[solved]Help in Autotune crash - what went wrong

Hi,

Test flight after upgrade to Copter 3.6.10 (with a complete wipe out of params).

After initiating autotune, everything was fine until the quad just plunged into the ground.

vehicle:
680mm frame
2.6kg auw
380kv motors
T-Motor F55A 4in1 ESC using DShot 150 (rcout9-12)
15x5 folding props
6s power

Any help in analyzing the log would be much appreciated.

Thanks,
Gal

Hi @gnitzan
in your log everything looks good and unfortunately there is no any data of motors status (RCOUT)
does your copter suddenly started turning around ? or losing altitude to a side with a angle more tan 40 degree ?

1 Like

Hi @hosein_gh,
Actually, there is a lot of data, sorry I forgot to mention I use T-Motor F55A 4in1 ESC using DShot 150 so RCOut is 09-12.

Also available is ESC1-ESC4 telemetry, I just don’t understand what I am looking at.

another view using APM Planner

it seems that there is a motor failure when your copter was in 4 meter altitude

1 Like

Hi @hosein_gh,

I believe you are right.

I have plotted the RCOut of motor 1 vs ESC1.RPM and it seems like the motor rpm is not rising. Let me know if my reading is correct.

you are corrent @gnitzan
also after replacing burned motors or ESC try to test all your motors and ESCs in mission planner motor test to make sure all of your motors are ok

1 Like

Thanks @hosein_gh, actually I did motor test last night and all was well with no problems, I set test up to 50%.

Actually, I am not 100% sure what went wrong, because from telemetry, the motor was not stressed out, its temperature was like the other motors, so maybe ESC?

I will take it apart and check everything.

I have done nothing at home except removing the props and broken legs.

Here is some test I did which shows nothing is wrong, any input would be highly appreciated
Test on workbench

Without propellers you can’t put you motors under full pressure

1 Like

Hi, I have tested with props on (upside down and rotated like for comp motor calibration) and all seems fine.
Log seems fine

so it seems I am still at square one.

I did a video in my test, it is still uploading to DropBox, I shall link to it when ready.

yes now your log is good
so check all ESCs and motors wires and do a test flight with 1m altitude

1 Like

ok, thank you @hosein_gh for all your help, I shall test it tomorrow.

Agree with motor- esc failure;
if you cant find it; what I do: don’t try autotune until you be shure your failure was solved, put an extra protect thing under your quad, a X like student pilots use or a soft extra improvised legs near the motors to avoid flip and broke props and then try to fly near the floor over soft terrain for some time loitering, then doing soft movements pitch and roll, use all the flying time your batt allow, see what happens and share your log to see if any of Us can detect something after the next autotune, good luck.

1 Like

For all who are interested.
I did yesterday in perfect weather a Autotune with a newly build Hexacopter ( 4kg).
FW: ArduCopter V3.6.10 FC: Pixhawk 1.0
Carefully balanced and trimmed, but using default PID,s
Takeoff in stabilize, switch into PosHold and than Autotune. Setup not aggressive but all 3 axis in one go.
Total time exactly 10 minutes. After saving the Autotune values I tested the Hexacopter in all modes and it is flying perfectly.
Need a test in windy conditions only.
I cannot find any problem with AC v 3.6.10.

2 Likes

I am relatively new at analysis, but does the log analyzer reveal anything of interest to your initial experience? I note the change of mag field, or is this perhaps an artifact from the crash?

log

1 Like

Hi Jack, actually I didn’t look at the log analysis, my bad. As you mentioned, it’s probably from the crash since I did a calibration right before takeoff.
However, THR_MIN not found and current not available I will. Anyway, good catch.
Thank you
Gal

Crashed again.

I was flying in LOITER at about 1-2 meters, very slow. Initially pitching and rolling harder and harder without any problems.

The second I touched yaw, it dropped to the ground. Luckily because of the height, nothing broke.

Attached my log on Dropbox.

Any insight would be appreciated.
Gal.

1 Like

yes unfortunately your motor one stopped after changing yaw

lets test this problem on ground
arm motors in stabilize mode and rise throttle a little bit (we are not going to fly) when your motors rpm rised , change yaw stick and see whats happening (motor on will stop sniping or not)

1 Like

Hi @hosein_gh, something doesn’t make sense to me. Rcout is the pwm command sent to the motor, is it not?
In that logic, why did the fc sent this command?

Or I am totally don’t understand?

Thanks

RCIN is the pwm input by your radio and RCOUT is pwn output by FC to ESCs that trimmed by radio calibration (and some other params only in auto modes)

in the picture above when a motor or ESC stopped working FC automatically will put burned,stopped or damaged motor to max throttle and opposite motor to lowest throttle to keep the balance of copter but this method are not working with a quadcopter but in hexacopter and octocopter working well for example if one of the motors of a hexacopter stop working FC will stop the opposite motor and now hexacopter is converted to a quadcopter and can fly and land without problem

1 Like