Unexpected Behavior during motor test

After configuration and calibration of all components I tried to make a motor test.
At the QGroundControl motor test page i test the motors using the sliders for each motor.
The speed of the motors was related to the position of the slider. The direction of spinning was also OK. But after moving the slider the motor spin only for about 1,5 - 2,5 seconds and than stop the moving (no forces applied to motor, the motor behave in the same way as without power).
I expected that the motor should spin continually with a speed which is in relation to the position of the slider.
Does somebody has any idea related to the unexpected stop of spinning ?

I use the current Pixhawk 4 firmware and the current version of QGroundcontrol for Windows.

Best regards

Andreas

Hey there and welcome to the community. :slightly_smiling_face:
Actually no, they’re not supposed to keep spinning. This is the exact expected behavior from the QGC motors test page.
cheers

Hello Christian-Born,

Thank you for the hint!

Now I have another problem related to my ESC’s.
I have a Quadro Copter X configuration and the battery is a 6S Lipo.
I have 4 Motors from neumotors, the model name is 3825 (545).
I selected 4 x APD Brushless F series Motor controllers, model 80F3[x] (40F3 was not available) and a neumotors colleague explains that I could use the 80F3[x] .
But now I’m not sure that the selected ESC was a good choice.
After ESC calibration and the motor test I make e test with the remote control (FrSky X10S), without rotors to test if the motors are then spinning continually.
The motors are not spinning ing normal way, are spinning shortly in one direction and then spining very short in the other direction.
The test was only 3 - 4 seconds long. I made a second test but in this case the capacitor (330uF) was burning and I expect that the ESC is damaged.
Do you have a idea what is going wrong ?
Do I have configure the ESC’s before using ?
Are the ESC’s not compatible with Pixhawk4 and power distribution board ?
The 80F3[x] could be used with 8S but I use only 6S - is this a problem ?
I plan to use other ESC’s. I thought ESC’s from FrSky could be a good choice.
I want to use FrSky Neuron 60S ESC 3S-6S 60A. What do you think about this ESC’s.

Thanks in advance

Best regards
Andreas

Hi^^

I looked into your choice of ESCs and the notable thing about them is their firmware.
From their website:

The APD F-Series ESCs bring forward a ground-breaking new design utilising a Powerful 32-bit F3 processor running brand-new firmware (APD Proprietary) developed from the ground up.

So they’re not necessarily a bad choice nor unsupported by PX4 but they’re certainly uncommon and I can’t tell how they handle their protocols and everything.

Where did you do your calibration from? A groundstation or an external PWM loop like an ESC tester or an R/C receiver etc.?
Either way you should make sure to get some kind of feedback from the calibration process (I know that BLHeli ESCs have a distinguished beeping sequence).
First verify the motor protocol which the ESCs are utilizing inside their currently flashed firmware and then set PX4’s motor-outputs protocol accordingly.

Judging from your description (calibration successful; *motor weirding- and then burning out from *
R/C throttle signal) points to either a defective motor (does it work on a different ESC? - Or well, did work…^^) or your radio is outputting unexpected signals to its channels.
Can you verify those like with a groundstation?

I hope I sound inquiring and not condescending or anything LOL, as I don’t know about your experience with all the various involved systems, you know. :wink:
cheers

1 Like

Hi,

Thank you for the hints,

 *Where did you do your calibration from? A groundstation or an external PWM loop like an ESC tester or an R/C receiver etc.?*

I use QGroundcontrol programm

*Either way you should make sure to get some kind of feedback from the calibration process* 

I could see that the LED’s on the ESC’s are blinking and QGroundcontrol report that the calibration was successful.

*First verify the motor protocol which the ESCs are utilizing inside their currently flashed firmware and then set PX4’s motor-outputs protocol accordingly.*

I have decided to use other ESC’s. My next try are ESC’s from FrSky - model Neuron 60S .

*udging from your description (calibration successful; *motor weirding- and then burning out from *
R/C throttle signal) points to either a defective motor

I expect that the motors are OK because the motor are running fine during the motor tests.

or your radio is outputting unexpected signals to its channels.

How can I detect that the radio is outputting unexpected signals or the the signals from the redio is OK ?
The signals from the radio are received by the receiver R8 Pro (FrSky) and goes then to the Pixhawk 4. Are there debugging features or specific logfiles which are useful for trouble shouting for this case ?

What motor protocol should I use with the new setting - Pixhawk 4 and Neuron 60S ?
I bought a USB BLHeli32 Adapter so I can program the Neuron 60S.

Do you have some other hints related to degugging and troble shouting the Radio / Pixhawk / ESC setting ?

Thanks in advance

Best regards
Andreas

You check the radio screen in QGC and confirm the channels with your radio on.
That’s basically it, if it looks okay here there’s no need for further looking into the radio link.
It might just be your channels in the radio itself are mapped differently from what PX4 expects so you go through the calibration process and then afterwards do the ESC calibration using QGC and your radio.
Tbh I think going for different ESCs isn’t the worst choice either way… ,)

1 Like

Hi,

now I have replaced the 80F3 ESC’s with the FrSky Neuron 60S ESC’s. I have checked the ESC’s with BLHeliSuite32. After that I used QGroundControl to select DShot600. I also changed the wiring between Pickhawk4 and the power distribution board (described in PX4 documentation).
I used the QGroundControl / Analyze Tools / MAVLink Console to test all motors.
I used the motor_test command and test all motors with 5,10,20,30,40,50% of power.
The rotation of the motors was without vibrations.
After that I tried to make similar tests with the remote control FrSky X10S Express.
I used this remote control with my Tarot 650 in the past and the remote control works as expected.
The tests with the remote control works not in the way I expected.
The propellers was not mounted and I tried to use only the throttle Stick to control the motors.
The flight controller Pixhawk 4 was in manual flight mode.
After arming the motors are rotation with low speed in normal way without vibration.
After moving the throttle Stick to give more power (I made this normal / slowly) the motors generate very high vibrations - I think to motors are not rotating evently.
I’m not sure the reason for that.

Is the remote control sending wrong or faulty signals ?
Could I debug this by programming a test program using MAVLink API ?
Could the behavior initiated by the Pixhawk control program (the pixhawk tried to control the motors so that the drone lift up) ??

Do you have a hint to solve this problems ?
Thanks in advance

Best regards
Andreas

Hey Andreas,
it sound like this might be just the pid-loop reacting to the quad not reacting (following a setpoint). Did you fixate it during the test? You can just kinda hold and press it down with your hand.
It should definitely be fixed to a position.
Try that and see if it still gets shaky.
I’m not even entirely sure if a “messy” pwm signal would even make the motors shake in that way…
Only thing else to try might be a lower dshot but normally any somewhat recent hardware should handle 600 no problems.
cheers man

1 Like

Hi Christian,

. Did you fixate it during the test? You can just kinda hold and press it down with your hand.
It should definitely be fixed to a position.

Do you mean the drone ?
My Drone lies down (without “legs”) in a fixed position.
I have create a Logfile and tried to analyse the situation. It looks like that the radio signals are completly correct.
The diagram for position moving shows (GPS) that the drone “beliefs” she is moving around some meters (about 8 meters). I’m not sure if that is normal because the drone has definitly a fixed postion. On the other site I know that a normal GPS Signal is not really accurate.
The PID analysis page shows in all Step Response diagrams a zero flat line.
Is this the normal behavior or expected signal diagram in this “fixed position” throttle controled motor test ?
Best regards

Andreas