Hello,
I am hoping someone can help as I am completely stumped!
I have self-built a hexacopter with the following specs:
Emax MT2216- 810KV motors
Spider 20A ESCs re-flashed with BLHeli Blu20A firmware
Pixhawk clone flight controller
Turnigy i10 Transmitter and receiver running PPM, plane mode
I have carefully calibrated accelerometers, compasses, radio and ESCs using both mission planner and without. Checked the range independently using an Afro ESC USB interface to confirm it is between 1020 and 1996ms, which is what mission planner gets.
When calibrating the ESCs (straight through mode) the motors operate uniformly and accelerate proportionally to the throttle input.
The problem occurs when the Hex is armed normally. The motors start to spin on arming (slightly faster than I would like), but as soon as I introduce any throttle the motors start to accelerate disproportionately and then it flips over and usually trashes a propeller or two. I have tried various settings for throttle but nothing seems to help. I have attached the flight log of the copter tethered to a board and only just enough throttle to start (ie maybe a couple of mm on the stick) and the motors accelerate to probably 1/2 -3/4 max speed? I will also attach the parameter file as it is currently. This hex certainly will not fly as it is and I’m at a loss what to do next. Please help!
Good morning. The problem occurs with or without GPS lock I’m afraid. No idea what an EKF2 issue is or how to even start to address it. Suggestions would be gratefully received.
Thanks
Hello,
Is there any way of turning the Extended Kalman Filter off so I can test using legacy DCM filtering? I believe AHRS_EKF_USE is not available in Copter 3.3.
Kind regards
Double check the wiring from the ESC is giving you CW & CCW on the relevant motors. I remove the props and tape a little flag to the motor shafts. A short burst of throttle lets you see the direction of rotation.
Also check the wiring from the pixhawk to each ESC i.e.1 to 1, 2 to 2, etc. and the correct props are on each motor . . .
Mission Planner has a motor test page which can help verify a lot of this.
Many thanks for the feedback. The motors are rotating the correct directions and plugged in to the Pixhawk in the correct order. Checked several times and confirmed. The Pixhawk still ramps up the motors to a high level when a small amount of throttle is introduced after arming. Mike suggested turning the EKF off to test if that was the issue but this feature is not available on Copter 3.3! The parameter yo do so is not there. Still stumped here. This seems to be a fairly recurrent theme on several threads so is it a 3.3 firmware issue??? Nobody has suggested a solution that I can find. I will attempt to get the hex airborne at some point and upload the log files for review.
Thanks
Right, 3.4 is being worked on so the website has already been updated with the parameter names for that version and previous version names are lost. Would have to load up a 3.3. version to see what it was called.
I turned throttle Thr-min and THR-mid parameters down (a lot) and the throttle is now manageable. Motors still accelerate by themselves on introduction of a small amount of throttle, but to a much lower level than before. However as soon as the motors approach lift off power the Hex just flips over suddenly. It seems to be randomly forwards or backwards. I’ll upload the logs tomorrow as I am travelling today. Still scratching my head in bemusement here!
Thanks Max,
I have already started from scratch twice and re-loaded Copter 3.3. I think I may give up on 3.3 and load theCopter 3.4 Beta firmware. At least I can switch off the Extended Kalman Filters and see if it will fly on the legacy systems.
I’m having the same issue with the current 3.4 master code.
Everything calibrated, and the moment I increase the minimum amount of throttle, the drone just flips.
I did the same, but holding the drone with my hands, and yeah, sometimes it goes in one direction, and sometimes on the other.
OK I’ve now tested on 3.3, 3.4beta and then 3.2.1 all carefully set up and calibrated from start to finish. The hex responds the same with each firmware. I introduce throttle to almost takeoff power and then it flips. Sometimes forwards and sometimes backwards, but seemingly random. Never seems to go sideways though! I have log files and I will upload them when I return from Northern Ireland tomorrow. I was thinking that it was a duff pixhawk clone???
Best wishes
Patrick.
Here’s a thought- what if I load the quad firmware, unplug the two lateral motors from the FMU and see if it flies in that configuration? Might be a little less stable in roll because of the geometry but no reason why it shouldn’t work…
