Understanding of motor mixing table

[r0sewhite]

Getting my AQ and ESC32 hopefully tomorrow I read a little in the wiki in the past two days. What I don't understand is the hexa x example: Why should the outer motors (L and R) run with 75% on roll while the inner ones run with 100%?

Taking a look at mixers from other FC's like MultiWii or DJI, they all work the same: The most outer motor to the appropriate axis is set to 100% and all other motors will be set to their percental distance to the axis. Though there is a 2nd fact to consider they all base on the fact, that a motor which is further away from the axis will have to run a further distance to gain the same angle than a motor which is closer to the axis. These are examples of a octo + and a hexa x:





Considering the roll axis, the most left motor on both copters have to speed up most and thus are set to 100%. On the octo +, the front left and rear left motors are 70% of the distance. For example, if each arm is 100cm long, the FL and RL motors are 70cm far away from the roll axis. On the hexa x, the FL and RL motors are only half the distance of the left motor to the roll axis, so they are only set to 50%.

The 2nd fact which is usually not considered is the leverage: The outer motor has a longer lever and thus needs less power than a motor with a shorter lever. But can this deviation be so much that an outer motor needs less power though it has to move a further way than an inner one? I would really like to understand the hexa x example from the wiki but it only says the outer motor "don't need to have 100% output". What is the calculation behind that or is it just a guess?

[LPR]

r0se

I have an octo that is flying OK and I'm using a table I found for Openpilot's CC board.
The table is for a Clockwise number flat + octo. The first motor in the front of the octo is number one.
Motor Thro Pitch Roll Yaw
1 100 -33 0 -100
2 100 -33 33 100
3 100 0 33 -100
4 100 33 33 100
5 100 33 0 -100
6 100 33 -30 100
7 100 0 -33 -100
8 100 -33 -33 100

I'm going to try some new settings, like instead of all the same values for pitch or roll I would use 33 and then 66 and than back to 33. Also 35, 50, 35.

If there is anyone with experience I would like to hear what their using.

Larry

[r0sewhite]

Hi Larry,

I know about the theory to set all motors to the same value because it's not only the angle (distance) but also the lever (power) which has to be considered. But I'm not sure if setting the values to only 33% is a good idea. This way you created a limiter which you probably won't notice in normal flights. But if it ever happens that a motor fails, the limiter can lead to an instable copter.

I experienced the effect of a limited mixer some months ago when I had a motor failure on my X8: With MultiWii nothing happened when the motor stopped working. It was still absolutely stable, even in gyro only mode. With the Wookong, the copter was hardly controllable: A strong drift (even with position hold) and unwanted yaw rotations when giving a little throttle.

[LPR]

r0se

My octo has ESC32s and they are calibrated but if a prop hits the grass or a twig a motor will stop. With the AQ board I have no ability to stay in the air if one motor is not working, I've tried. The AQ is very stable because the inner loop does not use joystick inputs. I've read in the wiki that your frame has to be balanced for the AQ.
I don't think that the octo will be able to fly with a motor off but with the right motor mix I might get a more aggressive control.
I had a CC board on the octo with the ESC32s using the same 33 mix and found the control was very very good. I don't think the AQ works the same way as the CC. I'll just test a number of mixes and let you know what I find out.

Larry

[r0sewhite]

[LPR]

Did a couple of test with motor mixes. Tried all 75s and found it too high a number to work. That mix must generate too high of PWM ms because before you can get to a hover one or more of my fully calibrated ESC32s turn off and have a red led. Then I tried all 50s and found I could hover in the basement but I'm just a little worried about trying it outside and having motors shut down when I use higher throttle. Could not feel any difference in control between all 33 and all 50s. That was a surprise.

[Max]

[LPR]

Max

Everything is at defaults for PWM settings. My guess is that PWM pulses get too long for the ESC32. I calibrated the ESC32s and have a lot of flight time on them. The ESC32s have never had this behave your before.
Because I don't see any noticeable change in control from all 33s to all 50s I can't see any reason to change to higher numbers than 50 for an octo.
I've read in the wiki that the OM PID setting is used to limit the length of the PWM pulse. Maybe there is no such limit for the mixing table. I it could be that the mixing table is set up for quads and as you increase the number of motors you need to divide up the total settings (200) for pitch and roll.

Larry

[epyonxero]

[LPR]

It's good to hear that an AQ multi can keep flying with a motor stopped. I'm using default PID settings but I not sure the octo is performing as it should. I know an AQ multi is not supposed to be able to be an aggressive flyer but the octo seems slower to tilt than what I've seen in some videos of quads. Hard to tell really. To keep the octo in the air while doing a mission I have to set the navi speed to 4 instead of the default 7. I try to fly with a motor stopped after I change the motor mix table to 50s instead of 33s.

Larry