strange behaviour of pixhawk-request help to understand


I am building a 1300mm octocopter with 30AH lipo. I want to be using a tegra tx1 for onboard visual processing coming from a sony a6000. 

I have built the drone and assembled it. I have configured it as usual. I was using the external compass on the m8n gps. and when the compass calibration happened both of them had similar reading all under 150. I have also done a compass motor calibration for the offset.

however when I took it to the field and wanted to test it in autonomous mode, it started behaving weirdly. below is the auto diagnosis of the data log and also please find attached the flight log.

I fail to understand what could have gone wrong. I would appreciate any help solving it. I wish to fly the machine and then work the tegra part.

Size (kb) 856.1298828125
No of lines 10433
Duration 0:00:28
Vehicletype ArduCopter
Firmware Version V3.5.3
Firmware Hash 1a85c237
Hardware Type
Free Mem 0
Skipped Lines 0
Test: Autotune = UNKNOWN - No ATUN log data
Test: Brownout = GOOD -
Test: Compass = FAIL - Large change in mag_field (257.75%)
Max mag field length (1112.76) > recommended (550.00)

Test: Dupe Log Data = GOOD -
Test: Empty = GOOD -
Test: Event/Failsafe = FAIL - ERRs found: CRASH GPS_GLITCH
Test: GPS = UNKNOWN - join() takes exactly one argument (2 given)
Test: IMU Mismatch = WARN - Check vibration or accelerometer calibration. (Mismatch: 0.81, WARN: 0.75, FAIL: 1.50)
Test: Motor Balance = FAIL - Motor channel averages = [1722, 1684, 1360, 1314, 1480, 1441, 1754, 1650]
Average motor output = 1550
Difference between min and max motor averages = 440
Test: NaNs = GOOD -
Test: OpticalFlow = FAIL - FAIL: no optical flow data

Test: Parameters = FAIL - 'THR_MIN' not found
Test: PM = GOOD -
Test: Pitch/Roll = UNKNOWN - 'BarAlt'
Test: Thrust = GOOD -
Test: VCC = WARN - VCC min/max diff 0.563965v, should be <0.3v

thanks in advance 

Views: 224


Replies to This Discussion

I have thoroughly reviewed your log. You have major magnetic interference from your power wiring being too close to your magnetometer. That's why it couldn't fly. I get the felling that this is the first time this machine has flown, so it should never have been put into Auto mode at all,, irresponsible and dangerous considering that it is a large machine hovering at 60 amps. That's a big expensive aircraft.

OK enough of that. Your primary magnetometer in the GPS module needs to be at least 20 cm from induced magnetic fields from high current conductors. That's why you always see them mounted on a mast. Its not for the GPS, it's for the mag. 

I don't know what's going on with your power system, but power to your flight controller was marginal and close to a fatal brownout reset. The voltage dropped when the motors spun up. That should not happen.

Vibration levels looked very good until the props started beating the ground.

Looked like a very short and scary flight. I hoop you learned a lesson. Drones are hard. A lot of things must be right for it to fly, but only one thing wrong can bring it down.

You should first fly in stabilize mode with a new machine. It is safer to get some manual flight experience on a less lethal machine.

If stabilize works then try althold mode. When that is working well and tune you can try loiter mode. If that works, then you can try an auto mission.

Good Luck!

1200 watts to hover , wow !

What is your AUW ?

moriati, please do not attempt to fly this machine until it is fixed. I noticed something. You are using a Pixhawk power module on a big copter and totally overloading it. You are lucky not to have a total power failure. The standard power module is only rated for a 4 cell and only measures up to 60 amps. The data shows you were above 60 amps(flat line at 60). That also explains the vcc drop when the copter stars flying. That is the problem that didn't yet cause a crash. The other is the magnatometer interference. When the current increases the mag readings change before the copter moves. That is the cause of the crash.

The current sensor is a shunt resistor that has a maximum rating of 100 amps. If your machine hovers near 60 amps it will easily exceed 100 amps often during windy or fast flight. You really must redo your power power monitor. And get the mag/gps away from current conductors and your current indication can't go above 60 amps.

Hi Greg

thank you so much for the support. I have actually tested the drone yesterday again. I have moved the compass away from the battery, I have put the battery below the drone and the compass on a mast. and yet it had the same result although I was lucky this time not to have a crash.

regarding the power module, can you suggest any power module I shoud buy. plz find the flight log from yesterdays flight


my AWP now is about 9 Kilograms

Luc Maximilien said:

1200 watts to hover , wow !

What is your AUW ?

I looked at your latest log. There is still a strong magnetic influence that follows current on both mags, one more than the other. there is also a large motor imbalance like one of the motors is not working or loosing sync.

Regarding the power module, you could use this one, from 3DR:

Hi Greg,

thank you so much for the support through out. I apologise for the late reply. I have recently changed my autopilot, Compass, shifted the battery a little further down and moved the autopilot also a little above its previous position. I have performed a bench test tying the drone down to ensure everything is alright before flight, Also as you suggested I have purchased the attopilot 180 Amp current sensor, and coupled it with a BEC to make a power module.I will be bench testing it with the power module changed and uploading the flight log later in the day.

However I am uploading the last flight log when I bench tested the drone after I made reasonable changes. whiel I was going thru the log on auto analyse, I found a couple of things that don't seem right right to me:

Test: Parameters = FAIL - 'THR_MIN' not found

also when I try to arm the drone, it always doesn't respond to the command. I found it weird. it arms at times and at times does not respond to the command even if I hold the sticks centred for a while.
Greg Fletcher said:

I looked at your latest log. There is still a strong magnetic influence that follows current on both mags, one more than the other. there is also a large motor imbalance like one of the motors is not working or loosing sync.


Thanks Javier for the suggestion. I found an interesting 180 amp attopilot current sensor on sparkfun and got that instead. 

Javier Cerna said:

Regarding the power module, you could use this one, from 3DR:


© 2018   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service