Brownout fail ?!?

Hi there
I was testing my huge new build quad-octo, the motors are powered by 4 6s 10000mah, and the Pixhawk and electronics by a 3s 5000mah.

I successfully did the compass dance and it flew for some 10 min in Stabilize, AltHold and Loiter.

I was pretty happy beside some sloppiness in Stab, for which I thought I would try to low a bit D, Alt Hold was extremely stable and in Loiter it showed a tendency to move in a circle which I am not sure it is because the GPS is mounted on the frame and gets interference, so I was planning to try move it or to mount it on a mast.

I downloaded the logs and it seemed to me the vibrations are perfectly in specs, DesRoll vs Roll and DesPitch vs Pitch follow each other pretty smoothly.
It came as a shock as I tried to Auto analyze the log

Log File C:/Program Files/Mission Planner/logs/OCTOROTOR/1/2015-10-02 12-41-35.log
Size (kb) 10621.4677734375
No of lines 142870
Duration 0:10:34
Vehicletype ArduCopter
Firmware Version V3.2.1
Firmware Hash 36b405fb
Hardware Type
Free Mem 0
Skipped Lines 0

Test: Autotune = UNKNOWN - No ATUN log data
Test: Balance/Twist = GOOD -
Test: Brownout = FAIL - Truncated Log? Ends while armed at altitude 7.75m
Test: Compass = FAIL - Large change in mag_field (52.11%)

Test: Dupe Log Data = GOOD -
Test: Empty = GOOD -
Test: Event/Failsafe = FAIL - ERR found: FS_BATT
Test: GPS = GOOD -
Test: IMU Mismatch = GOOD - (Mismatch: 0.68, WARN: 0.75, FAIL: 1.50)
Test: Parameters = GOOD -
Test: PM = FAIL - 62 slow loop lines found, max 14.50% on line 838
Test: Pitch/Roll = GOOD -
Test: Thrust = GOOD -
Test: VCC = GOOD -

As you can see it shows a

  • Brownout FAIL which for my understanding it should have resulted in a crash and I never noticed any fast drop of the octo. What could possibly be the cause of this? I still didn’t put the capacitors banks on before the ESCs, would this help?

  • Compass FAIL, could this be because of the GPS mounted directly on the frame?

  • FS_BATT fail - I disabled all failsafe, where does this come from?
    plus from the logs and on the ground it seems to me the batteries still had some power left.

  • PM=fail ?? what is this?

I couldn’t find any explanation on these…
May please someone with a better understanding of the logs, shed some light for me?
I am now very scared to fly it again.

The link to the log below…7aB4CKyga?dl=0

Any assistance is, as usual, MUCH appreciated

First off, I can’t download the log file (says 404 file not found).

Secondly, if I’m not mistaken, a brownout can happen on the main board, and if there is a 5v power source from another method (maybe 5v on the power rail via one of the ESCs?) it can still fly. Also, it could be a bad barometer reading.

Magnetometer (compass) error is likely due to the compass/gps being mounted directly to the frame. I had this exact issue with a large hex I now use as my daily driver and found that mounting off the frame, just by 3 inches, reduced my interference levels from 180% down to an average of about 6%.

The PM (performance management) log errors I see everyone get those. I’ve disabled that logging so it doesn’t scream at me.

I can’t do much more until I can get to the log.

Hi BIgTulsa

Thanks a lot for your reply sorry for the link I tested it and now don’t know why it doesn’t work anymore please try again this link should be working … 5.log?dl=0

In the meantime I checked all the power circuits and added the APM power module so now the +5v and grounds are coming from the apm module (connected to a 3s battery) into the power port of the Pixhawk and the sensors are coming from the ATTOpilot (connected to the 6s batteries) into the power port.
I also kept the UBEC into the out rail as a back up power and added a zener diode.

I am waiting for a mast for the gps and I will elevate it before I fly again.

Yes I did read some posts about the PM so I will ignore it.

I would really appreciate if when you have sometime you could have a look at the log and see if you can find what triggered the battery warning and the brownout, I am still lost about it and want be comfortable in testing that big bird if I don’t have a clue about the issue.


Sorry I read again your answer, for main board you mean the Pixhawk I guess. Which logs should I look at to analyze the power to the board?

The PH was powered only by a UBEC feeding the out rail, without a zener as my understanding is that you only need the diode if there are servos attached to the rail.

There are two PDB on the octo, one distributing 6s batteries to the ESCs which are connected to the PH with only signal and ground cable, and one distributing a 3s battery to PH, OSD, VTX, rx etc each one with its own BEC.

The two PDB have a common ground.

I hope this clarify a bit the setup.
Thanks a million for the assistance

Hi There
I am still getting Brownout Fail in Auto Analyze logs, but cannot notice any anomalous behavour when flying and neither I can’t find in the log any recording of BO.
May please someone have a look at the log and point it to me?

I now modified the powering of the Pixhawk, it gets power from the power module to the power port, and the sensor are connected to the attopilot 180A attached to 6s circuits for the motors.
It also has a UBEC in the out rail and a zener diode.

This is a log of a short flight this morning in my garden, pretty windy condition, I just hovered for 4 min and than let the copter switched on on the ground for some test. … 6.log?dl=0

Any help much appreciated

Check the GPS time, it should ramp linearly, it has steps, this is an indicator of a potential brownout. Though it may just be faulty GPS in your case as VCC seems OK.

The analyser tool reports, i’m guessing no linearity in the GPS time as a brownout fail.


Hi billbonney
Thanks so much for your time and for pointing out those GPS glitches.
It seems to me they happen when the octo is landed and on the ground, while the auto analysis was detecting a brownout at some 5 mt altitude. :confused:

I see from your attachment you use Apm P on a MAc, I can’t find how to Auto Analyze a log on Mac.

Is there any way to read from the log if the Pixhawk has switched its power supply from the power port to the out rail back up power?