Ice_2k Posted at 1-22 05:43
The altitude drop issue was not introduced in 1.0.4 afaik. There are reports of it for 1.0.3 and below. I myself experienced imprecise altitude holding (flying forward 3-4m high over tarmac at full speed in S-mode) and the drone was going closer and closer to the ground. Once I stop, it starts climbing. Sometimes it was the other way around. Very strange behavior.

Sensors need good light to work, flying over tarmac “black” won’t help sensors working correctly, on top of that they need good texture which will also be a problem with tarmac . It also won’t work correctly at speeds over 10 m/s

Ex Machina Posted at 1-22 06:02
FWIW I'm pretty sure that kind of progressive altitude loss from horizontally forward flight at sports speeds is normal behavior in a quad -- the AC is tilted in order to move in a given direction, that results in loss of lift especially at high speeds which need to be compensated for by the pilot and/or ground sensors if within their range of operational conditions.

I don’t think the other DJI drones do this

hallmark007 Posted at 1-22 06:23
Altitude and dropping I think it was there all the time, but why is really the question, I tested for this in normal conditions and found no significant drop in altitude, so again I decided to test when weather got cold. So at minus 3/4 I tested and there are noticeable drops and even some increase in altitude involuntary on the Mavic mini, however advice is clearly drawn out in the manual NOTto fly in such conditions, and the reason for this and the conclusion I came to was IMU was feeling the affects of the cold so causing aircraft difficulties in maintaining its correct attitude, I have come across this with the Mavic pro orignal , so am assuming temperature may have something to do with this, also I see many calibrating their IMU indoors if this is done in high temperatures it could have an effect on what temperature IMU is happy flying at.

Just my two pence worth .

Good catch with the IMU being adjusted indoors, I'd like to see that tested and see if that makes any significant difference.

ABeardedItalian Posted at 1-22 07:10
Good catch with the IMU being adjusted indoors, I'd like to see that tested and see if that makes any significant difference.

I’m not certain it will when flying, but I have done extensive testing on the rate of start up time when calibrating IMU in a warm environment .

So if you calibrate indoors and go outside IMU will take longer to warm up as it understands it needs to warm up to temperature it was calibrated in, and if you calibrate it closer to temperature you are likely to fly in it will warm up and be ready to fly that much quicker .

Ice_2k Posted at 1-22 06:33
I don’t think the other DJI drones do this

Yes they do and they all carry the same warnings

WrongWay Feldman Posted at 1-22 04:13
As someone who has been tech support for a number of family members and quite a few business customers, I can't count how many times a customer has told me they already tried that (that could be any steps I tell them to preform) and then when I show up and do exactly what they say they did it solves the problem.

In order for me to properly troubleshoot, especially over the phone or through email, it's important that the customer follows my steps exactly even if the first step is something as simple as restart the device.

While I fully understand that, I also work as tech support for music industry gear... However like some mention before a 500euros gadjet that flies and can crash or cause damage to others, the "reset FW" and see what happens seems vague.

There were known problems with the gimbal for many of us and this was fixed in FW 1.0.3 without DJI ever acknowledging it, many were asked to return their minis even though they were fine, so that's why I would expect some more feedback from support, rather than automated responses.

Ex Machina Posted at 1-22 06:02
FWIW I'm pretty sure that kind of progressive altitude loss from horizontally forward flight at sports speeds is normal behavior in a quad -- the AC is tilted in order to move in a given direction, that results in loss of lift especially at high speeds which need to be compensated for by the pilot and/or ground sensors if within their range of operational conditions.

Well, since the drone’s objective is to maintain altitude unless otherwise instructed by the pilot, your reasoning is a bit flawed. If the pilot *can* compensate for the drop, then the drone could also do it on its own. Unless something was preventing it from correctly measuring the altitude while flying in S mode. Bottom sensors are clearly disabled during S-mode full speed because of the tilt so it’s down to the barometer. I remember a topic on this forum where for a similar situation the user was told by DJI he had a defective barometer and was issued a replacement.

InspektorGadjet Posted at 1-22 09:35
While I fully understand that, I also work as tech support for music industry gear... However like some mention before a 500euros gadjet that flies and can crash or cause damage to others, the "reset FW" and see what happens seems vague.

There were known problems with the gimbal for many of us and this was fixed in FW 1.0.3 without DJI ever acknowledging it, many were asked to return their minis even though they were fine, so that's why I would expect some more feedback from support, rather than automated responses.

The main reason you are asked to refresh FW is simply when installing FW first many don’t preform it correctly, I read in the posts here all the time many saying when updating FW it only reached 95% / 99% and they left it at that.
I don’t think there is anything suspicious going on, so when asked to update it’s just their way of confirming that you have correctly updated.
I have on one occasion been guilty of this myself and at the second time of updating my issue was sorted .

Ice_2k Posted at 1-22 04:57
... not if the problem you're troubleshooting involves an aircraft losing altitude uncontrollably. I don't think "restart and try again" is the smart thing to recommend at that point. What if the user follows your advice and the aircraft falls out of the sky?
The only correct way this should be handled by DJI support is "ground your aircraft & send us the logs to investigate what happened".

I was just addressing the question about why the user would be instructed to refresh the firmware as the initial troubleshooting step.

There have to be steps that are followed in a specific order to properly troubleshoot anything.

I did not realize there was a specific problem with a specific drone that was being asked about. I thought it was a general question about tech support.

If your drone is falling out of the sky for no reason then maybe you should open a support ticket and follow their instructions to the letter.

Ice_2k Posted at 1-22 10:00
Well, since the drone’s objective is to maintain altitude unless otherwise instructed by the pilot, your reasoning is a bit flawed. If the pilot *can* compensate for the drop, then the drone could also do it on its own. Unless something was preventing it from correctly measuring the altitude while flying in S mode. Bottom sensors are clearly disabled during S-mode full speed because of the tilt so it’s down to the barometer. I remember a topic on this forum where for a similar situation the user was told by DJI he had a defective barometer and was issued a replacement.

It happened with the P4 aircraft. It was reasoned at the time that there was an air pressure partial vacuum caused by flying the aircraft at speed close to the ground. When you stopped the aircraft, pressure returned to normal and the aircraft would gradually climb up and resume hover.

Geebax Posted at 1-22 13:38
It happened with the P4 aircraft. It was reasoned at the time that there was an air pressure partial vacuum caused by flying the aircraft at speed close to the ground. When you stopped the aircraft, pressure returned to normal and the aircraft would gradually climb up and resume hover.

I've worked on and with barometric pressure transducers for laboratory use.  These were used in part, to log barometric pressure for compensation of dissolved oxygen measurements (as in BOD and COD analysis).

These transducers cost more than the entire Mini.  So imagine what a bargain the Mini is!

Anyhow, I can certainly imagine what you're talking about happening easily.  It might be particularly bad when zipping along close to the ground, but I can imagine it also happening at altitude, too.  The air rushing over the body of the mini could create pressure differences outside and within the housing that the sensor would report.

It doesn't take much of a difference to appear as an elevation change of a number of feet.  If the drone cannot rely on its downward-looking sensors, and must rely on its barometer (altimeter) readings, you'd really notice these shifts when flying close to the ground.

There might be nothing that can be done in firmware to completely eliminate this effect.

Sigmo Posted at 1-22 15:59
I've worked on and with barometric pressure transducers for laboratory use.  These were used in part, to log barometric pressure for compensation of dissolved oxygen measurements (as in BOD and COD analysis).

These transducers cost more than the entire Mini.  So imagine what a bargain the Mini is!

"There might be nothing that can be done in firmware to completely eliminate this effect."

I would suggest you are correct in that assumption. Unfortunately a lot of drone owners expect more accurate altitude measurementsa, but with current technology, it is not really possible to improve on the barometer principle without spending ridiculous amounts of money. If some company comes up with a low cost accurate radar ranging device, then it would change dramatically.