Catastrophic design flaw of Phantom 3

[Meta4]

Moderator,
It is obvious that you are most likely a DJI employee.

GM .. it's obvious that you don't know that the forum is completely independent of DJI.
They have their own forum. And you probably don't realise that the mods here don't get paid.
We have to buy our drones just like you and get no special deals.
My interpretation is my own and based on what evidence I've seen.

 

[Larry D.]

Its 50USD inc delivery for the strong arm bits. I haven't decided yet but I think I might save up for the revised shell. My bird has never been crashed and always hand caught.
I'm in Aus so warranty is pointless. And I think DJI warranty is also pointless.

I opened her up last night - the hidden clips are painful and you really need to be patient and work and rework around the bird.
I used hobby glue gun to prevent further cracks and hopefully it would prevent the motors from ripping out. This same fix was done on the P2 with success. Hopefully it would yield the same results on the P3A.

Wipe off the excess glue near the screw holes.

Fairly new P3 here and I have 4 screw hole (not the 2 hole kind) prop guards on - maybe 30 practice flights around home. Reading these treads about stress cracks made me start looking at mine. I haven't unscrewed prop guards to look underneath yet bet I do see one crack in one of the inner screw holes already - UGH! Hopefully not as bad yet as some, but will see after guard removal.

Choices?
Call vender on Monday...
Call DJI om Monday...
I really don't want to send her back but.... Will see...

What type of plastic is this shell made from? Poly?........? Or ?

Question for someone that knows plastics. If I decide to remove top shell and fill (reinforce) the motor mounts and screw hubs where crack is. What would be BEST to use?
Hot glue like above? or some kind of plastic epoxy? If epoxy, what would be the ultimate brand and type to use, to fill the motor mount area up to top of screw hubs. Would this be a bad idea? Fill all 4 the same for weight balance etc...

I would then put prop guard mounts back on as they are the quick removal guards and the guard mounts can stay on.

Opinions please!

Sorry, but my OCD won't let me fly again until I am SURE I did all I can to prevent what has happened to others over stress cracks etc...

It is bad enough, that I am nervous over everything else that could happen to my baby.

Thanks.

 

[jcampen]

About 15 years ago when I switched to digital cameras there were alot of issues with the Nikons and others. We started a thread with serial numbers and patterns in the issues related to production runs began to be obvious (database was online in a forum) and the manufacturer stepped up to the plate to take care of the repairs. Wondering if anyone would be willing to do that here? Identity the location of the cracks, serial number of the bird and let's start a database? About 50 hours on my bird, SN P76DCF15014174 no signs of any cracks on near the motors or on the arms but will be installing a kit as soon as it arrives.

Andrew

Just off topic here, because I am curious about your remark here. . You have done about 50hrs on your Ph3 which if on average each flight is 15 min, means you have done 200 flights?. How many batteries do you have?

There is a big thread already on RC Groups DJI owners thread somewhere on this topic and many find the same as I do. We still have a running dispute with DJI that their batteries are not safe after 40 cycles (even if that, we now don't fly them in any built up areas now after 30 to be really safe) where as DJI claim they are good for 200 cycles which in my view is a lie. This are with our 4 Ph2=v2 though. I wonder how well we will go when the Ph3-A/P batteries start getting over 40 cycles?
.

 

[jcampen]

Looks to me that the higher torque motors and sending vibrations through the plastic arms. I don't think it has anything to do with hard landings etc. It is primarily a high frequency vibration from the more powerful motors. I am kind of surprised that this wasn't Engineered into the motor mounts in the arms. There always needs to be a dampener between motors and solid mounts, kind of like rubber motor mounts in a car. My Phantom 3 is coming on Monday. I will bring to work and discuss with other Engineers next week and try to come up with something.....

Just a note also that the Ph3s motors are very slightly tilted outwards towards left and right (not forward and backwards strangely though) where as they were no so on the P2s. Not saying its related, but just might be something to look at?

 

[jcampen]

I would be concerned how the glue handles the heat skippy. Some glue gets soft.

Its 50USD inc delivery for the strong arm bits. I haven't decided yet but I think I might save up for the revised shell. My bird has never been crashed and always hand caught.
I'm in Aus so warranty is pointless. And I think DJI warranty is also pointless.

I opened her up last night - the hidden clips are painful and you really need to be patient and work and rework around the bird.
I used hobby glue gun to prevent further cracks and hopefully it would prevent the motors from ripping out. This same fix was done on the P2 with success. Hopefully it would yield the same results on the P3A.

Wipe off the excess glue near the screw holes.

 

[bschaub]

Seems like buying the prop guards and then cutting off the ring would work fine and be cheap.

 

[DDW]

This might be a stupid question, but if you have UAVBits strong arm plates installed, how would you even know about stress cracks underneath if they developed in or around the screw holes?

 

[Rod Carnochan77]

You put what on the P3 to stiffen the arms gully1965 ?

 

[jcampen]

Cheap is to add these to the udnerside of the motor arm mounts

This might be a stupid question, but if you have UAVBits strong arm plates installed, how would you even know about stress cracks underneath if they developed in or around the screw holes?

Firstly its a high probability that when you shift the stress away from the original motor mount screw posts to the outer plate the stress cracks will no longer occur.

Secondly, even if it does crack its not going to matter a lot, because the added plate will still keep it in place and the only way the motor can come free is if it rips the whole bottom and top side of the motor mount off from arms. I really don't think that going to happen in a normal flying situation.

Lastly, I myself would not be using any plastic under plates because again these are only as strong as the small area support directly under the motor screw heads.

Ask yourself these questions when considering to buy under supports made from plastic. .Is the heat from the motors contributing to these stress fractures we see? Highly likely in my opinion. Is plastic under support design specific for this problem and is it a suitable proven strength type of plastic and up to what temperature? The prop guard mod suggested by some might be a good idea and it may just work, but its not design specif to this situation and the heat just might affect this as well.

For the peace of mind , just use carbon fiber or aluminum under supports is my recommendation.

. .

 

[THG Aviation]

Just off topic here, because I am curious about your remark here. . You have done about 50hrs on your Ph3 which if on average each flight is 15 min, means you have done 200 flights?. How many batteries do you have?

There is a big thread already on RC Groups DJI owners thread somewhere on this topic and many find the same as I do. We still have a running dispute with DJI that their batteries are not safe after 40 cycles (even if that, we now don't fly them in any built up areas now after 30 to be really safe) where as DJI claim they are good for 200 cycles which in my view is a lie. This are with our 4 Ph2=v2 though. I wonder how well we will go when the Ph3-A/P batteries start getting over 40 cycles?
.

3 bateries

 

[cdronefly]

Looks to me that the higher torque motors and sending vibrations through the plastic arms. I don't think it has anything to do with hard landings etc. It is primarily a high frequency vibration from the more powerful motors. I am kind of surprised that this wasn't Engineered into the motor mounts in the arms. There always needs to be a dampener between motors and solid mounts, kind of like rubber motor mounts in a car. My Phantom 3 is coming on Monday. I will bring to work and discuss with other Engineers next week and try to come up with something.....

My guess is that these motors separating are a "special" case. Here's what I think.....that is, until other evidence comes in. There are approx 150K P3's out there - MANY in hot climates (even where they are made is sub-tropical). The DJI American crew all live and beta test in hot areas.

I have read about 2 or 3 like this - total separation. But let's add a bunch - say it's 100 of them out of 150,000 - still, wouldn't a lot more be failing if this was due to a basic design problem?

And so - I'm thinking that 600,000 of these brushless motors are on those 150K+ birds. The motors are nice - BUT - what is the expected defect rate in a toy motor?
Stay with me....
Let's say that a very few of the motors have defects which causes interior friction or interior short or heating of some type. The heat conducts through to the mounting bolts - and softens (melts) the plastic....and that's that.

Let me know what you think of my theory. If your engineering analysis doesn't consider the multiple possibilities of causation, then the suggested remedies may not be as relevant. That is - if the motor is getting to 500+ degrees because of these defects, not much is going to stop the bird from self-destructing in one way or another.

 

[notelpats]

I'm not sure this is heat related. Members from all over the world are getting similar cracks. This is, however, the first time I've seen a motor completely fly off the Phantom.

Seems like buying the prop guards and then cutting off the ring would work fine and be cheap.

You can get snap on prop guards. I have them but always fly with guards on just in case for extra protection.

 

[jcampen]

3 bateries

Ok so that's about 67 flight for each battery correct?

This is well outside our "very nervous" zone. On the 3 cell Ph2 batteries we find they can randomly drop the charge very quickly and suddenly without any notice after around 45-50 cycles and the copter just descends with no control.

We had one virtually dropped out of the sky due to this, it was at 90% and then withing 30 seconds dropped to 30% and by the time we could get it back it lost that much charge we lost connection and down it came. Complete lose of the copter.

Considering we are up to having bought around 50 or more genuine DJI batteries in the last 3 years all with this same level of performance average, its a long way off from the 200 cycles DJI claim.

We worked it out it costs us about $4 a flight so we are find to throw them out after 40 cycles considering we make around 2-3k a shoot. Its regarded as cheap fuel. But for the cost of these batteries and particularly if your just a hobbyist, its not cheap and DJIs claim of 200 cycles is just a joke.

Its does very much concerns me also how really hot these newer batteries get, so they are really working hard.

I would be very careful sir to push them over the 70 cycle mark and please keep a close eye on your flights when they are getting over 50 cycles mark . These new type 4 cells might be ok, but just a point to be aware of.

 

[jcampen]

My guess is that these motors separating are a "special" case. Here's what I think.....that is, until other evidence comes in. There are approx 150K P3's out there - MANY in hot climates (even where they are made is sub-tropical). The DJI American crew all live and beta test in hot areas.

I have read about 2 or 3 like this - total separation. But let's add a bunch - say it's 100 of them out of 150,000 - still, wouldn't a lot more be failing if this was due to a basic design problem?

And so - I'm thinking that 600,000 of these brushless motors are on those 150K+ birds. The motors are nice - BUT - what is the expected defect rate in a toy motor?
Stay with me....
Let's say that a very few of the motors have defects which causes interior friction or interior short or heating of some type. The heat conducts through to the mounting bolts - and softens (melts) the plastic....and that's that.

Let me know what you think of my theory. If your engineering analysis doesn't consider the multiple possibilities of causation, then the suggested remedies may not be as relevant. That is - if the motor is getting to 500+ degrees because of these defects, not much is going to stop the bird from self-destructing in one way or another.

I live in the tropics as well and have had Ph2 for around 3 yrs and most don't have any cracks, but I don't think we are referring so much to hot climates or a 5-10C difference here, but more about the hotter working temperature of these new motors and with using 4S batteries.

There seem to be a hell of a lot of current draw which Ive yet to check what it is and when you take the battery out after a flight its really "scary hot!", it worries me a lot. I take them straight out after a flight to cool them down and never put them in the car and drive off until its cooled off for 15 min.

 

[cdronefly]

but more about the hotter working temperature of these new motors and with using 4S batteries..

No doubt more power=more heat.
But I suspect it's still not the norm that is causing this tiny percentage of events.
I'm still sticking with my "motors outside of spec somehow heat to VERY hot temps" theory.

Then again - personally I never fly battery after battery. I am of the opinion that these are flying cameras and I get the shots or a dozen minutes of video and then I land and put it away. I see some guys just seeing how many miles they can put on them and recording every flight and scratch my head...because what the heck can one do with all those gigs of "flying around" video?

Anyway - my thoughts are that if something happens at an extremely low percentage that it may not be a design thing but a QC event (on the motors). We'll see. Not enough data with only a handful of these events.

I suspect future models may even have temp sensors in the engines to warn us against this kind of stuff - of course this is one more example of proof positive why flying over people and highways, etc. is going to be a problem with the current generation of toy machines.

 

[Phil Tuggle]

Looks to me that the higher torque motors and sending vibrations through the plastic arms.... I am kind of surprised that this wasn't Engineered into the motor mounts in the arms. There always needs to be a dampener between motors and solid mounts, kind of like rubber motor mounts in a car. .....

I do believe you and the poster "Bert OConnor" above nailed it, as far as dealing with the problem in the absense of a redesigned replacement shell. You are absolutely correct that throughout industry and in products, almost all rotating machines are mounted on some sort of vibration absorber. Snubbers we called them on the manufacturing floor.

And I believe Bert OConnor above nailed part two of the equation; though the motors are all mounted at angles NOT normal to Earth's surface (as noted in another post, they are all canted - and that is intentional), to anticipate torque and propulsion dynamics. But the arms DO probably still flex differentially along the arm axis, as Bert says.

So.....how about remounting the motors with tiny o-rings through which the screws pass, and applying two small opposing helical strands of carbon tow (saturated in epoxy liquid to cure, and lose pitch - just two or three full wraps - a double helix) along the arms. Maybe there is a strong bonding, narrow, low elongation fiber tape available. Of course, you will not be able to open the shell later without cutting those...and it will look a bit hideous, but I think crack worries might be terminated.

 

[Daninho]

I live in the tropics as well and have had Ph2 for around 3 yrs and most don't have any cracks, but I don't think we are referring so much to hot climates or a 5-10C difference here, but more about the hotter working temperature of these new motors and with using 4S batteries.

There seem to be a hell of a lot of current draw which Ive yet to check what it is and when you take the battery out after a flight its really "scary hot!", it worries me a lot. I take them straight out after a flight to cool them down and never put them in the car and drive off until its cooled off for 15 min.

yes those motors and the batterie are always hot after the flights, my P2 was much cooler in direct comparision. The P3 seems to drain a lot of power, also because of lightbridge. You can see the voltage dropping when the P3 is on the ground. The P3 standard is completely different.

 

[westlightaerial]

Do you think quick release prop guards would have prevented the failure? They only cover the 4 motor screws, but it seems like the guard mounts beef up that area significantly.

No , I have quick release prop guards and thought the same thing until I took them off to take a closer look. It would have been a matter of time before my copter failed if I did not check them on a regular basis. I tool the prop guards off only to find stress cracks on two arms. IT just bad plastic and dji will have to make a better plastic because they will all start falling out of the sky if people don’t double check the arms before they fly.

 

[43k]

No , I have quick release prop guards and thought the same thing until I took them off to take a closer look. It would have been a matter of time before my copter failed if I did not check them on a regular basis. I tool the prop guards off only to find stress cracks on two arms. IT just bad plastic and dji will have to make a better plastic because they will all start falling out of the sky if people don’t double check the arms before they fly.

This makes me wonder if people are aware that when disassemble and reassemble the copter/motors, they should use proper torque when tightening the screws.

It is easy to over-tight those tiny screws. With so many people adding and removing prop guards and what-not, there sure will be over tightening by some, with resulting cracks.

 

[Fplvert]

This makes me wonder if people are aware that when disassemble and reassemble the copter/motors, they should use proper torque when tightening the screws.

It is easy to over-tight those tiny screws. With so many people adding and removing prop guards and what-not, there sure will be over tightening by some, with resulting cracks.

Totally agree, I think many P3s came from DJI overtightened. I slightly loosen screws and just snug them. I check the airframe after each flight, no cracks yet.