Plane V3.3.0 Crash on takeoff

Hi Guys,

I have been building a flying wing running a Pixhawk. Up until today, I have been very successful with the build and tuning. The machine has been operating well in all modes and both auto takeoff and landing work well. I have followed the online guidance religiously (Zener diode, ESC/BEC for rail power) and modified the power module or use with 6s. All testing has been done without a camera until today. The wings have small digital servos (one in each) and analogue servos for the flaps (one in each). All current draw ratings are below the BEC MAX.

I and using a Sony A6000 with a stratosnapper v2 as the trigger. The camera wifi is off. During an Auto takeoff, as soon as the machine left my hand it held the requested pitch angle of 15 degrees for about half a second, then did a 180 degree roll and crashed. I rebuilt the machine, changed the servo's over, thinking it may have been the compass as I did not do a compass calibration before takeoff. My last compass cal was done about 50Km from this site.

Today I did another launch in MANUAL mode, without the camera. The stratosnapper was plugged in to the pixhawk and external BEC. The aircraft performed flawlessly. We conducted a manual landing.

We put the camera in again and performed a MANUAL takeoff. The same crash happened again, this time to the left. 

I think it could be the stratosnapper, or the way it has been wired. Currently I have only a signal and earth wire on RC port 10 on the pixhawk which connects to SERVO 1 on the SS. The SS then has an external BEC on SERVO 2 providing it with 5.1 Volts. The Sony is then connect on the WIRE port of the SS.

Looking at my data logs for the flight, the Vcc looks to be ok, fluctuating between 5.05 and 4.87.

During both crashes, the CH1 In and CH 1 out appear to be momentarily reversing (going in the wrong direction). The CH1 in and Ch1 Out in other datalogs vary in opposite directions on the graph, but in the two crashes they vary in the same direction momentarily.

I have rewired the SS to be powered by the pixhawk, thus removing the BEC from the SS Servo port 2. I am thinking that this may fix the issue, as this is the way the tutorial suggest it should be wired, I interpreted it incorrectly.

Has anyone had this issue before? Have I overlooked something?

Any help much appreciated!

 

2015-07-23 16-04-53 99.bin.log

2015-07-23 17-10-35 82.bin.log

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

Join diydrones

Email me when people reply –

Replies

  • 97 is the successful flight log

    2015-07-23 16-59-23 97.bin.zip

    https://storage.ning.com/topology/rest/1.0/file/get/3702057509?profile=original
  • Successful flight long is 97

  • Developer

    Hi Chris.  We have looked at your logs.   It looks like the plane is trying not to roll over so we aren't sure.  Any chance you could attach a dataflash log (downloaded from the pixhawk) of a successful takeoff so we can compare the two?

    Thanks, Grant.

    • Thanks Grant, Attaching a successful flight log now. This flight was conducted between the two crash flights, using manual mode for takeoff and landing.

      On the second crash, the machine went in nose first, but I noticed a broken control horn on the right hand elevon (at the time of the crash I thought it could just be post crash damage). I have a suspicion that this may have been cracked and broken just on launch, leaving the left elevon at nearly full up defection and the right elevon flush with the wing, thus causing a rapid roll to the left. If this is the case I just need to figure out why the auto launch on the previous day did a rapid roll to the right. Note that before the auto launch I didn't do a compass or accelerometer calibration (I usually do but didn't on this flight), and I was getting a 'compass variation' warning on the HUD intermittently, but I had the PC volume down so i couldn't hear it. Therefore i'm thinking it could just be a coincidence of the two crashes happening so close together, the first one being just a compass and gyro cal needed and the control horn breaking on the second day.

      Thanks,

      Chris

  • (original)Pixhawk PWM output are buffered, and works of 3,3volts, not the servo rail.  So your camera trigger cannot influence, and certainly not alter another channels PWM.

    what is SS  ? -after 1945 anyway :)

    • Hi Andre

      I just used SS to abbreviate 'stratosnapper', the trigger I am using for my camera. The SS has been plugged into the pixhawk for many flights now, without any issues. Now that I have installed a camera and plugged the SS into it (the camera), the drone is rolling very rapidly on takeoff, with makes me think it has to be th camera trigger causing it.

      My SS has a 5.0 V input from a BEC (castle creations Bec pro). This BEC has two outputs, one is being used for the SS and the other forms the 6S power module I made. The only connection the SS has to the pixhawk is via a signal and earth lead to RC10. Could it be the SS affecting the Bec powering it? Maybe this is causing power module interruptions?

      Thanks for your help

      Chris
      • Like I said, rail power is not related to PWM output - PWM signal is made using 3,3v which, if failed, would result in a brownout, no logging, crash.

        there is no way I see any reversing happen in manual mode, check everything else, especually CG, like gary says. Flying wings are not very good at handling bad CG.

        • The camera is near the I2C splitter, could it be causing interference with that?
        • Hi guys,

          Yeah I have been very carefull with the cg. Before the camera was installed I put ballast in place of it for weight and cg test. Throughout all of my flying the cg has been exactly the same, and the aircraft is well under its max weight
      • Moderator

        You sure its not a CG issue now you have added the camera

This reply was deleted.

Activity