Welcome to PhantomPilots.com

Sign up for a weekly email of the latest drone news & information

Post crash mystery

Discussion in 'Phantom 2 Discussion' started by wkf94025, May 4, 2014.

  1. wkf94025

    Joined:
    Jan 19, 2014
    Messages:
    608
    Likes Received:
    8
    Location:
    St. Helena, CA
    Looking for insight from anyone out there that can help me understand what is going on with my new P2 post crash. I took delivery of the new P2 with H3-3D a month ago. Assembled fine, and other than gimbal tuning, had no issues. Got some great footage in Cabo before first mishap: following startup, while P2 was warming up in its famous TBE circle, I looked away to adjust the Lilliput on tripod, and ship TBE'd into an oak tree ~40 feet up. Tumbled hard, bent gimbal main arm quite a bit, and then sat on bench for a week while I flew backup old P2+H3-2D. Got the H3-3D bent carefully and precisely back in alignment (somewhat to my surprise), and resurrected new P2+H3-3D config. 4 new props, deep IMU calibration, compass dance, home lock, etc., etc. Bird flew very erratically and I ended in a low altitude / low impact tumble in the grass. I have since logged half a dozen such prop-busters over the past 24H trying to figure this mystery out. I stripped off the H3-3D, and reverted to original gains in P2 Assistant. Still somewhat erratic. At times, wouldn't hold altitude. Played with gains (default and factory), with and without H3-3D, and even using compass from my old P2 (deep IMU and compass dance after that change). Upgraded from nearly latest FW to latest FW. No changes in behavior with that. Latest flight was with default gains, no gimbal on board, but still carrying iOSD Mini and 600mW ImmersionRC VTx; Ground Station installed, but not powered, and GS antennas taped to hull out of the way of other antenna. Even flew the old P2 with Futaba radio to insure it wasn't a problem with Futuba. Old P2 rock solid hover and behavior. New P2 erratic in both GPS and ATTI mode, though my sense is more skittish in GPS. I opened the shell last night and visually inspected for broken solders, loose connections, etc. Nothing found.

    Three clues for you Aviation Clouseau's to consider: (1) sometimes I am getting compass warning when I connect to Assistant 2.0; all 3 compass values are < 500, and the mod is in the mid teens range; (2) there are times when it simply wont hold altitude; I apologize I didn't record payload and gains that resulted in those can't-hover conditions; can retest this afternoon with more logging; (3) when hovering (or trying to hover), when I then apply forward left stick to climb, the new P2 acts as if it's not getting equal lift from rear left prop/motor. The ship jumps southwest as it tries to climb. Very curious. It will climb, and it will stabilize eventually. I tried descending several times to see if there was asymmetry there, but none. Straight down level and smoothly. The asymmetrical climb happens in both GPS and ATTI modes. I check all four motors post flight, and didn't find any noticeable difference. I suspect the problem is all "intellectual" not "physical", meaning not a weak motor or marginal ESC solder. Very curious, eh? None of this behavior was there before the 40' fall from grace, and now it's consistently erratic. With no gimbal on board, and factory gains, I can maintain reasonable hover and small maneuvers nearby. Trying to keep flights at low altitude over soft surfaces, for obvious reasons. I have gone through 8 props since the fun began yesterday.

    Thoughts on other experiments later today: (1) swap SW motor and/or ESC with NE motor and/or ESC to see if behavior remains same, or moves to NE from SW; (2) swap GPS module from old P2 to new P2; (3) degauss compass, then repeat deep IMU, compass dance, etc.

    Appreciate any theories as to where the ghosts in the machine are, and any suggestions as to additional experiments to isolate the source. Worst worst case I will dive into the H3-3D install in the old P2, and ship new P2 off to the experts.

    Kelly
     
  2. goldfishrock

    Joined:
    Jan 7, 2014
    Messages:
    427
    Likes Received:
    0
    Location:
    Devon, UK
    Hi Kelly, sorry to hear about all the problems you've been getting. Unfortunately having fun with these things carries the risk of something like this happening every now and then and it can be fustrating trying to rectify any damage.

    I have a crash a couple of months ago and it took me weeks to get the bottom of the problems caused by it. Have you checked the motors for any sign of damage? One of the other members on this forum gave me some advice about checking motors.... just run your P2 up and throttle the motors up and down for a while, then shutdown and check each one with your fingers (or a laser thermometer) to see if one is significantly warmer than the others. This can indicate a problem.

    After many hours of going over my P2 I found one of my problems to be a cracked engine mount (something I discovered by chance). Take the top of the P2, hold each arm firmly in one hand and gently bend the motor one way and then another, all the time watching around the holes the motor bolts are located in. On mine I had two cracks that under normal circumstances weren't visible and only appeared when the airframe was genly maipulated. It's also worth checking the rest of the airframe using this method to look for any other cracks that may have propogated from weak spots.

    You may also want to check the GPS modile in the lid. The actual antenna has a ceramic ring around it which can sometimes crack upon impact if it's hard enough.

    The other problem I found was on the main board, it seems that one of the tracks that supplied motor power had lifted slightly under the solder joint meaning that in theory the +ve and -ve could have come together (albeit for a split second) which would be enough to cause majoy problems in flight.

    I the end my repairs were extensive. I had a new shell, new Main Board, 2 new ESCs and two motors.... but now it's back and happy. I hope you are able to get to the bottom of this :) Keep is posted hey?
     
  3. wkf94025

    Joined:
    Jan 19, 2014
    Messages:
    608
    Likes Received:
    8
    Location:
    St. Helena, CA
    GFR,

    Thanks for those tips and insights. I will try the warm motor check on the bench, then inspect airframe and solder joints carefully. After that, probably diagonal motor swap followed by flight test. Fingers crossed thisisn't weeks of debugging.

    Kelly
     
  4. ianwood

    ianwood Taco Wrangler
    Staff Member

    Joined:
    Jan 7, 2014
    Messages:
    4,929
    Likes Received:
    1,800
    Location:
    Lost Angeles
    As you suggested already, you can swap out parts to help diagnose the issue but my guess is you have a damaged motor bearing or lose e-clip. Check for play in the motor bell housings, either side-to-side or up and down. Do they feel crunchy when you turn them by hand? And as GFR suggests, check for one motor being hotter than the others. I'd put the gains back to default.
     
  5. wkf94025

    Joined:
    Jan 19, 2014
    Messages:
    608
    Likes Received:
    8
    Location:
    St. Helena, CA
    Thanks Ian.. What is an e-clip?

    Kelly
     
  6. ianwood

    ianwood Taco Wrangler
    Staff Member

    Joined:
    Jan 7, 2014
    Messages:
    4,929
    Likes Received:
    1,800
    Location:
    Lost Angeles
    At the bottom of the motor, there is a 3mm clip on the end of the main shaft that holds the bell housing to the rest of the motor and keeps the bottom bearing in place. Common for these to come off after crashes.
     
  7. EMCSQUAR

    Joined:
    Sep 5, 2013
    Messages:
    1,692
    Likes Received:
    6
    Location:
    N 44.895 W 93.354 Minnesota
    Kelly,

    Things I would check:
    a) first RELOAD default settings from DJI
    b) check craft orientation - make sure it's set for quad - not some other config. (I've had this happen after a reset)
    c) do another advance cal & STICK recal (compass #s between 1200 -1700s)
    d) compass recal/dance
    e) double check your settings in your Futaba (I find myself always bumping digital trim buttons)

    from the symptoms it almost sounds like a GPS or compass conflict
     
  8. wkf94025

    Joined:
    Jan 19, 2014
    Messages:
    608
    Likes Received:
    8
    Location:
    St. Helena, CA
    Thanks Gents. Will employ both your suggestions in the debug jihad. I did rule out (e) in EMCSQUAR's post by flying my other P2 with the Futaba, and all was well there. If I don't find the source soon, I will upload a video of the behavior. It's quite amusing, and consistent.

    Kelly