RepKK Posted at 2018-2-5 15:28
Here are two videos. I put my Spark and Mavic Air outside in 22 degree weather (no batteries) for about 10 minutes. So 10 degrees below freezing. I then put batteries in and raise them up and see the difference. I am going to agree with the cold drone theory for the Mavic AIR. But why doesn’t it do same thing to the Spark?
Yeah, my MA does exactly what yours does in the video...when it's cold. I sort of haven't been able to fly again between rain, the Superbowl, and the acquisition of an MP that someone was getting rid of for cheap. Will try to fly the MA some more tonight.
While Mavic Air's motors have brakes: those vibrations you're seeing is just the motor quickly speeding up and braking constantly. Probably in cold conditions, the motors are slower to be detected and to react..
Have the same issues with shake and vibration - tried changing the gain settings for Roll and Pitch to 90% but still see the issue - after approx 30 sec. it seems to disappear, like the electronics has a comfort operating temperature. (Outside temperature in Denmark -4dg celcius.)
I don't understand that it will take-off when it is not warmed-up. This warming-up is important to reduce gyro drift and overall stability. It is the one thing i like the best about the Naza-M, and the only reason i still use those. With the warming-up the performance is rock solid.
So mostly what we know is this happens in the cold...and is related to gain, imu... I don't know if it's related to so much the new ESC's user a few pist back DJT suggested brakes were the reason well guess what the mpp does not do this soooo....dont think thats it. All I know is something is up and ALL other dji aircraft do not do this. Yes the manual does State not the fly below 32 degrees Fahrenheit but all of my other DJI aircraft have been flown in way lower temperature
For clarity on my post above mpp and mavic air have same esc's ....mpp does not do this in cold temps mavic air does. So "brake theory " is out the window
Happens as the IMU is located differently in it's design compared to the spark or pro. The heat-sink is not near the IMU so it warms up slower. This is apparently better for the longevity for the unit. Regardless, it's int he manual to warm it up so it's an intended design that DJI is aware of I'm sure.
$gambino$ Posted at 2018-2-6 12:53
So mostly what we know is this happens in the cold...and is related to gain, imu... I don't know if it's related to so much the new ESC's user a few pist back DJT suggested brakes were the reason well guess what the mpp does not do this soooo....dont think thats it. All I know is something is up and ALL other dji aircraft do not do this. Yes the manual does State not the fly below 32 degrees Fahrenheit but all of my other DJI aircraft have been flown in way lower temperature
No shake in the promotion video - assume it was cold here too ;-)
I hope one of the moderators forwarded this issue to their engineers for analysis. I believe a little feedback from the company can build strong customer relationships.
So does anyone have a conclusion on this issue? Is this ok in cold weather or should we be concerned? It seems the majority of people are having the same issue.
Epyonz Posted at 2018-2-6 13:02
Happens as the IMU is located differently in it's design compared to the spark or pro. The heat-sink is not near the IMU so it warms up slower. This is apparently better for the longevity for the unit. Regardless, it's int he manual to warm it up so it's an intended design that DJI is aware of I'm sure.
Ok... so the next question is this: If this is just how the Mavic AIR works and the units are NOT to be run in cold weather and it is just about allowing unit to warm up before flying... then why is everyone that opens up a issue report with DJI being told to send the unit back? If we are all just not following the directions - why are they taking the units back when they just should be saying "don't do that and follow the written directions!"
Similar initial vibration here in New England where it was in the 20's (F) here. This is my first drone so I didn't think much of it. It only lasted a few seconds and everything was fine afterwards.
Hi there, for those who have the issue, could you please provide me with the flight data from the drone so I can forward to our engineers to confirm this? Please connect the drone to DJI Assistant 2 v1.2.2 and export the Flight Records and Sensing, upload to Dropbox and paste the link here. Thank you for your help.
SSFRIH Posted at 2018-2-5 17:22
It seems to be a IMU problem causing this but the manual said you should not fly in such cold but a Mavic Pro and Spark and Phantom 4 pro can with no problems. Why?
Hey, can you link where it says this in the manual? I'm having the same issues with the mavic air. Never had this issue with my spark or mavic pro.
I did not notice any shaking before but after reading the post, I did a test in my house again. I took it off and let it hovered. The Air was not freezed but move a bit continuously (like rebalancing), not sure if it is normal in a small indoor room in which the drone is disrupted by the airflow it generated. Or it is the shaking problem you guys experienced too.
And yes the props with white mark can be moved a very bit if I hold the motor but it is tight and force is needed to move
DJI Thor Posted at 2018-2-6 18:28
Hi there, for those who have the issue, could you please provide me with the flight data from the drone so I can forward to our engineers to confirm this? Please connect the drone to DJI Assistant 2 v1.2.2 and export the Flight Records and Sensing, upload to Dropbox and paste the link here. Thank you for your help.
Here is one of my logs that has this issue. Its about the first 30s of my flight or so.
So I have a replacement ticket open on this...considering it seems like all Mavic Airs are affected, im assuming there is not point in returning the drone for just this issue alone?
Just wanted say I experience the same thing. I can normally just 'fly through' it. Meaning , get a touch of air, move around a bit then things seem to stabilize. I did try a second set of props and they def suffer less from this issue.
Thinking of canceling my order. If this is due to cold weather, then why are all other DJI products dont have the same issue? On my P4 I have never experienced this.
I hadn't seen this on my MA, but up until I saw this thread yesterday I hadn't started the MA and taken off straight away. Doing a manual takeoff I haven't seen this phenomena but then its probably because the MA was warming up a bit first. Yesterday I did a test and turned the MA on and straight away did a auto takeoff and just after it launched it did do the shake. I will try again when I have more time over the weekend perhaps and see if I can nail down what circumstances it happens in.
I tested warming up the MA prior to flying and It worked. I experienced no shaking whatsoever.
- I let the MA internals warm up inside for a couple of minutes, I waited till I heard the higher RMP fan speed engage
- Turned it off and went outside
- Turned it back on and was able to fly immediately without the scary wobble
I am ok with doing this, and I havent tested letting the MA warm up while outside, I may try that today
Crazyant Posted at 2018-2-7 04:25
Thinking of canceling my order. If this is due to cold weather, then why are all other DJI products dont have the same issue? On my P4 I have never experienced this.
It's a bit forward to be thinking of this IMO. It's a drone that has just been released, only has one or two firmware updates, is still being worked on and no doubt will have an error or two.
The Mavic pro had an issues or two when it also came out, and they got fixed by DJI. Pretty sure the same will happen here with the Air, as DJI would be risking too much loss if they didn't sort this out pretty quick. I'd say hold for now and see what happens. But of course, its your call
Single Lens Posted at 2018-2-7 04:59
It's a bit forward to be thinking of this IMO. It's a drone that has just been released, only has one or two firmware updates, is still being worked on and no doubt will have an error or two.
The Mavic pro had an issues or two when it also came out, and they got fixed by DJI. Pretty sure the same will happen here with the Air, as DJI would be risking too much loss if they didn't sort this out pretty quick. I'd say hold for now and see what happens. But of course, its your call
I agree, if this is an issue that can be fixed with a firmware update. If this is not a firmware fixable issue but more of a hardware issue, then it means that we would have to wait for the next production batch.
I will test mine as soon as possible (just delivered). It really sounds like an IMU / gain issue. Has anyone tried it with a lower gain? It currently is around 0c here so it should be testable. A lower gain could mean a slight reduction in power consumption also....
i remember when i had to put my p3p in the fridge for an hour then do an imu calibration soon after. otherwise the drone would take a few minutes to warm up before being able to fly, agonizing when even a battery swap means waiting again.
i just did my first flight test indoors and hovered and learned to fly when the drone didn't want to move anywhere because of obstacles lol... i will try outdoors when the sun comes up , only -10c right now.
Tadango Posted at 2018-2-7 05:56
I will test mine as soon as possible (just delivered). It really sounds like an IMU / gain issue. Has anyone tried it with a lower gain? It currently is around 0c here so it should be testable. A lower gain could mean a slight reduction in power consumption also....
DJI Thor Posted at 2018-2-6 18:28
Hi there, for those who have the issue, could you please provide me with the flight data from the drone so I can forward to our engineers to confirm this? Please connect the drone to DJI Assistant 2 v1.2.2 and export the Flight Records and Sensing, upload to Dropbox and paste the link here. Thank you for your help.
Oh dear, this is bad news. I didn't counted the people having issues in here, but to me this thing seems like a common behavior in the cold.
It could be either software or hardware related. Maybe software, because the regulation of sending power to the motors is not calibrated. Hardware issue might be that the IM unit (IMU) is allergic against cold temperature. That would mean, the electronics are too cold to compute the right way. After they got warm, they are working fine. I'm very very scared that this can't be solved by firmware updates. This seems to be like a hardware flaw (let's hope I'm wrong).
Another video I saw is leading me to my assumptions:
here he flies in very cold weather. But if you know this youtuber, he knows how to deal with cold weather and is preparing his drones before flying in the cold. So basically he puts the batteries and drones in a warm place until the point he starts to fly. So when he was taking of, the drone and IMU was fine. But after some minutes of flying he got various IMU errors, so I think the drone cooled down too much.
So for me, the IMU is the one to blame, and this seems hardware related. Please proof me wrong.
Like already said, DJI lists their drones with operating temp of 0° up to 40°C. This would mean, we can't really blame them if the drone is acting weird outside that corridor. But for sure, I will return it if it's acting that way. Glad I bought the drone now to make sure it flies in low temperature. Imagine the drone would be released in summer and you'd need half a year to find out, there is such a flaw. So basically I'm happy to be able to test this.
Ya. My first flight after I landed because of the 'wobbling' - the app asked me to calibrate something, I'm pretty sure it was the IMU calibration, but my memory could be amiss. And second take off didn't show the issue, but that could also possibly be attributed to the craft being warmed up at that point.
I wonder if a cold imu calibration would solve this...like we did in our p3 and p4 days basically just let the aircraft sit outside in the cold for 20 minutes then bring it inside to the warm and calibrate it although I don't know if this will work worth a shot..... honestly I think it's more of a gain issue so you guys should be trying to lower it 10 or 20 points on the roll and Pitch axis..... I do also think this is software related and can be fixed in firmware