Hi There,
I need a bit of help looking through one of my logs, as I had a crash a couple of days ago due to a single motor stopping during landing, but I can't work out why it should have occured (the motor is seemingly working fine now!!)...
A bit of background... I am doing a PhD, and part of it is using a UAV to monitor invasive species, in this case Rhododendron within woodlands. I am using a 3DRY6 with a pixhawk, standard 3DR accessories etc, Sunnysky 980kvs, nothing special really, but I am carrying a fairly heavy camera payload (twin cameras, ~500g) and using two 5200 3S multistar batteries.
The aircraft has been auto tuned and run various missions already without incident, but I think it is a bit overtuned (definitely some twitching going on just beore takeoff), but it seemed to be flying fine and I had not had the time to go back and tone it down.
So to the mission... I was flying a series of waypoints at 100m AGL, taking pictures along the way with digicam control (see below, landing point marked in yellow) and two things were noticed during the mission that were not as expected.
First off it was slow to fix its YAW to point towards its intended heading for the first series of waypoints, as can be seen within the yellow circle below:
Secondly and more importantly, at the end of the mission and at point of initiating its landing decent (so at 100m above the landing zone), one of the motors stopped and stayed stopped. However the aircraft descened seemingly under control and at first I thought it would land just fine, but once it reached the 10m alt mark it pitched forwards, clipped a tree branch and landed on its back doing various bits of (thankfully) minor damage.
As far as I could work out visually at the time it was the rear lower motor (so number 4) that had stopped. However in testing after the event (just using the motors command from Mission Planner), all motors appear to be working, and I can't see anything in the logs as to why a motor would have stopped. RCOUT commands seem to all be fine, current doesn't change much (note the power module is only recording against one of the batteries, so current should be doubled for real values) ... so a bit stuck really.
One thing that could have caused this is the batteries, as allthough being almost brand new they have a massive volt drop on them which is annoying, but has not caused issue before (I have had mission planner indicate voltage of 8.5 and yet the aircraft is flying fine with still with more than 20% capacity left after the flight) ... Could one of the 20A Afro Slim ESC's cutout due to it thinking low volts?
I am not certain however if this low volts is just a configuration issue with mission planner as the hardware battery monitors are not warning me (set at 3.3 volts), even when mission planner is indicating much lower voltages (something else I need to find the time to re-check).
So, sorry for the above essay, but if anyone could take a peak at the attached log and work out why one of the motors seemed to stop around the point of starting its landing decent, that would be great :)
Thanks,
S.
ps: I did get some good data though despite the crash!
Replies
Simon,
You've done some good analysis on the logs it seems. I struggle to find more answers but..
It's hard to say exactly why the motor failed but the vehicle has/had quite badly imbalanced output the motors even before that. The top motor (#6) is near full throttle while the bottom one (#3) is near it's minimum. From looking at the end of the log I agree it's #4 that fails (it's throttle goes to full which is the flight controller trying to get the most out of that motor) so I don't think the motor failure was because it was being run too hard during the flight.
As you said, the battery voltage is also quite low at the end, dropping to 9V. It drops this low earlier in the flight as well which may just mean the battery's C value is too low and should be replaced with a higher "c" number battery. I'll indirectly ping the 3DR RTF guys to ask them to double check their choice of batteries for these kits although it's really up to them.
The ESC cut-off voltage is a very strong possibility. The other items are the ESCs not calibrated or the THR_MIN is too low. Maybe it's a combination of low-voltage plus thr-min meaning a motor doesn't spin. So the motors spin fine when the voltage is high but the thr-min is too low when the voltage is low.
Nice to see a copter used for this kind of research.
Thanks for taking a look Randy, and yeah I am thinking it was most likely an ESC failure or perhaps auto cutoff... I will replace that one ESC just in case. I had manually calibrated the ESC's but I will re-calibrate this time through the pixhawk itself rather than just straight off of the rc receiever. I will also increase the THR_MIN from 130 to 250.
I can see now from the RCOUT that the tuning was very bad and I had not realised that motor 6 was running so high in relation to the others... its even clipping its maxium output for a while and that explains the poor YAW behavoir at the begining of the flight (see in yellow below).
Its annoying about those batteries as they are low C ones (10C supposedly) and seemed good value with weight vs mAh. As I am using them in parallel I was not expecting any issues, as at max draw with all motors at full throttle ~90 amps is required, and at 10C those bateries should have be able to supply ~52 amps each without issue (and with a 10 second burst at ~104 amps each) ... I was not expecting such a large voltage drop, but I can't change them at the moment as I don't have the funds for that right now.
I also definetly need to re-tune and this time make sure that its actually balanced (the weather gods need to give me a few more low wind days!!). I am probably trying to pull a little too much weight with this little UAV but it is essentially giving me a cheap aerial platform so I can get results like the below when using a couple of (very cheap) second hand, hacked Canon cameras :)
Thanks again,
S.
Hmm, saying that just identified the below as well. It seems that once the aircraft reached its take off point during the RTL phase of the mission,it then intitated a heading change (by the looks of it to match the heading it had during takeoff). This combined with low volts etc could well have been the catalyst for why the motor stopped.
Do you happen to know the setting for why Pixhawk makes this final heading change, as I have just checked and this does not occur with my APM 2.5 quad using the same firmware (3.2), as it simply keeps it last heading as per the final waypoint ...
Regards,
S.