P3 went spinning out of control and crashed!

[0DRK3RT]

How is it that he had 95% battery at take off, but each cell's voltage readings are approx. 3.4-3.5 after only 30sec. Into the flight?

Seems odd

 

[Luap]

Had a look at the pictures.
So Left front prop is the one with missing piece.

So the question is why did it break off in flight?
Ever had any tip overs with them or the likes that might have damaged them? Or could this be a manufacturing defect?

 

[TacAir42]

Same as mine...just happened on 1 JAN 2016. Look at my log...

Strange Crash: P3P Lost, Compass Went Haywire | DJI Phantom Forum

HealthyDrones.com - Innovative flight data analysis that matters

 

[BudWalker]

I'm still looking at the .DAT but it's clear the compass was not an issue
View attachment 39617

The time axis is offset 192.04 (motor start - battery on). Until the incident that starts happening at 210 the compass is normal, in fact very normal. The incident appears to be a propulsion issue with the Left Front motor/propeller. At 220 the A/C is spinning and the magMod value is to be expected.
View attachment 39618
Starting at 210 the LF is being commanded higher with a small increase in speed. Starting at 220 commanded is maximum (10000) and more increase in speed. I think this is a broken prop issue. Although, it seems that there wasn't a abrupt separation. Rather, it seems like the prop broke in the interval 210 to 220.

I'm still looking at this.

Look at these plots from another flight.

Here it was the Left Back with the propulsion issue. The commanded goes to the max, speed increases (probably to the max), but the load decreases. Just what to expect if the prop is gone.

In Ksavoie's flight it's not so clear cut.

Although LF gets commanded to the max it cuts out occasionally to be replaced by the RB being commanded to the max. This oscillation behavior is consistent with partial propulsion from the LF prop. In addition, the LF load doesn't change which indicates that the prop isn't entirely gone.

@Luap pointed out that this could be a motor or ESC problem and not necessarily a prop problem. I have to agree this is also possible. Anyway, it does seem like it's a propulsion issue with the LF.

 

[ksavoie]

After the crash, I am getting an ESC error. But that's probably from the crash.

 

[BudWalker]

After the crash, I am getting an ESC error. But that's probably from the crash.

Actually, seems like too much of a coincidence. If so, it would seem there is a good chance DJI will cover it under warranty.

 

[AlexSP]

I'm with @BudWalker here (again). Although it's hard to point what exactly could have caused the ESC failure (assuming it failed before the crash of course), it seems an unlikely coincidence. I'm not sure in which Phantom version DJI changed to a one-in-all (central type) of ESC, they had one for each motor at some point before the P3.

Not that it could cause this or any other problem - though some engineers think that is a better option - I for myself have no formed opinion about this and think that both are fine if properly designed and made. But it's one more aspect to consider...

Overall, DJI created a rather easy-to-fly, easy-to-care quad with the P3 in my opinion. Very, very advanced, sophisticated and capable, yet rather simple to control and also to maintain. Especially when compared to other RC ACs, some even less capable than it. It's cheaper, safer, easier and funnier (IMHO) than most copters and planes I've owned or own (or maybe I just like the newness LOL).

But it's still a complex aircraft with lots going on with the hardware and firmware. And that demands more and more interaction and intervention from the pilot with the passing of time. Like all these updates, calibrations... It can reach a point where anyone can screw up something, even something small that could lead to a crash or failure at some point. Like the ones we've been reading about here.

Just another theory, of course, more food for thought.

 

[BudWalker]

@ksavoie would it be possible to get the .DAT file from your P3 whilst it's saying ESC error. I'll use it to locate the ESC error indication in the .DAT file and include it in some future version of DatConverter. Just turn the P3 on and as soon as it says ESC error do the download procedure to transfer the .DAT. As soon as you initiate the download procedure it'll create the .DAT for that "flight".

 

[Mark The Droner]

I would very much disagree with your reasoning. Lets look at the facts.

All the reasons you have cited are necessary for GPS flight and RTH function. Under low GPS lock condition P3 goes into ATTI mode.

Many people purposely fly under ATTI mode for one reason or another.

Neither is OP leveraging RTH or triggering RTH.

I fail to see how lack of GPS signal would contribute to essentially an ATTI flight. Can you elaborate?

Clearly my opinion was wrong after Bud's post. So yeah, it's easy to disagree with it.

I was referring to the pre-flight procedure which apparently was done when I thought it was not done. Many fatal flights are made due to an improperly working compass or an invalid home point position. I'm happy to elaborate further if you need me to.

 

[ksavoie]

Ok....I will do this tomorrow evening

 

[ksavoie]

@BudWalker @AlexSP

Here is the .dat file. Today the beep stopped and actually the props started to spin except for the front left.

P3A

 

[BudWalker]

Thanks, I sent you a PM with some questions about these files.

 

[KyleMaxxUAV]

New or old style of motors ??


Sent from my iPad using PhantomPilots mobile app

 

[ksavoie]

probably old....I pre ordered this one.

 

[KyleMaxxUAV]

Thanks for the quick reply ! Sorry to hear the bad news .. Great photo though !!


Sent from my iPad using PhantomPilots mobile app

 

[BudWalker]

I looked at the extra flights that @ksavoie provided to find out how a .DAT reports information about a ESC error. In his P3's current state it will try to start the motors and then quit after the initial two burst sequence because the LF didn't start. In particular, it was commanded to start, no speed was detected, and there was no load. At this point the P3 has declared an ESC error, but it waits until a subsequent motor start attempt before it reports the ESC error to the tablet. That is, it deduces that there was an ESC error from the fact that one of the motors didn't start. This works only before the flight begins, i.e. it can't be used to determine an ESC error during flight.

I was hoping to find some other signal that would indicate an ESC error after motor start. Without this it can't be 100% concluded that an ESC error occurred in flight leading to the incident in FLY232. It's possible that the crash is now causing the ESC error. I personally believe that an ESC error occurred in FLY232 and was the cause of the incident, but I can't honestly say that this has been 100% demonstrated. As I said in the previous post the loss of propulsion on the LF was gradual, not abrupt like would be expected with the loss of a prop.

 

[Luap]

I was looking again at these charts

The motor load Left front is reaching 0 value on x axis more and more with time. Non of the other motor load shows this behaviour.
Could be that this is result of a bad solder joint.

 

[BudWalker]

I was looking again at these charts

The motor load Left front is reaching 0 value on x axis more and more with time. Non of the other motor load shows this behaviour.
Could be that this is result of a bad solder joint.

Absolutely. A bad solder joint, or more generally, something that causes an interruption of current being delivered to the motor. This started before before the incident. Furthermore, as the load spends more time at 0 the commanded gets higher which is consistent with an attempt to keep speed up as the motor is becoming disconnected.

I had been looking in unexplored corners of the .DAT for some evidence that there was a problem before the incident. Fortunately, @Luap could see what was already evident in these plots (which, BTW I had made )

 

[ksavoie]

Sent my P3A back to DJI today. I will keep you guys informed on what DJI says about this incident and warrantee issues. I would like to give may thanks to @BudWalker and @Luap for their analysis on this issue. Without them, I would have zero insight and would still be thinking that the crash was caused by a battery or a compass issue.

 

[John N Kronyak]

I warmed up the P3 for a few seconds and then took it up to around 170 feet and then almost immediately it started behaving crazy. So I decided to take it down and it got worse, eventually hitting a tree on the way down......I put the flight log in healthydrones.com and it shows something is up with the battery when I got to 173 feet. Can someone decipher this for me?

HealthyDrones.com - Innovative flight data analysis that matters

I lean towards the compass calibration. The only scary crash I had(No damage, luckily)was not calibrating in a new location. Took it off and it went out of control at about 25' Agl and crashed flat out into some tall grass.Learned a lesson from that one.