Hello. Today it happened to me.
I decided to fly for a while. The beginning was all right. I checked Kp: 1 Satellites: Ok. Clear sky. No wind. Perfect weather. I was at such a tiny local modeling airport where local RC modelers meet. There it is clean and without electromagnetic interference.
I started calmly, then flew off 200 meters away from me, low in sport mode. I turned and when I stopped the AC information about the compass appeared. I also saw that Spark is not stable in the air. I did a compass calibration. I restarted AC. I started again, at a low ceiling of 1 - 2 meters high. I flew off 16 meters away and I screwed. When I came back I noticed that the spark is still not stable. I decided to stop flying because something bad is happening. I wanted to come to me and land. Then I realized that Spark was starting to speed up! in my direction. I released the steering rods and he was still accelerating in the same direction. I jumped back because he would hit me - he did not react. I pulled the bar down so it would not fly away. At that time, Spark was constantly accelerating - he hit such roles with grass and crashed. It also turned out that 3D Sensing Technology on the DJI SPARK did not work.

http://www.phantomhelp.com/LogViewer/HS4QDLCCC3XGCBW2X0LU/

Any suggestions? I can not understand this. Thanks and best regards.

Looks like you had IMU issues

that's why there were messages: compass error and speed error ? There were no other errors

tomaszmm76 Posted at 2018-1-20 11:44
that's why there were messages: compass error and speed error ? There were no other errors

How high were you flying?  According to log less than 5 feet

I flew low because I saw something is wrong. And I did not want him flying away for a few hundred meters. I decided to fly low to me and land then. I was convinced that Spark can fly at the height of one meter and it will be a safe, slow return. Then he accelerated. I will also add that there was no error on DJI GO 4.

tomaszmm76 Posted at 2018-1-20 11:58
I flew low because I saw something is wrong. And I did not want him flying away for a few hundred meters. I decided to fly low to me and land then. I was convinced that Spark can fly at the height of one meter and it will be a safe, slow return. Then he accelerated. I will also add that there was no error on DJI GO 4.

Spark flew ahead. Spark flew ahead - why Obstacle Avoidance System did not work ???

First of all I am wondering why your flightrecord starts at 4m 5.7s. Secondly you write above that you were in sport mode but this is not reflected in the flightrecord. If you were in sport mode, obstacle avoidance would be automatically switched off. Many question marks in this case.

So I'm no expert. But I have been doing a lot of study and reading here on the forums, low level sport flight with fast turns, may cause errors according to a few posts I've read. Also obstacle avoidance does not react quickly compared to how fast the Spark is able fly in sport mode, what I'm saying is, it is very possible to sport fly the Spark smack into a building if you wanted to. But it could still be a drone problem, just letting you know what I've learned.

Obstacle avoidance is turned off in sport mode, hence the sensor icon will be red in the DJI Go 4 flight screen.

One thing that I figured out for myself is that sport mode and cold weather is not coming along very well, as there compass troubles kicks in.
Yes in sport mode the visual system is turned off, but in your log there is only P-GPS mode, so I'm confused as well.
You definitely should contact support, as there was a crash with not very pleasant outcome.

Your flight log.
You calibrated compass, why?
The first 4 minutes of your log is missing
While you got compass problems how and where did you calibrate
Your Aircraft was responding to many if not all of your stick movements, because of compass problems actual stick movements may not have been correct.
3 D sensing will not work at speeds your Aircraft was travelling
You at almost all times were pushing on at least one stick , if you had of completely lifted off sticks your aircraft had gps and should have been able to lock and you could have made a better landing.

Last stick movements show you pulling almost 100% right stick down and 100% left stick down causing aircraft to crash land.

Wachtberger Posted at 2018-1-20 14:32
First of all I am wondering why your flightrecord starts at 4m 5.7s. Secondly you write above that you were in sport mode but this is not reflected in the flightrecord. If you were in sport mode, obstacle avoidance would be automatically switched off. Many question marks in this case.

It starts from 4m 5.7s because this is the second flight after compass calibration and AC restart.                .... ''sport mode. I turned and when I stopped the AC information about the compass appeared. I also saw that Spark is not stable in the air. I did a compass calibration. I restarted AC. I started again, at a low ceiling of 1 - 2 meters high.'' - this is a quote from my post In the second flight I was not in Sport mode.  Here is the first flight:  
http://www.phantomhelp.com/LogViewer/4LJW0ESHZWDLZ4C8KT00/

tomaszmm76 Posted at 2018-1-20 15:16
It starts from 4m 5.7s because this is the second flight after compass calibration and AC restart.                .... ''sport mode. I turned and when I stopped the AC information about the compass appeared. I also saw that Spark is not stable in the air. I did a compass calibration. I restarted AC. I started again, at a low ceiling of 1 - 2 meters high.'' - this is a quote from my post  In the second flight I was not in Sport mode.  Here is the first flight:  
http://www.phantomhelp.com/LogViewer/4LJW0ESHZWDLZ4C8KT00/

Very difficult to say what might have happened there. Were you prompted by GO4 to do a compass calibration and are you sure that there were no electronic devices nearby when you did it (e.g. your mobile phone, Apple watch or whatsoever)?
Edit: I only now saw hallmarks analysis above and you can trust that he knows about it much better than I do!

tomaszmm76 Posted at 2018-1-20 15:16
It starts from 4m 5.7s because this is the second flight after compass calibration and AC restart.                .... ''sport mode. I turned and when I stopped the AC information about the compass appeared. I also saw that Spark is not stable in the air. I did a compass calibration. I restarted AC. I started again, at a low ceiling of 1 - 2 meters high.'' - this is a quote from my post  In the second flight I was not in Sport mode.  Here is the first flight:  
http://www.phantomhelp.com/LogViewer/4LJW0ESHZWDLZ4C8KT00/

I can see how you had some compass errors appeared 15 ft away from you, so there could well have magnetic interference in the area you were flying in, your second log shows many more errors so could calibration been the cause only you would know this.

hallmark007 Posted at 2018-1-20 15:15
Your flight log.
You calibrated compass, why?
The first 4 minutes of your log is missing

I calibrated the compass - because in the first flight I saw a compass error. Then I landed and calibrated him. Then I started a second time to check if everything is fine. But it was not.
I've already joined the second flight log.
Calibration I made so that the RC with the phone I put on the table - (there is such because it is the airport of the modeling club) - and I left 2 or 3 meters with Spark and did the calibration.
I have no idea why Spark was accelerating in my direction - I did not do it - when he passed me he did not stop accelerating.
He could make a better landing but he did not want to stop. I did everything to stop him.

tomaszmm76 Posted at 2018-1-20 15:46
I calibrated the compass - because in the first flight I saw a compass error. Then I landed and calibrated him. Then I started a second time to check if everything is fine. But it was not.
I've already joined the second flight log.
Calibration I made so that the RC with the phone I put on the table - (there is such because it is the airport of the modeling club) - and I left 2 or 3 meters with Spark and did the calibration.

Yes , but if you look at your log you can see when you push stick up your aircraft moved up an same when you pushed left right and yaw movements, you can see clearly at the end of your log you were pushing left stick down 100% and right stick down 100% this is what drove your aircraft into the ground.

Yes because of bad compass these commands may not have been 100% correct but log shows using yaw pitch and roll that they were pretty close to pilots commands.

Wachtberger Posted at 2018-1-20 15:25
Very difficult to say what might have happened there. Were you prompted by GO4 to do a compass calibration and are you sure that there were no electronic devices nearby when you did it (e.g. your mobile phone, Apple watch or whatsoever)?
Edit: I only now saw hallmarks analysis above and you can trust that he knows about it much better than I do!

Nearly a few hundred meters, there were no electrical devices. I left my smartwatch at home because I recently read that it causes disruptions. Go 4 did not ask, but there was a standard message that there is a problem with the compass and to go away from the sources of stabs or do a calibration. So I did Calibration of the compass because I was sure it was clean. Thanks for the answer.

By the way. In DJI advertisements flying Spark is a pleasure. You take out the box, you pair with the controller, connect with the phone and fly.
But now I noticed that it is not so easy.
You must first examine the area where you want to fly very precisely. Do not have any electronic devices. Look Kp. And much much more. And be lucky because Spark can fly away anyway.
It's a bit sad.

They are a pleasure.  My wife had fun flying mine.  However, radio control models are subject loss.  None are foolproof against this loss.  Never send into the open skies what you can't afford to lose.  With the compass errors (smart watch or nitro engine starter nearby?) some error happened where GPS stayed on but was not accurate, maybe?  You have to pull back on the sticks to stop forward motion of a heli or quad when it is not being GPS stabilized (ATTI mode equivalent).  Can't just release the sticks to stop like it does when GPS/IMU stabilized.

On the up side, you might know something about RC models since you were at a RC field.  That frame you broke is only $30 and you just need to follow some youtube videos to swap the parts over.  Easy repair if DJI doesn't warranty it.

hallmark007 Posted at 2018-1-20 15:55
Yes , but if you look at your log you can see when you push stick up your aircraft moved up an same when you pushed left right and yaw movements, you can see clearly at the end of your log you were pushing left stick down 100% and right stick down 100% this is what drove your aircraft into the ground.

Yes because of bad compass these commands may not have been 100% correct but log shows using yaw pitch and roll that they were pretty close to pilots commands.

Thank you for your answer. I can not understand why Spark was accelerating to the very end. When I noticed what was happening, I thought that if he approached the obstacle, he would stop by himself. But he did not do it, only by accelerating he fastened the right front motor. He was a little too low. And did not he get too fast in the usual flight mode? I would like to remind you that AC accelerated by itself. I did not do it.

Gunship9 Posted at 2018-1-20 16:36
They are a pleasure.  My wife had fun flying mine.  However, radio control models are subject loss.  None are foolproof against this loss.  Never send into the open skies what you can't afford to lose.  With the compass errors (smart watch or nitro engine starter nearby?) some error happened where GPS stayed on but was not accurate, maybe?  You have to pull back on the sticks to stop forward motion of a heli or quad when it is not being GPS stabilized (ATTI mode equivalent).  Can't just release the sticks to stop like it does when GPS/IMU stabilized.

On the up side, you might know something about RC models since you were at a RC field.  That frame you broke is only $30 and you just need to follow some youtube videos to swap the parts over.  Easy repair if DJI doesn't warranty it.

Thank you for your message. This is not about money. I want to understand what happened and why Spark flew on me constantly. It would be a fear to think if my or somebody's child was on the way to Spark. I jumped back and I want to know why it happened.

tomaszmm76 Posted at 2018-1-20 16:40
Thank you for your answer. I can not understand why Spark was accelerating to the very end. When I noticed what was happening, I thought that if he approached the obstacle, he would stop by himself. But he did not do it, only by accelerating he fastened the right front motor. He was a little too low. And did not he get too fast in the usual flight mode? I would like to remind you that AC accelerated by itself. I did not do it.

You could check your flight playback in your app, if you check your gps bar graph to see the strength of your gps, (not the number of satellites ) but the graph, you must have 4 bars lit up green any less and you didn’t have enough gps lock, I’m thinking this might be the case, if Aircraft is not holding position then it is a result of weak gps.

tomaszmm76 Posted at 2018-1-20 16:53
Thank you for your message. This is not about money. I want to understand what happened and why Spark flew on me constantly. It would be a fear to think if my or somebody's child was on the way to Spark. I jumped back and I want to know why it happened.

Fair enough.  I think putting it into the ground is what I would have done.  Better than it flying away if it can't be controlled.  It is durable and I think most parts are fairly cheap.  

I think it lost accurate compass info but did not drop into ATTI mode as it should have.  It seems to have stayed in GPS without knowing where the aircraft was pointing making it impossible to stability hover.  You would be fighting the autopilot's efforts at that time.  Something around the launch area that pulled the compass slightly away from true north?  Super magnet holding a RC plane's radio hatch closed?

Did you lose orientation with the front of the drone and push the sticks the wrong way (newbie RC pilot type mistake)?

Gunship9 Posted at 2018-1-20 17:03
Fair enough.  I think putting it into the ground is what I would have done.  Better than it flying away if it can't be controlled.  It is durable and I think most parts are fairly cheap.  

I think it lost accurate compass info but did not drop into ATTI mode as it should have.  It seems to have stayed in GPS without knowing where the aircraft was pointing making it impossible to stability hover.  You would be fighting the autopilot's efforts at that time.  Something around the launch area that pulled the compass slightly away from true north?  Super magnet holding a RC plane's radio hatch closed?

I have not lost my orientation. Spark flew past me in a steady direction. I detached him and tried to stop him.

Let’s assume that you did not do anything wrong. There is possibility that Spark behave that way. There was something wrong with the IMU that cause the front of the aircraft lower. The meaning is the aircraft was not leveling horizontally but laid head down and that cause the aircraft to fly forward in quad copter. According to you, you encountered compass error. In many reports the IMU error is connected to compass error for Spark. When the speed is high enough the obstruction avoidance is not working.

Spark is a high-tech aircraft. But not everyone of them is perfect. When an incident occurs, we need to investigate for all possibilities. Only a few of the possibilities can be explained via flight log and pilot story. The rest of them require laboratory examining. Your case is an example. What cause the Spark to fly forward by itsself can be wrong sensor value reading, propeller degradation, motor mal function, speed controller malfunction, flight controller error, etc, It is easy if the error can be recreated. But if it is tempolary and reset buy system restart then it is hard to find. In most case the tempolary error may not happen again forever or until the proper condition occurs. So try it again under suitable control to see if this may happen again. Then decide to take chance or send it for examination,

hallmark007 Posted at 2018-1-20 15:55
Yes , but if you look at your log you can see when you push stick up your aircraft moved up an same when you pushed left right and yaw movements, you can see clearly at the end of your log you were pushing left stick down 100% and right stick down 100% this is what drove your aircraft into the ground.

Yes because of bad compass these commands may not have been 100% correct but log shows using yaw pitch and roll that they were pretty close to pilots commands.

In sum. If the DJI Spark control system analyzing all data ie compass, IMU, battery, RC, etc. during the AC start, displays information that everything is normal and that we can start. Does it not mean that it is so? Can not trust?  According to you, what should I do if something like this happens to me?  Do not lower the AC flight and try to land? Because you are suggesting that I was crushed by my fault. Can I, according to you, raise the flight higher and let Spark fly away? What then is associated with many hours of searching for it? I think I was still broken because I would not count on Home Point anyway

When you received compass error early on, did you move around and observe the compass interference screen.?it's 100 per cent essential to do this before recalibrating. You may have recalibrated in the problem area, then flown out of it with the
error programmed in.

tomaszmm76 Posted at 2018-1-20 23:16
In sum. If the DJI Spark control system analyzing all data ie compass, IMU, battery, RC, etc. during the AC start, displays information that everything is normal and that we can start. Does it not mean that it is so? Can not trust?  According to you, what should I do if something like this happens to me?  Do not lower the AC flight and try to land? Because you are suggesting that I was crushed by my fault. Can I, according to you, raise the flight higher and let Spark fly away? What then is associated with many hours of searching for it? I think I was still broken because I would not count on Home Point anyway

Ok if you look at your flight log that starts at 4.20 sec.
You took of hovered for approx 10 sec , you received many compass warnings at 20 seconds into that flight.
When this happened you were at altitude of 1.3 metre, you had a problem and you had time to land aircraft , but decided to carry on flying hoping compass problem would go away.
As the pilot that is your decision and now your responsibility. You may have your own reasons why you carried on flying but this is always the pilots choice. So my question to you is, who is responsible?

Regarding end of flight crash, well it’s similar again it was you who decided to push down on both sticks resulting in aircraft crash, again your decision your responsibility.

Regarding the problems I’m not equipped to tell you there was no problem with aircraft or malfunction, that will be decided by dji and while they may or may not give you warranty , I was reading your log and looking at what you had wrote and making my points on this.

Regarding pilot error, in most of these cases there will always be an element of pilot error and while most are maybe not experienced to handle some situations simply because of lack of experience, it doesn’t mean that they should not receive warranty simply because if there was a problem with the aircraft this should be paramount to any mistakes made by the pilot.

So I’m not saying this was your fault, but a Pilot always has a responsibility.
Below is a link that might help when doing a preflight check.

https://forum.dji.com/forum.php? ... &fromuid=260008

Was it flying forward, backwards or sideways when it crashed?

20 warnings I did not see on Go4 during the flight. This can only be seen on phantomhelp - at home and after all. So do not tell me about responsibility. I would be irresponsible if I let the drone fly away and who knows what, where or whom he would hit. Then what ? This is my opinion. You do not have to agree with this. Peace

I would advise to use a mobile device screen recorder all the time! Very usefull to playback a flight and see what happened on the screen while looking at the drone.
Could be great proof to DJI when indeed some messages are not shown on the App, mayby there should be a loud beep when an warning error shows up the screen....
cheers JJB

JJBspark Posted at 2018-1-21 02:54
I would advise to use a mobile device screen recorder all the time! Very usefull to playback a flight and see what happened on the screen while looking at the drone.
Could be great proof to DJI when indeed some messages are not shown on the App, mayby there should be a loud beep when an warning error shows up the screen....
cheers JJB

Yes. I agree with you 100% I will definitely record everything from now on. Although they will probably try to blame us. Seeing what is happening.

Radio controlled models need a slightly philosophical approach.
If you fly a lot, some get wasted.
Some immediately blame the radio gear etc.
But if you do not wish to improve, it's not for you.
Learn from it, move on.

Hello, everyone again. So the whole situation with Spark ended positively. The little drone was sent to the Netherlands, I described the whole case that happened to me. And after a few days DJI sent me a completely new Spark. I do not know what was wrong with the one who crashed. But in my opinion it was too much compass errors, and I paid particular attention to fly in places clean from electromagnetic interference. I have no information about what caused Spark to break up. So most importantly - I regained confidence in DJI.
PS. I do not fly anymore with Spark, I decided to sell.  I added and bought Mavic PRO Platinum. This is what I needed. Flying with Mavik is a great experience.    And I'm very happy.
Thank you DJI for your support and help.

tomaszmm76 Posted at 2018-1-20 13:55
[view_image]

Spark flew ahead. Spark flew ahead - why Obstacle Avoidance System did not work ???

I   read your    Problem and   when I saw  word Sports mod that  did that your drone Crash Obstacle avoidance   did not work because your drone     was in Sports Mod  you can read manual there you can find that Obstacle avoidance will no work when Spark in Sport Mod  You should turn Sport mod Button off    for obstacle avoidance to work another way it will not work or go directly to  DJI go 4 apps and activate it  

Its to bad that you drone Crashed   There  are  3 thing that  Contribue to your drone to crash  1 . Compass  Erro  2  IMU error  3  Sport Mod  

tomaszmm76 Posted at 2018-1-20 13:55
[view_image]

Spark flew ahead. Spark flew ahead - why Obstacle Avoidance System did not work ???

-"Spark flew ahead. Spark flew ahead - why Obstacle Avoidance System did not work ???"-
Sport mode!

o.o.T
Any changes to phantomhelp?
Clicking the links to the flightlogs gets me a switch to https://www.phantomhelp.com/Phantom-4/ every time.
On several browser

edit:
clicked on other links, they opened.
so whats different with this ones?
Did OP take 'em down?

tomaszmm76 Posted at 2018-1-20 11:58
I flew low because I saw something is wrong. And I did not want him flying away for a few hundred meters. I decided to fly low to me and land then. I was convinced that Spark can fly at the height of one meter and it will be a safe, slow return. Then he accelerated. I will also add that there was no error on DJI GO 4.

Nope.
flying low is a pilot error especially @1 meter height. The GPS doesn't always work at low heights. You need to be atleast at 30 to 40 meter height for a good flight ...