Hi folks,

We had a crash a few weeks ago and DJI are refusing to repair the damaged X5 gimbal because they say the crash was our fault.

Here is the crash log file: https://www.dropbox.com/s/1tv0ad ... 1-21-00%5D.txt?dl=1
Here is the HealthyDrones link: http://healthydrones.com/main?share=EPYnPW&page_id=GENERAL

Now, the drone took off with 33%, which is low but the flight was just a simple hover for 1 minute and so I would have imagined that 33% was sufficient.

The battery temperature was always above 75 f (24 c) during the flight.

DJI said "battery voltage is very low, when it is under 3.45 for 15s. it will desend." and refuse to provide any further explanation / help....

Here is another version of the log file, showing battery voltage: http://www.phantomhelp.com/LogViewer/KVTPGO1AF7NEFEM103AT/

You can see it dipped below 3.45 for 40 seconds in fact, before descending...

Any ideas on what happened and why this is pilot error? Do we void our warranty if we fly using a battery that is not at 100% level?

"Now, the drone took off with 33%, which is low but the flight "
That was your mistake, DJI strongly warns never to take off with less than a full battery.I return home at 35%. Why chance it?

Sorry mate, I would never take off in cold weather with a battery at 33%.  You need to remember in the best of conditions that aircraft want's to land with about 12% battery.  I think you took a risk and lost.  I am sorry that your crashed.

good luck - donnie

Donnie Posted at 2016-4-23 18:12
Sorry mate, I would never take off in cold weather with a battery at 33%.  You need to remember in t ...

Well, in an ideal world we would only fly on a fully charged battery. Unfortuately we do a lot of filming in the wild where we only have limited supply of batteries and charging equipment. The battery was sufficiently warm as you can tell in the log file. I just want to know why the battery failed and didn't start a return to home - it knew its exact voltage during the entire flight after all...

Healthy drones show the landing battery was at 3%. Doesn't that answer your question? Sorry I'm an old cranky guy but... I'm an old cranky guy.

flighttime1 Posted at 2016-4-23 19:24
Healthy drones show the landing battery was at 3%. Doesn't that answer your question? Sorry I'm an o ...

Yes, but after only 1 minute of flying! My question is why? And why the drone did not return-to-home as expected?

DroneAlps Posted at 2016-4-24 05:12
Yes, but after only 1 minute of flying! My question is why? And why the drone did not return-to-hom ...

I guess the answer to your question is that it ran out of fuel, sorry but your expectations of performance at 3% battery are a little high. Hope that extra shot was worth it, I always ensure landing around 30%, voltage drop after this is exponential.

Was that 33% from the same days flying or from a previous day ?  ..   

DroneAlps Posted at 2016-4-24 03:12
Yes, but after only 1 minute of flying! My question is why? And why the drone did not return-to-hom ...

The battery will close down if the voltage drops below 3.3 for a extended period .. you could have a weak cell in your battery... it is a pre flight must to bring up the battery display and check for any abnormality ..

Wolfman Posted at 2016-4-24 00:22
I guess the answer to your question is that it ran out of fuel, sorry but your expectations of per ...

Sorry but it had 33% at launch, and 3% a minute later after hovering - that is not exactly expected behavior!

skycam Posted at 2016-4-24 00:25
Was that 33% from the same days flying or from a previous day ?  ..

Same day flying

skycam Posted at 2016-4-24 00:36
The battery will close down if the voltage drops below 3.3 for a extended period .. you could have ...

No weak cells were detected before or during the flight

DroneAlps Posted at 2016-4-24 09:14
No weak cells were detected before or during the flight

It's strange going from 33% to 3% in 1 minute, I hope for you that DJI will have an explanation or do something about that repair. If you film different scenes behind each other it feels normal that you can go up with a 'safe to fly' for just a shot that takes no more than lets say a minute. Using the battery only with 100% is not like you expect from a pro or semi pro tool?

Siep

DroneAlps Posted at 2016-4-24 02:44
Well, in an ideal world we would only fly on a fully charged battery. Unfortuately we do a lot of f ...

The numbers returned to the craft and the app are suspect under some firmware versions. DJI say they fixed the issue with this current version, but then they said that same thing 3 or 4 times before ;-)

This led to the advice to fly with 100% charge, especially if the battery has gone into self-discharge mode.

Common practice has people not flying in the manner you describe - we understand that your circumstances led you to launch, but the system is just not consistent or reliable at those charge levels.
Sorry to hear about the issue. Hope you get back in the sky shortly.

Farnk666 Posted at 2016-4-24 09:54
The numbers returned to the craft and the app are suspect under some firmware versions. DJI say th ...

So this means..only 'take off' once? Not going to a next scene 3 miles further and start again with lets say 60-80% battery?
Is that the official advice?

Siep

siep Posted at 2016-4-24 01:51
So this means..only 'take off' once? Not going to a next scene 3 miles further and start again wit ...

I get about 12 or 13 minutes per battery. At 60% that means I have about 4 minutes till I hit the 30% warning. If I was taking stills, then yes, 60% is workable for takeoff. Anything else and I would grab another battery just so I didn't waste time if I ran out of battery and had to land and restart.

DroneAlps Posted at 2016-4-24 04:14
No weak cells were detected before or during the flight

The more important question: was the battery properly calibrated? How resent?  If not, that 33% does not mean much.  

DroneAlps Posted at 2016-4-24 15:14
No weak cells were detected before or during the flight

Flying always requires a balance of risk assessment versus reward. Trying to squeeze every last bit of fuel from our batteries requires conservative piloting strategies that take into account the possible reduced propulsion output and slower response at these low fuel levels. Situationally, unless pilots actually practice flying under these "low fuel" situations and understand the flight envelope, they should not continue flying below 30% other than returning home and landing.

For a short period of time, DJI aircraft were dropping from the sky due to inteligent battery cell shutdowns. DJI incorpoated improvements so that propulsion output is automatically adjusted based on battery temperature and cell voltage. Horizontal and vertical speed are also decreased to avoid triggering battery protection.

You originally posted the video of this flight which showed your aircraft climbing above a skier who tracked from the bottom of the frame to the top of the frame with the Inspire looking straight down. As the skier exited the frame top, it appears that the pilot began decending to pan up and behind the skier. It appeared that the Inspire settled into the snow behind the skier and then power was applied to climb out of the snow and return to the pilot's location for immediate landing.

Based on your RC logs, the pilot stick inputs correspond to the previously posted video recording:


The six cells of the Inspire battery appear to have operated as stated above. Realizing that between 3.4 and 3.3 volts, the battery is essentially on "low fuel light", the power reduction kept you flying without battery shutdown:

Battery Cells

If you review the healthy drome summaries, it appears you were operating in an area that may have contained quartz deposits or some other interference was reducing your RC-> aircraft communication and reducing your GPS signal:



From your RC log, it appears that the decent/throttle stick input was excessively long, perhaps due to poor control communications, and the aircraft settled into the snow. The batteries operated surprisingly well to keep the aircraft powered even after settling into the snow.

This conjecture can ony be confirmed by reviewing the aircraft Flight Data Recording logs and comparing the information the aircraft was actually receiving.

If you want to share the aircraft log for this flight, I'd be glad to take a look and compare the flight from the Inspire' viewpoint.

It really sucks that the camera was damaged, sorry!

Actually, I don't think that DJI Inspire1 is not perfect drone. Maybe It still has many kinds of bugs and problems. Only DJI knows them or DJI don't know them and we don't know them.
Because almost customers don't have enough time and enough many samples to test.
And normally, almost DJI Inspires are falling down by user's mistake before enough testing and finding bugs or problems. It's not easy to proof bugs or problems by customers. Maybe normally DJI will refuse.

We are care of our drone more carefully than DJI warn.
Never trust almost Drone. If you trust drones like DJI, you will be foolish.

Only I hope that DJI don't sell Products with bugs or problems.  

terrylewis Posted at 2016-4-24 23:40
Flying always requires a balance of risk assessment versus reward. Trying to squeeze every last bit ...

Thanks for looking into this Terry. I agree, we aim to start landing at 30% - and so for a simple shot like this I felt the risk/reward was totally acceptable. I took the video down because it didn't really add anything and is property of the client we were filming for. You're correct that throttle down was applied to get that shot. On the flight log in the App it shows that there was a big delay between throttle down and then the aircraft basically drops out of the sky. Maybe the "reduced power state" caused this delay and subsequent fast descent? Unfortunately the aircraft log file was overwritten because internal memory was full.

terrylewis Posted at 2016-4-25 07:40
Flying always requires a balance of risk assessment versus reward. Trying to squeeze every last bit ...

Hi Terry,

what a remarkable review of the events.

I don't want to bomb Dronealps thread but it would be appreciated if you can outline what you used to collate the information as you have presented it here, it is fantastic! Maybe I have missed something as  I have not seen anything like this before.

Regards Wolf

siep Posted at 2016-4-24 16:51
So this means..only 'take off' once? Not going to a next scene 3 miles further and start again wit ...

Hi Siep,

So my procedure goes roughly as follows.
I store batteries between days flying at approx 50%.

When I am due to fly within 24 hours, I charge everything up to 100%.
I fly each battery down to 30% on the day, either in one flights or in a series of flights.
I land with 30% minimum on each battery.

Every 10 or so times I go out flying, I setup in the backyard and take each battery down to 0% in a close hover to the ground. The Inspire will try to auto land during this, but I manually hold it up off the deck just a few cm as I watch the battery level fall. 0% and 3.3v are the numbers I am after.

When the batteries are cooled, I charge them back to 100% and fly them down to 50% for storage.

I started the procedure after a number of instances where I saw immediate drop of percentage from flying at mid 20s down to 4-6%. My experimentation with my Inspire has me doubting the accuracy of the capacity tracking functions of the TB batteries once they fall below 30%.

YMMV

Keep safe

F   

Wolfman Posted at 2016-4-25 23:07
Hi Terry,

what a remarkable review of the events.

I too would be very interested in the tools used to generate these graphs!

DroneAlps Posted at 2016-4-24 03:14
Sorry but it had 33% at launch, and 3% a minute later after hovering - that is not exactly expecte ...

That's expected behavior when it's cold out.

R&L Aerial  Posted at 2016-4-26 20:09
That's expected behavior when it's cold out.

Care to explain?

DroneAlps Posted at 2016-4-23 11:12
Yes, but after only 1 minute of flying! My question is why? And why the drone did not return-to-home as expected?

The best thing to do in future is never to trust the machine readings. Try to be more patient. Try to fly with 50 power left and land it at 30. if the machine by mistake telling you it has 20 power to go. you still are safe, if it only has 10 to go.. hope you get what i am trying to explain. sorry for the lost. I hope you did managed to get the best shot to recover the losses. sometimes during our bad time. this machine can give us fault reading only to cost losses to us. so, never again trust machines. Its like a accident that we never thought about it. Mistakes always happend when we are toooooo excited or toooo haste or toooo over confident. sorry.

DroneAlps Posted at 2016-4-23 23:14
Sorry but it had 33% at launch, and 3% a minute later after hovering - that is not exactly expected behavior!

by right.. at 30 percent the drone will give you warning.. only at 10 percent it will land itself.  As i said before.. during our bad time...the machine will give wrong readings. DJI is huge company.. its not to say they dont know its something went wrong with firmware or other things.. but they are big business team. if they help you..there will be many more.. hope you get me clear. sorry for the lost.

siep Posted at 2016-4-23 23:40
It's strange going from 33% to 3% in 1 minute, I hope for you that DJI will have an explanation or do something about that repair. If you film different scenes behind each other it feels normal that you can go up with a 'safe to fly' for just a shot that takes no more than lets say a minute. Using the battery only with 100% is not like you expect from a pro or semi pro tool?

Siep

In extreme cold weather. it can.. it will...

flighttime1 Posted at 2016-4-24 01:04
I get about 12 or 13 minutes per battery. At 60% that means I have about 4 minutes till I hit the 30% warning. If I was taking stills, then yes, 60% is workable for takeoff. Anything else and I would grab another battery just so I didn't waste time if I ran out of battery and had to land and restart.

Anyone can take off as much times they like..e
if the battery are above 30. even below you still can do that...but normally at below 30... if the weather is extreme cold..the machine may give wrong readings.. so, the best is always 50. try to bring it down at 30. play safe play longer.

DJI is acutely aware that battery voltage indications are unreliable in the lower spectrums, much like automobile gas gauges. Thats why they overcompensate by requiring; "never takeoff less than 100% battery". This requirement resets the gauge error and covers them judicially as you took off outside of warranty conditions. It sux.

The DJI logs, no matter the circumstance, are used more against you... than for you. Thats why the feature can't be turned off, easily.