Hi.
I’m build custom mixer with 16 (8 cw and 8 ccw) motors wich working with can line. Have two gps can. Flight controller cuav x7 pro.
Reciever with crsf protocol.
latest_config_origa.param (17.2 KB)
Problem with altitude and yaw twitches. From logs i’m cant see any big problems or maybe i’m wrong.
Please help me - what can be done to fix the problem.
Log - https://drive.google.com/file/d/1jqUNAQgyY1u5QqfYXHUYVz1bqlecCBQn/view?usp=sharing
It looks like there’s a big physical yaw bias problem that gets bad as soon as the motors start to take some weight, like a significant amount of frame flex.
There’s a lot of prop wash affecting the barometer too, so this might affect landings and altitudes later.
Ok.
aluminum frame - very strong.
And several other flights went without such problems - which is also strange. Maybe compass interference?
In flight compas drift very quickly but some time stands clearly
Upgrade to firmware 4.3.3, there’s some fixes that might be related.
Have you got a log of a better flight?
Here - have some althold problems but wihtout big yaw drift.
https://drive.google.com/file/d/1UkNy5zZQOM-IvQIL78s7MN39arGkknLV/view?usp=sharing
I’m not sure which are your CW and CCW motors, so there’s either a significant weight imbalance towards 2,3,6,7,10,11,14 (and maybe one or two more, I couldnt graph them) or it’s twisted motor mounts giving a phsyical yaw bias.
To me the yaw bias makes some sense because, even though attitude control is actually quite OK, the yaw is all over the place (comparatively)
Some other slightly unrelated things that wont be helping you:
You are clearly using poshold and expecting that to work, which is OK, but you’ve disabled some GPS related checks. IMU calculated position and GPS position is generally bad the whole time.
This is usually caused by flying before there’s a good GPS Fix, poor GPS reception, and most often vibrations. So if there’s something you can do to slightly improve vibrations, that will help a lot.
I would set:
ARMING_CHECK,1
BRD_BOOT_DELAY,3000 ← ensure CAN devices boot up before the FC
FENCE_ENABLE,1 ← forces you to wait for a good 3D Fix and Home is set
GPS_AUTO_SWITCH,4
INS_ACCEL_FILTER,10
INS_HNTCH_FREQ,90
INS_HNTCH_BW,45
PSC_ACCZ_I,0.4
PSC_ACCZ_P,0.2
After you update firmware you can set
INS_LOG_BAT_OPT,4 ← log pre- and post-filter data
INS_HNTCH_FM_RAT,0.7
You can probably also set
ATC_THR_MIX_MAN,0.5 to increase prioritization of stability over throttle.
EDIT:
Set COMPASS_USE,0 to disable the internal compass.
If you can do those fixes, then do a test flight with lots of yaw we can run that through magfit and improve the compass calibration.
I’m disable GPS checks because i’m have GPS unhealthy messages which related to GPA.delta abnormal - maybe related to CUAV NEO 3 GPS which working with CAN line or maybe other problems - dont know… at the same time, both gps always have 3dfix.
My mixer for all motors
{ 22, AP_MOTORS_MATRIX_YAW_FACTOR_CW, 1 },
{ 22, AP_MOTORS_MATRIX_YAW_FACTOR_CCW, 2 },
{ 53, AP_MOTORS_MATRIX_YAW_FACTOR_CCW, 3 },
{ 53, AP_MOTORS_MATRIX_YAW_FACTOR_CW, 4 },
{ 114, AP_MOTORS_MATRIX_YAW_FACTOR_CW, 5 },
{ 114, AP_MOTORS_MATRIX_YAW_FACTOR_CCW, 6 },
{ 164, AP_MOTORS_MATRIX_YAW_FACTOR_CCW, 7 },
{ 164, AP_MOTORS_MATRIX_YAW_FACTOR_CW, 8 },
{ -164, AP_MOTORS_MATRIX_YAW_FACTOR_CW, 9 },
{ -164, AP_MOTORS_MATRIX_YAW_FACTOR_CCW, 10 },
{ -114, AP_MOTORS_MATRIX_YAW_FACTOR_CCW, 11 },
{ -114, AP_MOTORS_MATRIX_YAW_FACTOR_CW, 12 },
{ -53, AP_MOTORS_MATRIX_YAW_FACTOR_CW, 13 },
{ -53, AP_MOTORS_MATRIX_YAW_FACTOR_CCW, 14 },
{ -22, AP_MOTORS_MATRIX_YAW_FACTOR_CCW, 15 },
{ -22, AP_MOTORS_MATRIX_YAW_FACTOR_CW, 16 },
Big number - upper motors
Ok.
I’m try what you say and test again. thanks.
OK, for the GPA.Delta issue try reducing the number of constellations.
You dont have SBAS there and maybe these will be more stable:
Set GPS_GNSS_MODE and GPS_GNSS_MODE2 to 65 or 8 or 5 or 4 - whichever works best for your area.
Keep both the same, whichever you choose.
New log but i’m forgot enabling imu fft log ((
But now i’m have real RPM from all 16 motors.
Some better but again yaw jumps - maybe this is because need tuning pids?
i’m think Copter have poor or slow stabilization for yaw…
https://drive.google.com/file/d/1owyyFkmT35NURruNyUl9NC5ah3W9w5Lh/view?usp=share_link
First GPS only with two gps systems (65) work better but no second GPS((
INS_HNTCH_FREQ,140 should be 80
INS_LOG_BAT_MASK,0 should be 1
INS_LOG_BAT_OPT,6 should be 4
It looks like you changed to disable the external compasses and only left the internal compass enabled. I would change to:
COMPASS_USE,0
COMPASS_USE2,1
COMPASS_USE3,1
For yaw try:
ATC_RAT_YAW_I,0.08
ATC_RAT_YAW_P,0.8
For pitch and roll try:
ATC_ANG_RLL_P,6.0
ATC_ANG_PIT_P,6.0
ATC_RAT_RLL_P,0.12
ATC_RAT_RLL_I,0.12
ATC_RAT_RLL_D,0.005
ATC_RAT_PIT_P,0.12
ATC_RAT_PIT_I,0.12
ATC_RAT_PIT_D,0.005
New logs
https://drive.google.com/file/d/1NBMkM8kc-yl-nsvWL2IsfLoytbOT8dOn/view?usp=sharing,
https://drive.google.com/file/d/1UYRI-Or_yncfOKTfoYR5aGQnit9XYf_Z/view?usp=sharing
YAW pids super overshut - i’m reduce to 0.2 and 0.02
Using external barometr from NEO 3 module - make more better results - but noise from internal ms5611 very big - why this? Bug in Cuav X7 pro ?
And i’m think frame have disbalance between pitch and roll and very slow yaw… i’m think this is make some problems while stabilising… i’m think need increase roll pids and need tuning yaw again…
Harmonic Notch Filter is now working perfectly
The barometers in the flight controllers can be subject to prop wash. The Neo3 GNSS unit is great because it includes a baro and you can mount it on the stand up above all the prop wash.
The second log looks quite good for pitch and roll, yaw is nearly OK. But I’d say there is still some underlying issues with PIDs
This is quite high for yaw, it’s usually lower than the pitch and roll values.
ATC_ANG_YAW_P,8
Did you try:
ATC_RAT_YAW_I,0.08
ATC_RAT_YAW_P,0.8
and the other ATC_RAT values I suggested?
In both those logs you have very high D terms - check for hot motors!
There’s also a few of battery parameters that need updating too, But I cant quite work out what battery you actually have.
How many cells and what chemistry?
This is definitely too low
BATT_LOW_MAH,100
I think it should be 2400 = 20% remaining charge of 12000
Yes im trying all what you say.
With yaw 0.8 0.08 and after yaw movement with stick the copter tilts very strongly and seems to fall but then levels out. After reducing the pids, everything immediately returned to normal. Maybe filtering needs adjustment for YAW?
In my opinion, this can be seen in a large log at the very beginning - I pulled the YAW sticks and then sat down because of problems.
For some reason, the ardu glues the logs - very bad - why can’t the log be made separate after each disarm?
After a few flights I tried to adjust the PIDs - increasing the D and P and listening to the oscillations - the motors are cold but the temperature is low outside.
And about barometers - cuav x7 pro have two internal ms5611 which jumps -+2m on the table without flying. Neo 3 pro have same 5611 (maybe 
) but jumps as normal -+5cm on the table…
Im have mamba h743 with dps310 - jumps -+5cm.
Maybe i’m find where problem with barometers - from datashet for ms5611 -
“The best noise performance from the module is obtained when the SPI bus is idle and without communication to other devices during the ADC conversion”
But each ms5611 on x7 pro share SPI lines with two gyros!
Try this:
LOG_FILE_DSRMROT,1
Two lipo 6s each 6A (12A).
Copter have weight about 9kg. 10x45 inch props.
About BATT_LOW_MAH - yes need correcting.
I’m disable all prearm checks while testing and i’m have many messages - one internal error 400 related to can esc converter uc4h when enabled telemetry - maybe ardu dont understand some params which send converter with status messages…
Second - gps unheatlhy warnings - gnss mode 65 dont help - GPA.delta jumps to 2s - very bad situation - dont understand reasons…
Some time i’m have battery error messages - maybe related to CAN PMU…
Maybe CAN line overloaded and ardu incorrectly processes all messages in the CAN bus?
Gameover 
Copter die when i’m start mission - copter make takeoff but began to toilet and collapsed into the wall of the building. compasses calibrated…
But before i’m flying 4 time with poshold, althold and stabilise and one mission.
Any body help me understand why?
https://drive.google.com/file/d/1jLHoqOOUl6qRtOvijAlvezd4nRaWOhs2/view?usp=share_link
You used Guided mode to move in steps (repeated pitch and roll movements) until there was a collision.
Contributing factors:
Arming checks disabled
PID’s and tuning far from finalised
Voltage levels and failsafe actions dont make sense
Fence not enabled
Vibrations not totally sorted ??
Propwash over baros ??
You’re right - all need fix. I just did not expect that it might happen when he flies normally in manual modes and poshold.
I have another question about the ADIS16470 gyroscope - it alone has such strong vibrations - the other two are very different - maybe it is too sensitive?
