My video quad keeps having an infuriating major anomaly and as I'm currently vacationing in a very remote spot in Southern Africa I have very limited internet access so is very hard to try debug.
The quad flies perfectly normally in all attitudes and modes for a while but then suddenly goes out of control, tilting briefly then wobbling down with almost no control until it hits the ground, full throttle makes no difference and roll and pitch commands seem to have no effect, if a take-off is attempted immediately after there is not enough power to get off the ground and almost no control. If disarmed and armed again on the ground it'll take off fine and fly OK for a while and then suddenly do the same again again. This is a quad which has had perhaps 200+ flights without showing this behaviour. Power for the APM2 is (and always has been) from the 4 BEC's the Turnigy Super Simple 18A ESC's so I don't believe this to be a brown-out. The out of control bit has happened at earliest 1 minutes and as late as 9 minutes into the flight so I can't think overheating could cause it.
This problem showed up first with versions 2.7.1 and 2.7.2 then I tried with 2.4, 2.6 and 2.7 and its present in all.
The logs (attached) show a few strange things and the out of control bit shows up very clearly.
Can anyone explain what's happening here and how I can get rid of this infuriating behaviour/bug?
Image QC32.1 shows roll_IN and roll of 4 occasions of the problem – you can see how it rolls the wrong way and opposite to my inputs.
Image QC32.2 shows Thr_IN and baro_alt, you can see how the throttle increases to try stop the descent
Image QC37.1 shows another couple of strange traces around a crash.
Image QC36.1 shows normal flying with the sudden divergence at the around the crash.
Replies
Did you manage to solve this issue?
A couple more logs with RAW and MOTORS enabled, look at motor #2 at the end of each log
2012-09-14 14-35 11.log
2012-09-15 15-59 2.log
Chris, I'm using all 4 BEC's in parallel and as I mentioned this quad's components have had 200+ flights but never showing these symptoms.This is a new frame but I started having this problem just before I changed from the old one. I intitially suspected an ESC overheating but seeing as it occurs any time between a minute (example: once early morning) and nine minutes (example: once in midday heat) I ruled that out.
Jason, the only time I reduced throttle is at the point of impact to try limit prop breakage, otherwise I always increase throttle to maintain height as soon as I detect the start of the fall. The quad does seem to stay more or less level as it falls so thankfully prop breakages have been minimal. I had a crash landing this morning (and had to go fetch the quad from very near to a large bull elephant!) where the problem was very pronounced and I was a lot higher up so the increase in throttle will show up longer.
I've done an 'erase' & 'reset' & radio 'calibrate' after every code change. Going back to 2.4 was just desperation as I happened to have it on the laptop I have with me. I will do the CLI tests, report back and post the log from this morning's incident which may be clearer, will also try load 2.7.3. My radio BTW is a Hitec Aurora 9 with Optima 7 2.4 GHz receiver.
Graham,
I can see from the first log that the throttle input keeps dropping to 0. Are you intentionally lowering it?
Can you recalibrate the radio and run the CLI tests for PWM and Radio input?
Jason
That does sound like a power situation, since it shows up on all firmware versions, but just to be sure can you check it on 2.7.3, which is a significant advance in firmware and really seems to be flying without issue?