/dist/images/branding/favicon

Had my first Major Crash today

Discussion in 'Cinestar 8' started by David N Atkisson, Jan 17, 2013.

  1. David N Atkisson

    Joined:
    Nov 24, 2012
    Messages:
    73
    Likes Received:
    0
    Hey Folks. I had my first major crash today. I was flying with my new camera operator and was hovering with Alt Hold on around 50 feet in the air and the CS8 started to plummet. I had no audio warning and when I went to try to throttle up nothing happened. After picking up the pieces I started my post mortum on the wreckage. At this point the tally is up to 2 grand for repairs. I know these things happen without warning and accepted that going into this. I tried to get flight data off the micro sd card but nothing was on the card. It must not have been inserted correctly.

    Time to rebuild. I'm so glad I built it the first time so at least I know how to approach this. I am certainly going to make sure that I am getting flight data in the future because not know how this happened is going to make me nervous in the future.

    Thanks for all help you folks have given me to get to this point. I am very thankful that it was only the CS8 that was damaged and that no people where hurt during this incident. Now it's time to rebuild and this has made me even more aware of my responsibility to maintain high standards in safety. I will try even harder in the future to take all precautions in the future.

    David
    In IMG_9013small.jpg IMG_9018small.jpg
     
  2. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    Oh, Jeez, David -- that's a really sad sight.

    Could you see any data files on the micro-SD card (e.g. SETTINGS.INI and a LOG folder?)
    I did have a couple of flights last year in a new location where I obviously didn't wait long enough for a GPS fix and thus didn't get any log files.

    Did you get the sense that the motors all just shut down and the bird came down flat, or did it roll over? From the second image it looks like it landed upside down suggesting it tipped over as it fell (or it bounced)?

    Either way, sorry to learn of this.
    Andy.
     
  3. David N Atkisson

    Joined:
    Nov 24, 2012
    Messages:
    73
    Likes Received:
    0
    It fell behind my house on a golf course. The sd card had no files on it at all. It looked like maybe I didn't push it in and lock it into the tiny drive. It was in the slot but possible not pushed and engaged or something. I think it must have bounced because what I saw before it went behind my house was it falling straight down flat. I did noticed that the booms I lost where booms 1, 2, 8. They where shattered and must have taken alot of the crash energy. My 5D mkII was lying a ways away but still functional without any damage as far as I can tell.

    I lost the bottom half of my 3 axis gimbal, essentially the 2 axis part. It's still has salvageable parts but the rear roll axis mechanism and radian are messed up enough I'm going to just have to replace them. It seems the pan axis part is okay but I won't know on some of the parts until I get things back together and can start testing to find out.

    Most props where broken. The thing that makes me nervous is not knowing what part of the system failed so I can be alert to failure in the future. This was pretty sudden and I had no feeling that anything was wrong with my system when I started to fly.
     
  4. Steve Maller

    Steve Maller UAV Grief Counselor

    Joined:
    Oct 30, 2012
    Messages:
    3,981
    Likes Received:
    807
    Oh man. So sorry about this. Can you check the voltage on your battery? What kind of battery were you running? Was there any audio warning? My sympathies! :(
     
  5. David N Atkisson

    Joined:
    Nov 24, 2012
    Messages:
    73
    Likes Received:
    0
    I always check voltages before I lever the ground. I have the nice lady in my ear giving my altitude and voltage warnings. Nothing, no warnings at all. Thats whats so weird. It fell flat with blades moving until out of site. I just had no thrust when I went to tried to get her to stop falling. Nothing at all. I'm assuming she bounced. Parts where scattered in the 20 foot radias. Unfortunately I had no data of any kind on my micro sd card. Not from any of my flights. I will be checking for data in the future rest assured.

    I have the beginning list of parts ordered for repair but won't begin fixing until I get back for a PPA convention next week.
     
  6. Gary Haynes

    Gary Haynes Administrator
    Staff Member

    Joined:
    Jul 6, 2012
    Messages:
    5,211
    Likes Received:
    460
    David it didn't help that you landed on grass. Mine hit on frozen ground. If you are using an MX-20 please pull the data card out and grab the folder of log files. That might help give some data. Don't turn the MX-20 on so that we make sure which file is the last one recorded.

    For others try to remember to set the date and time on your MX-20, or maybe any other transmitter you are using. Graupner's all seem to record to the data card. Not sure about the other brands. Using an MK stack and the down linked data there is good information that might help determine what went wrong if the SD card didn't record.

    And a bit of luck with the camera. I was only flying my NEX-7 and totally smashed the LCD screen. Thanking that it wasn't my D800.
     
  7. Jason Smoker

    Jason Smoker Active Member

    Joined:
    Oct 23, 2012
    Messages:
    881
    Likes Received:
    70
    David -- that's a really sad sight. I have a feeling it was battery problem too!! where you taking stills or movie??

    If you where taking movie we may be able to pick up something maybe
     
  8. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    Good point, Jason. if there is audio that might give a clue.
    Andy.
     
  9. David N Atkisson

    Joined:
    Nov 24, 2012
    Messages:
    73
    Likes Received:
    0
    Sorry guys, I checked the video CF Card. I was training a new Camera Guy and he hadn't switched on the Record function. I thought he was recording the whole time but unfortunately he wasn't. Also I also checked my micro SD on my mx-20 and didn't see any flight data on that card either. I will be much more careful and work out these problems before I fly again. It's weird because the card in the mx-20 was installed correctly and yet I didn't see any flight data on it either.
     
  10. David N Atkisson

    Joined:
    Nov 24, 2012
    Messages:
    73
    Likes Received:
    0
    It strange because when I go to the mx-20 card and click on the LogData folder their are no files in it at all. Was there something I needed to do in the settings of the Graupner to enable it to log the data? If so I need to know so in the future all flights will be logged in.
     
  11. Marc BRACELET

    Marc BRACELET Active Member

    Joined:
    Jul 10, 2012
    Messages:
    204
    Likes Received:
    27
    Sorry for your Cinestar David, i don't know what to tell you (english is not my language) but I can imagine your rage not knowing what happened ...

    Marc.
     
  12. Alex Smart

    Alex Smart Member

    Joined:
    Dec 5, 2012
    Messages:
    100
    Likes Received:
    10
    Hi,
    I m very sorry about the accident Marc...nasty one..
    Im just thinking loud here,if it voltage was low enough to make the okto plunge you would of heard the alarm even after the crash if it woesnt disconected by it,but having another look at the site I see the boards are all over the place..
    Are you using any switches between bateries and FC?
    To me it sounds like a bad conection,maby between boards?
    I would rebuild from scratch involving all solder points and gluing all conection cables..
    May the force be with you!
     
  13. Steve Maller

    Steve Maller UAV Grief Counselor

    Joined:
    Oct 30, 2012
    Messages:
    3,981
    Likes Received:
    807
    David, you were flying with AH on, right? Were you able to shut off AH when she started her descent? The fact that the props were still spinning (to me) indicates that maybe this wasn't a power loss, but rather a serious AH malfunction. I also had a problem once with my transmitter where I was wearing a neck strap that I'd attached with a carabiner, and the carabiner was too big and was resting agains the nick trim on my transmitter. I didn't know what the beeping I was was hearing was, but all I knew was that the quad (what I was flying when this happened) only wanted to nose over instead of taking off. I wonder if it's possible that the throttle trim was bumped or something like that.

    My heart goes out to you, David. This is a terrible situation.
     
  14. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,712
    Likes Received:
    311
    HI David

    I feel your pain. It sounds like a very similar crash to mine if you saw my post and video. I know that you don't have much to go on but I have a few things for you to think about. When you go up in the air, do you engage the motor cut off safety switch via a switch on your MX20 that prevents the motors from being turned off by accident in case you take the left joystick to the down left position? IF you normally do, could you have forgotten to engage it? This is what happened to me although I wasn't in altitude hold. I Yawed left and gassed all the way down and my motors just stopped and the bird fell out of the sky like a brick. Luckily for me I had 3 inches of soft snow on the ground to cushion the blow and I didn't have a camera gimble on. Also another thing I noticed by the pictures is that one of your propeller motors has moved forward on the boom toward the hub (see attachment). This should not happen if they are tightened down really tight even with a crash. Did you use torque seal? If you did I would inspect the torque seal and see if anything looks like they moved. If you didn't use torque seal you can still check all the bolts to see how tight they are. Maybe you had something just tight enough and when it got up in the air the vibration kicked it loose to where it got unstable. Maybe this could have started the bird to go unstable and then you gave it serious gas/yaw corrections that could have cut the motors completely?
     

    Attached Files:

  15. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    David:
    On reflecting about what happened, one possibility, given that your bird was in AH at the time, might be that the air pressure sensor malfunctioned.

    This is pure speculation, of course, but can you have a look at the air pressure sensor and see if any of the solder joints or legs are fractured. I keep thinking, "if the bird's in AH, what could cause a sudden plummet other than a battery failure." The answer might just be that the air pressure sensor failed, indicating that the aircraft was at some arbitrarily high altitude, at which point, what would the flight controller do given that the setpoint altitude to hold was several tens/hundreds of meters below the bogus "indicated" altitude?

    It may be that the impact damage renders my thinking moot -- if you do find that there's a problem with the air pressure sensors legs/solder joints, that could have been impact damage.

    Out of curiosity, was the air pressure sensor hot glued to the FC board and thus less likely to vibrate and stress the legs/solder joints?

    I'll try and make time to look at the Flight Control source code to see how the code would react if the air pressure sensor suddenly shows a dramatic change -- I'm curious to know if there are any "sanity checks" that would prevent a sudden altitude change triggering destructive corrections? For example, if the pressure sensor suddenly were to read at time T, "I'm at 50 meters" and then T + 1 second, "I'm at 500 meters" does the AH logic say, "Holy crap, cut the throttle?" Or does the code effectively react as in "Nahh, that cannot be correct -- the delta altitude just cannot change that fast, the sensor must be wrong?"

    Of course, over-arching this kind of thinking is the fundamental question: When the pressure sensor fails, what readings does it give? Does it show that the bird is down below sea level, or approaching the stratosphere?

    If there are any other forum members with electronics backgrounds, the sensor is the Motorola MPX 4115A. The datasheet is here: http://www.datasheetcatalog.org/datasheet/motorola/MPX4115A.pdf

    Figure 4 suggests that the output voltage is inverse to altitude (the higher you are, the lower the pressure and the lower the voltage).

    The Transfer Function is stated as Vout = Vsupply * (0.009*AirPressure-0.95) +/- Error. Vsupply is presumed to be constant at 5.1v. So, as AirPressure gets smaller, the right hand side of the equation will decrease.

    Thus, I infer that a failure that leads to zero volts output would indicate falsely high altitude. Unless the FC board has sanity check/rate-of-change logic in it, it will indicate sudden altitude gain and thus cause the sudden drop in throttle setting to try and return the bird to its setpoint altitude.

    Right now, I think we might be more accepting of this kind of inexplicable accident, but I have a funny feeling that a few years hence, once the FAA has laid down flight regulations for commercial operations, I think we will either need to be our own accident investigators, or we will be talking to folks from the National Transportation and Safety Board.

    Copters dropping out of the sky is an unacceptable situation and we need to try and get to the bottom of this.

    Bottom line: How badly damaged is your FC board, David? Could you take some images, especially of the pressure sensor legs, please? Could you even connect up your FC board, by itself, to the MK USB adapter and attempt to see what the pressure sensor is reading? You'll need to put the 5v jumper on the MK USB adapter so that it shorts out the two pins.

    If you're uncomfortable doing this, please email me at ajohnsonlaird@me.com and send me the FC board for analysis. I'll publish a report and return the FC board to you ASAP.

    Damn. My copter work is taking on the same characteristics as my day job as forensic software analyst. :)

    Andy.
     
  16. Jei Swan

    Jei Swan Member

    Joined:
    Jan 6, 2013
    Messages:
    102
    Likes Received:
    0
    I actually had a problem similer to this a few days ago. was flying straight line then all of a sudden i loose control of the heli. luckly i was only about 20 ft of the ground and was able to kill it with no damage.
    video of the flight is posted below. its in processing at the time of this post.

    The control loss is near the middle/end of the video

     
  17. Steve Maller

    Steve Maller UAV Grief Counselor

    Joined:
    Oct 30, 2012
    Messages:
    3,981
    Likes Received:
    807
    Very astute, Andy. I hope we can all help David get to the bottom of what happened.
     
  18. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    The video is still showing as unavailable, Jei -- it's now five hours after you posted....so I think something bad has happened.

    Andy.
     
  19. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,712
    Likes Received:
    311
    Andy

    I have 20 years of electronics experience. I started out repairing VHS machines, moved to repairing broadcast equipment, and now oversee AV installations and overall support of college classrooms. I took a look at the information you posted on the sensor and yes the Vout is inverse to altitude. What I found most interesting is that according to the data sheet pins 4, 5, and 6 for Package MPX4115A should not be connected to any external circuitry. But the sensors on the MK FCB have all the pins soldered to the board. Maybe they don't do anything at the board and they just terminate there and maybe they do it just to provide extra strength for the other legs but I found it to be a little odd.

    I understand your thoughts about the pressure sensor, this is why I was so careful to make any mods to my sensor. However I really think that a full inspection can help determine the cause of the wreck. I think coming to the conclusion of sensor failure is very premature and can cause a major panic for others that read this post. There are already many people that have scrapped MK for other options because of the perception that MK electronics are not reliable.

    When I troubleshoot I ALWAYS start with the most obvious things first and then work my way to the least obvious. In my 20 years of troubleshooting here's the way everything always pans out.

    95% of all failures results in from something simple, well documented or obvious.
    5% of all failures result in from odd or random failures or defects.

    I always eliminate the most obvious problems and work my way to the LEAST obvious. 9 out of 10 times its something that is obvious or documented as a problem like (these caps become really leaky from heat due to poor design of where it was placed on the board)

    Let's try to look at the most obvious things first (I see a couple) IF you look at the picture I attached here from David's crash, you can see the motor bracket I point to has moved up the boom off of the end. Even with a crash, if the bolts were tight preflight it's not going to move up the boom. I can see it snapping in half, cracking, or have other signs of physical damage but the motor mount should not move up the boom. From my own experience of disassembling and reassembling the boom clamps need to be dead nuts tight or they will move. I also have noticed that is very easy to tighten one side of the boom clamps too much leaving the other side way too loose. After assembling my own C8 I sure have come to the appreciation that anything that is not completely tight can come back and cause a major crash. That's why I loctite the snot on everything and I double check all the torque on just about every bolt before I go out and fly. I also would strongly look at the connectors. If hot glue wasn't used in several spots on the boards loose connections could have also caused problems. (I can't tell you how many times loose connections are the cause of an electronic issue. Also please excuse my english but I hot glue the piss out of everything. I don't care how pretty it looks there's no way the connectors from the nav and GPS to the FCB are coming loose and I eliminated the molex connector and hardwired it. Nobody will care how pretty my boards look, they will care about the footage and how impressive the stability and reliability of the C8 is. At the very least that molex connector needs to be hotglued because it will come loose fairly easy as I have experimented when I was assembling. All these things need to be carefully inspected one by one and we can only offer suggestions on what to check. I know you know all this as I learned the details from you and your DVD's. Sometimes its easy for us to gravitate to one specific component before looking at the overall picture and systemically eliminating one thing at a time.

    Lastly I know that the bird was in AH at the time of the crash, but AH can be overwritten by the stick commands. Last week I flew in AH and PH and was still moving the bird around with my stick controls because it was much more stable doing that then in freeflight. Could the motor kill safety switch not been engaged or even setup? At my school, over half of the reported problems come from operator error or lack of experience. Not to say this has happened here because I don't know David or his experience but I know its also a factor that shouldn't be eliminated. Even careful and experienced operators make mistakes or overlook a checklist. There is a crapload of things to check and need to remember when you setting up on location, and while you are in flight.

    Sincerely

    Dave
     

    Attached Files:

  20. Jei Swan

    Jei Swan Member

    Joined:
    Jan 6, 2013
    Messages:
    102
    Likes Received:
    0

    Yeah that was my fault put the computer to sleep for the night before it finished up loading. its up and working now.

    The loss of control happens at the 2:54 mark

     

Share This Page