AQ board after repair going crazy, dangerous accident

Info and discussion about the original AQ v6 flight controller

AQ board after repair going crazy, dangerous accident

Postby LordPikaChu » Sun Jan 04, 2015 2:41 am

Hello everyone,

so, my AQ board has been repaired by Jussi (MCU replaced) and I've tested it today. Unfortunately the results are miserable (again).

After getting the board from repair it seemed to work again (the lights were up, new firmware flashed without issues, everything working alright). I made the setup as usual (motor mixing, radio type, quick level and MAG calibration etc.) and here's what happened:

Accident one

Testing everything at home, without props. Radio connection is good - motors are arming and powering up with throttle stick. The copter is responding to all radio commands. Suddenly it stops responding to radio inputs so I've checked radio connections - everything is as it should be, the Rx is powered up and connection to the board is okay. I couldn't disarm motors or actually do anything via my Tx, but the lights on the board didn't indicate that anything was wrong with the board - the green light was solid. Power reset made the board responsive again. The copter was then powered up for 20 or more minutes until I finished setting up all Ultra ESCs and everything seemed to work without issues again, so I proceeded with testing the hex handheld with props on.


Accident two

The firmware and ESCs are set up, props mounted. We power up the copter, let it initialize. My friend holds the copter above his head, I arm the motors and proceed to increase the throttle slowly, until the copter starts leveling itself. It all took maybe 10 seconds, and then, suddenly, the copter aggressively rolls 45deg left and applies really high throttle, much higher than it should. It would probably flip over if it wasn't for my friend preventing it from rolling any further. I lower the throttle to 0 - no response from the copter - it is still applying almost full throttle. Disarming the motors doesn't work either. Only powering the copter off completely by pulling the battery connector did the trick.

It's only a luck that I've already lost my confidence in AutoQuad during all previous failures of this FC and we performed the test handheld instead of letting the copter hover by itself. Otherwise we would have crashed expensive equipment.

Now, the funny part - what's in the AQ log? Absolutely nothing, the log just cuts off right before the accident happened. The entire accident was not logged, and that leads me to believe that the AQ board just froze, and before it did it applied random throttle to each motor.

Any ideas what's wrong? I think that the board is fucked up and needs replacement.
LordPikaChu
 
Posts: 93
Joined: Fri Aug 09, 2013 12:42 pm
Location: Poland, Lodz

Re: AQ board after repair going crazy, dangerous accident

Postby JussiH » Sun Jan 04, 2015 5:24 pm

You are gonna have to provide more details, especially what FW version you where using and equally important, provide details about your Radio link

it sounds to me more like you have a problem with your R/C link - most likely some PPM sum setup error or a failure of the Receiver.

I am willing to replace the board, but somehow I dont think this is HW related and I would like to rule out problems with the R/C link first.

Your board was thoroughly tested before shipping it out and logging worked as it should.
JussiH
 
Posts: 2103
Joined: Thu Jun 21, 2012 9:25 pm

Re: AQ board after repair going crazy, dangerous accident

Postby LordPikaChu » Mon Jan 05, 2015 12:50 am

Hello Jussi,

I can give you all the details that you want, as well as video proof that it is indeed the AQ board that is freezing mid-flight.

Today I have been able to reproduce the issue without props 100% of the time. I have a dozen or so videos, but will shoot a few better ones tomorrow and post here.

I am running r442 as firmware version (suggested by download assistant). I am using Graupner MX16 as my Tx with GR16 as Rx in SUMD mode, connected to the AQ via Spektrum port (or is it SBUS port?).

This config has worked without any issues in the past, before I have started having issues with this board.

Basically what happens every time is:

1. I connect the battery to the copter
2. Start telemetry in QGC on a laptop
3. Arm motors
4. Increase throttle
5. Now everything seems to work as it should - I can hear the motors responding to radio inputs*
6. I take the copter with my hands and start moving it around a bit
7. After 10 seconds to 2 minutes the AQ board freezes completely. The telemetry (and SD logging) stops, AQ either applies random throttle values to motors or stops them altogether. The debug LED will either be solid or gone (seems random). The green LED will either stay solid or respond to motor arm/disarm commands, but motors won't (this is a proof that my radio connection works). The AQ board will not respond to any other inputs (radio inputs or moving the copter around) and has to be reset or power cycled to start working again.

Now, in step five I said that everything seems to be running alright, except one other issue, but I think it is separate from the board freezing. When I move the pitch/roll stick around, some motors will respond as they should, but some of them will be commanded to go from low to full throttle several times a second (I can see that in the telemetry motor plot). Since I'm using active freeweling and ESCs can handle such a fast RPM change (especially without props) this results in some motors giving off really loud, low screeching sound and vibrating like hell. Keep in mind that this has all been done without props, so there's no excessive vibration affecting the AQ and I have no idea why it would change the motor RPM so rapidly with amplitude ranging from 200 to 2000 (in telemetry view).

Again, as I said before, I can reproduce these issues 100% of the time so I will post a video explaining everything tomorrow.
LordPikaChu
 
Posts: 93
Joined: Fri Aug 09, 2013 12:42 pm
Location: Poland, Lodz

Re: AQ board after repair going crazy, dangerous accident

Postby LPR » Mon Jan 05, 2015 7:49 am

I had the same problem today while build a few unique small quads. I know that your sure that the motor connection are in the right order but twice today I messed it up and I've build over 50 quads.

When ever I have a strange problem like your having I try to simplify that's wrong by only connecting one motor and then tilt the copter and listen to the motor react.

An easy way to turn off a motor is to change in the motor mix the type of control to Can or PWM.

Good luck

Larry
LPR
 
Posts: 1323
Joined: Tue Jun 26, 2012 1:29 pm
Location: MN, USA

Re: AQ board after repair going crazy, dangerous accident

Postby aBUGSworstnightmare » Mon Jan 05, 2015 9:20 am

LPR wrote:An easy way to turn off a motor is to change in the motor mix the type of control to Can or PWM.


1+ from me too!
This 'easy trick' should make it in the wiki since it will save many many crafts from flipping, props being damaged, fingers being cut....
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

Re: AQ board after repair going crazy, dangerous accident

Postby LordPikaChu » Mon Jan 05, 2015 2:22 pm

Hey guys, thanks for your responses, but I think that you've completely missed the point. I have issues with AutoQuad board freezing (hanging, bugging out) mid-flight, not with motors being misconnected. The motor screeching is a separate problem that I've encountered only now and all motors are connected as they should to the AQ.

Anyways, here's the video that I promised earlier that will show you how the AQ board hangs and will also give a proof that my radio connection is working good:



You can clearly see the board bugging out and being unresponsive. Please follow the telemetry plot on my laptop's screen to know the exact moment when the board freezes. You can also see that the radio is working correctly, as even though the board has bugged out, I can still perform arm/disarm action on my radio and the green LED on the AQ will react to it. The green LED is the only thing that responds to any input, and it only happens sometimes. Other times the green LED won't react and the debug LED will be off. It seems random.

Also the issue is not caused by vibrations - I have recorded two incidents: one where we shake the copter and the board freezes, another one where the copter sits still. In the second example you will see the board freezing in the same moment that Tx throttle is lowered, this is, however a coincidence.
LordPikaChu
 
Posts: 93
Joined: Fri Aug 09, 2013 12:42 pm
Location: Poland, Lodz

Re: AQ board after repair going crazy, dangerous accident

Postby sandmen » Mon Jan 05, 2015 4:25 pm

Are you able to connect SWD?
Or are you able to Debug?

It would be interesting where it hangs.
sandmen
 
Posts: 997
Joined: Fri Jun 22, 2012 7:25 am

Re: AQ board after repair going crazy, dangerous accident

Postby LordPikaChu » Mon Jan 05, 2015 4:30 pm

sandmen wrote:Are you able to connect SWD?
Or are you able to Debug?

It would be interesting where it hangs.


I don't know how to do either of this things. :(
LordPikaChu
 
Posts: 93
Joined: Fri Aug 09, 2013 12:42 pm
Location: Poland, Lodz

Re: AQ board after repair going crazy, dangerous accident

Postby Florian75 » Tue Jan 06, 2015 8:44 am

Hello,

please try an older Firmwareversion. Do you remember wich Firmware runns correct at last on your AQ?

Are you sure, that you have downloaded the correct Firmware for your Board? (AIMU or DIMU ? Board Version 0 or 1 is only for the correct Bat. Voltage.)
Your video shows, that the Controller is running (or the "supervisorTaskCode()"). The green LED reacts to you stick-commands (Arm/Disarm). Do you use CAN or PWM?

regards
Florian
Florian75
 
Posts: 48
Joined: Wed Oct 09, 2013 7:35 am

Re: AQ board after repair going crazy, dangerous accident

Postby LordPikaChu » Tue Jan 06, 2015 12:57 pm

Florian75 wrote:Hello,

please try an older Firmwareversion. Do you remember wich Firmware runns correct at last on your AQ?

Are you sure, that you have downloaded the correct Firmware for your Board? (AIMU or DIMU ? Board Version 0 or 1 is only for the correct Bat. Voltage.)
Your video shows, that the Controller is running (or the "supervisorTaskCode()"). The green LED reacts to you stick-commands (Arm/Disarm). Do you use CAN or PWM?

regards
Florian

Hi Florian,

Yes, I have the correct firmware version for DIMU. I am using PWM.

Actually the green LED is the only thing that seems to run at all, and it only happens sometimes. Other times the green LED won't even react and the DEBUG LED will be off. You can clearly see that the telemetry stops and the board is not even logging to SD card and otherwise not reacting to anything at all.

I will try a different firmware (last time it worked on r440) but still I want to have a board which can run any firmware and which I can trust to not crash my expensive equipment.

I will report back what happens with older firmware.
LordPikaChu
 
Posts: 93
Joined: Fri Aug 09, 2013 12:42 pm
Location: Poland, Lodz

Next

Return to AutoQuad 6 Flight Controller

Who is online

Users browsing this forum: No registered users and 2 guests

cron