Could someone point out what went wrong with my flight today?
I’ve had three successful flight before this and during my last flight, quad copter went into a crazy side drive and ended up breaking two of its arms.
I flew three flights prior to this and they all finished perfect. Upon finishing the last flight, I took Iris+ into my truck, drove to a point where I planned on doing next flight and let it sit out in the open for over 5 mins,
Meanwhile, I uploaded the mission and then I went on to remove the used battery, plug in a new one and wait for a GPS fix again. Once I’ve had green light, I then flew the drone, put it in loiter mode and ensured that it remained in fixed position and then let it go towards the designed waypoints. This is when something went off and it crashed.
Size (kb) 20944.3720703125
No of lines 285547
Duration 0:08:48
Vehicletype ArduCopter
Firmware Version V3.2-Iris
Firmware Hash b212c020
Hardware Type
Free Mem 0
Skipped Lines 0
Test: Autotune = NA -
Test: Balance/Twist = NA -
Test: Brownout = GOOD -
Test: Compass = FAIL - Large change in mag_field (119.59%)
Max mag field length (587.89) > recommended (550.00)
Test: Dupe Log Data = GOOD -
Test: Empty = GOOD -
Test: Event/Failsafe = FAIL - ERRs found: CRASH GPS_GLITCH
Test: GPS = FAIL - GPS glitch errors found (14)
Min satellites: 4, Max HDop: 52.24
Test: IMU Mismatch = GOOD - (Mismatch: 0.58, WARN: 0.75, FAIL: 1.50)
Test: Parameters = GOOD -
Test: PM = NA -
Test: Pitch/Roll = NA -
Test: Thrust = NA -
Test: VCC = GOOD -
Could someone point out what could’ve gone wrong with my setup?
BIN File is too big to be uploaded here, dropbox link below
It’s really tough to say, but looking at this log I see quite a few Baro errors and GPS glitch errors. Had this Iris been in a crash before?
Looking at the auto-analysis, the first thing that really jumps out at me is the compass error. Large change in mag field. That can for sure cause problems with autonomous flight. Your vibrations look ok (I would expect that since it is a pre-fab copter after all).
All those GPS and Baro errors concern me though. Were you around power lines? A lot of metal? Had you done any compass calibrating recently?
It appears to me using APM Log File Analyser this is series of events:
You armed in stablize took off, then switched to loiter after 10 seconds.
After that, you wait 6 seconds and arm auto mode.
You navigate 7 waypoints in about 1:40, then switched to loiter again (you may or may not have done this; I’m just analysing the log file).
At this point the shit starts hitting the fan.
I see a baro glitch that clears almost immediately, however it triggers a failsafe RTL it appears.
Thirty seconds after this, I see a GPS glitch. Clears itself almost immediately but the failsafe kicks in again
Then it appears you fly for another five minutes or so with nothing but constant GPS glitches. I don’t see anywhere where it went into failsafe. However, the program I use to parse the file indicates the log file might be corrupt. It was able to handle the over 30,000 lines of log it considered error. After that five minutes, the flight terminates with a crash.
Since this is an Iris, it’s tough for me to know what is normal for it. I’m sure there are many people on here who can read this log better than I, but at first glance the compass issue is a big problem, and then obviously the baro and GPS glitches. That will sure cause a crash if it cannot engage failsafe properly.
Also, if there is a GPS glitch, your best bet is to have failsafe set to Land instead of RTL (my assumption based on the logs is that you have it set to RTL). Obviously a GPS glitch will not be saved by an RTL since the GPS would be needed to guide you back home. Land or AltHold are better options.
If someone else reads something different, please, chime in. I’m by no means a great log file analyzer but I am getting better.
I was flying over an open pit area and I did 3 flights pior to this one. They worked out solid.
Your flight analysis is correct, except, after 7 waypoints navigation, the drone went in a crazy side dive and to prevent that I switched it to loiter (hoping that it would recover itself). This is when everything went sideways and it decided to land in a bush(not the pit thank god)
The last 5 minutes of flight is actually where it crashed. It landed with GPS facing up and it was able to get a solid 3d lock for me to find it.
One thing I do notice is that AHRS_EKF_USE was set to 0 and I wasn’t utilizing the advanced methods at detecting faults. And like you said, I’ll go ahead and switch the failsafe action to LAND. That may help save the drone in near future.
I have two more IRIS+ as a backup and I am scheduled to wrap up the rest of the work at the site tomorrow.
Hoping that now updated settings will aid in keeping the drone safe.
Also, one thing I noticed is a flight prior to the main flight. When I installed a new battery and flew it, the IRIS went around in circles this is when I put it back in stabilized mode and manually landed it. It looked exactly similar to what this video describes when AHRS_EKF_USE is set to 0.
All in all, I strongly believe that if AHRS_EKF_USE was set to 1 then I could’ve prevented the crash from happening.
Let me know if this agrees with your theory. The previous (short) flight log is attached, hoping that it confirms with the conclusions we’ve come up with.186.BIN (899.4 KB)
Can you update the Iris to 3.3? It requires AHRS_EKF_USE to be 1 (actually it’s set to that and can’t be changed). If not, then obviously changing to 1 on the 3.2 is perfectly fine. After seeing that video I would for sure suggest that enabling it would be a good idea. I think on 3.2 the failsafe for that is called GPS_FAILSAFE as opposed to later versions where they changed it to EKF_FAILSAFE (same parameter pretty much but new name for it).
The problem with Auto Analysis is that it analized everything including the crash itself. All of the error occured after it hit the ground so those don’t mean anything. We need to consentrate between 55 and 65 after waypoint 5 to see what happened.
I can see the copter was facing West and that it was flying East at a good pace. Motor 3 was at max for a short time here at which point you switched to Loiter mode. At the same time you raised the Throttle to help it stay in the air but it continued to fall. The copter had good voltage and current at the time so it looks like motor issue at this point.
Yes that sounds like what I did when I saw Iris go off the waypoint course and decrease in altitude. Loiter and Throttle were two commands that I thought may make the drone recover but, if it was a motor issue then I guess it was unavoidable. Could it be either motor or one of those 4 in 1 ESCs? I’ll perhaps disassemble the drone and see for any burn marks on the ESC board.
I have the functioning wreckage. Front right motor is completely torn apart but the other 3 managed to function alright.
I’ll give it a quick run later today to see if they are functional.
Here is the picture of the fallen one. It served well
