Pixhawk takeoff problems

Hi everyone,

Got into problem with my pixhawk and cannot find what happened.

Problem is that on take of seems like ESC descalibrating and motor starts to spin different speeds, drones likely will overturn so i need to cut emergency motor stop…

I did mane things like:

  1. everything wired correctly for sure, as it was working fine, i do changed bearing for one T-motor U5 (it is very unlikely but it happened next day…) and next day went to film and was unable to take off.

  2. recalibrated esc lot of times, everything went ok on small speeds with propellers or full speed without them everything spinning as must, until i do restart and try to take off

  3. same problem on stabile, alt hold, pos hold modes.

  4. i do even changed esc from simon-k to hobby wings x-roto pro, to be sure everything is compatible.

  5. i upgraded firmware to latest one but no change on this…

now passed 2 weeks, i desoldered everything from scratch everything checked but no luck…

Any ideas?

Dainius

@Infinitadrones,
Please provide tlogs and/or dataflash logs to help troubleshoot your issues.
Regards,
TCIII GM

Thanks for getting back!

Attaching log just downloaded last one.

Dainius

Seems like found a problem, must be lot of vibrations after bearing change… If of course it can be, will try to put old bearing or in worst case order new motor…

1 Like

Seems like something wrong anyway, as if i go to calibration mode, everything spins perfect and even tried with props on until getting to takeoff, no any anomaly occurred, but if i try in stabilise mode after arming all motors they all start to spin perfectly, but if i starting to rise throttle after reaching maybe 5-10% of throttle one motor stops (and not one which was repaired…).

So case remains still open, as no solution and whats wrong yet…

Dainius

One more log, i did more aggressive and less aggressive try to takeoff, but same remained.

Will be great if someone will give feedback and maybe possible solutions.

Also will try to put old simple nasa v2 to see if everything apart from controller can be problem or it will stay with pixhawk…

The attached logs are not valid…

Mike

I haven’t looked at the logs but my guess is that the ESCs have a slightly larger than normal deadzone at the bottom. Try increasing the MOT_SPIN_ARMED value from 70 to something higher like 90 or 100 (don’t go above 130).

If the motors stop spinning once the pilot throttle is above zero then the THR_MIN may also need to be increased from it’s default of 130. Maybe try 150 or 170. Going over 200 shouldn’t be necessary.

[quote=“iseries”]The attached logs are not valid…

Mike[/quote]

Strange as i can read them in APM Planner v2.0.19 rc4. How i need to export so you can read them?

Dainius

[quote=“rmackay9”]I haven’t looked at the logs but my guess is that the ESCs have a slightly larger than normal deadzone at the bottom. Try increasing the MOT_SPIN_ARMED value from 70 to something higher like 90 or 100 (don’t go above 130).

If the motors stop spinning once the pilot throttle is above zero then the THR_MIN may also need to be increased from it’s default of 130. Maybe try 150 or 170. Going over 200 shouldn’t be necessary.[/quote]

Looks like there is no more MOT_SPIN_ARMED in 3.3.1 arducopter or i am wrong?

Still no luck with moving forward.

Attaching video how it happens and at same print screen with, acc for vibrations, raw output and throtle channel.

https://youtu.be/cyldoVs3YWw

Are you 100% sure the motors / blades are the in the correct configuration. I know video makes blades look odd due to shutter timing but the initial startup in slow motion rely looks as if they are wrong.

Also it it the FC shutting the motors off at the point it powers up or is this you? (OK read entire post again, emergency stop)

Hello,

I haven’t analysed the log file ( can’t right now ) but by the video it looks like motor #1 ( if the copter was facing away from you… not sure ) is not producing enough thrust.

Did you recalibrate your escs ? Do you have a bench where you can measure the thrust and see if they are all the same?

Did you also calibrate the accelerometers, by clicking on “Calibrate accell” and “Calibrate level” in mission planner?

If you move your right stick when the throttle is about to takeoff , would the copter respond right to your inputs ( roll / pitch ) ?

If you did all the tests, and you are pretty sure that all the setup are right, I would try to takeoff by compensating the movement using the right stick for that.

If you consider to try that, you need to be confident in your flight skills and need to have a place that would minimize any damage to the model ( i like soft glass or mud) and to yourself , in case something bad happens.

If you are able to takeoff this way, then you could trim the model to be level and hover still. After that, next take off would be ok.

Other thing… you said that you have upgraded the firmware. What was the version you were running previously?

Did you changed your pids? last time you try to takeoff , it looks like your copter were starting to build high frequency vibrations ( mostly by the sound… )

Regards,

  1. MOT_SPIN_ARMED is there @ default 70
  2. The log you posted is incomplete / corrupt Can you re-post it
  3. You will have a newer flight log now I guess, try posting that
  4. The screen shot is unclear, cant reed the data, can you re-post at an attachment.
  5. Blades, are they new? i.e. have you flown with them before. From the video they look awfully like HK carbon blades. (correct me if I’m wrong) if they are the ones I think they are they kill the U5 motors. They cause them to ‘lose step’ and play havoc with the ESCs. Also they are 48% less efficient than the T-motor blades. (I have done a LOT of testing)
  6. From video it also “sounds” like PIDs are too high for those motors / blades. Again I may be wrong at its only a quick snip of sound just before you kill motors.
  7. Fatly Pixhawk? Unlikely, they (in my experience) are reliable; the pre-arm checks should catch any drastic failures. (256 hours of flight logged with them over 6 rigs, and only one crash so far, which appears to be a power fail)
  8. Any more info you could provide albeit insignificant in your mined could reveal something.
  9. Bearing fail on a u5! How many hours has it flown? Unbalanced props?

edit( 6:-[quote]tabascoz:- Did you changed your pids? last time you try to takeoff , it looks like your copter were starting to build high frequency vibrations ( mostly by the sound… ) [/quote]

[quote=“ACPUK”]Are you 100% sure the motors / blades are the in the correct configuration. I know video makes blades look odd due to shutter timing but the initial startup in slow motion rely looks as if they are wrong.

Also it it the FC shutting the motors off at the point it powers up or is this you? (OK read entire post again, emergency stop)[/quote]

Hi, 100% sure they really rotate all like must be, yes in video it is hard to find out real movement, but they all rotate as it must be, same configuration was everything ok, until one morning i went to filming and this started, if i do takeoff in Stabilise mode it gets even more worse…

arducopter before upgrade was 3.2.9 rc4 i guess. not 100 % sure but not 3.3.x for sure.

Yes it seems that one or other motor is not producing enough power, but most strange that it happens sometimes with one another try with different motor.

All esc recalibrate and everytime it happens i do it again.

Accelerometers calibrated to.

I am mostly sure that if i will do hard takeoff it will go up and hopefully under control, but anyway it is somewhere problem i do need to find, as do use it for work and with this state this drone sits on shelf…

Will try today a bit later (now only 6am in Spain) with different modes, log everything and again upload, as it is strange but another thing, one day i can open .bin logs and analyse them, another day it opens empty…

Dainius

[quote=“tabascoz”]Hello,

I haven’t analysed the log file ( can’t right now ) but by the video it looks like motor #1 ( if the copter was facing away from you… not sure ) is not producing enough thrust.

Did you recalibrate your escs ? Do you have a bench where you can measure the thrust and see if they are all the same?

Did you also calibrate the accelerometers, by clicking on “Calibrate accell” and “Calibrate level” in mission planner?

If you move your right stick when the throttle is about to takeoff , would the copter respond right to your inputs ( roll / pitch ) ?

If you did all the tests, and you are pretty sure that all the setup are right, I would try to takeoff by compensating the movement using the right stick for that.

If you consider to try that, you need to be confident in your flight skills and need to have a place that would minimize any damage to the model ( i like soft glass or mud) and to yourself , in case something bad happens.

If you are able to takeoff this way, then you could trim the model to be level and hover still. After that, next take off would be ok.[/quote]

Attaching todays log, before i did esc calibration, after pos hold and stabilise modes trying to takeoff. Hope now will be ok as uploading as .tlog file.

I saw your tlog. Is it possible to upload your .bin or .log?

I understand your point of view regarding the current state and the shelf, but my opinion is that there is still some work to be done in your setup. After that it will fly smoothly since the very beginning.

All the drones that I’ve build had a similar behaviour on the very first flight. I would rather take-off and feel what needs to be setup than trying to set it on the ground.

How about your propellers ? ACPUK has raised some interesting points. What esc/firmware are you using?

How is your center of gravity ? is it well balanced?

How was the test regarding pitch/roll response on the ground? In stabilize, with 25% of thrust, if I command roll or pitch, I can see my X6 raising the landing gear a little bit off the ground accordingly. This is an important test to see if it is reacting well.

Yes you right, i will try just at beginning of all this problem, i do tried and copter was intended to roll out one side straight on takeoff.

Esc i was using before was 40AMP RCI with SimonK, again day before it happened drone was flying just perfect (in this video: youtu.be/DqeCDAG_gww i was tested circle mode and was really happy with everything, next morning i went to film another object and was unable to takeoff).

So i thought maybe it can be problems with Simonk ESC and bought new 4 Hobbywings Xrotor Pro´s 40AMP. But have had no change at all. Of course i do calibrated and old one and new ones.

On the ground it responds to all directions, but seems to the right responds less than other sides, on this .tlog log file must be seen this test to.

Is it something wrong with .tlog file, i can upload .bin file but last time i did it was said non readable.

Props are carbon 15" propellers 15/5.

Centre of gravity is well in the centre, as i said day before it happened it was raising very very smoothly without any drift at all.

Really got confused, as almost every part i do remounted and every wire resoldered…