Hello, I'm from Germany and I've some problems with my "new" DJI Spark Combo.My english is very terrible, but I hope you can understand me. ;-)

My first copter is a DJI Phantom 3 Adv., but a few month ago, I thought about a smaller, second copter, that I always can use or take with me.

So I decided to buy the DJI Spark as second copter.
One or two month all is ok.
I always made the updates or calibrated the copter if the dji go app wanted it.

Now a few weeks ago, I wanted to take pictures near a empty stadium in Germany (no game, no people at sunday morning).
It's a no fly zone (yellow), but I talked with the "operational line" from the "fire Department" and it was ok.
So I opened the yellow fly zone on the DJI Homepage for 3 days.
On Sunday morning I had also to open the fly zone in the DJI Go App, ok.
Then I wanted to start, but first there was to much metal (compass error), so I went to an other place nearby and started the copter successfully (homepoint ok).
I flew 9:43 min., took some nice pictures, but then I got many error-messages on my display (i flew with remote controller an my iphone 6s, all things with the newest firmware).
But first the conditions of the flight in the last minutes: Sport-mode, 20 Satellites, 29% accu, flight-altitude 49 m, distance 50 m.
Now the errors: "your flight goes through a no-fly zone, be carefully...", "GPS-Error (with 20 satellites?)", "weak GPS-signal", "compass-error. exit P-GPS-mode", "working IMU encounters heading exception,please switch to atti mode if craft behave abnormally", "weak GPS...fly with care?" and so on.
Then the copter wanted to fly away over the tall trees and another problem was, that the position lamps are all red, if there are errors, so I put the sticks right and left and forwards and back and then - I was very happy about it - the copter wanted to return home because the accu was empty.
My fault, that I started under a tree, so he came back and I had forgotten, that he started there.
He fell down and broke one leg.
Ok, so I gave the copter to a colleague (his hobby: he builds kopter and repair them) and he change one motor and repair the leg.

But now, 4 days ago, I started the spark copter again.
Only fields, 250 m away a wind turbine. I could start without error-messages. Only in normal-mode, 20 m away, 50 m altitude. After 2 Minutes I had the best position to take pictures from the city far away in the evening-sun, but suddenly many errors appeared on my display! I look in the sky, the spark was away!
It was so fast, I couldn't see him!
I try to connect the DJI Go-App again with a restart, nothing!
I restart the remote controller, nothing!
I went 3 ours with my family and friends over the fields, nothing!
The next day, I asked my boss for one holiday and search 4 hours, nothing.
Then I didn't want to search anymore.
In the evening I read some threads in our german copter forum and I found the option to read the flight-records from my handy in http://www.phantomhelp.com/logviewer
The are more details about the flight and the place where the copter landed! More than 300 m from the place I started!
It was in the night, but I phoned with a friend and we drove to the field and after one hour we had found the spark.
This is the good part of the second fly away.
The spark was wet (one short rain), but no defects, ok the broken leg was broken again.
After 2 days in a rice dip (?) I started the copter without propeller. no errors.
Then I repair the leg with epoxide resin and fiberglass tape.
And now?

In this forum and in the german forum I collected so many informations, that I can understand some more details.
I could read the whole flight record, I saw some threads with lost sparks and I'm pretty uncomfortable or better: At this moment I have no confidence in the technology of the spark!!!
I had made all updates (IOS, DJI GO-App, copter, controller, accu), I always check the environment conditions, there are no errors, I can fly more minutes and then the spark fly away? Why?
In the flight records it often/always began with "Yaw Errors" (and you didn't see it on the phone-display!) or "compass + yaw errors", always these yaw errors!
What is it or better: Where do they come from?
Here I can read from a "kiss of death" (beautiful name) or a fly away and one time dji exchanged the copter without comment.
But this is not my desire/wish!
I got the spark combo cheaper from china, so I must pay for my faults, it's ok.
But when I read about all the problems with this YAW-Error, then I think, the 2 fly aways ... what's this? Were these my faults?
I only want one thing:
It would be enough for me if someone would say to me: "Do this or that and then the yaw errors will not occur anymore! You can trust in the technology of your spark, if the connection lost, the spark came back to his homepoint."
But it is noch solution, if the spark copter lost the signal with yaw-errors and gps-errors and you see 20 satellites on the phone display BEFORE AND AFTER lost signal!
Because I want to be able to fly somewhere in the city without fear anytime!
With my phantom 3 no problem, with this spark or with this yaw-/gps-problem impossible.

Thanks for reading this long text. And maybe somebody can help me...

The link to the last fly away flight record: https://www.dropbox.com/s/iyb755 ... 6-11-14%5D.txt?dl=0

Your link does not take you to your log, think you may have put up wrong link.

Hello, I try it and the link take me to the log. I think you must download the txt-file? Or can I try another way?

gkrause Posted at 2017-11-17 06:48
Hello, I try it and the link take me to the log. I think you must download the txt-file? Or can I try another way?

When I try to reach phantomhelp log it just brings me to upload page.

Oh, I don't know what I have to do. Here is the link to the folder in my dropbox to read the txt-file:
https://www.dropbox.com/sh/w6cfk ... jLUMv0aKjTeega?dl=0
Perhaps you can work with this?

Personally I think to fly with a glued up arm let alone motor replaced unless it was a genuine spark item is crazy. You dont know if that repair failed in flight, thus making diagnosis even harder.

gkrause Posted at 2017-11-17 06:57
Oh, I don't know what I have to do. Here is the link to the folder in my dropbox to read the txt-file:
https://www.dropbox.com/sh/w6cfknacc2afj9w/AADPVKOe4HdjLUMv0aKjTeega?dl=0
Perhaps you can work with this?

The problem is ou haven’t uploaded your log which you must do from your app. I can’t read other file it just comes up encrypted.

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

hallmark007 Posted at 2017-11-17 07:30
The problem is ou haven’t uploaded your log which you must do from your app. I can’t read other file it just comes up encrypted.

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

Here it is: http://www.phantomhelp.com/LogViewer/39SJZ205YKL9YRKWJH9O/
(I have downloaded it from his dropbox and then uploaded on plantomhelp)

Again it starts with a YAW error, in 0.5 secs to ATTI, followed by the other well knows messages....
Flying at 200 ft in kind of open area...must be that magnetic interference flying into position...., yeah ofcourse...

Thank you Wachtberger!!!!
@JJBSpark: My English is not very well, I don't understand. Can you help me and describe it a in other words? The google translator can't help me, there are open questions.
You mean, it's the wind turbine?
I forgot to tell you, that I was at the same place 4 weeks ago and there was no problem with the spark.
I flew near the wind turbine til the spark says, he didn't go further. But no problems and I can take it back.

What I wonder is, how accurate the indicated flightpath in the logfile reading can be since for the larger part of the flight the Spark was in ATTI mode already. If there was some wind couldn't it easily have been pushed over the area behind that is covered with trees and would make it extremely difficult (if not impossible) to retrieve?

Bright Spark Posted at 2017-11-17 07:25
Personally I think to fly with a glued up arm let alone motor replaced unless it was a genuine spark item is crazy. You dont know if that repair failed in flight, thus making diagnosis even harder.

Possible, but very unlikely, I can't imagine. The 3 persons, who repaired the copter, are really professionals and do it since 30 years and more.

Wachtberger Posted at 2017-11-17 08:51
What I wonder is, how accurate the indicated flightpath in the logfile reading can be since for the larger part of the flight the Spark was in ATTI mode already. If there was some wind couldn't it easily have been pushed over the area behind that is covered with trees and would make it extremely difficult (if not impossible) to retrieve?

Yes, of course, I'm happy that he had beautifully kept this straight line! A little bit more wind or a small bend and I would never have found it again, everyone agreed!

Wachtberger Posted at 2017-11-17 08:51
What I wonder is, how accurate the indicated flightpath in the logfile reading can be since for the larger part of the flight the Spark was in ATTI mode already. If there was some wind couldn't it easily have been pushed over the area behind that is covered with trees and would make it extremely difficult (if not impossible) to retrieve?

Yes, of course, I'm happy that he had beautifully kept this straight line! A little bit more wind or a small bend and I would never have found it again, everyone agreed!

Sorry GKrause, my post was little cynical. This because i read a lot of this Spark mishaps and it always starts with a YAW problem.
I do not believe that all these cases happens because of an external cause, in my opinion something must be wrong with the Spark. (bad batch made on a monday...)
But only DJI knows this and ofcourse they  won`t tell us.

Cheers,
Hans  

JJBspark Posted at 2017-11-17 09:15
Sorry GKrause, my post was little cynical. This because i read a lot of this Spark mishaps and it always starts with a YAW problem.
I do not believe that all these cases happens because of an external cause, in my opinion something must be wrong with the Spark. (bad batch made on a monday...)
But only DJI knows this and ofcourse they  won`t tell us.

https://forum.dji.com/thread-103989-1-1.html

Here’s one form July I don’t think it’s a bad manufactured batch.

JJBspark Posted at 2017-11-17 09:15
Sorry GKrause, my post was little cynical. This because i read a lot of this Spark mishaps and it always starts with a YAW problem.
I do not believe that all these cases happens because of an external cause, in my opinion something must be wrong with the Spark. (bad batch made on a monday...)
But only DJI knows this and ofcourse they  won`t tell us.

There is a lot that they do not tell us.

The compass drifting as the spark accumulates flights (42 flights in this case) as described in #13 of https://forum.dji.com/thread-114121-1-1.html

FW update 6 totally resolved that, 120 flights and found a deviation of 12, im sure that is the contributor to the reduced compass errors found.

Still have to try FW update 7, waiting for better weather.

JJBspark Posted at 2017-11-17 09:15
Sorry GKrause, my post was little cynical. This because i read a lot of this Spark mishaps and it always starts with a YAW problem.
I do not believe that all these cases happens because of an external cause, in my opinion something must be wrong with the Spark. (bad batch made on a monday...)
But only DJI knows this and ofcourse they  won`t tell us.

Hi Hans, I also don't believe only in external causes! The same flights, the same locations with my DJI Phantom 3 make no problems.

hallmark007 Posted at 2017-11-17 09:32
https://forum.dji.com/thread-103989-1-1.html

Here’s one form July I don’t think it’s a bad manufactured batch.

Thank you Captain, I will read it carefully!

Hecate Posted at 2017-11-17 09:46
There is a lot that they do not tell us.

The compass drifting as the spark accumulates flights (42 flights in this case) as described in #13 of https://forum.dji.com/thread-114121-1-1.html

Having a compass error while flying doesn’t necessarily mean you have a bad compass, you could have a good compass getting interference causing the error.

Hecate Posted at 2017-11-17 09:46
There is a lot that they do not tell us.

The compass drifting as the spark accumulates flights (42 flights in this case) as described in #13 of https://forum.dji.com/thread-114121-1-1.html

I almost fear it!
But I use always the newest firmware 4 days ago (13. Novembre). Is it Number 7?
Thank you for your link and tests!!!

hallmark007 Posted at 2017-11-17 10:27
Having a compass error while flying doesn’t necessarily mean you have a bad compass, you could have a good compass getting interference causing the error.

I agree, and i am sure that all possible combinations can occur.
One of my concerns was a compass in the green that had drifted  205 points which chipped  40% of the max allowable interference during flight away and was causing compass errors in flight. But that's solved now with FW 6.

gkrause Posted at 2017-11-17 10:31
I almost fear it!
But I use always the newest firmware 4 days ago (13. Novembre). Is it Number 7?
Thank you for your link and tests!!!

Yes 7 is the latest

gkrause Posted at 2017-11-17 06:57
Oh, I don't know what I have to do. Here is the link to the folder in my dropbox to read the txt-file:
https://www.dropbox.com/sh/w6cfknacc2afj9w/AADPVKOe4HdjLUMv0aKjTeega?dl=0
Perhaps you can work with this?

Looking at your flight log, your environment certainly should not have caused you any problems, I’m not sure if your first crash may have been instrumental in causing your problems, patching up the arm of your spark doesn’t fill me with confidence.
I’m not sure if it’s possible for you to arrange to ship your spark to Netherlands’s for repair but at this stage I would think this might be the best course of action. They will also be able to check your dat file from your spark this will give a much better understanding of what happened on your first crash and second one. If it’s only body parts to be replaced and to get compass and IMU checked out I don’t think it would be to expensive.

I have a P4Pro I have never calibrated compass or IMU and I have over 300 flights, I do check compass and IMU before every flight and always check small triangle on radar to make sure heading is correct and matches heading of Aircraft.

Hecate Posted at 2017-11-17 10:42
Yes 7 is the latest

Do you fly in FCC-Mode? I think the signal between RC and copter is better. Or do you think, it‘s equal for the signal-quality?

gkrause Posted at 2017-11-17 11:16
Do you fly in FCC-Mode? I think the signal between RC and copter is better. Or do you think, it‘s equal for the signal-quality?

Yep i am in Canada so fcc much farther range than in EU

hallmark007 Posted at 2017-11-17 11:15
Looking at your flight log, your environment certainly should not have caused you any problems, I’m not sure if your first crash may have been instrumental in causing your problems, patching up the arm of your spark doesn’t fill me with confidence.
I’m not sure if it’s possible for you to arrange to ship your spark to Netherlands’s for repair but at this stage I would think this might be the best course of action. They will also be able to check your dat file from your spark this will give a much better understanding of what happened on your first crash and second one. If it’s only body parts to be replaced and to get compass and IMU checked out I don’t think it would be to expensive.

Yes, I also think so. The environment not have caused the problems.
And I don't think the arm/leg of my spark is the problem.
You think the fall from 4 m height was responsible for the second accident? Vibrations, then defects?
Maybe - but the error messages (compass and yaw) were already present before (!) the first crash.
But if I don't get any idea how I can control the copter safely in the city... then I agree with you!
Only DJI itself can find and fix the problem.
And my colleague would still be able to exchange the middle frame, he had already offered it to me.
But the costs are of course already an issue, because I do not (yet) see that I have made a mistake.
Thank you for your opinion, tomorrow I try to fly again.
At first only in beginner-mode and before starting I calibrate the sensors.

gkrause Posted at 2017-11-17 14:40
Yes, I also think so. The environment not have caused the problems.
And I don't think the arm/leg of my spark is the problem.
You think the fall from 4 m height was responsible for the second accident? Vibrations, then defects?

Sorry for you experience. For your first crash, pre-flight check is necessary and you should fly in a place with good GPS signal. Also, I would not recommend you repair the drone without sending back as no professional worker do the fly test for your drone.  Please contact our support first and start a case: http://www.dji.com/support.

DJI Elektra Posted at 2017-11-17 20:25
Sorry for you experience. For your first crash, pre-flight check is necessary and you should fly in a place with good GPS signal. Also, I would not recommend you repair the drone without sending back as no professional worker do the fly test for your drone.  Please contact our support first and start a case: http://www.dji.com/support.

Thank you for your participation! Yes, I always look for a good GPS-Signal (I had 17-20 satellites before the first crash). No errors, no wind, no sun storms, no calibration necessary. At the beginning of "my life with copters" one year ago, I had make a checklist before every flight. I think I have minimized the error rate.
I knew that the Spark is a little bit silly (is this the right word?).
So if I get errors (compass error), I try another place 10 meters next to it.
Only if I get no errors I'm starting.
But what to do in the air, if the sticks not work after the compass- and yaw-errors?
And that was the first time, before the first crash.
I don't know, I lost confidence in technology. After 2 months - from today to tomorrow - came the mistakes.
Anyway, I'll test a few more days with the new tips (thanks @Hecate: measuring the magnetic field at the starting point, calibrating the sensors every time if the compass value is too high) and today I'll get antenna amplifiers.
And if all this does not work then I'll check the spark at DJI.

From what I have seen in this case and many others, it always starts with Yaw Error warnings quite before the other error messages and a possible loss of connection kick in. From this my conclusion is not to wait, but to immediately land when I see the first Yaw Error message. Any other views on that?

gkrause Posted at 2017-11-17 23:55
Thank you for your participation! Yes, I always look for a good GPS-Signal (I had 17-20 satellites before the first crash). No errors, no wind, no sun storms, no calibration necessary. At the beginning of "my life with copters" one year ago, I had make a checklist before every flight. I think I have minimized the error rate.
I knew that the Spark is a little bit silly (is this the right word?).
So if I get errors (compass error), I try another place 10 meters next to it.

Since your spark crashed two time, I strongly recommend you send it back for a stable flight performance. Thanks for your support.

Wachtberger Posted at 2017-11-18 01:04
From what I have seen in this case and many others, it always starts with Yaw Error warnings quite before the other error messages and a possible loss of connection kick in. From this my conclusion is not to wait, but to immediately land when I see the first Yaw Error message. Any other views on that?

It takes only less than a second (most times about 0.5 sec) after the first YAW message that the Spark goes automatically in ATTI mode.
Than after very short moment Spark tells us to switch to ATTI mode if problem remains....
Real "fun" because i cannot switch to ATTI mode and Spark is already in ATTI!

I agree that mayby the best reaction is to land the craft, if it is nearby flying. It become a challenge if the craft is more than, lets say, 300 feet away....

So i think DJI should come with a solution the the YAW failures.
Its always magnetic interference they suggest, well i did fly yesterday starting 2 meters next to my car (1500 kg of pure metal) and no interference at all. I think that all SPARKS are not build to the same quality level. (if the sum of all the electronic parts-deviation is too great you will face problems...)

Cheers
Hans

'Magnetic interference' is often mentioned like a mantra in threads concerning yaw errors. But the point is that physical laws are not some kind of magic and it is quite unlikely for a magnetic interference to just drift in the air. Out of top of my head I can imagine just two cases that you may encounter a magnetic interference when you are high in the air (i.e. away from any objects and structures): either you fly into a storm cloud or in a radioactive area. In both cases you have much more serious problems at hand than yaw error. Any other idea how you may encounter a magnetic interference in open air, anyone?

_pk Posted at 2017-11-18 04:36
'Magnetic interference' is often mentioned like a mantra in threads concerning yaw errors. But the point is that physical laws are not some kind of magic and it is quite unlikely for a magnetic interference to just drift in the air. Out of top of my head I can imagine just two cases that you may encounter a magnetic interference when you are high in the air (i.e. away from any objects and structures): either you fly into a storm cloud or in a radioactive area. In both cases you have much more serious problems at hand than yaw error. Any other idea how you may encounter a magnetic interference in open air, anyone?

Just one more natural cause is solar activity, all the rest would be anthropogenic.
Magnetic interference cannot drift in the air as EMF requires a current, shut off the current and the EMF is gone.
In my case the yaw errors i had was not caused by magnetic interference as i used a magnetometer and RF multi band scanner prior to launch and the error started 10 feet away from the launch location. If interference was present at 10 feet from me, my measuring equipment would have picked it up.

Today I started Spark for the first time after the fly away at the same location!!!
But first I calibrated the compass (without prompting).
Unfortunately it was very, very windy, so I only let fly him for 3-4 minutes.
No error messages, except the wind warning on my phone display.
If you feel like it, feel free to look at the flight record file.
Here is the link to the folder in my dropbox:

https://www.dropbox.com/sh/w6cfk ... jLUMv0aKjTeega?dl=0

gkrause Posted at 2017-11-18 06:36
Today I started Spark for the first time after the fly away at the same location!!!
But first I calibrated the compass (without prompting).
Unfortunately it was very, very windy, so I only let fly him for 3-4 minutes.

According to the phantomhelp flightlog viewer you have had five connection losses although flying in very short distance at low speed (beginner mode) only:

http://www.phantomhelp.com/LogViewer/FU37COBTA6XXADWJW3KB/

This is also confirmed by the Airdata logfile view (although it doesn't display the first very short connection loss):

https://app.airdata.com/main?fli ... ENERALNotifications

Please correct me anyone who is more experienced than I am! But for me this would be a reason for concern because it is not normal behaviour and I would rather get my Spark checked after all what had happened before already as reported here in this thread.

Do you have any idea, where the powerlines from the windmill got digged in?
Asking myself, it if they could be below your starting point ....
I am flying sometimes near a windmill, mostly in resting position.
And in case it is moving, I try to stay away from it.
What do you mean with this, please:
"I flew near the wind turbine til the spark says, he didn't go further."

S-e-ven Posted at 2017-11-18 19:42
Do you have any idea, where the powerlines from the windmill got digged in?
Asking myself, it if they could be below your starting point ....
I am flying sometimes near a windmill, mostly in resting position.

I think, the app said there was a "no fly zone", perhaps because power generation?

Wachtberger Posted at 2017-11-18 15:02
According to the phantomhelp flightlog viewer you have had five connection losses although flying in very short distance at low speed (beginner mode) only:

http://www.phantomhelp.com/LogViewer/FU37COBTA6XXADWJW3KB/

Wow, the airdata app is very good! It shows mistakes better than the logviewer from phantomhelp AND explains the values!
Ok, it would be better to send the Spark to DJI to get it checked.

Only one question remains, how did it happen after 2 months and how did the first yaw errors and compass errors come about? But that DJI will certainly find out.