About to perform a maiden; motor duty cycle & navigation

ArduCopter Copter 3.1
1

Hi,

I’m a new multirotor user (but have had experience in fixed wing RC & UAVs) and have built my own quad which I intend to take on it’s maiden soon.

To ensure that things go smoothly, I’ve been running ground tests to ensure that the vibration is fine, that the temperature predictions from eCalc were matching with real life, etc.

However, one thing I’ve noticed and which I’m a bit confused / nervous about is stability and how it’s going to perform on it’s maiden.

Specifically, when attempting to double-check whether I had the motors in the correct orientation / sequence, I had it set to loiter mode and (without propellers obviously) picked it up and tilted in various directions to see how the motors / gyro respond.

Ultimately it was hard for me to tell much (it’s difficult to judge RPM visually and the motors are close enough that sound doesn’t help), so I checked the RC in / servo out readings, with no transmitter attached and with RC failsafe disabled (to allow for Mission Planner arming).

What I found was that motors 3 and 4 (which in my configuration are the front left and rear right motors) had a relatively high output signal (roughly 60%), while motors 1 and 2 (front right and rear left) were only at something like 20%.

Off the top of my head I can think of 3 potential causes:

  • That this is normal and part of the ArduCopter / Pixhawk control scheme; 2 opposite engines maintain hover thrust and counter each-other’s torque, the other 2 are kept low to permit a greater generation of net torque or directional thrust when required.

  • RC1-4 are not what I think they are (do not indicate the throttle for each motor)

  • The Pixhawk thinks it’s generating equal thrust on all engines, but that motors 1 and 2 do not require as much output signal (I have no idea why that would be however).

On a side note as well, I’d like to inquire as to how accurate I need to be with my accelerometer configuration; alone it’s not easy to use a tablet and hold my quad exactly on it’s side, nose, etc and so I’ve only been able to get it within a few degrees of properly level. Will this cause a consistent drift in flight, or is this something that will / should be mostly negated by the quad referencing it’s GPS location? I could always enlist the help of someone else to make this calibration more accurate, but if it’s unnecessary it’s not going to be worth the hassle.

Regards,

Josh

2

I wouldn’t recommend doing your initial setup testing in Loiter. I use Stabilize for that. Stabilize just works for self leveling, which is what you are testing for, Loiter throws in trying to hold a GPS position, so it could be causing your motors to do all sorts of unexpected stuff.

I do what you are doing, but in Stabilize. I also put some strips of tape on the motor shafts to help visualize rotation. No props for this.

Bring it up to about 1/3 throttle and put the pitch forward. Do the rear motors speed up? Good. Now pitch backward. Do the front motors now speed up? Good again. Now try right and left rolls. Do the appropriate right and left motors speed up. Good once again. If you want you can pick up the quad and see if the motors speed up to counteract your tilting it by hand.

After it passes this test, I put the props on for an “almost hover” test. Still in Stabilize I bring up the motors so that it doesn’t quite lift off. By carefully pitching forward I should be able to lift only the rear end. I try the same thing in roll. This will show up if the frame configuration is correct (X or +).

If it passes this you shouldn’t have any trouble popping it off the ground in any flight mode you’d like.

Your leveling sounds OK. I’ve often had motors under no load conditions spin up more than I would have expected, but with props on they don’t do that.