Flip and spiral death from 100m, solved

3689517144?profile=original

I recently had my 4th in-flight motor shutdown on my old 3DR Quad - I think I've finally figured it out so I thought I'd share the story so others can check for the same problem as well as show some logfile diagnosis.


My earlier flip deaths were from quite low so I didn't have a chance to see what was going on, and I didn't have the motor logging on to see which one shut down. I also suspected I could have been having brownouts in the past as I was powering my gimbal servos from the signal pins for a while until I discovered that was *BAD*.


This one was from high enough that I could observe and try a couple of things on the way down. It definitely still had power to the APM and some of the motors as you could hear it trying to correct itself, I switched out of auto mode into stabilize, pulled the throttle back to idle then up again in the hope it would restart the 4th motor, but it didn't. I heard the other three spin up again and whine dramatically but it never corrected. It thumped into the ground, broke all the props, bent the arms, but all the electronics survived (even the S100 camera). Reinforcing the message: NEVER FLY OVER PEOPLE! For sure that would have caused more than a bruise if someone was below it.

From the logs it looks like it was the #4 engine that quit (rear-right on a quad), as you can see it suddenly pitches up and rolls right, commanding more power for the #4 and less power to the opposing #3 engine (to no avail, as there's no RPM feedback in our current hardware so it doesn't know the motor is out).

3689517252?profile=original

3689517154?profile=original

The frustrating thing was that after this crash when I tested the motors they all ran fine, just like the previous flip crashes, so I was at a bit of a loss to diagnose the problem. I thought it could be low power and bad ESC programming that would cause the #4 ESC to cut out early, so I ran an old pack right down and all 4 motors slowed down in sync together as the voltage dropped off.


So I knew it was something with the #4 motor and it was unlikely to be the ESC as I'd replaced that before after one of the crashes. I removed the props, ran the motors and pulled and jiggled all the wires around to check the connections and...

BINGO!

The motor glitched then restarted. I pulled the cable some more and it stopped altogether. Here's a vid showing the problem:

https://www.youtube.com/watch?v=_iB5r6-AE-o

Turns out a bad wire connection to this motor had caused me 4 crashes and lots of headscratching. One more for the 3DR Quality Assurance department I think (well, 4 more...)

3689517215?profile=original

So a lesson for all: when you're finished a new build and periodically thereafter, remove your props, run your motors and jiggle all the wires around to check you have solid connections!


AC.

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • Developer

    I agree with you Rob: the single core enamel wire needs to terminate at a post with the multi-stranded flexible wire connected to that (and not soldered). Soldered joints are prone to instant failure as they just break. Multi-stranded wire fail slower as each strand tends to break one at a time, and it's easier to see problems before failure

  • Thanks for this excellent post-crash analysis.  I really like to see this type of thing, we can all learn from eachother.

    The issue here I believe is not so much quality control, I'm guessing those were built to spec.  But maybe the spec needs working.  The solder joint is rigid, and then it transitions to a flexible piece of wire.  I'm guessing it broke right where the flexible portion starts.  This really needs strain relief.  

  • Developer

    This has happened to me 4-5 times. Usually a minor crash triggers it. The motors start to dip or pulse for a few flights, then bam it drops out of the sky. You need to remove all mechanical stress/vibration around the solder joints to prevent it, otherwise it is bound to happen eventually. 

  • It hasnt taken me very long to experience this problem too.  My broken wire was in the motor bell on Turnigy motor.  It took some work to get it apart enough to see it and fix it.  I used the open/short feature on my meter to diagnose it, but I still had to get the motor half way apart to get at it.  I have a crap soldering iron and that worries me when I read about cold solder joints.

  • Bad solder joints are a common problem in our hobby. After my own bad experience i always check everything on my own, peel off the ESC heatshrink and directly solder everything my self. Even my RX is soldered directly. I stay away from motors that lead out the stiff copperwires to the connectors - they will break after time if you don't put some epoxy on the lead out of the motorcase.

    But thats only me...

    Cheers Kraut Rob

  • I have been in this same situation before....F'in motor joints used to snap .....even though they were soldered well, even then cold solder joints formed when the solder was just solidifying and the movement of the two wires which were being soldered caused the crystal structure to deform...that's when I learnt how badly I needed to have a pair of helping hands to support the two wires I am trying to solder...I am sure that would have helped you too..Really sorry for your crash

  • I've had this happen 3 or 4 times on my hexa now!  never had to re-solder one of my joints, but the origional ones fail from time to time...  I've actually made this test part of my 'rebuild' procedure... probably wouldn't hurt to re-solder all of the wires on the motor, if one has failed, the other two won't be far away... and if I think of it, I'd put glue filled heat shrink on the leads to take the strain instead of the solder joint...

  • I've had this happen with a cold joint in a bullet connector.

    It was an impressive quadruple flip. Somehow got lucky with a flat landing. 

This reply was deleted.