Inverted Crash during First Autotune Cycle, APM 2.8

Hi all,

My name is Jeremy, and this is my first post. I am totally new to APM, GPS flight controllers, and to this forum. I would like to apologize for not having a genuine 3DR APM 2.6, and being that I have a clone 2.8 version I hope that this doesn’t mean I won’t be able to get help here. I honestly had no idea that 2.8 was not a genuine 3DR product when I purchased it. Forgive me.

I am hoping to get some help in understanding how to read logs and find a cause for my recent catastrophic crash. I can pull up the logs but it seems like some of what data would be useful either only has a dead flat line for the duration of the flight, or there isn’t data for that parameter. I’m hoping someone can make sense of my data file.

I’m trying to not be too devastated by this crash, as I have had my share of quadcopter and heli crashes, so I’ve come to expect these things and I’m just hoping to avoid whatever caused this one in the future. That being said, this one is more disappointing being that with the 4 successful flights I had prior to this crash, I hadn’t even gotten through a successful tuning period to even begin to see the capabilities of this flight controller before I’m in the process of rebuilding from a crash! Ugh…

Onto the crash.

(X525 Quadcopter with extended arms to 680mm with Sunnysky 2216 800kv and APC 12x45 props, Boscam VTX, eBay gimbal, and GoPro on board)

Took off in stabilized mode, selected Alt Hold mode within a few seconds, gained approx 25’ of altitude, then selected Autotune. It began its process and did so for a couple of minutes. The inputs it was making were rather drastic I noticed, quite pronounced. I brought it back near the starting point, at about 25’ probably 3 times, then on the 4th time of it being some distance away I began the input to steer it back once again. At 2:34 into the flight it went completely inverted and fell from about 25 feet. Once it was inverted it didn’t seem to roll any more, just went flat upside down and fell. But I only had about 2 seconds to watch it happen so that’s only what I can remember.

I am suspicious of my on board GoPro possibly causing some interference. Here’s why: On my previous catastrophic crash (same airframe, different motors and KK2.1.1 flight controller), which put the whole craft into a shallow pond, my gimbal mounted GoPro 3 Black was recording video just fine, then at the moment it fell from the sky (in essentially the same exact manner) the video stopped recording suddenly. The same thing is noticed on this crash; rather than having footage of the descent and impact the video just stops the instant before the sudden change in attitude.

Could it be that the GoPro is putting out some sort of signal that just sends these flight controllers out of control? I mean, I, and many others run these tiny cameras on miniquads all the time and I’m not aware of this being a known issue.

Anyway, hopefully someone can decipher the log and see if anything raises any flags. I mean this was during Autotune, but I can’t seem to find any Autotune messages. All the data I’ve been able to graph looks pretty consistent all the way up to the stop point. But that is also puzzling because the APM was still powered on until I unplugged it, some 20 seconds after the impact… They don’t show any sudden change in ATT or ALT from what I can tell. But that’s what I’m probably not understanding in reading the logs.

And the one question I’d ask directly about the log is: does the “time mark” along the bottom of the graph represent a physical time, or what do those numbers mean? All my data ends at one specific point, I’m assuming that’s when I unplug it.

Once home I downloaded the log file and checked to see if the GPS could still see satellites, mostly being concerned because the GPS antenna was the first point of impact. Although the mast was destroyed and the GPS case scratched up a bit it still found 8 satellites under my car port, so that’s a good sign. I think the APM is okay, being that the GPS antenna and my motor nuts took the brunt of the impact, but I’ll have to get into it again to see if the sensor are all working properly.

Thanks for your help. Sorry for the lengthy post. Hope to get to the bottom of this one or at least further my education on this whole Mission Planner and APM system.

Jeremy

To hopefully make it easier I am attaching all the screen shots of the data I can view. You will see some of the data lines are flat, while others don’t seem to make much sense. In the end, however, none of the data lines seem to represent any sort of sudden change in attitude like going inverted an falling from 25’. Maybe that’s because the data isn’t recorded that quickly so it just didn’t show up. I’m not sure. I’m hoping some experts here can recognize a likely cause.

Thanks

This one is Satellites visible. I would expect to see somewhere between 8-14 as I had seen at my house. The scale doesn’t make sense to me.
[attachment=1]Screen Shot 2015-10-23 at 12.04.28 PM.png[/attachment]

More GPS data
[attachment=0]Screen Shot 2015-10-23 at 12.05.38 PM.png[/attachment]

Here we see attitude, which again doesn’t prove to be very useful or an accurate representation.
[attachment=1]Screen Shot 2015-10-23 at 12.06.16 PM.png[/attachment]

AHRS
[attachment=0]Screen Shot 2015-10-23 at 12.06.41 PM.png[/attachment]

Heartbeat
[attachment=1]Screen Shot 2015-10-23 at 12.07.18 PM.png[/attachment]

HW Status
[attachment=0]Screen Shot 2015-10-23 at 12.07.43 PM.png[/attachment]

Nav Control, which I assume doesn’t mean much because I wasn’t flying an autonomous mission.
[attachment=2]Screen Shot 2015-10-23 at 12.08.23 PM.png[/attachment]

RAW IMU. The image after this one is more scaled towards the end of the flight.
[attachment=1]Screen Shot 2015-10-23 at 12.09.36 PM.png[/attachment]

[attachment=0]Screen Shot 2015-10-23 at 12.09.53 PM.png[/attachment]

The RC Channels. The image after this one is more scaled to the end of the flight.

Please note: I am using a DX6i TX with my Ch 5 set to choose between Stabilized and Alt Hold, with my channel 6 switch programmed to place into Autotune when switched and I have that channel from the receiver plugged into CH7 on the APM.
[attachment=1]Screen Shot 2015-10-23 at 12.10.46 PM.png[/attachment]

[attachment=0]Screen Shot 2015-10-23 at 12.11.03 PM.png[/attachment]

These are the sensors. Another graph that isn’t very useful or seems to misrepresent the actual data.
[attachment=2]Screen Shot 2015-10-23 at 12.11.50 PM.png[/attachment]

The following two images show the power, or System Status. All lines are flat except for the Load line, which is represented by a different color on the following two images, but it is the same line.
[attachment=1]Screen Shot 2015-10-23 at 12.12.52 PM.png[/attachment]

[attachment=0]Screen Shot 2015-10-23 at 12.14.07 PM.png[/attachment]

Here is where I was hoping to see some sort of spike or drop from one of the ESC’s or motors or just some indication of abnormality. But, along with the rest of the data the flight appeared to have gone just fine.

That’s why I’m stumped and I’m not really sure any of this data will help me identify the cause of this problem. I might just have to go out and get her flying again and see what happens on the next time it does this.

I’m wondering if the autotune should even be trusted. The 4 prior flights went well, although I needed to tune some PID’s to get it to behave a little better. At this point I will probably baseline it back to stock and start over, assuming the flight controller sensors all work properly as indicated on the Mission Planner Flight Display and so forth.

Sorry for the overly detailed posts, I’m hoping this is a place where this kind of detail is appreciated and comprehended by the forum members, unlike other forums where it is often wasted effort only met by unconstructive criticism and off topic banter.

Thanks to anyone who takes a few minutes to study this event with me. It is greatly appreciated!

Just to round out the conversation I’ve had with myself on these two threads, and to close them outright…

In the end I just ended up getting a used APM 2.5 Hobby King version from a local guy.

I concluded that somehow there was a fault with that specific 2.8 clone board. Whether it was the cause of the crash in the first place, or whether it was a result of the impact I’ll never know.

I’m happy to report that my 2.5 board is working flawlessly and I’m very much looking forward to fine tuning it! Never again will I attempt an auto tune cycle however.

Jeremy A