Hi,

I purchased a phantom 4 pro and a phantom 4 as backup to do aerial filming in Greenland for an online magazine.

As soon as I started flying the 4 Pro I encountered multiple issues:

- the loss of GPS signal was so frequent that it was impossible to fly in GPS mode and I had to fly in Atti almost all the time: I reckon that it may be due to the northern location although a local Phantom 3 owner told me it was not  nearly as bad with his phantom.

However I found out that there were many more problems:

- I frequently had gimbal problems with the camera getting locked on one side with the effect of the landing gear of the drone showing in the display. Not only does it destroy your footage but when you fly without GPS it is is very disorienting as you can't even clearly see in which direction you're flying

- this problem was often accompanied -though not always - by the message "gimbal motor overloaded". It could be that the temperature played a part but I am not fully convinced by this explanation. Other local phantom owners didn't encounter this problem or very seldom.

- the horizon rarely stays stable for a long time as soon as you start turning the drone. I found similar complains on the internet so it seems like a rather common issue with the Phantom 4 (pro).

- but the CRAZIEST thing of all was that from time to time the drone started flying in a random direction at full speed and became UNCONTROLLABLE. I am not talking about the drone drifting with the wind in the absence of GPS ; no it just flew very fast, loosely flying in loops or something like that before getting back to normal after maybe 30 seconds or so. It is very scary, the only thing you can do is try to gain altitude to avoid any low obstacles.

- because of all these problems I unpacked the "backup" Phantom 4. We first tested it flying very gently in beginner mode. It looked good with no GPS loss or anything. But suddenly the same problem occurred with the Phantom 4 getting completely crazy as well ; after a few seconds it hit a roof violently! The outer shell cracked in one part near one of the motors. I didn't try to fly it again after that.

To me it looks like these drones are completely unreliable and untested. I would at least like to get a replacement for the broken Phantom and ideally a refund since I feel the crash was caused by a malfunction for which I am not responsible at all. I have opened a ticket and uploaded all flight log files - both .dat and .txt - for the Phantom 4.


Has anybody here experienced similar issues?

oh and I forgot that a battery stopped working after a few flights. It shows 4 green bars, you can turn it on and off but when you turn it on it doesn't power the aircraft. Nothing happens.

ukaleq Posted at 2017-4-23 14:03
oh and I forgot that a battery stopped working after a few flights. It shows 4 green bars, you can turn it on and off but when you turn it on it doesn't power the aircraft. Nothing happens.

The GPS issues could have been due to the Kp index has been fairly high in the last few days which means there is some solar activity.  That could have caused compass issues as well being that far north.  Also I am not sure temps you were in.  That could have effected your IMU.  You IMU may have needed to be warm before flying.

As for the gimbal.  Yeah, I am on bird number 2 now with a P4P.  My first one did exactly what happened to you.  I swapped it out at the local retailer.  My current one is not as bad.   But it can not keep a horizon to save its life.  I kind of feel like the roll motor is under powered, which makes me wonder if the roll motor design was flawed and is too weak for this gimbal.  It seems like they lowered the power to that motor to keep it from getting damaged.  A good gimbal should not be losing the horizon line that bad.  Wind should not be an issue.  These things should be able to go into the wind at fairly high speed and maintain level with maybe just a little bit of roll, but not 5 degrees!

Thanks for your comment. My feeling was that the gimbal problem was somehow unrelated to the environment and it tends to confirm that.
The biggest problem though remains the wild behavior which resulted in a broken drone. Even if the GPS and compass outputs are wrong the software should filter out aberrant data and not move the drone like crazy in the absence of joystick input.
Obviously the customers are used are beta-testers. The problem is that it's not just a software it's a physical thing and all these bugs have real-world implications.

Did you do any test flying at home before packing everything off to Greenland?
Do you have any Phantom experience prior to this?

If you can find a flight record from the P4 pro the data might help explain some of yhe issues you describe.
Go to http://www.phantomhelp.com/LogViewer/Upload/
Follow the instructions to upload your flight record.
Come back and post a link to the report it provides.

I did a short test flight of the 4 Pro and all went well. I did not test the 4. Here is a link to the data of the crash. It doesn't show that many details unfortunately.
http://www.phantomhelp.com/LogViewer/SOSLPI06N0MAO7KX9X9P/

Right now I am having yet another problem. If I try to synch the flight data of the pro plus remote it gets stuck and never finishes uploading the data...

Did you calibrate the compasses?  Being that far north, the magnetic field lines are pointing downwards rather steeply, if they were expecting them to be nearly flat then there could be some difficulties.

Sounds like you had faulty GPS on the Pro but to get two unreliable ones would be very unlucky, they may not be well tested at the factory, they are consumer items, but they do normally work perfectly straight out of the box.

ukaleq Posted at 2017-4-23 15:26
I did a short test flight of the 4 Pro and all went well. I did not test the 4. Here is a link to the data of the crash. It doesn't show that many details unfortunately.
http://www.phantomhelp.com/LogViewer/SOSLPI06N0MAO7KX9X9P/

More about this. The beginning of the flight path (that boring straight line) was controlled by the user.
If I remember correctly the drone had already gone into wild mode when it did the loop around the small house. And of course the final acceleration in the direction of the big building was totally uncontrolled. It is fortunate that it hit the roof with no apparent damage to the roof and not a window, which could have harmed someone inside. The drone bounced off the roof and was found on the other side of the building!

ukaleq Posted at 2017-4-23 15:26
I did a short test flight of the 4 Pro and all went well. I did not test the 4. Here is a link to the data of the crash. It doesn't show that many details unfortunately.
http://www.phantomhelp.com/LogViewer/SOSLPI06N0MAO7KX9X9P/

Your gps looks ok in that flight, but you seem to have a problem with your battery.

Nigel_ Posted at 2017-4-23 15:34
Did you calibrate the compasses?  Being that far north, the magnetic field lines are pointing downwards rather steeply, if they were expecting them to be nearly flat then there could be some difficulties.

Sounds like you had faulty GPS on the Pro but to get two unreliable ones would be very unlucky, they may not be well tested at the factory, they are consumer items, but they do normally work perfectly straight out of the box.

Yes I did calibrate both units. I calibrated the 4 pro multiple times. The 4 was also calibrated before the very short first flight and before the second (and last) flight.

CelticWarrior Posted at 2017-4-23 15:44
Your gps looks ok in that flight, but you seem to have a problem with your battery.

The battery was not fully charged but then again we did not plan to fly it long. It was supposed to be a very short, uneventful flight in beginner mode.
I think the voltage drop seen at the end is caused by the acceleration of the aircraft and the cold may have compounded the effect. Ironically that battery survived the crash and was used several times on the 4 pro with no issue!

ukaleq Posted at 2017-4-23 15:46
The battery was not fully charged but then again we did not plan to fly it long. It was supposed to be a very short, uneventful flight in beginner mode.
I think the voltage drop seen at the end is caused by the acceleration of the aircraft and the cold may have compounded the effect. Ironically that battery survived the crash and was used several times on the 4 pro with no issue!

How is your compass did the P4 just preform that circle just before crash on its own.

ukaleq Posted at 2017-4-23 15:45
Yes I did calibrate both units. I calibrated the 4 pro multiple times. The 4 was also calibrated before the very short first flight and before the second (and last) flight.

Why were you flying in beginner mode?
What's your experience level before the Greenland trip?
You mention calibrating the P4pro+ multiple times ... why multiple times?
Was there any steel close to where you calibrated ... including reinforced concrete?

The big curve in the flight path is a clear indication of compass issues but it's not clear if the problem is due to the magnetic environment at that location or issues due to the way you calibrated it.

ps ... it didn't matter here but never launch with a partially discharged battery, even for short flights.
It can seriously reduce the probability of coming home with a Phantom.

ukaleq Posted at 2017-4-23 15:46
The battery was not fully charged but then again we did not plan to fly it long. It was supposed to be a very short, uneventful flight in beginner mode.
I think the voltage drop seen at the end is caused by the acceleration of the aircraft and the cold may have compounded the effect. Ironically that battery survived the crash and was used several times on the 4 pro with no issue!

Because you're having the same problems with both craft, could mean the area you were flying in is giving you bad compass, or you calibrated both craft in the same bad area.

ukaleq Posted at 2017-4-23 15:26
I did a short test flight of the 4 Pro and all went well. I did not test the 4. Here is a link to the data of the crash. It doesn't show that many details unfortunately.
http://www.phantomhelp.com/LogViewer/SOSLPI06N0MAO7KX9X9P/

Bit puzzled by the stick movements in the P4 log.  I don't have a way of plotting what it should have done to compare with what it did, but those are not the stick movements I expected.  

Biggest puzzle is why didn't you gain a bit of height so that there wouldn't be a collision with a roof?

Second, when things went wrong,  I would expect you to have left go of the sticks to see if it stopped moving, but you appear to be driving it in circles using only the right stick, ie without actually turning the aircraft, or taking a pause to evaluate what was happening.  You have to fly forwards for the obstacle avoidance to work...

Maybe you should have a play with the simulator sometime.

Having said that, there was obviously something wrong since it flew beyond the max distance.

Nigel_ Posted at 2017-4-23 16:53
Bit puzzled by the stick movements in the P4 log.  I don't have a way of plotting what it should have done to compare with what it did, but those are not the stick movements I expected.  

Biggest puzzle is why didn't you gain a bit of height so that there wouldn't be a collision with a roof?

Where do you see the stick movements? I would like to see that myself.

Gaining altitude is indeed what was key to my avoiding this type of crash with the P4Pro when it behaved the same. I was not piloting the P4 myself when it happened, rather it was someone who had actually more experience  than me with drones (an early adopter of the P1).

He told me that he tried to increase the height but that the aircraft did not respond.

Like I said similar behavior happened with the P4Pro and after calibrating the compass in several places, sometimes miles away from that one!

The bottom line is: there is a major issue with those drones. As you pointed out it should not have accelerated beyond the maximum distance when flying in beginner mode, even if the operator had actually ordered it - which he didn't.

I would also add that you do not have a whole lot of time to think as it happens very quickly. It is easy to panic in such situations making the situation even worse.
With the P4 Pro I experienced similar issues multiple times. I  got used to it and immediately positioned the aircraft high enough after takeoff  to avoid potential obstacles in case of uncontrolled behavior. When it happened I was also more relaxed and observed that the aircraft followed the same circling patterns even in the absence of stick input. I am not sure if I have a record of that though since the logs do not seem to go very far back and I didn't experience the problem during the last flights.

What if the problem was precisely that the system receives fake input sticks? Just speculation of course.

ukaleq Posted at 2017-4-23 17:20
Where do you see the stick movements? I would like to see that myself.

Gaining altitude is indeed what was key to my avoiding this type of crash with the P4Pro when it behaved the same. I was not piloting the P4 myself when it happened, rather it was someone who had actually more experience  than me with drones (an early adopter of the P1).

When the crash occurred, the throttle hadn't been pushed up (how you gain height) for over 40 seconds, my understanding of the trace is that it started behaving oddly about 20 seconds before the crash.

If you want to see the sticks, it is probably best to go into the Go App and replay the flight, there is an icon on the control bar to the left of the pause/play that turns a display of the sticks on/off.

Labroides Posted at 2017-4-23 16:02
Why were you flying in beginner mode?
What's your experience level before the Greenland trip?
You mention calibrating the P4pro+ multiple times ... why multiple times?

Some of you guys seem to be very biased.

Beginner mode is the default mode when you start the aircraft for the first time. I am glad it was in beginner mode because at least there is something clear: it should not have continued past the maximum distance.

Why did I calibrate it multiple times? Precisely because I assumed there could be compass issues due to local variations in the local magnetic field. I would say the fact that the problem repeatedly happened after calibration in different places pretty much rules out the fact that it is a calibration issue.

As a matter of fact the first time it happened to me with the P4 Pro I thought it was my mistake. After a couple of times I had no doubt that there was a serious problem. Then when we tested the P4 I already encountered problems with the first flight. It is then that the other guy (a P1 owner) took the remote for the second flight as, like you, he thought I might be the problem. And then it crashed and all he said was "I couldn't control it".

Nigel_ Posted at 2017-4-23 17:39
When the crash occurred, the throttle hadn't been pushed up (how you gain height) for over 40 seconds, my understanding of the trace is that it started behaving oddly about 20 seconds before the crash.

If you want to see the sticks, it is probably best to go into the Go App and replay the flight, there is an icon on the control bar to the left of the pause/play that turns a display of the sticks on/off.

I do not have the iphone that was used to run the app any more unfortunately... If there is a way to reimport the data e.g. in the P4 pro plus remote it would be nice.

ukaleq Posted at 2017-4-23 17:35
I would also add that you do not have a whole lot of time to think as it happens very quickly. It is easy to panic in such situations making the situation even worse.
With the P4 Pro I experienced similar issues multiple times. I  got used to it and immediately positioned the aircraft high enough after takeoff  to avoid potential obstacles in case of uncontrolled behavior. When it happened I was also more relaxed and observed that the aircraft followed the same circling patterns even in the absence of stick input. I am not sure if I have a record of that though since the logs do not seem to go very far back and I didn't experience the problem during the last flights.

You started off saying that "To me it looks like these drones are completely unreliable and untested"

I suspect that in this case, the operator was unreliable and untested and this is likely to have been at least part of the problem.
If you read your Phantom manual, you'll find in the Flight Environment Requirements it says:  The Phantom 4 cannot operate within the polar areas.
Elsewhere DJI have stated: all Phantom series aircraft cannot be operated in Polar Regions in P mode
Your flight was located only 1500 km from the North Magnetic Pole.
This can cause compass problems as well as issues with GPS as the satellites are only going to be low on the horizon.
You've said you calibrated the compass multiple times but not given any details.
This is also concerning.  If you had problems calibrating but persisted, there's a good chance you gave it a bad compass calibration.
If you screw up your compass calibration, it's going to cause problems back in low latitudes.
Up in the high arctic, it's likely to be even worse.

Given the environment and your experience level, it's probably not right to blame the Phantom for the problems you encountered.

ukaleq Posted at 2017-4-23 17:56
I do not have the iphone that was used to run the app any more unfortunately... If there is a way to reimport the data e.g. in the P4 pro plus remote it would be nice.

If you used the same account and synchronised the flight logs then you should find that it is already available on your P4P remote - you can select which aircraft to the left of the flight list...  Maybe do another sync...

ukaleq Posted at 2017-4-23 17:53
Some of you guys seem to be very biased.

Beginner mode is the default mode when you start the aircraft for the first time. I am glad it was in beginner mode because at least there is something clear: it should not have continued past the maximum distance.

Biased?
One has to ask questions to find out the information necessary to understand what was going on.
Your initial description lacked much relevant information.

Asking why you were flying in beginner mode doesn't indicate bias.

It's very unusual for anyone to fly in beginner mode unless they are taking their first couple of flights with their first Phantom.
I wondered if there was a particular reason you were doing this.

We feel sorry for your crash, since you've contacted us and provided all the flight records, our data analysis engineers will help you analyze it, then contact you for the result.

Nigel_ Posted at 2017-4-23 17:59
If you used the same account and synchronised the flight logs then you should find that it is already available on your P4P remote - you can select which aircraft to the left of the flight list...  Maybe do another sync...

ok good to know, unfortunately those flights were not synchronized, I just saved the data on my hard drive...

Labroides Posted at 2017-4-23 17:56
You started off saying that "To me it looks like these drones are completely unreliable and untested"

I suspect that in this case, the operator was unreliable and untested and this is likely to have been at least part of the problem.

Ok I got the message, I am an idiot who screwed up the calibration and then pushed all sticks in the wrong direction. But what do you do of the phantom going beyond the distance limit in beginner mode? Oh yes I should not have been in beginner mode.

How can one screw up the compass calibration, please explain? You rotate the phantom until it tells you it's ok or a problem occurred and you have to do it again.  Is there any particular knowledge required for that? I removed all electronic equipment from my pockets and performed it in places where there was only sea ice only to find out that it did not make it any more reliable.

How do you define polar areas? Which latitudes/longitudes exactly? Anything above the arctic circle? That's a lot. If the phantom can get wild at any time in "polar areas" why is there only a note buried in the manual and not a warning flashing as soon as it gets a GPS lock and knows where it is? Actually it should simply refuse to take off if it is impossible to operate it in those regions.

Are you paid by DJI to defend their products?

ukaleq Posted at 2017-4-24 02:08
Ok I got the message, I am an idiot who screwed up the calibration and then pushed all sticks in the wrong direction. But what do you do of the phantom going beyond the distance limit in beginner mode? Oh yes I should not have been in beginner mode.

How can one screw up the compass calibration, please explain? You rotate the phantom until it tells you it's ok or a problem occurred and you have to do it again.  Is there any particular knowledge required for that? I removed all electronic equipment from my pockets and performed it in places where there was only sea ice only to find out that it did not make it any more reliable.

I don't think you did get the message.   He is trying to help diagnose the problem, that normally requires asking for more information and it is not always easy to ask for information without it sounding like you are accusing the pilot, especially when you are writing in Australian!  We do know that it can be difficult to do the right thing when things start to go wrong fast, especially when you haven't experienced that problem before.
  
Curved flights like that almost always indicate a compass problem.  It is hard to diagnose from the logs because half the information we have is derived from the sensors that are having problems, thus we are working with the same inaccurate information that the aircraft had, so it looks like the aircraft was doing the correct thing.

Did the same person calibrate the P4 and P4P?  Maybe he was wearing a self winding watch with a strong magnet in it on the hand that was holding the aircraft?  Was there anything added to the aircraft that may have affected it?  What was in the buildings that you flew over, you were very close to them?

There is an explanation for the problem, very unlikely that both aircraft had the same manufacturing fault.   Was it caused by the location, the operator, the calibrator, the aircraft, the weather/magnetic conditions on the day, something else?  Can you do any tests to remove some of these as a possible cause, maybe by flying again in a completely different location?

ukaleq Posted at 2017-4-24 02:08
Ok I got the message, I am an idiot who screwed up the calibration and then pushed all sticks in the wrong direction. But what do you do of the phantom going beyond the distance limit in beginner mode? Oh yes I should not have been in beginner mode.

How can one screw up the compass calibration, please explain? You rotate the phantom until it tells you it's ok or a problem occurred and you have to do it again.  Is there any particular knowledge required for that? I removed all electronic equipment from my pockets and performed it in places where there was only sea ice only to find out that it did not make it any more reliable.

You're getting messages that I'm not sending.
You came with a very vague description of your incident that needed a lot of questions to work out what might have happened.
Without actual data for the P4pro+ there are still a lot of unknowns and a bit of guesswork to fill in some blanks.
Add to that, the fact that you are flying in a very rarely flown area and the effects of flying there can be tricky but are poorly defined for those of us that haven't been there to find out.

Compass calibration.
This is a critical point of Phantom flying.  Without a good compass calibration, you cannot fly straight and your Phantom will fly big curves rather than straight and slowly spiral rather than hover in place.
You screw up your compass by calibrating it in a bad location, one where the earth's normal magnetic field is not normal.  Like on top of reinforced concrete or next to a car or powerline.
Usually attempting to do this will fail but sometimes you can "successfully" calibrate in a distorted magnetic field.  This is fine while your Phantom is in the distorted field but is a big problem when you move further away into the earth's normal field, which the compass has been adjusted to not recognise.

The manual for the P4 and P4 pro advise NOT to calibrate the compass unless there is a problem that needs recalibration.  I would have been very interested to find out how the P4pro flew without tinkering with the calibration but it's too late to find out now.
It might have been fine - I can't be sure though.

DJI don't define polar areas.  There are several things that DJI don't explain well and we gradually work out what's best and what works from experimenting.
Your flight was the closest to the north magnetic pole that I've heard of and I would think it definitely falls within the polar region.
You weren't just 11 degrees north of the arctic circle, you were also at one of the closest settlements to the north magnetic pole.

I would guess that DJI don't stop you from flying there because it's quite possible to fly there as your local Phantom flyer does.
The catch is that you need to know what you are doing and not make it even trickier than the local conditions make it.

It takes only 5 minutes to learn to control the Phantom but that can be a trap for beginners, giving them a false confidence.  There are lots of things that can go wrong and it takes a lot of experience and time to find out about them and how to make sure that they don't cause problems for you.
There quite a learning curve to become proficient.

You still haven't explained your experience level but there are enough clues that suggest you are very new to the Phantom.

Am I paid by DJI?    I wish I was.
I'm just a user with a lot of experience, trying to help people with problems.
I don't defend DJI or their products but their products are very good and often much better than some users might think.
Where mistakes cause problems, learning from those mistakes helps make everyone a safer flyer in the future.

I'm going to say again that your adventure isn't reason to say that DJI drones are unreliable and untested.
There must be over a million sold by now and they've been pretty well tested and found to be very reliable.
The Phantom is advertised as ready-to-fly but that doesn't mean that the owner is too.

Labroides Posted at 2017-4-24 03:42
You're getting messages that I'm not sending.
You came with a very vague description of your incident that needed a lot of questions to work out what might have happened.
Without actual data for the P4pro+ there are still a lot of unknowns and a bit of guesswork to fill in some blanks.

Here's what the manual says about compass calibration.

"IMPORTANT: Always calibrate the compass in every new flight location. The compass is very sensitive
to  electromagnetic  interference,  which  can  produce  abnormal  compass  data  and  lead  to  poor  flight  
performance or flight failure. Regular calibration is required for optimal performance."

So this is exactly what I did. After flying the Phantom in Europe, I calibrated it again once in Greenland. Since I had problems I recalibrated it in different locations precisely to rule out possible bad calibration due to an improper environment. But the same problems kept happening again and again.

And what you have to understand is that there was no systematic drift or anything like that.  Typically the Phantom would fly in a very normal and stable way -whether in GPS or Atti - until it suddenly got "crazy" for a brief period of time.  In Atti mode you see some drift caused by the wind but I understand this is normal and you can still control the aircraft. This is not what I am talking about here.

Since the same problem affected both aircrafts, I agree that it is very likely related to the environment and to the compass. To me it looks like a software issue that shows up in that kind of environment.

In any case I see no good reason why the same aircraft should behave normally one minute and suddenly become crazy the next minute. If the GPS receiver or other sensors suddenly deliver some aberrant data, the software should not respond by making the aircraft go in circles at full speed, period.

And yes I am a beginner but I do not think it is relevant when dealing with that kind of issue.

Also, the gimbal problems show that there are also hardware issues with this drone. See the comment of a fellow user above.

ukaleq Posted at 2017-4-24 04:47
Here's what the manual says about compass calibration.

"IMPORTANT: Always calibrate the compass in every new flight location. The compass is very sensitive

If you check an up to date manual for the P4 (p51) or P4pro (p57), you'll find that they do not say that at all.
Here's what they say:  Only calibrate the compass when the DJI GO 4 app or the status indicator prompt you to do so.
The compass is a critical system.
Experience investigating a large number of incidents has shown that unnecessarily recalibrating a perfectly good compass has caused the loss of many Phantoms.
"In Atti mode you see some drift caused by the wind but I understand this is normal".
That's all very well with a Phantom that knows which way it's pointing ... but screw up the compass calibration and it's a very different matter.

I've looked at lots of lost and crashed Phantom stories and I do understand that behaviour of an affected Phantom can be erratic.
I also know that issues can also affect the functioning of the gimbal.

That you are a beginner is highly relevant when you go beyond your experience in an environment that has significant issues with a Phantom with a likely badly calibrated compass.
When you aren't aware of the issues and possible consequences, it's very easy for things to come unstuck quickly.

"In any case I see no good reason why the same aircraft should behave normally one minute and suddenly become crazy the next minute. "
How far out was magnetic north from true (GPS) north at the location you were flying?   I'm, suspecting close to 30 degrees out?

When there is a huge conflict between compass direction and GPS direction, there is going to be a problem flying.  The compass is what is used for directional stability since it is considerably faster and more consistent, when it's working properly, and operates in three dimensions rather than just 2 for the GPS.

Many people who have flown in Greenland say that it is Atti mode only to avoid flying in curves.  Very different to Iceland or Norway because for them the two poles are almost in a straight line so there is little angular error.

Nigel_ Posted at 2017-4-24 05:26
"In any case I see no good reason why the same aircraft should behave normally one minute and suddenly become crazy the next minute. "
How far out was magnetic north from true (GPS) north at the location you were flying?   I'm, suspecting close to 30 degrees out?

What do you mean with "flying in curves"?
Do you mean problems similar to the one described here or simply that the aircraft does not follow a straight line if you push the right stick forward?

After firmware update, my bird is moving automatically  towards up and down left and right till 2-3 feet, help me out, already done the IMU, COMPASS, VPS, REMOTE LINKING AND RC, PLEASE help  me out

fansf4b02c46 Posted at 2017-4-24 07:04
After firmware update, my bird is moving automatically  towards up and down left and right till 2-3 feet, help me out, already done the IMU, COMPASS, VPS, REMOTE LINKING AND RC, PLEASE help  me out

Is it too dark for the VPS to work, or maybe you are too close to the ground?   If you are only 2 feet from the ground then you are too low for the VPS to work, it then has to use GPS which does have errors, sometimes of a few meters, so the aircraft will move about a bit while hovering.

ukaleq Posted at 2017-4-24 06:20
What do you mean with "flying in curves"?
Do you mean problems similar to the one described here or simply that the aircraft does not follow a straight line if you push the right stick forward?

It's a bit hard to see what happened in your P4 log since you kept moving the right stick all over the place!
But I think the answer to this question is: yes in both cases.

You want to fly north so you follow the compass north, you see an error developing on the GPS position because magnetic north is 30 degrees away from true GPS north, you have to assume the error is  due to wind so you correct for it, but the error keeps increasing so you have to correct more and more giving a curved flight.  Switch to Atti mode and you are not correcting for errors due to wind so it flies OK.

Note that when you calibrate the compass, you are not calibrating it against GPS so there is no correction for true north.  The calibration is mainly to correct for the magnetic field of the aircraft.  I think DJI could probably do a bit better and take into account the position of the magnetic north pole when GPS is being used.

fansf4b02c46 Posted at 2017-4-24 07:04
After firmware update, my bird is moving automatically  towards up and down left and right till 2-3 feet, help me out, already done the IMU, COMPASS, VPS, REMOTE LINKING AND RC, PLEASE help  me out

You'll get more help if you start your own thread rather than adding your post to someone else's thread about an unrelated incident.

Nigel_ Posted at 2017-4-24 07:25
It's a bit hard to see what happened in your P4 log since you kept moving the right stick all over the place!
But I think the answer to this question is: yes in both cases.

Hope this helps understanding

About magnetic and geographic north...

ukaleq Posted at 2017-4-23 17:53
Some of you guys seem to be very biased.

Beginner mode is the default mode when you start the aircraft for the first time. I am glad it was in beginner mode because at least there is something clear: it should not have continued past the maximum distance.

Some stuff to take into consideration when calibrating compass.

Many things can distort the earth’s magnetic field in the area you are flying:

Steel framed or reinforced concrete buildings, bridges and roadways, iron pipes and culverts, high power electric lines, heavy equipment, trucks and automobiles, steel tanks, electric motors and even computers.
Flying between steel framed or reinforced high rise buildings will distort the magnetic field in addition to causing GPS multi-pathing.
Safe distances for compass calibration

6” (15 cm) minimum: Metal rim glasses, pen/pencil, metal watch band, pocket knife, metal zipper/buttons, belt buckle, batteries, binoculars, cell phone, keys, camera, camcorder, survey nails, metal tape measure.
18” (50 cm) minimum: Clipboard, data collector, computer, GPS antenna, 2-way radio, hand gun, hatchet, cell phone case with magnetic closure.
6 ft (2 m) minimum: Bicycle, fire hydrant, road signs, sewer cap or drain, steel pole, ATV, guy wire, magnets, chain-link fence, bar-wire fence, data collectors

that use a magnet to hold the stylus.
15 ft (5 m) minimum: Electrical box, small car/truck, powerline, building with concrete & steel.
30 ft (10 m) minimum: Large truck, metal building, heavy machinery.

CelticWarrior Posted at 2017-4-24 11:55
Some stuff to take into consideration when calibrating compass.

Many things can distort the earth’s magnetic field in the area you are flying:

Ok In fact I remember we did 2more flights with an android phone before we switched to the iphone with which the crash occured (the video link was not working properly presumably because it's a low-end phone with a small processor).

The same thing happened. Look at what happened at ca. 5:30. You see the speed go up and this is when the aircraft runs along the curve. There is hardly any joystick input at this time in the Go App replay. If you look at the csv file it shows RC.aileron and RC.elevator to be zero when it occurs (I understand that these values correspond to the right stick horizontal/vertical moves). Yet you see roll values change significantly. The yaw stays more or less constant. So we have erratic roll changes that are not related to the joystick input.  

http://www.phantomhelp.com/LogViewer/3EWHZ843YQP7YXOTJNDL/

Ideally there would be a viewer allowing us to see all these values change without having to scan the CSV file in an excel spreadsheet. Does that exist?

ukaleq Posted at 2017-4-24 16:55
Ok In fact I remember we did 2more flights with an android phone before we switched to the iphone with which the crash occured (the video link was not working properly presumably because it's a low-end phone with a small processor).

The same thing happened. Look at what happened at ca. 5:30. You see the speed go up and this is when the aircraft runs along the curve. There is hardly any joystick input at this time in the Go App replay. If you look at the csv file it shows RC.aileron and RC.elevator to be zero when it occurs (I understand that these values correspond to the right stick horizontal/vertical moves). Yet you see roll values change significantly. The yaw stays more or less constant. So we have erratic roll changes that are not related to the joystick input.  

"So we have erratic roll changes that are not related to the joystick input."
It looks to me like the roll changes are to correct for wind, however when correcting for wind it flies in the wrong direction due to the difference in angle between GPS north and compass north which DJI don't appear to take into account and which doesn't get corrected by calibration.  Because it corrects position in the wrong direction it then has to correct the correction which then produces the curved flight.

I expect that if you fly these aircraft back home in Germany there will be no issues.  If you want to fly in this location then you need to fly in atti mode and correct for wind manually.

I would be worried about the poor communications in that log file though, half the data is missing and with RTH not working that seems too risky for me to fly!

I may be wrong about this, I don't know the details of how the navigation firmware works, but there is clearly a problem using GPS mode in that location, if it works OK in another location away from the Earth's magnetic poles then it is certain to be the location that is the problem, or how DJI's firmware deals with the unusual magnetic situation in that area.

Get too close to the magnetic north pole and all compasses become useless, it does look to me like DJI could do better in this location though.