Quad is an IRIS with a front mounted GoPro3 and TP g8 light 5000 battery
This was my third suscessfull mission of the day, I was working on establishing the best speed/power draw for long missions. I am going to set this IRIS up for FPV
Simple 5 waypoint mission with a total distance of 1.3km or so, hover over my friends house then return. The problem happend when the quad overshot the landing waypoint by 15meters or so and started a “falling leaf” type decent over the top of my house. I switch out of Auto with full throttle applied, first to loiter, then to stabalize. I was not able to stop the descent, and the IRIS impacted the roof and fell to the ground.
After installing new propellers, one motor was lazy when armed. it would take much more throttle than the other motors to spin up. I ran the ESC calibration and that fixed the lazy motor for now. This is the second time this has happened. Any Ideas??
I have the log file if anyone is interest in reviewing. In the attached graph, you can see the quad was pitching and rolling uncontrollably.
looks like the PWM on chanel 3 saturated several times. I can’t fault the PixHawk, looks like it was doing all it could to maintain flight with a weak leg. When I tried to save it (switching to stabilize and full power) I most likely made things worse. Im not as fast as the flight controller.
Now I am really looking at the ESC. The motor on channel 3 has allready been replaced.
strange, why would the ESC lose its calibration on one motor? It looks like it was doing fine before it started the descent to landing?
I sign up at droneshare.com and the server said that it determined that my flight was too boring. Nothing like a server with an attitude. I will try do something more exciting before I crash next time. 12meters/second is not fast enough. Iv’e gone 17m/s but that is flat out and altitude control is marginal, I will post that one and see if it allows it. Anyhow, I guess an awesome crash into my roof and sliding off into the flower bead is boring. I guess the exciting flights the server is looking for are 4meters/second and below.
Anyhow, If someone would like to help I am will post here and I appreicate your time . I have a video link that will be good for 12 days. the file is large but the end of the video is awesome. The flight is well out of radio range, enjoy.
Thanks for the feedback on Droneshare. I had not expected you to post so quickly as it won’t even take dataflash logs yet. I suspect it made gave that error because it did not load that file… I’ll let Kevin know about the boring comment.
You definitely have your front two motors saturating at full throttle during the descent, but prior to that they were lower than the rear motors.
I’m inclined to say that something went wrong in hardware-land. The pixhawk was outputting nearly full throttle to the front and minimum throttle to the rear to track the pitch demand and it wasn’t succeeding. What is really odd to me is that the roll never stopped tracking and the pitch does this very strange cyclical pattern. I can’t think of an effect that would cause this to happen.
Droneshare is under development and is a major undertaking - be nice
No worries on the droneshare. It is a great idea and I like the sarcasm, just tell me what I did wrong too. People are too stuffy these days, my fault I tried to load the .log not .logt file. Honestly, I am new to the Pixhawk and Adrino controllers, but I have jumped in full force. I have 4 machines running the PixHawk now. I am a 3D heli guy so at first it was very uncomfortable to hit the auto switch. Now, I have two 700 size traditional helis and two 3DR IRIS one with gimgal. So yes I am out of control and addicted to this stuff.
Jschall,
I agree, with your comment. I can’t figure this one out. The PixHawk did a better job than me flying with whatever it was dealing with. Im in Arizona (hot) and the only thing I can think of is the ESC went into a thermal protection mode and ramped down the power?
This IRIS (I call it number 2) had a similar problem right out of the box. On the first flight, I swtched into loiter and all was well for a few seconds, and then it started a fairly rapid messy decent. I switch to stabalize and punched it and it flipped over and broke a leg on the concrete. in the post mortum, the same motor was lazy when armed. I replaced the motor and ran the ESC cal and all seemed normal at first. It soon lost the cal again and I could not lift off without it wanting to flip over (same motor). I sent it back to 3DR and they “fixed” it. I never got a ticket on what was repaired. I now feel I need a new ESC, that was my prognosis and conversation when I sent it back to them. I want to try FPV with this machine (got some fatshark goggles coming) but this one is not ready yet. I might just pull the ESC if 3DR does not help me, and replace it with somthing else. On a positive note, my other IRIS has the Tarot gimble and legs and I have had no problems with it, amazing product.
I flew the IRIS as light as can be with the stock 3700ma battery and no GoPro and no PoPro mount (light as can be). I just hovered it around in loiter. After about 4.5 minutes of flight, it started to have problems holding altitude and became unstable and started to descend. I left it in loiter and made one attempt to climb, but it became more unstable. This time I have learned, and left it in loiter and it land in the grass and disarmed. Motor on channel 3 did the same thing.
IThe powers crossed over like the previous graph posted. I thought it could be a signature of effective translational lift from forward flight similar to a traditional heli, but this is not the case.
I contaced 3D Robotics and they are great, They have been working with me and yesterday they e-mailed me to send it back. They gave me an RMA# and said they will replace the IRIS. Having this data logging capability is so fantastic to identify problems and keep you sanity. I am so glad I chose 3DR
