Hi. I am really disturbed about the new P4P I got 3 days ago (a replacement of a P3P). I spent 2  days doing all the initial preparation: IMU calibrations, firmware  updates, all initial DJI Go 4 settings, Props balancing and so on.  Yesterday after the usual compass calibration I took it off and raise it  at 2 mts height; this is what I always do as a regular pre-flight 15 sec hover to see stability  before flying away. Well, I could not do that, after 6 sec hovering it  just turned off and felt on the asphalt pavement. I was really  disappointed because my previous flight, with a Phantom 3 Pro (18  months flying ok), the drone just got crazy and felt down from 120 mts  and was totally damaged (I will post a new thread about this issue).  Well after initial checking I saw no huge damages except a bent leg and  some scratches in that leg and the camera, the drone was off and I tried  to turn it on, but no way, the P4P does not turn on again, changed for  another fully charged battery and the same. Conclusion the brand new P4P  is broken after the initial flight of 6 sec. I already initiated the  process to send it back to DJI but I live in Caracas, Venezuela and that  is not a simple thing due to the political and social issues we are  struggling with nowadays. I wonder what kind of testing DJI does before  sending this kind of very sensitive products to clients. This crash  could have been avoided if they would have it tested in advance. When I  bought the P4P, with redundancy in IMU and compass I thought it was  going to be a very trustful and secure flying camera to be used for  serious work. I am a pro photographer and it is very disappointing how  much I have spent in DJI drones and both has crashed by technical  malfunctioning. I was hired to to a job of some aerial photos tomorrow and I will tell my client I will not be able to do it due to a technical issue.  

I'm not picking on you but  I never understood the idea behind having only one camera or drone for a paid job.  When I was shooting weddings I always had a back up camera in case something went wrong and if your bread and butter is taking aerial photography or video then you should always have a backup drone.  If you're a professional you owe it to your client to complete the job rather than giving the having technical problems excuse.

The Phantoms are consumer items, not professional aircraft, you can't expect them to be given 24 hours and 1000Km of flight tests at the factory given the prices we pay for them.  There are a few that have problems when the consumers flight test them.

Good luck with getting a speedy replacement.

I don't think it is necessary to do an IMU calibration on a brand new P4P. And also a compass calibration is not something that is "usually done"...  both only when the system prompts you to do it. What was the reason for all this?

Antonio, is right about calibration, if you want some answers maybe try putting up your flight log although I know it going to be short.

hallmark007 Posted at 2017-4-29 12:10
Antonio, is right about calibration, if you want some answers maybe try putting up your flight log although I know it going to be short.

Thanks for your suggestion. Can you let me know how to put the flight log here.  It is already under my account.

ibuyufo Posted at 2017-4-29 11:01
I'm not picking on you but  I never understood the idea behind having only one camera or drone for a paid job.  When I was shooting weddings I always had a back up camera in case something went wrong and if your bread and butter is taking aerial photography or video then you should always have a backup drone.  If you're a professional you owe it to your client to complete the job rather than giving the having technical problems excuse.

I have spare cameras for my regular photography work. I did a lot of pro work using a Phantom 3 Pro without any glitch in many different environments for more that a year. I expected a much more advanced aircraft with redundancy systems. Yes, I have postponed the work until I get a borrowed drone.  

Antonio76 Posted at 2017-4-29 11:32
I don't think it is necessary to do an IMU calibration on a brand new P4P. And also a compass calibration is not something that is "usually done"...  both only when the system prompts you to do it. What was the reason for all this?

Compass calibration before a flight in a new place is recommended by DJI. Also it is recommended is case a drop on the drone. After traveling to my country by all kind of courier services you never know how rude that box was treated. It is a very sensitive equipment and it is also recommended in those situations by DJI.

apomares-Vzla Posted at 2017-4-29 16:32
Thanks for your suggestion. Can you let me know how to put the flight log here.  It is already under my account.

http://www.phantomhelp.com/Phantom-4/#DataLog just go to this link and follow instructions then come back here and post the link.

Nigel_ Posted at 2017-4-29 11:23
The Phantoms are consumer items, not professional aircraft, you can't expect them to be given 24 hours and 1000Km of flight tests at the factory given the prices we pay for them.  There are a few that have problems when the consumers flight test them.

Good luck with getting a speedy replacement.

Up to a point I agree with Nigel. In this case only 10 seconds would have been enough. The P4P felt down from 2 meters after 6 seconds in hover on the very first flight. A basic factory test would have saved me to pay from-to my country a couple of hundred bucks in courier and customs taxes. Something else, I always mention to my students in the university that there are no professional cameras there are professional photographers using cameras; also, I wonder who is going to risk a 30.000% Hasselblad camera on board a M600 PRO drone.

hallmark007 Posted at 2017-4-29 16:45
http://www.phantomhelp.com/Phantom-4/#DataLog just go to this link and follow instructions then come back here and post the link.

ok, thanks for the link Hallmark007. I managed to upload the flight log here:

https://healthydrones.com/main?f ... 23782e880c8593b4817

Let me know what you think.

Antonio76 Posted at 2017-4-29 11:32
I don't think it is necessary to do an IMU calibration on a brand new P4P. And also a compass calibration is not something that is "usually done"...  both only when the system prompts you to do it. What was the reason for all this?

I wrote it down Antonio but here it is in response to your post: "Compass calibration before a flight in a new place is recommended by DJI. Also IMU calibration is recommended is case a drop on the drone. After traveling to my country by all kind of courier services you never know how rude that box was treated. It is a very sensitive equipment and it is also recommended in those situations by DJI."

apomares-Vzla Posted at 2017-4-29 16:50
Up to a point I agree with Nigel. In this case only 10 seconds would have been enough. The P4P felt down from 2 meters after 6 seconds in hover on the very first flight. A basic factory test would have saved me to pay from-to my country a couple of hundred bucks in courier and customs taxes. Something else, I always mention to my students in the university that there are no professional cameras there are professional photographers using cameras; also, I wonder who is going to risk a 30.000% Hasselblad camera on board a M600 PRO drone.

I have to correct something in my post, the flight only lasted 4 seconds. I saw it in the healthydrones data analysis included in the link.

apomares-Vzla Posted at 2017-4-29 16:40
Compass calibration before a flight in a new place is recommended by DJI. Also it is recommended is case a drop on the drone. After traveling to my country by all kind of courier services you never know how rude that box was treated. It is a very sensitive equipment and it is also recommended in those situations by DJI.

It should not have made your Phantom fall but .....
Compass calibration before each flight was never necessary but poor wording in the old manuals made some people believe it was.
Check the manuals for the newer Phantoms and DJI have changed the wording completely.
You don't need to recalibrate the compass at all - even new out of the box.

I never do a compass calibration or IMU calibration every time I fly.  I never had a problem.

apomares-Vzla Posted at 2017-4-29 16:50
Up to a point I agree with Nigel. In this case only 10 seconds would have been enough. The P4P felt down from 2 meters after 6 seconds in hover on the very first flight. A basic factory test would have saved me to pay from-to my country a couple of hundred bucks in courier and customs taxes. Something else, I always mention to my students in the university that there are no professional cameras there are professional photographers using cameras; also, I wonder who is going to risk a 30.000% Hasselblad camera on board a M600 PRO drone.

I would expect an M600 to be significantly better tested since it costs a lot more and is probably considered a professional aircraft, but even so, on receiving it I would want to do some test flights without the Hasselblad on board!

The fault you had may have required two flights to show up, not just the basic factory test flight, they always seem to arrive in working condition so presumably some testing is done.  I think yours is the shortest lived one I've see, you have been unlucky, but maybe best to find the problems on the test flight, not on the first work flight and crash land on top of the customer!

You didn't follow hallmark007's link, his link is a much more useful site for diagnostics, although I don't expect we will find much anyway.   Why was the battery so hot?  Was it just the climate...

(PS: Corrected link- http://www.phantomhelp.com/LogViewer/upload/ )

As for the compass, on a P4 it is generally considered that there is more risk of a crash from doing a bad calibration than there is of a crash due to anything that happened in transport.  Best to only calibrate if you experience poor flight control, and then only do it well away from the place you experienced the poor flight control so that you don't just correct for the environment and then cause a crash when you fly in the next place.   I haven't done a compass calibration for about 8 months and have flown in more than one country.

Sorry to hear about your unpleasant experience.
Since you've created the case number and the drone is on the way to DJI service center, please do not worry about it, our engineers will analyze the data for you to figure out the reason for the crash.
For any further questions, please feel free to contact us.

apomares-Vzla Posted at 2017-4-29 17:27
ok, thanks for the link Hallmark007. I managed to upload the flight log here:

https://healthydrones.com/main?flight=6dfbf4232684d23782e880c8593b4817

4 seconds, best leave that up to the experts, those logs from airdata show very limited data, so it's certainly nothing I can help with, but from what you say I would say you have a good case for warranty replacement or repair, hope you get it back soon .
Good luck...

There is nothing in the flight log out of place.  This will need to be examined by DJI.  Normally you would sync your flight log with them and you would request they examine it for you.  That uploads all the flight telemetry, including parameters we can't see.  DJI can access all of it including motor speed/load/performance. If a prop flies off or is broken in flight they can determine it from this information.  I suspect in this case it will be related to the motors, it is possible one seized or stalled, perhaps some debris has fallen inside.  I assume the props were attached properly and sufficiently tight.  There seems to be some rotational twitching  of up to 10° as it ascended, which shouldn't happen if the opposing motors and props are balanced and operating correctly.  They should be in a brand new aircraft!  There was no lateral or rotational input from the sticks to account for this.  Make sure you request that DJI examine the motor parameters for you.  They will extract the flight log from the internal recorder so there is no need to upload anything in this case, as it has also been sent in for repair.

DJI Natalia Posted at 2017-4-29 23:55
Sorry to hear about your unpleasant experience.
Since you've created the case number and the drone is on the way to DJI service center, please do not worry about it, our engineers will analyze the data for you to figure out the reason for the crash.
For any further questions, please feel free to contact us.

Thanks for your response Natalia, still not under way. I first need the shipping label and it will take some while I do all the arrangements to send it to the USA from Venezuela.

hallmark007 Posted at 2017-4-30 01:47
4 seconds, best leave that up to the experts, those logs from airdata show very limited data, so it's certainly nothing I can help with, but from what you say I would say you have a good case for warranty replacement or repair, hope you get it back soon .
Good luck...

Yes, I did not know it was too short until I saw the data in the healthydrones site. Maybe I should  get a "Guinness Record" for the shortest flight ever on a P4P. By the way, I got confused and ended in healthydrones . I will follow again the instructions to get to the right place to upload flight log data. Thanks a lot for your help.

hallmark007 Posted at 2017-4-30 01:47
4 seconds, best leave that up to the experts, those logs from airdata show very limited data, so it's certainly nothing I can help with, but from what you say I would say you have a good case for warranty replacement or repair, hope you get it back soon .
Good luck...

Thanks for your help anyway hallmark007!!!

Nigel_ Posted at 2017-4-29 23:34
I would expect an M600 to be significantly better tested since it costs a lot more and is probably considered a professional aircraft, but even so, on receiving it I would want to do some test flights without the Hasselblad on board!

The fault you had may have required two flights to show up, not just the basic factory test flight, they always seem to arrive in working condition so presumably some testing is done.  I think yours is the shortest lived one I've see, you have been unlucky, but maybe best to find the problems on the test flight, not on the first work flight and crash land on top of the customer!

Agree Nigel. Who knows when this drones are going to fail anyway. This is not the only case I have. My previous drone a P3P, used perfectly for more than a year, was working perfectly until crashed after flying away last month. I guess this is also another record I have: 2months, 2 drones, 2 last flights, 2 crashes. The P4P was the replacement for the P3P. I was very confident about this technology but now I don't know what to think about it.
When I get the next aircraft I will make it hover at 1 mt over grass, several times, several batteries, before thinking to get it any higher (not kidding)
About the compass and IMU calibrations it is good to know is not needed. I was doing and following what I read and saw in DJI manuals and videos from more that 1 year ago. As I see things has changed.
I will follow the link you and hallmark007 provide and hope to get to the right place this time .
Thanks for your help and comments!!

Labroides Posted at 2017-4-29 20:51
It should not have made your Phantom fall but .....
Compass calibration before each flight was never necessary but poor wording in the old manuals made some people believe it was.
Check the manuals for the newer Phantoms and DJI have changed the wording completely.

Not only in the old manuals but in DJI videos as well. I have been flying for more than 1 year with my late P3P and that was the information I had. Good to know that is not need it anymore.

Nigel_ Posted at 2017-4-29 23:34
I would expect an M600 to be significantly better tested since it costs a lot more and is probably considered a professional aircraft, but even so, on receiving it I would want to do some test flights without the Hasselblad on board!

The fault you had may have required two flights to show up, not just the basic factory test flight, they always seem to arrive in working condition so presumably some testing is done.  I think yours is the shortest lived one I've see, you have been unlucky, but maybe best to find the problems on the test flight, not on the first work flight and crash land on top of the customer!

I just managed to upload the flight log to phantomhelp.com the last link took me directly to it, I include it here:
[url]http://www.phantomhelp.com/LogViewer/18AMY1I7WK4CWR2YIHRQ/[/url]

Question: is this info much better than healthydrones?. I think both are complements, correct?

apomares-Vzla Posted at 2017-4-30 12:36
I just managed to upload the flight log to phantomhelp.com the last link took me directly to it, I include it here:
http://www.phantomhelp.com/LogViewer/18AMY1I7WK4CWR2YIHRQ/[/url]

Well for starters they have recorded 11.3 seconds of flight time, although it doesn't tell us much, those logs give good battery info you can see there was deviation in battery cells but not enough to be concerned, they also give very accurate stick movements.

I wish you luck with dji I can see from your log or what you say that this should be a warranty repair.

apomares-Vzla Posted at 2017-4-30 12:36
I just managed to upload the flight log to phantomhelp.com the last link took me directly to it, I include it here:
http://www.phantomhelp.com/LogViewer/18AMY1I7WK4CWR2YIHRQ/[/url]

The main advantage of phantomhelp is that it gives us more information including compass and gyro data, plus what the pilot was telling it to do, but also allows us to access the original log file to get all the information.

I can see from the log that it dropped, nearly fell over and went into a spin without any input from you, indicating a power failure to the motors, but not a battery failure since it was able to transmit the information to the controller to be logged.   DJI will see the same and more so that makes a warrantee replacement fairly certain.

Something I've not noticed before is a voltage in the 5th battery cell at the end of the log as it is falling - there are only 4 cells!

endotherm Posted at 2017-4-30 11:38
There is nothing in the flight log out of place.  This will need to be examined by DJI.  Normally you would sync your flight log with them and you would request they examine it for you.  That uploads all the flight telemetry, including parameters we can't see.  DJI can access all of it including motor speed/load/performance. If a prop flies off or is broken in flight they can determine it from this information.  I suspect in this case it will be related to the motors, it is possible one seized or stalled, perhaps some debris has fallen inside.  I assume the props were attached properly and sufficiently tight.  There seems to be some rotational twitching  of up to 10° as it ascended, which shouldn't happen if the opposing motors and props are balanced and operating correctly.  They should be in a brand new aircraft!  There was no lateral or rotational input from the sticks to account for this.  Make sure you request that DJI examine the motor parameters for you.  They will extract the flight log from the internal recorder so there is no need to upload anything in this case, as it has also been sent in for repair.

Thanks for your input. For what I saw the P4P turned off while in hover, that is why it felt down. There was no way to turned on anymore.

Nigel_ Posted at 2017-4-30 14:01
The main advantage of phantomhelp is that it gives us more information including compass and gyro data, plus what the pilot was telling it to do, but also allows us to access the original log file to get all the information.

I can see from the log that it dropped, nearly fell over and went into a spin without any input from you, indicating a power failure to the motors, but not a battery failure since it was able to transmit the information to the controller to be logged.   DJI will see the same and more so that makes a warrantee replacement fairly certain.

There are a lot of new things about all this Nigel, I never need to get into it until now. Thanks!!

hallmark007 Posted at 2017-4-30 13:13
Well for starters they have recorded 11.3 seconds of flight time, although it doesn't tell us much, those logs give good battery info you can see there was deviation in battery cells but not enough to be concerned, they also give very accurate stick movements.

I wish you luck with dji I can see from your log or what you say that this should be a warranty repair.

Thanks hallmark007. That is a brand new battery with just two recharges. Well, it is under warranty (i got it 2 days before the crash)  plus DJI Care refresh and I expect DJI will not charge this issue under the Care Refresh.

apomares-Vzla Posted at 2017-4-30 17:32
Thanks hallmark007. That is a brand new battery with just two recharges. Well, it is under warranty (i got it 2 days before the crash)  plus DJI Care refresh and I expect DJI will not charge this issue under the Care Refresh.

"I expect DJI will not charge this issue under the Care Refresh." - I would certainly hope not ... but you are sending it to the USA so they might try to!  

apomares-Vzla Posted at 2017-4-30 17:26
Thanks for your input. For what I saw the P4P turned off while in hover, that is why it felt down. There was no way to turned on anymore.

It didn't turn off (disconnect from power) while hovering.  It simply failed to stay in the air.  Instead it recorded everything as it fell to the ground, and upon impact it stopped recording.  It fell faster than 1.3ft in 100milliseconds (which was the last line recorded) when it hit the ground and stopped before the next line of data could be recorded in the log.   It likely dislodged the battery upon impact, which is pretty typical but somewhat surprising at such a low altitude.  Perhaps it broke off a battery terminal inside at the same time, which is why it won't start up again.

Regarding the difference between AirData and Phantom Help, they are very similar and nearly as useful as each other for determining the majority of incidents, providing you analyse the full csv data each supplies.  Otherwise it is just a preference thing.  Both sites start with the same raw data uploaded, but it is up to each which data is translated and returned to the user.  PH returns slightly more data relating to pitch/yaw/roll and in the P4, VPS height.  On the other hand AD only supplies compass heading, but returns information in relation to gimbal movements, camera tilt and when/where images were captured (much more useful if you are fixing camera issues).  The rest of the information is very similar and generally sufficient to conduct a crash analysis.  Both sites choose not to return other information which is collected, e.g. temperatures, motor speeds and loads... which would be useful at times (AD does return some of this as graphics on their various sub-pages, but does not supply the figures in the csv data).   DJI has access to all of this info and they can make better informed conclusions as a result, if they can be bothered to study it correctly.

Nigel_ Posted at 2017-4-30 14:01
The main advantage of phantomhelp is that it gives us more information including compass and gyro data, plus what the pilot was telling it to do, but also allows us to access the original log file to get all the information.

I can see from the log that it dropped, nearly fell over and went into a spin without any input from you, indicating a power failure to the motors, but not a battery failure since it was able to transmit the information to the controller to be logged.   DJI will see the same and more so that makes a warrantee replacement fairly certain.

Where did you see the data for a 5th battery cell?  Some of these sites do return 5th and 6th cell data in the csv (for Phantom 3 model flight logs at least) which invariably contain "0".  I believe it is a parameter that is collected and returned in the larger aircraft like the Inspire which have more cells, but it is an oversight presenting this absent information for lower spec aircraft.

endotherm Posted at 2017-5-1 02:45
It didn't turn off (disconnect from power) while hovering.  It simply failed to stay in the air.  Instead it recorded everything as it fell to the ground, and upon impact it stopped recording.  It fell faster than 1.3ft in 100milliseconds (which was the last line recorded) when it hit the ground and stopped before the next line of data could be recorded in the log.   It likely dislodged the battery upon impact, which is pretty typical but somewhat surprising at such a low altitude.  Perhaps it broke off a battery terminal inside at the same time, which is why it won't start up again.

Regarding the difference between AirData and Phantom Help, they are very similar and nearly as useful as each other for determining the majority of incidents, providing you analyse the full csv data each supplies.  Otherwise it is just a preference thing.  Both sites start with the same raw data uploaded, but it is up to each which data is translated and returned to the user.  PH returns slightly more data relating to pitch/yaw/roll and in the P4, VPS height.  On the other hand AD only supplies compass heading, but returns information in relation to gimbal movements, camera tilt and when/where images were captured (much more useful if you are fixing camera issues).  The rest of the information is very similar and generally sufficient to conduct a crash analysis.  Both sites choose not to return other information which is collected, e.g. temperatures, motor speeds and loads... which would be useful at times (AD does return some of this as graphics on their various sub-pages, but does not supply the figures in the csv data).   DJI has access to all of this info and they can make better informed conclusions as a result, if they can be bothered to study it correctly.

I got your point. It was all very fast and I thought it had turned off since none of the propellers were damaged, not even a scratch, and I thought were not turning at the impact.
I looked into the csv data and the P4P recorded the fall as you mentioned. I looked at the battery connectors inside the P4P and there are no signs of damage it must be inside or something else.
Thanks again!!

endotherm Posted at 2017-5-1 03:04
Where did you see the data for a 5th battery cell?  Some of these sites do return 5th and 6th cell data in the csv (for Phantom 3 model flight logs at least) which invariably contain "0".  I believe it is a parameter that is collected and returned in the larger aircraft like the Inspire which have more cells, but it is an oversight presenting this absent information for lower spec aircraft.

I loaded the txt file into CsvView https://datfile.net/CsvView/intro.html and exported a csv file.

As it falls, Cell5 changes from blank to "0.032" - could be a power diagnostic with the column incorrectly labelled rather than an incorrect value?

Nigel_ Posted at 2017-5-1 03:53
I loaded the txt file into CsvView https://datfile.net/CsvView/intro.html and exported a csv file.

As it falls, Cell5 changes from blank to "0.032" - could be a power diagnostic with the column incorrectly labelled rather than an incorrect value?

Hi Nigel, could this be a loose battery or maybe just a bad connection?

hallmark007 Posted at 2017-5-1 04:18
Hi Nigel, could this be a loose battery or maybe just a bad connection?

Since it sent the information back to the controller to be logged, it can not have been a complete battery failure.  A battery failure of any sort seems unlikely, and IIRC a second battery was tried after the crash and it wouldn't start up which almost confirms an aircraft fault.

Probably a loose connection inside the aircraft, endotherm pointed out that the takeoff wasn't completely smooth which would tie up with a loose connection, maybe to an ESC since it appeared to fall first rather than spin first which would better indicate 1 motor failing rather than 2.  I'm not sure on this though since on most power failures we see, the log just stops and we don't see what happens.

After the crash, did you try to start it up with a different battery?

IflewbyU2 Posted at 2017-5-1 06:35
After the crash, did you try to start it up with a different battery?

Yes, I inserted a full new one and same result, the battery shut off every time; the P4P is broken.

Nigel_ Posted at 2017-5-1 03:53
I loaded the txt file into CsvView https://datfile.net/CsvView/intro.html and exported a csv file.

As it falls, Cell5 changes from blank to "0.032" - could be a power diagnostic with the column incorrectly labelled rather than an incorrect value?

Nice!  They've improved csvview a lot since last time I saw it.  Over 150 parameters recorded, but it still leaves out a few important ones like individual motor speed/load/performance data.  Some of these fields will be very useful for me in future crash analyses.  Thanks.

Anyway, my result was different from yours.  It had no data at all for the cell 5 column, and nothing but "0"s for cell 6.


Could be a glitch double-handling the data in and out of the Phantom Help engine.  I've seen a few glitches appear in their csv data from time to time that does not show up on Healthy Drones/AirData.