Phantom 3 fly away and a strange logfile

[Flying Dutchman]

First: Excuse me for my very bad grammer.

Yesterday morning I made a video of the sun rising. I did not calibrate the compass because I did not change position more than 10 mile sins the last time I flew my supermachine. I took off and had the Phantom advanced at the same spot without moving for some time.. Suddenly I noticed that my Phantom changed position and had a very radicale flight path.. I tried to take back the controles but it seemed that the quadcopter did not listen. I made an affort to get back to the original position, gave full throttle forwards but again the Phantom did not listen like she used to. There was so little progress and left and right turn did not work also. I switched the GPS off and after a few seconds back on. After a very scary ten minutes II finally got her back on her position and landed.

So now I want to now what was going on there. So healthydrones here I come. But by my suprice, when I wanted to download the file from my Ipad it says that its deleted....WTF! (screenshot below) al my other flight records are there, but the one I need with the problems got deleted.. How is this possible? Is there a way to repair the file? What went wrong up there? There is a warning ( i did not notice while in flight) what kind of warning was that? What could it be?

I landed and send her right away up there again, two more times I emptied a battery without any problems. I really would like to now what was going on. Maybe you guys have an idea?

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

The video is 2x

Thx for your time in advance

 

[moswissa]

It sounds like pilot error to me no offense. glad you did lose her.

why would you change from your original position and how far did u walk?
i always feel like that will throw something off.

you should always calibrate even tho you didnt change a position, its just reassurance and it only takes literally 8 seconds.

 

[CloudsNeverDie]

Sync your flight logs using the DJI GO app, then use a tool like DJI Flight Record Converter to download your flight log from SkyPixel. Then post the flight log file (preferably raw) here. We can all analyze it and figure out what went wrong.

 

[JohnK]

It sounds like pilot error to me no offense. glad you did lose her.

why would you change from your original position and how far did u walk?
i always feel like that will throw something off.

you should always calibrate even tho you didnt change a position, its just reassurance and it only takes literally 8 seconds.

This is one of those hot topics of disagreement for which a sort of superstition overrules factual information. Simply no need to "always recalibrate", and calibration under disadvantageous conditions may carry some risks, if you can believe all that you read out there. I personally don't. It's not a matter of how long it takes. Check your mods before each flight. I just took a road trip and drove almost 6,000 miles in a rectangle, and calibrated my compass only a few times in 40+ flights without an inkling of an incident.

 

[Flying Dutchman]

It sounds like pilot error to me no offense. glad you did lose her.

why would you change from your original position and how far did u walk?
i always feel like that will throw something off.

Its because of my bad englisch that you did not understand it. I did not change position while in flight. I mean compared with my last flight.

you should always calibrate even tho you didnt change a position, its just reassurance and it only takes literally 8 seconds.

thx for your answer, its not possible that its a pilot errror. Because the Phantom was holding position for already 4 minutes, it started to drift after I pushed the left throttel up. The only error I made is that I did not notice the drifting right away.

About the calibration, I used to do the calibration every flight. But that its not necessary and can do more harm than good. Just like JohnK says. You find the information on page 44 from the manual.

 

[Flying Dutchman]

This is one of those hot topics of disagreement for which a sort of superstition overrules factual information. Simply no need to "always recalibrate", and calibration under disadvantageous conditions may carry some risks, if you can believe all that you read out there. I personally don't. It's not a matter of how long it takes. Check your mods before each flight. I just took a road trip and drove almost 6,000 miles in a rectangle, and calibrated my compass only a few times in 40+ flights without an inkling of an incident.

Thx JohnK, by the way..Wow, you have a great site, really pro photographs

 

[Flying Dutchman]

Sync your flight logs using the DJI GO app, then use a tool like DJI Flight Record Converter to download your flight log from SkyPixel. Then post the flight log file (preferably raw) here. We can all analyze it and figure out what went wrong.

Thx Matthew, your the best!

 

[Chris P Duck]

you should always calibrate even tho you didnt change a position, its just reassurance and it only takes literally 8 seconds.

Please don't suggest this, listen to what those with experience or even what DJI themselves state in the manual. You may get away with it for the entire lifetime of the P3, but that doesn't make it best practice.

Honestly it will increase the chance of something going wrong.
The more you do it the greater the chance of overwriting a good calibration with a bad one (when the good one would have been fine unless you moved a great distance)
Check the mods and recalibration when needed, not every flight.

 

[JohnK]

Thx JohnK, by the way..Wow, you have a great site, really pro photographs

Thanks,amigo ... you just made my day ... and it's my birthday. Will be posting drone vids soon.

 

[flyNfrank]

Thx Matthew, your the best!

I have looked a great deal of flight logs. And what you posted here is doesn't come close to the flight you described.

 

[CloudsNeverDie]

This flight log does seem to be the same as the one from our original post. I'm seeing high xSpeed and ySpeed values even though you had the stick central. Was it very windy?

 

[CloudsNeverDie]

Hmm, I answered my own question, looks like winds were about 5mph in Franeker at the time of your flight.

If I had to guess what caused this, I'd say a bad IMU calibration, or an IMU calibration that didn't cover the full range of temperatures. Did you perform an Advanced IMU Calibration yet? I note that it was quite cold at the time (about 35 deg F) and the change in temperature as the electronics warmed up during your flight might have been quite large, which could have caused bad gyro readings, leading to a bad roll and pitch estimate. In other words, the quad thought it was level and stationary, but actually it was tilted 10-20 degrees and flying off in one direction, yet at the same time fighting that and trying to stay in one spot.

 

[Flying Dutchman]

Hmm, I answered my own question, looks like winds were about 5mph in Franeker at the time of your flight.

If I had to guess what caused this, I'd say a bad IMU calibration, or an IMU calibration that didn't cover the full range of temperatures. Did you perform an Advanced IMU Calibration yet? I note that it was quite cold at the time (about 35 deg F) and the change in temperature as the electronics warmed up during your flight might have been quite large, which could have caused bad gyro readings, leading to a bad roll and pitch estimate. In other words, the quad thought it was level and stationary, but actually it was tilted 10-20 degrees and flying off in one direction, yet at the same time fighting that and trying to stay in one spot.

What you describe make sense, because thats exactly what happend. I had the feeling while I tried to steer the Phantom home, that I was batteling against heavy winds. But there was no wind at all. Atleast on the ground there was non. Maybe in the upper layers at 200 meters, but that sounds also wrong because the differ with the ground wil be too big. I flew in heavy winds, and this was the same feeling but only there was no wind at all. It was strange and scary at the same time. You can see that I panic a little bit, because of my rapid stick movement.

If you put my log file threw healthydrones.com there is a new chapter called Wind. It says the average wind speed is 43.8 km/h (27.2 mph) and the max gust speed 69.8 km/h (43.4 mph) And the avarage wind direction was North West. If you watch the youtube video at 3.01 minutes you see that the Phantom suddenly takes a stiff left corner with full speed. And then with the stick full throttle it recovers at 3.12 and goes straight again. Around that time I let the stick sit in neutral and the Phantom starts going backwards. It all adds up.

The solution you give is plausible if not the correct one, Thank you very much I feel much better now that I now the reason for the irradicale flight.

So if I understand it correct, you advice me to do a Advanced IMU Calibration? I never did one. Because (same like compass calibration) if its ok I do not touch it. But if the advice is a Advanced IMU Calibration than I will do that. Again thank you very much!

 

[Flying Dutchman]

I have looked a great deal of flight logs. And what you posted here is doesn't come close to the flight you described.

Hi flyNfrank, thx for responfing. Englisch is not my native language. I tried to explain it as good as possible. And the solution "Cloudsneverdie" gives seems the correct one. How often do you do a Advanced IMU Calibration?

 

[CloudsNeverDie]

You're welcome! Thanks for your UAVForecast donation, much appreciated.

IMU calibration instructions here:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[flyNfrank]

Hi flyNfrank, thx for responfing. Englisch is not my native language. I tried to explain it as good as possible. And the solution "Cloudsneverdie" gives seems the correct one. How often do you do a Advanced IMU Calibration?

I didn't pay any attention to your wording or misspelled words. I only looked at 3 areas in the longs that should have had plenty of data matching the flight you described. But none of the areas I looked at showed ANY of what was mentioned. Example; You said you switched flight mode a few different times. But the log's do not show any switching taking place whatsoever. I'm not saying you were making things up, but more so based on the fact that some really strange stuff took place kind of like a conspiracy, if you will. When there is flight mode switching during a flight, it will show in a couple of different categories. There is no real way around this to not display in those categories when it actually takes place.

I'm curious if the .TXT file you attached in this thread is not the same as the flight before it, or after?

As for IMU Calibrations... For my aircraft, I keep an eye on the displayed data of the sensors when they are live/active. As long as they are with in my acceptable range, I no longer show any concern to the IMU. When a time comes that I would do a IMU Calibration, it would be by only doing the Advance Calibration. There is 3 reasons why I would do a IMU Calibration. 1. is after all firmware updates. 2. is if the sensor range is out of spec. 3. is if the aircraft is not flying normally, such as in a straight line, or not hovering properly(a non calibrated R/C can cause this as well).
*Btw, for IMU Calibrations I try to get my aircraft as cold as possible without putting it in the refrigerator or freezer. In most cases I use the air conditioning from my tahoe. or house. The freezer and refrigerator can cause condensation which is a big enemy for the electronic components and such inside the aircraft.

I know you did not ask me this, but I'm going to share it with you all the same. But I calibrate my compass prior to every flight. I have never had any weird or crazy stuff take place, like others do including yourself. I also have never inspected a flight log from a crash, or fly away that showed a compass calibration prior to flight. I'm a dji product tester and I go by what is dji's preferred method, except during certain points in a beta test. I then purposely do things out of the preferred method as some one else might to see if I can simulate a issue that may come about later on, and which would be using equipment near the same.

 

[moswissa]

Please don't suggest this, listen to what those with experience or even what DJI themselves state in the manual. You may get away with it for the entire lifetime of the P3, but that doesn't make it best practice.

Honestly it will increase the chance of something going wrong.
The more you do it the greater the chance of overwriting a good calibration with a bad one (when the good one would have been fine unless you moved a great distance)
Check the mods and recalibration when needed, not every flight.

i see, Well I used to recalibrate every flight with my P2 and never had a problem but what you said makes alot of sense. I guess I am just always paranoid about having a flyaway then coming back to the forums and have everyone say "you should have recalibrated" lol

good luck OP

 

[Flying Dutchman]

I didn't pay any attention to your wording or misspelled words. I only looked at 3 areas in the longs that should have had plenty of data matching the flight you described. But none of the areas I looked at showed ANY of what was mentioned. Example; You said you switched flight mode a few different times. But the log's do not show any switching taking place whatsoever. I'm not saying you were making things up, but more so based on the fact that some really strange stuff took place kind of like a conspiracy, if you will. When there is flight mode switching during a flight, it will show in a couple of different categories. There is no real way around this to not display in those categories when it actually takes place.

I'm curious if the .TXT file you attached in this thread is not the same as the flight before it, or after?

As for IMU Calibrations... For my aircraft, I keep an eye on the displayed data of the sensors when they are live/active. As long as they are with in my acceptable range, I no longer show any concern to the IMU. When a time comes that I would do a IMU Calibration, it would be by only doing the Advance Calibration. There is 3 reasons why I would do a IMU Calibration. 1. is after all firmware updates. 2. is if the sensor range is out of spec. 3. is if the aircraft is not flying normally, such as in a straight line, or not hovering properly(a non calibrated R/C can cause this as well).
*Btw, for IMU Calibrations I try to get my aircraft as cold as possible without putting it in the refrigerator or freezer. In most cases I use the air conditioning from my tahoe. or house. The freezer and refrigerator can cause condensation which is a big enemy for the electronic components and such inside the aircraft.

I know you did not ask me this, but I'm going to share it with you all the same. But I calibrate my compass prior to every flight. I have never had any weird or crazy stuff take place, like others do including yourself. I also have never inspected a flight log from a crash, or fly away that showed a compass calibration prior to flight. I'm a dji product tester and I go by what is dji's preferred method, except during certain points in a beta test. I then purposely do things out of the preferred method as some one else might to see if I can simulate a issue that may come about later on, and which would be using equipment near the same.

The text file is the right one. I made three flights that morning. The first one gave the problems. About changing flight mode, the youtube video. 3.55, you see I change flight mode. But I change back right away. Funny you think conspiracy, I am just a guy who bought a Phantom 3, 2 months ago. I can understand that if you are a tester and did not see this ever it is strange. I don't no why. But if you want you can check the other two flights also, so maybe you find something.

HealthyDrones.com - Innovative flight data analysis that matters (THE flight)

HealthyDrones.com - Innovative flight data analysis that matters (flight 2)

HealthyDrones.com - Innovative flight data analysis that matters (flight 3)

 

[Flying Dutchman]

i see, Well I used to recalibrate every flight with my P2 and never had a problem but what you said makes alot of sense. I guess I am just always paranoid about having a flyaway then coming back to the forums and have everyone say "you should have recalibrated" lol

good luck OP[/QUO

i see, Well I used to recalibrate every flight with my P2 and never had a problem but what you said makes alot of sense. I guess I am just always paranoid about having a flyaway then coming back to the forums and have everyone say "you should have recalibrated" lol

good luck OP

I used to calibrate also every time. But than sombody told me to read the manual about compass calibration. Thats why. I also never had a problem, and most of the time I fly over water so the thingie what happend to me was scary. And I wanted to now what was going on. Its a great piece of technic and a fun machine to fly with, but every time you start your Phantom and lift off you have a change thats the last time you ever saw it. Lets hope it will never happen.

 

[Flying Dutchman]

Update: If you check the healthydrones.com readouts which I posted before than all the readouts from the sensor erea are a mess. Looks like a rainbow, beautiful, but not really a necessity.

If I control the readings from my flights before they are all green. Exampels:

Flight 1
Flight 2
Flight 3
Flight 4

I just calibrated my Phantom and its night now, so tommorow ill try and do the same flight, straight up for 15 minutes. I hope its all green now. Thx all for reading and responding.