33935 [L-RC]rc_func_disable @ [684342 ms]
502.585 : 34163 [L-FMU/LED]action changed. rc completely lost0)
504.185 : 34243 [L-FLYMODE][Ctrl<6>] REQ_RC_LOST AUTO_LANDING ctrl_auto_landing
504.985 : 34283 [L-RC]Start landing. rc vib!
525.805 : 35324 [L-RC]rc_connect:0=>1
525.805 : 35324 [L-RC]rc_resume @ [712122 ms]
525.805 : 35324 [L-RC]set rc_resume_lock

 

[AnotherMavicPilot]

Thank you both for your input, I've never experienced a problem with the Atlas previously and have had it for two years, but I understand your concern. I will double-check all connections on the antenna to make sure all connectors etc are good.

Just curious, if the power or connections to the antenna system failed whilst the drone was a couple of feet in front of me, would I get a disconnect? I would assume that since the drone is still right next to me, there would still be an adequate signal from the controller to the drone, would I be correct in that assumption?

I should also add that while the drone was hovering in front of me, I noticed quite considerable drift, it seemed to be drifting up and down by about 4 to 8 inches, I've never noticed this behavior before. I then tried to grab it lightly by its legs and pull slightly to test if it was responding to my resistance and to see if it attempted to hold its position, which it did. I now wonder upon reflection if this was enough for it to cause interference with the signal, as I understand the antenna for the drone is situated in the front legs.

Another thing to add would be that the sun was setting, so the ambient light wasn't all that great, would this be a possible cause for the false positive of the object avoidance warning and the slight drifting of the drone?

Thanks again to you both for taking the time to look at my logs, it's much appreciated!

 

[new2mavic]

Just curious, if the power or connections to the antenna system failed whilst the drone was a couple of feet in front of me, would I get a disconnect? I would assume that since the drone is still right next to me, there would still be an adequate signal from the controller to the drone, would I be correct in that assumption?

Yes, at that close distance, the radio signal leaking out the RC should probably be strong enough to reach the drone although this has not be tested. It is possible that the transmitter on the RC side or the receiver on the drone side got a glitch of some sort during that 30 second black-out period.

I should also add that while the drone was hovering in front of me, I noticed quite considerable drift, it seemed to be drifting up and down by about 4 to 8 inches, I've never noticed this behavior before. I then tried to grab it lightly by its legs and pull slightly to test if it was responding to my resistance and to see if it attempted to hold its position, which it did. I now wonder upon reflection if this was enough for it to cause interference with the signal, as I understand the antenna for the drone is situated in the front legs.

Another thing to add would be that the sun was setting, so the ambient light wasn't all that great, would this be a possible cause for the false positive of the object avoidance warning and the slight drifting of the drone?

I doubt very much if holding the leg can cause a drop in radio signal from 100% to zero.

The VPS sensors ( ultrasonic sensor + camera at the belly ) did detect some drifting in altitude ( about 60 cm ) when there was no throttle input. Just don't understand why the flight controller did not react.

Intresting enough, when you pulled it with your hand, the altitude was maintained unexpectedly well. It seems that the drone reacted better to quick change in vertical acceleration but not slow drifting in altitude. I cannot offer any explanation here.

 

[FoxhallGH]

Another thing to add would be that the sun was setting, so the ambient light wasn't all that great, would this be a possible cause for the false positive of the object avoidance warning and the slight drifting of the drone?

Just a question about this aspect ... Was the drone flying 'toward' the setting sun when the obstacle avoid warning triggered? That part of all this could have just been the sensors reacting to the sun???

 

[AnotherMavicPilot]

Just a question about this aspect ... Was the drone flying 'toward' the setting sun when the obstacle avoid warning triggered? That part of all this could have just been the sensors reacting to the sun???

No, it wasn't. In fact, the sun was already down behind a hill by the time I started my flight (I live in a valley)

 

[AnotherMavicPilot]

Yes, at that close distance, the radio signal leaking out the RC should probably be strong enough to reach the drone although this has not be tested. It is possible that the transmitter on the RC side or the receiver on the drone side got a glitch of some sort during that 30 second black-out period.



I doubt very much if holding the leg can cause a drop in radio signal from 100% to zero.

The VPS sensors ( ultrasonic sensor + camera at the belly ) did detect some drifting in altitude ( about 60 cm ) when there was no throttle input. Just don't understand why the flight controller did not react.

Intresting enough, when you pulled it with your hand, the altitude was maintained unexpectedly well. It seems that the drone reacted better to quick change in vertical acceleration but not slow drifting in altitude. I cannot offer any explanation here.

View attachment 96635

Thanks for your detailed analysis. I think the best thing for me to do, would be to update the firmware on both my drone and controller, then do a new calibration, then do a small test flight in the open and not fly it far, if I do this with all 3 of my batteries and there's no more disconnects or drifting, I should be good to go.

Once the weather gets a bit better here, I'll give it another go and I'll report back here my findings.

Again, many thanks for your help.

 

[FoxhallGH]

No, it wasn't. In fact, the sun was already down behind a hill by the time I started my flight (I live in a valley)

Another idea then ... A valley may reduce the GPS sat's available for an accurate positional fix. That may have contributed to the 'drift' you have mentioned ... Do we have any indication of how many GPS sat's were locked during that part of the flight??

 

[AnotherMavicPilot]

Another idea then ... A valley may reduce the GPS sat's available for an accurate positional fix. That may have contributed to the 'drift' you have mentioned ... Do we have any indication of how many GPS sat's were locked during that part of the flight??

Yeah, it's not that sort of Valley haha.

I mean I'm surrounded by hills living in Wales, the hill the sun went behind is prob 20 miles away. I connect to plenty of sat's (15) every time I fly, have you looked at the logs? It's not a GPS related issue, I'm almost certain of that. I've never had any issues with the number of satellites I connect to and certainly wouldn't attempt a flight if I didn't have good sat coverage.

But thank you for your suggestion, none the less.

 

[Slade]

Lots of helpful info. I’m always wary of non-OEM accessories. These toys are too expensive for me to take chances. Happy flying?

 

[AnotherMavicPilot]

Lots of helpful info. I’m always wary of non-OEM accessories. These toys are too expensive for me to take chances. Happy flying?

Just to add to this, I've never experienced any issues with my antenna in the 300+ flights I've made with this modded antenna.

The antenna cost me £379, so this was an expensive mod to fly my expensive toy...

 

[Sixgun89]

Just remember anything added reduces the MTBF of any system. Mean Time Between Failure will go down when parts are added to the equation. Try to take the antenna upgrade out of the system to test its impact. 300 flights doesn’t mean it can’t fail.

 

[Galeforce1]

Why don't you try taking everything back to stock? Remove the add-on antenna, update the app and maps if necessary, reset the app settings, and re-calibrate everything. These are highly complex machines and there are always bugs in the software and the possibility of errors in the settings. It is a flying computer + camera, after all. After 300+ flights, it becomes more of a probability.