Trouble tuning new props for hexa X

[chschmid]

After changing my props from 10X4.5 CFK to 12x4.5 CFK due to more weight of the new cam, I am in trouble with the hexas attitude.
I changed the mixing table as well from 25/50 to 50/100 (hexa X).

In calm conditions it hovers (very) nice, in gusty-windy conditions its wobbly and tends to loose yaw. (once it was yawing 90° to the right and then come back after 2secs.
When I look at my logs, i am frightened. Inputs and outputs are close to clipping (or clipped).
I was playing with tilt rate D from 6900 down to 4000 but I could not see really improving Log values.
The logs attached are from todays flight in gusty wind.

I really need some advice to go in right direction.

Thanks
Christof

[leopold1]

Christof,

I'm using the 33/66 motor mix approach and must say its more responsive.
Try it and you'll see.

[chschmid]

Over the Weekend I was trying to adjust the attitude and some nav PID's. I am not sure if it is ok to reduce the F param, but it seemed the only way to reduce the control outputs and the mot inputs.

The altitude hold shows still some sinking the first minute of flight but I think its more related to GPS height that gets lower within the first minute after lock.

The yaw lock moves a littlebit back after turning and releasing the sticks. But not always. Whats the remedy here?

The result "logwise" looks ok. Any hints are very welcome.

//Hexa Params
#define NAV_SPEED_P 5 //Standard 7 m/s
#define NAV_ALT_POS_P 0.205 // Standard 0.20000000298
#define NAV_ALT_SPED_P 225 // Standard 200
#define NAV_ATL_SPED_I 3.55 // Standard 2.85
#define NAV_ATL_SPED_IM 700 // Standard 600
#define NAV_ATL_SPED_OM 750 // Standard 600

#define RADIO_TYPE 0
#define CTRL_TLT_ANG_P 55 //STD 60
#define CTRL_TLT_ANG_I 0.0005 //Std 0.0005

#define CTRL_TLT_RTE_D 3800 //STD 7180
#define CTRL_TLT_RTE_F 0.19 //STD 0.25

#define CTRL_YAW_RTE_P 250 //STD 300
#define CTRL_YAW_RTE_I 0.32 //STD 0.15
#define CTRL_YAW_RTE_F 0.22 //STD 0.25

#define CTRL_FACT_THRO 0.710
#define CTRL_DBAND_THRO 80
#define CTRL_DEAD_BAND 40
#define RADIO_TYPE 0 //11Bit

Best
Christof

[LPR]

Christof

Have you changed motor mix to 33, 66?
I've found that if the motor mix pitch and roll add up to over 100 you take away from yaw control and it shows up as you carry more weight.

Larry

[chschmid]

Hi Larry

I used to use 25/50. Tested 33/66 for two flights. But now I took the challange to try 50/100 as discussed many times here. But I see exactly the behaviour you describe. Without the PID changes i did, the system showd as underpowered and all outputs where clipped. In gusty wind I was really loosing yaw.

Maybe I should give it another try with 33/66 and go back to standard PID's and compare the results.

Cheers
Christof

[LPR]

Christof

I don't have a Hexa but from what I have seen with the octo and two quads 33, 66 and default PIDs would be my guess.

Larry

[teramax]

reopen this tread, still playing with the Hex-tuning.
The question is -how should a motor-LOG look like?
like so:


or more like so:

example_01.JPG (114.37 KiB) Viewed 12646 times

i am asking me this cause i also feel that a high stable pitch and roll steals yaw autority AND i have also the feeling that a wrong pid set sucks your battery up more than necessary.

I can hover, glide and fly smooth like no other FC but only when i command like so.
Agressive manuvering is a nogo
The AQ FC answeres hard stick-input or heavy trottle with twitching and shaking the frame.

My setup, yea 100/50/50/100 and nearly default PID´s exept TRD, played D down meanwhile to 6000


Some boring flyby´s but turn on the speakers, when the gimbal begins to shake you can hear the twitching sound.

[Max]

Video looks pretty good to me (3 cams, wow! I can barely remember to get one rolling). I think you just have some vibes that set up when the motors are at certain RPMs or thrust points. Someone else here mentioned having better luck with wooden Gemfans (or was that you?). The gimbal shaking and whatever camera you have hanging on the arm there probably aren't helping when that happens, and I'm assuming it's a bit out of balance (with the arm cam). The fact that you need to turn down TR-D also suggests to me that something could be smoother. Maybe the long arms are moving a bit under high torque? It's hard to tell from the video because of the CMOS jello effect.

I can't really tell what is going on in the graphs, but I think the less noisy ones look prettier.

For more aggressive flying, you will want all the possible power available to each motor... so use the right motor mix.

-Max

[Max]

[teramax]