Had an I2C fail mid flight. Molex was solid on the FC to Power distro. The up side is our BM4k survived!
Would you be willing to share the GPX from that flight? It’d be extremely valuable to try and see if it’d be possible to determine the cause of the failure. In particular, there might have been something that precipitated (or at least preceded) the i2c failure. If not, that’s understandable.
NP=) It was just a straight I2c fail. Nothing preceded it. Kopter had about 30 hours on it. 2801 is the crash 2800 is the first flight we did that day.
Thank you. First blush doesn’t reveal anything. The video of the fall sure does sound like a catastrophic motor cutoff, which is consistent with a collapse of the i2c bus. But what caused it is obviously going to take some detective work. Which ESCs/BLs are you using, may I ask?
BL 3.0. MK already knows what happened. It was a I2C fail. Board got there yesterday and that was their call.
From the video it looks like your using the dual quadro 2xL board. The first version boards have been documented to shut down all motors when there is a distruption in the I2C buss line. I will be testing the redundant Flight control board setup where if there is a I2C failure a backup FCB takes over. Can I ask what motors you are running?
Dave, That is correct. this is the 2nd board we have had do this. First one was a bench test and motor 8 died with the rest going about 1-2 sec after and ESC8 catching fire (I was having some issues with low voltage/undervoltage LED blinking and wanted to make sure it was working when this happened). Running Tiger motors http://www.quadrocopter.com/T-Motor-Heavy-Lift-Motor-MN4012_p_1007.html I am not to worried about this crash. These things happen. Going with the updated Dual quad2xl board this time around
Ryan: Looks like the GPX file for the crash is truncated (as often happens), so we can see the I2C bus error happening, but then the log entries stored in the NC's RAM never get flushed out to the microSD card. It's a great shame that MK cannot find a fix for that as the most important entries (for post-crash analysis) are lost. But, as SteveM says, there's nothing to indicate that a failure is about to happen. Ryan, when you looked at the wreckage, did you happen to see if any Micro-Molex cables were dangling loose -- it'd be hard to pinpoint any such cable as the cause of the accident as, of course, it might have come loose on impact. Its also a shame that there's no "fail safe" mode that the NC board could go into when it sees that the I2C bus is littered with malformed data -- if only it could ignore the trash on the I2C bus and execute a controlled descent/landing. The problem is that the FC cannot even control the motors when the I2C bus fails (as you found out, Ryan) as the FC uses the I2C bus for everything, especially FC to PDB/BL-Ctrl communication. I suspect we may have to move away from the six wire Micro-Molex connectors wherever possible and hard solder the C and D lines that make up the I2C bus. It makes it a pain to remove boards, but at least there's a reduced chance that the I2C bus will collapse because of a dodgy connector. Andy.
Andy, It was very truncated. the LAST I2c fail corresponds with the flip. What Steve said is 100% correct. Honestly not to worried, Black Magic reached out, as did MK and Quadrocopter. Our M5 is on its way to you guys just to make sure it works and I am excited to build another kopter. Been flying my hex non stop. (sorry doorbell). We will be back up in less then two weeks
Ryan: I commend your attitude to the incident -- my concerns are that others may not be so lucky with their experiments at center-punching Mother Earth with their cameras! Or that someone or something underneath the copter might have experienced a Cinestar-sized compression fracture! Please let us know if you get any additional data that gives a clue to the causes of the I2C bus failure. Andy.
Andy, Holgar at MK has the board and is testing it. Thanks. I was a bit freaked out the day it happened but after looking at the GPX and talking to Quadrocopter it's honestly a non issue. Just glad noone got hurt=)
Here are a few pics of the damage to the Movi and the Kopter after I salvaged what I could. Also lost our spektrum receiver in the crash too ><
Fastest insurance payout ever. If I may ask, which insurance company do you use that has covered your copter damage so quickly?
http://www.transportrisk.com/uavrcfilm.html. Check is in the mail and I pulled from Volant's (my company) savings to get back up asap=)