Crashed, fixed (thought so), test flight ... then wierd behavior! HELP!

Discussion in 'Cinestar 8' started by Sebastian Meredith, Dec 1, 2014.

  1. Sebastian Meredith

    Sebastian Meredith Active Member

    Joined:
    Sep 30, 2012
    Messages:
    257
    Likes Received:
    41
    Hi guys,

    Its been a while since I posted anything ... all has been running so smoothly without incident ... until yesterday. I'll give the run up to the weird behaviour to give a bit of history incase any of it is relevent to the cause.

    We're shooting some footage on a remote Island in the indian ocean and had to strip our mostly stock CS8 (using all MK electronics - PDB, BLs, Navi, FlightCntrl, GPS) into spare bits to pack for our flight. On arrival I put it back together, did the relevent ACC and GPS recalibrations. All good. Took off for a test flight which behaved as per normal ... that's when the drama started.

    I got pushed but a gust of wind towards a palm tree while acending, I thought I managed to avoid the leaves but alas they caught me, and sent me tumbling down. CS gimbal all twisted, landing gear wrecked and a few carbon props shattered. I brough enough spares with just in case and managed to rebuild in a few ours. I did an inital power test on the PDB with the BLs using a 9v battery and all green lights. I then connected the 4S 8500mah lipo and poof 2 BLs went up in smoke. I luckily brought 2 spares, replaced them and out it back together. Connected lipo and the familiar 8 chirps followed. Smile on my face ...short lived ... in MK tools I tried a motor test and prop 4 would not spin, but instead it only managed to twitch its way around at a constant speed no matter the throttle. I had this behaviour once before when after a flight the next restart had one motor fail to spin when fired up and also reulted on the twitching rotation. The CS8 was locked down to an idle only. MK Tools reported no fault on the BL and it had a green light. Same diagnostic as yesterdays incedent. I then replaced that BL with my last spare and she fired up perfectly. Did some test on MK Tools. Redid all the calibrations - ACC, GPD ... rechecked with MK Tools and all seemed ok. All 8 motors were spinning correctly and in the correct directions. We also held it down and tested the response to a little thrust and some nick and roll movements and all seemed 100%.

    I took a deep breath and took off ....

    That's when my heart dropped into my stomach. I had put the CS8 into a hover about 2 meters high and after about a second she bucked towards me and then away from me and I brought her down immediatly. Only lost a prop as she tipped forward on landing. I replaced the prop and tried again. Same behaviour. As soon as I was about a meter off the ground in tried flipping over ... I recalibrated the GPS again just incase ... tried again and same thing.

    I am now clueless as to what to check. I had a quick look at the GPX file with my basic knowledge thereof and the motor currents seem to be all over the place. We island locked for another 7 days, so getting any spare parts is a no-go.

    Any suggestions on what is going on under the hood? GPX file attached (14120100.GPX is the first and the longest flight) from the first flip and then the few after as well. I'll upload the video shortly of the flipping taken by a bystander (internet is sporadic on the island)

    I appreciate any and all advise, suggestions and feedback!

    Cheers,
    Sebastian
     

    Attached Files:

  2. Sebastian Meredith

    Sebastian Meredith Active Member

    Joined:
    Sep 30, 2012
    Messages:
    257
    Likes Received:
    41
  3. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,371
    Likes Received:
    1,162
    Hi Sebastian:
    It almost looks like the ACC calibration (as distinct from the gyro calibration) has inadvertently been reset without the flight controller board being level. It's all too easy to do -- all it takes is moving the left stick to the top right hand corner and it will cause the Accelerometers to think "level" is whatever angle the flight controller is when the reset happens.

    I would first start by ensuring the flight controller is level -- this does not mean just place the copter on level ground as the vibration isolators can allow the frameset (booms and motors) to be canted over. What you really need to do is remove the flight battery, and use a spirit level (or an iPhone with a spirit level app.) to level the battery plate -- then you can be reasonably certain that the FC board is level.

    I'll have a look at the GPX files too....I'll post again once I've had a look. Did you you see anything out of the ordinary with the GPX files? Sorry...you already answered that in your posting.

    Andy.
     
  4. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,371
    Likes Received:
    1,162
    Firstly, given that you already did the ACC recal, that, on thinking about would tend to exonerate the ACC.

    Yeah, just looking at the first GPX file, 14120100.GPX, the motor currents and the total current from the flight batteries are really strange.

    My next guess would be that something in the I2C bus wiring (the small Molex six-wire connectors) is bad and it's causing problems.
    Either that or you have sick FC board. And I see what you mean, the Ile du Nord is a leetle bit remote -- and yet you have cell coverage?

    As you have the GPS board on the copter, I would disconnect the six wire Molex connector from it to the Nav board -- just try to sort out the problem without the GPS board. There *is* a chance that *that* is the bad cable, of course.
    You could also consider replacing the six-wire Molex cables by hard wiring the C and D pads between the flight controller, navigation controller, and the power distribution board. That C pad to C-pad on each of the boards.


    Hope this helps.
    Andy.
     
    Sebastian Meredith likes this.
  5. Jason Smoker

    Jason Smoker Active Member

    Joined:
    Oct 23, 2012
    Messages:
    850
    Likes Received:
    67
    Sebastion this might sound stupid but I would check that your fc and Bl boards have the arrow pointing the same way? You might have moved or put it in wrong in your attempt to fix it quickly?
    Also if you replaced your bl boards did you check motor rotation?

    I know this simple stuff but you got to start somewhere right?

    Sorry about your crashes!! :(
     
  6. Steve Maller

    Steve Maller UAV Grief Counselor

    Joined:
    Oct 30, 2012
    Messages:
    3,979
    Likes Received:
    807
    Needless to say, this sucks big time.
    First of all, you should find a way to tether this maniacal machine to the ground before you test it again.
    As Andy mentioned, it certainly looks like a gyro/acc problem.
    Tether it to your PC with the MKUSB and open MKTools, and initialize things (no props on!), then lift the copter and move it around and see if the 3D view of the copter matches the movements as you physically move the copter around. Take careful note of the "front" boom and make sure you're seeing movements that exactly match what you're doing. And that when you place the copter back down you're seeing it level and still.
    Also make sure your mixer is the correct one for the configuration you're flying. That can get horked pretty easily, too.
    Good luck, mate!
    :(
     
    Sebastian Meredith likes this.
  7. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,371
    Likes Received:
    1,162
    Good point, Jason. Given the re-build that could well be an issue.
    Sebastian do you have enough Internet bandwidth to post some images of the board stack so other eyes can have look -- especially in light of Jason's comments.

    As Mr. Maller The Eloquent put it: This sucks big time.
    Andy.
     
  8. Sebastian Meredith

    Sebastian Meredith Active Member

    Joined:
    Sep 30, 2012
    Messages:
    257
    Likes Received:
    41
    Hi guys ... As always your speedy responses are awesome!

    I've got pretty phenomenal internet connection here via satellite! But no Cellular mobile connection.

    I've check the boards, all seem to be in the correct direction. I checked the BL addresses for number sequence and all good. Green lights all around. I'm about to reinstall the FC and NAVI firmware just in case it's brains got scrambled in the first tree collision. After the first crash the one thing I noticed is the motor start up chirp is not in the same sequence that it usually is. It use to be 1 through 8, not it seams a bit random.

    Attached a photo of the board stack. If you need another angle let me know.

    Steve ... I've hooked up the Bluetooth dongle and will start up and do the 3D view test.

    Thanks again fellas. image.jpg
     
  9. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,371
    Likes Received:
    1,162
    The BL-Ctrl chirp is actually a random sequence on start up of the different BL's -- so there's no deep meaning there.

    Let us know if the little simulated 3D quadcopter is behaving rationally. It will twitch a bit, but it should not be spazzing out completely....

    What on earth is actually on Ile du Nord apart from you guys?

    Andy.
     
  10. Sebastian Meredith

    Sebastian Meredith Active Member

    Joined:
    Sep 30, 2012
    Messages:
    257
    Likes Received:
    41
    Andy ...

    The FC is already hard wired into the PDB. had too many issues with those silly little wires on the molex cable.
    Apparently there is no way to hard wire the gps to the navi to eliminate that molex. When flying in warehouses we've disconnected the 2 ribbon cables from the navi and flown completely manually. Is this better than just removing the molex cable to remove the gps from the equation?
     
  11. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,371
    Likes Received:
    1,162
    Ok. The hard wire should eliminate that. The GPS/Navi connection doesn't appear to be the standard I2C -- we'll need to contact MK to find out more.

    I'd certainly just remove the connections from the FC to the NC/GPS for testing. If nothing else this falls under the general diagnostic strategy of take out one thing at a time and see if it works. When/if it does, then whatever you last removed was the problem.

    Let us know re: 3D model being stable or spazzing.

    Andy.
     
  12. Sebastian Meredith

    Sebastian Meredith Active Member

    Joined:
    Sep 30, 2012
    Messages:
    257
    Likes Received:
    41
    Also known simply as North Island ... It's an exclusive luxury island. 10 villas only where the super rich and famous come to relax. It's only about 4000 euros a night per person. And most stay about a week. They call it barefoot luxury. Check out http://www.wilderness-safaris.com/camps/north-island
    I'll give feedback on the 3D view in a few minutes.
    Some eye candy from my room .... Just in case you are wondering ... The gimbal has been removed while working on it. Just got some landing legs in place.

    image.jpg image.jpg image.jpg
     
  13. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,371
    Likes Received:
    1,162
    Nice. Hope we can get the copter fixed otherwise you may be forced to stay there and imbibe....
    You don't happen to have a spare FC board do you?

    Andy.
     
  14. Sebastian Meredith

    Sebastian Meredith Active Member

    Joined:
    Sep 30, 2012
    Messages:
    257
    Likes Received:
    41
    Nope. That's the one thing I don't have a spare of!
     
  15. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,371
    Likes Received:
    1,162
    OK. Well, let's see how the 3D model looks without the NC/GPS hooked up the I2C bus.
     
  16. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,371
    Likes Received:
    1,162
    Just a random thought, I wonder whether the BL's you're currently using are all using the same firmware version and whether that might explain things -- don't take any action on this right away other than to tuck that thought away in the back of your brain...

    If the 3D quadcopter is spazzing, then I'd definitely try re-calibrating the ACC one more time. There's no downside and there's just a chance...

    Andy.
     
  17. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,371
    Likes Received:
    1,162
    Sebastian -- I'm exactly 12 hours behind your time zone, so I'm going off line until tomorrow morning Pacific. Let me know what you find.

    Andy.
     
    Sebastian Meredith likes this.
  18. Sebastian Meredith

    Sebastian Meredith Active Member

    Joined:
    Sep 30, 2012
    Messages:
    257
    Likes Received:
    41
    Ok. Done the the 3D test. All seams ok. Video below are some reading from the indoor test. We first did an outdoor test after redoing the ACC calibration and GPS calibration - all props off. During the test I manually lifted the CS up above my head, did some nicks and rolls. Some random stick inputs and flicked a few switches - PH and AH. the 3D test from the outdoor showed no out of the ordinary movements. I'll upload the GPX file for that. Then the indoor test I let it sit on the counter and did the same kind of movements. I recorded the data and are in the videos below.

    Sorry could trim them to eliminated the time wasting in the beginning.





     
  19. Sebastian Meredith

    Sebastian Meredith Active Member

    Joined:
    Sep 30, 2012
    Messages:
    257
    Likes Received:
    41
    A random thought. If any of the capacitors (pointing to in the photo attached) were not making decent connection to the board or if one of the were fried, would that come up as an error in MK tools and the GPX log? Just reaching for anything right now? If I do and Ohms test on the power leads when unplugged ... What sort of reading should I get as the capacitors load up?

    Another ... The pressure sensor on the flight control board ... Could that be playing up and actually over compensation the repose actions? Or does that only come into play when using PH and AH? Reaching for any ideas now.

    View attachment 6815
     
  20. Sebastian Meredith

    Sebastian Meredith Active Member

    Joined:
    Sep 30, 2012
    Messages:
    257
    Likes Received:
    41
    GPX files for the outdoor (..200.gpx) and indoor (..201.gpx) 3D tests ...
     

    Attached Files:

Share This Page