Mav3 gains altitude in spite of downward stick

[akdrone]

Today I was flying my Mav3 down a country road (no cars). It was very foggy at 100 feet but mostly clear at the ground. I got up 75 feet or so at one point - high enough to realize the camera lens was being coated with ice ( it was +8F) and decided to return. I was out maybe 500feet. Upon returning my Mav3 would gain altitude slowly even with no left stick input. If I pushed the stick down to descend it would continue to rise. When I stopped it would descend when I told it to do so, then I'd move forward with no left stick and it would rise on it's own. This was repeated 4 or 5 times until I got back. The lens was indeed covered with ice and there was some minor build up on the props and legs. Very little. I'm curious if anyone else has experienced this kind of an issue. I will fly the Mav3 again tomorrow or asap with no fog...

 

[Phantomrain.org]

Today I was flying my Mav3 down a country road (no cars). It was very foggy at 100 feet but mostly clear at the ground. I got up 75 feet or so at one point - high enough to realize the camera lens was being coated with ice ( it was +8F) and decided to return. I was out maybe 500feet. Upon returning my Mav3 would gain altitude slowly even with no left stick input. If I pushed the stick down to descend it would continue to rise. When I stopped it would descend when I told it to do so, then I'd move forward with no left stick and it would rise on it's own. This was repeated 4 or 5 times until I got back. The lens was indeed covered with ice and there was some minor build up on the props and legs. Very little. I'm curious if anyone else has experienced this kind of an issue. I will fly the Mav3 again tomorrow or asap with no fog...

So what is most likely happening is the landing sensor on the bottom is getting confused with ice build up or just being covered by the snow or ice can cause this to happen.

This is why the Wet Suits we make for Mavic 3 Protect the Landing sensor so it does not get confused by heavy snow or Ice/ Rain causing the drone to not want to come down or forcing it to land .

If you want to fly in Extreme conditions you might consider Protecting the Sensors from getting confused this is why we put so much Focus with the Wet Suit on protecting the underside of the drone.

Unfortunately there is not much you can do to stop ice build up on the props so you do have to watch that but we made a video showing just how much build up you can have and still maintain your flight .

Phantomrain.org
Gear to fly in the Rain. Land on the Water. Flying in the Ice and Rain.

 

[old man mavic]

@akdrone ,most likely cause was the sensors being fooled,by the weather conditions you were flying in

 

[Vic Moss]

As others said, weather (& likely icing on the sensors) would very likely cause this.

 

[Yorkshire_Pud]

I too thought "ice on the sensors" but have just had another thought, how much heat does the bottom shell get in flight i.e. would it be enough to keep the bottom sensors ice & condensation free?
I think the bottom of at least one of my drones has been warm after a flight but I can't remember which drone, if any, it was.

 

[old man mavic]

it would not necessarily be actual ice on the sensors lens, but just simply condensation misting ,because of the temperature ,difference ,because moving forwards caused the drone to rise this would indicate ,the system trying to compensate for what it thought was something below the drone ,the same way as it does if flying close to the ground coming up a slope ,the drone will rise up to follow the terrain ,even if it is not in APAS mode,

 

[Yorkshire_Pud]

Agreed, but if the bottom shell gets warm it is unlikely to suffer condensation.....isn't it?

 

[JoelP]

Today I was flying my Mav3 down a country road (no cars). It was very foggy at 100 feet but mostly clear at the ground. I got up 75 feet or so at one point - high enough to realize the camera lens was being coated with ice ( it was +8F) and decided to return. I was out maybe 500feet. Upon returning my Mav3 would gain altitude slowly even with no left stick input. If I pushed the stick down to descend it would continue to rise. When I stopped it would descend when I told it to do so, then I'd move forward with no left stick and it would rise on it's own. This was repeated 4 or 5 times until I got back. The lens was indeed covered with ice and there was some minor build up on the props and legs. Very little. I'm curious if anyone else has experienced this kind of an issue. I will fly the Mav3 again tomorrow or asap with no fog...

You post has me worried, since you are a very experienced drone pilot with lots of drones, so when you tell me there is an issue I take it as a fact. The Mavic 3 is rated for very low temperature and we see posts of it flying in very low temperature. It seems like you were high enough that it shouldn’t have been looking to the optical sensors for landing suitability. I understand that the IMU primarily judges altitude from a barometric sensor. Is it possible that that has an orifice that became clogged with ice? Is it possible that collision avoidance saw fog to be dense enough to avoid? Would a data log tell us something about what the IMU was doing?

 

[Crash Pro]

Just out of curiosity sake, what if you hit the return to home button. Will the drone come back on its own normally or would you encounter issues due to icing?

 

[akdrone]

I don't think the sensors on the bottom were iced over. I could have missed that because that does sound likely. The short flight was in temps clearly well below the specs that DJI sets since it was only +8F. I have, however, flown my M3 is worse icing conditions than this and in much much colder conditions which is why I was puzzled. We don't get fog or near ground fog very often but I next time I have a chance to fly in sub freezing fog conditions I'll do so and I'll roll my eyes if my bird flies away . Attached is the log of the flight in question. As I look at it it seems to me the Barometric pressure may have been fluctuating. I did get multiple "Unstable lighting conditions. Vision system and obstacle sensing unavailable. Fly with caution" notices. Forward sensors were clearly iced over. (Or NOT clear...) Either it was sensors or the rapidly fluctuating Barometric pressure was fooling things. I am attaching the log file and if you are bored you can look at 3' 3.8sec and see the left stick down while altitude goes up. At 3min 9.4sec the speed goes to zero (or so) and the drone goes down as requested. Rapidly fluctuating barometric pressure due to fog above or blocked sensors. Dunno. I DO know I don't like it when Mav does not behave as expected! The particularly odd things is the problem only existed while moving forward. When I stopped forward motion the drone stopped climbing and my left stick worked as expected. Very odd.... very odd...

 

[akdrone]

You post has me worried, since you are a very experienced drone pilot with lots of drones, so when you tell me there is an issue I take it as a fact. The Mavic 3 is rated for very low temperature and we see posts of it flying in very low temperature. It seems like you were high enough that it shouldn’t have been looking to the optical sensors for landing suitability. I understand that the IMU primarily judges altitude from a barometric sensor. Is it possible that that has an orifice that became clogged with ice? Is it possible that collision avoidance saw fog to be dense enough to avoid? Would a data log tell us something about what the IMU was doing?

I think all of that is possible...and the data log is now uploaded.

 

[akdrone]

Just out of curiosity sake, what if you hit the return to home button. Will the drone come back on its own normally or would you encounter issues due to icing?

thought about trying it but it would have taken me right up into the heavy fog as my return height was well into the fog and I didn't want to bother changing it.

 

[veryzebra]

I don't think the sensors on the bottom were iced over. I could have missed that because that does sound likely. The short flight was in temps clearly well below the specs that DJI sets since it was only +8F. I have, however, flown my M3 is worse icing conditions than this and in much much colder conditions which is why I was puzzled. We don't get fog or near ground fog very often but I next time I have a chance to fly in sub freezing fog conditions I'll do so and I'll roll my eyes if my bird flies away . Attached is the log of the flight in question. As I look at it it seems to me the Barometric pressure may have been fluctuating. I did get multiple "Unstable lighting conditions. Vision system and obstacle sensing unavailable. Fly with caution" notices. Forward sensors were clearly iced over. (Or NOT clear...) Either it was sensors or the rapidly fluctuating Barometric pressure was fooling things. I am attaching the log file and if you are bored you can look at 3' 3.8sec and see the left stick down while altitude goes up. At 3min 9.4sec the speed goes to zero (or so) and the drone goes down as requested. Rapidly fluctuating barometric pressure due to fog above or blocked sensors. Dunno. I DO know I don't like it when Mav does not behave as expected! The particularly odd things is the problem only existed while moving forward. When I stopped forward motion the drone stopped climbing and my left stick worked as expected. Very odd.... very odd...

That's what I think akdrone. I'm certainly no expert at reading logs but I've experienced the same on days when the barometric pressure was changing rapidly. No rain or fog, took off and should have been flying level at about 15M but instead it continued to rise and would be at 30M. Kept some downward pressure and it eventually worked itself out for the rest of the flight. Since Aug, I've had this happen on at least one flight on 3 different days. Based on where I was flying, the first time I was thinking it was interference but that made less sense and I narrowed it down to days with rapidly changing weather coming.

 

[lookout]

That's what I think akdrone. I'm certainly no expert at reading logs but I've experienced the same on days when the barometric pressure was changing rapidly. No rain or fog, took off and should have been flying level at about 15M but instead it continued to rise and would be at 30M. Kept some downward pressure and it eventually worked itself out for the rest of the flight. Since Aug, I've had this happen on at least one flight on 3 different days. Based on where I was flying, the first time I was thinking it was interference but that made less sense and I narrowed it down to days with rapidly changing weather coming.

Better than watching the weather girl on TV? My weather quad is detecting a change in the weather.

 

[veryzebra]

Better than watching the weather girl on TV? My weather quad is detecting a change in the weather.

Lol, couldn't tell you the last time I watched weather on the tv. Just not a tv guy.

 

[Jamesz]

The barometric pressure would have to fall massively over a really short time to affect a drones horizontal flight. Sensor icing or mist are a far more likely cause.

 

[akdrone]

The barometric pressure would have to fall massively over a really short time to affect a drones horizontal flight. Sensor icing or mist are a far more likely cause.

I wonder, would the white ground (snow) and white sky (fog) contribute to the confusion?

 

[MS Coast]

Unfortunately there is not much you can do to stop ice build up on the props so you do have to watch that but we made a video showing just how much build up you can have and still maintain your flight .

Ice buildup on an aircraft is a very serious issue that should be immediately dealt with. It can progress very quickly and cause a crash. Continuing to operate in icing conditions is not wise.

Manned aircraft have de-icing systems that melt or physically remove ice from the propellers and airframe. Descending immediately to warmer air is typically done at the first sign of any ice, even a patch the size of a fingernail.