Phantom 3 Professional sudden flip and crash (FlightRecord vid p.2, LogData .csv p.3)

[zincsoldier]

I doubt it is a battery. It is the same Lithium Polymer as you have in your cell phone. Nothing critical will happen after 200 cycles may be you will fly couple min less. It looks like a power cut. You probably did not insert battery till the second click.

I was familiar with Phantom 2 and I honestly know how to put the battery all the way securely in. I also know there is a second click with the later battery batch so I am very sure I put the battery in correctly. Did you know for a fact that no one has ever reached that magic 200 cycle mark for DJI Intelligent Battery, I am just saying Maybe they improve the battery for this new P3, we still don't know yet.

 

[JKDSensei]

Im sorry I was at work and did not really go into details but to went through what I thought to be critical for the battery. In this case, I think the battery is still slowly failing after a number of cycles which resulted in my crash. I also have been getting the intermittent red-colour voltage then it would go back to green or yellow depends on battery percentage, usually at 50%-70% battery in the DJI Pilot app. Overall, I am pretty sure it is the unstable battery at 30-cycle is causing this problem.

Zinc,

After watching the video a number of times, viewing your battery log and hearing your account, I'm pretty sure I see what happened now, but I need to know where you were standing and the orientation of the Phantom.

In the video on YouTube, relative to the white line which shows the course plot of the Phantom, where were you standing?

The orientation of the Phantom never changes and the two red stripes appear to be at about 350 degrees relative to the video screen. Is that correct?

 

[zincsoldier]

Zinc,

After watching the video a number of times, viewing your battery log and hearing your account, I'm pretty sure I see what happened now, but I need to know where you were standing and the orientation of the Phantom.

In the video on YouTube, relative to the white line which shows the course plot of the Phantom, where were you standing?

The orientation of the Phantom never changes and the two red stripes appear to be at about 350 degrees relative to the video screen. Is that correct?

I am not sure if I get what you meant, English is not my first language. As far as I could remember, I think I was standing behind and a bit to the right of the P3 when I first took off. Please point out if you need more information.

 

[PaulMuns]

Something else to consider. This comes from a repair site (usdronerepair.com) that someone recommended in a previous post. On their site (they are ex-military drone people) they list possible causes of DJI fly-aways. From their site...

"3. Oscillations and vibrations. Motor vibrations can induce emf signals that the esc may interpret as brushless motor EMF signals and stall your motor and even make it run backwards or overheat. These vibrations traveling through the body make their way to the flight controller and just imagine what kind of influence that has on a gyro stabilizer built into the flight controller. The dreaded flip of death that has claimed so many aircraft. Balance those props!!!!"

Makes me want to buy a prop balancer!

 

[JKDSensei]

Interesting. Gotta read up more on this.
After reading their page, it certainly sounds as though these folks are very advanced even beyond DJI.

Having read this, I rest my case....for now

 

[ScottyT]

I have contacted 24 hr support on DJI website. They seem to insist it was entire my fault according to the flight log as I was pullong down the stick so hard that it flipped during landing LOL. How was I supposed to land it right after starting the motors and went straight out, plus I had 100% battery, it is common sense I was ready to fly it, not to land it in anyway. Then, I implied on not doing business with DJI in the future, I don't know if they give a **** but they told me they would take this matter to technical support and will send me an email for further information. At the end of the day, I am not sure I am wiser on my problem but this definitely is not my fault.

Interesting response. I often descend at full speed, as would many I expect, and would hope there is not an issue in doing so. Sounds like a bit of a cop out on their part, although they're just 1st level support so they would likely be spinning a yarn.

 

[PaulMuns]

Interesting. Gotta read up more on this.
After reading their page, it certainly sounds as though these folks are very advanced even beyond DJI.

Having read this, I rest my case....for now

Yeah, they seem to know their stuff. When ever I need a repair, I think that I will ship it off to them. Did you check out the the DJI repairs and upgrades section?

 

[zincsoldier]

Something else to consider. This comes from a repair site (usdronerepair.com) that someone recommended in a previous post. On their site (they are ex-military drone people) they list possible causes of DJI fly-aways. From their site...

"3. Oscillations and vibrations. Motor vibrations can induce emf signals that the esc may interpret as brushless motor EMF signals and stall your motor and even make it run backwards or overheat. These vibrations traveling through the body make their way to the flight controller and just imagine what kind of influence that has on a gyro stabilizer built into the flight controller. The dreaded flip of death that has claimed so many aircraft. Balance those props!!!!"

Makes me want to buy a prop balancer!

I can claim that these props come with P3 are pretty much okay to me. I had no jello effect or problem with the camera because of the props from the beginning. One more thing, the previous firmware lets it hover like a rock so I think props are less of a problem.

 

[zincsoldier]

Interesting response. I often descend at full speed, as would many I expect, and would hope there is not an issue in doing so. Sounds like a bit of a cop out on their part, although they're just 1st level support so they would likely be spinning a yarn.

The flight controller with P3 now only allows you to do a maximum of 3m/s descend speed so that would save you from any VRS. Go full speed if you want, you would not get in the VRS, that is not the problem. In the past, P2 somehow there were a lot of VRS because Naza flight controller hasn't been adjusted to an allowed maximum descend speed And also don't worry about holding your throttle down for too long. It would know when it is on the ground or been hand caught after it stops descending for more than 3 seconds. Hope this helps!

 

[Luap]

Looking at the csv file I noticed I mapped the roll values to the pitch.
When your Phantom flipped did it flip forwards or sideways as per below video? From the csv its sideways =roll.

Below a screen capture of the last two seconds from the csv file. Total recorded flight time is 6.6 seconds.
The first seconds clearly show a CSC (not shown).
Between 5.4 and 5.5 seconds the OSD.roll goes from 2.4 to -172.3 (the sideways flip).
The Rc throttle is lowered to its max between 5.3 and 5.4 secs in 1/10 second.
You further say you hovered for a minute prior the crash but the logger logged only 6.9 seconds with a proper CSC at the start.
So the first question is why did the Phantom "reset" itself 6 seconds prior the crash? Do you have any video from the camera?

 

[cdronefly]

Someone correct me if wrong, but you can't manually flip a P3 without really really trying, right?

I'm going to use the Occams Razor method on this one and guess that there were IMU problems. My reasoning:
1. The Phantom rarely flips - and is probably programmed not to.
2. The IMU is the main device which keeps the Phantom level
3. Compass and GPS problems result in drifting or other flight problems, not in flips.
4. Battery problems usually lead to app warning and/or the machine powering off and landing
5. IMU's are known to need resetting and calibration regularly. Most P3's need a couple just to offset the bumps from shipping.

If that were my quad I'd clean it out and reset the IMU two or three times on a perfectly level surface - then do a compass calibration and fly again. If it does anything like this again - then the IMU may have internal or other problems.

The Dronerepair dude mentioned that some older P1's and P2's were "fixed" by banging a certain board against a table - I think he was referring to 100g of force being applied to unstick the inside of an IMU.

Remember, an IMU is a mechanical device - stuff inside can shift. Ideally they can be reset properly, but like anything else there could be a rate of failure.

 

[craig4855]

Yesterday I was hanging out at the beach and planned to do some footage. The location was around 100 km from my place. So, got down to the beach, turned on Tx, then the bird, then calibrated successfully, started the motors manually and took off. I let it hover for around a minute or so, everything was fine, light winds (5-6km/h), 19 satellites at the present. I hit forward like I normally would to fly it straight forward altogether with ascending as well then suddenly it flipped and crashed upside down to the sand The reason of the crash is still unknown, I have seen nothing wrong with the bird or Tx then it suddenly happened. Unfortunately, everything is still in good working order except for four motors were covered in a bit of sand. I went home and tested it again after spending 3 hours cleaning all the sand in all 4 motors, it worked and I still wonder if I need to take this matter to DJI warranty. I am definitely not confident enough to fly my P3P now. I am now on my 47 flight and took off at 100% battery (30-cycle) when this happened. I am still lucky enough to keep it in one piece.
Picture of the flight record for now, that was me trying to execute CSC when it flipped. I would try to pull the flight log later.

Some beaches have what is called black sand (looks almost black). This sand is very high in mineral content (plays havoc with metal detectors). I would calibrate it away from the beach. Doesn't matter if you are a mile ayay, magnet lines don't change that much.

 

[snowghost]

Never calibrated my IMU, out of the box nor with all the firmware updates. It hovers rock solid. Compass is another thing though. The P3 is very sensitive in that regard.

 

[zincsoldier]

Looking at the csv file I noticed I mapped the roll values to the pitch.
When your Phantom flipped did it flip forwards or sideways as per below video? From the csv its sideways =roll.

Below a screen capture of the last two seconds from the csv file. Total recorded flight time is 6.6 seconds.
The first seconds clearly show a CSC (not shown).
Between 5.4 and 5.5 seconds the OSD.roll goes from 2.4 to -172.3 (the sideways flip).
The Rc throttle is lowered to its max between 5.3 and 5.4 secs in 1/10 second.
You further say you hovered for a minute prior the crash but the logger logged only 6.9 seconds with a proper CSC at the start.
So the first question is why did the Phantom "reset" itself 6 seconds prior the crash? Do you have any video from the camera?
View attachment 24166

Unfortunately I don't have any footage from the camera as I was not ready for pictures and records at that moment. The P3 flipped forward not sideway. I was talking to DJI the other day and sent them my .dat log data. I'm looking forward to hear from them. I will keep updating.

 

[zincsoldier]

Someone correct me if wrong, but you can't manually flip a P3 without really really trying, right?

I'm going to use the Occams Razor method on this one and guess that there were IMU problems. My reasoning:
1. The Phantom rarely flips - and is probably programmed not to.
2. The IMU is the main device which keeps the Phantom level
3. Compass and GPS problems result in drifting or other flight problems, not in flips.
4. Battery problems usually lead to app warning and/or the machine powering off and landing
5. IMU's are known to need resetting and calibration regularly. Most P3's need a couple just to offset the bumps from shipping.

If that were my quad I'd clean it out and reset the IMU two or three times on a perfectly level surface - then do a compass calibration and fly again. If it does anything like this again - then the IMU may have internal or other problems.

The Dronerepair dude mentioned that some older P1's and P2's were "fixed" by banging a certain board against a table - I think he was referring to 100g of force being applied to unstick the inside of an IMU.

Remember, an IMU is a mechanical device - stuff inside can shift. Ideally they can be reset properly, but like anything else there could be a rate of failure.

I contacted DJI and sent them my log. I don't want to risk it opening the top cover and fix the IMU LOL, that might void the warranty as some members here said. My previous P2 had a crashed as well and I change all the little things myself such as shell change, esc, motors because that was my fault I crash it. However, I would just leave it to DJI now and see what happen.

 

[zincsoldier]

Never calibrated my IMU, out of the box nor with all the firmware updates. It hovers rock solid. Compass is another thing though. The P3 is very sensitive in that regard.

Yes I heard P3 compass is more sensitive compares to P2s. Well, if your IMU is okay just leave it that way Happy flying !

 

[zincsoldier]

Some beaches have what is called black sand (looks almost black). This sand is very high in mineral content (plays havoc with metal detectors). I would calibrate it away from the beach. Doesn't matter if you are a mile ayay, magnet lines don't change that much.

Yes I also heard that there are magnetic black sand which might cause interference with P3. Well, I calibrated just to be safe

 

[Dapering]

Some beaches have what is called black sand (looks almost black). This sand is very high in mineral content (plays havoc with metal detectors). I would calibrate it away from the beach. Doesn't matter if you are a mile ayay, magnet lines don't change that much.

Hello Craig
I am going to Tenerife soon and there is black sand there. I was planning to calibrate, how to do so there?
I did not understand your comment part about the distance to the beach.
Thanks

 

[RoyWalton]

On my first p3 flight, while descending at 10 ft, two right side props stopped dead, bird flipped. Two right side motors would not start for 10 min (have flown p2v+ for 18 months)

 

[zincsoldier]

On my first p3 flight, while descending at 10 ft, two right side props stopped dead, bird flipped. Two right side motors would not start for 10 min (have flown p2v+ for 18 months)

Another P3 with this problem Have you figure out the cause yet? I am still waiting from DJI for an answer.