Matek H743-Slim V1.5 "System Power unavailable"

[PrcsnFlyer]

Describe the bug

The Drone does not pass the Preflight Check. QGC Reports an RC-Receiver Error in the Pre Flight Check Section. If you try to arm Via QGC anyways, it gives you an "System Power unavailable Error"

To Reproduce

Steps to reproduce the behavior: 1.) Flash FC With a Fresh Bootloader and Firmware Image 2.) Connect RC Receiver, GPS, ESC etc. 3.) Power Drone on 4.) Go through Setup(Sensor calibration, RC-Stick calibration etc.) 5.) Try to Arm via QGC

Expected behavior

Motors Should Start spinning

Log Files and Screenshots

HW arch: MATEK_H743_SLIM FW git-hash: 283aad01fdffd39b8b82e58e3433acbdb6ea95c2 FW version: 1.13.0 0 (17629184) FW git-branch: master OS: NuttX OS version: Release 10.2.0 (167903487) OS git-hash: 91bece51afbe7da9db12e3695cdbb4f4bba4bc83 Build datetime: Jun 11 2022 08:50:40 Build uri: localhost Build variant: default Toolchain: GNU GCC, 9.2.1 20191025 (release) [ARM/arm-9-branch revision 277599] PX4GUID: 00060000000036353839333051170047003b MCU: STM32H7[4|5]xxx, rev. V INFO [param] selected parameter default file /fs/microsd/params INFO [param] importing from '/fs/microsd/params' INFO [parameters] BSON document size 1561 bytes, decoded 1561 bytes (INT32:29, FLOAT:48) INFO [param] selected parameter backup file /fs/microsd/parameters_backup.bson Board architecture defaults: /etc/init.d/rc.board_arch_defaults Board defaults: /etc/init.d/rc.board_defaults INFO [dataman] data manager file '/fs/microsd/dataman' size is 62560 bytes Loading airframe: /etc/init.d/airframes/4001_quad_x INFO [uavcan] Node ID 1, bitrate 1000000 Board sensors: /etc/init.d/rc.board_sensors icm42605 #0 on SPI bus 4 rotation 14 mpu6000 #0 on SPI bus 1 rotation 12 dps310 #0 on I2C bus 2 address 0x76 qmc5883l #0 on I2C bus 1 (external) address 0xD WARN [SPI_I2C] Already running on bus 1 INFO [init] Mixer: /etc/mixers/quad_x.main.mix on /dev/pwm_output0 INFO [init] setting PWM_AUX_OUT none ekf2 [565:237] Starting Main GPS on /dev/ttyS1 Starting MAVLink on /dev/ttyS5 WARN [mavlink] offboard mission init failed (-1) INFO [mavlink] mode: Normal, data rate: 1200 B/s on /dev/ttyS5 @ 57600B INFO [rc_input] RC scan: SBUS RC input locked Board extras: /etc/init.d/rc.board_extras atxxxx #0 on SPI bus 2 INFO [logger] logger started (mode=all)

NuttShell (NSH) NuttX-10.2.0 nsh> WARN [commander] Switching to Position is currently not available
INFO [commander] Failsafe mode activated
INFO [gps] u-blox firmware version: SPG 4.04 INFO [gps] u-blox protocol version: 32.01 INFO [gps] u-blox module: NEO-M9N ERROR [mavlink] FTP: Open failed - out of sessions

INFO [commander] Failsafe mode deactivated
ERROR [mavlink] FTP: Open failed - out of sessions

ERROR [uavcan] couldn't start parameter count: -2 ERROR [mavlink] instance 0: RADIO_STATUS timeout WARN [mavlink] Session was closed without activity WARN [ekf2] primary EKF changed 1 (filter fault) -> 0 INFO [mavlink] Starting mavlink shell INFO [commander] Connection to ground station lost
INFO [commander] Data link regained
WARN [PreFlightCheck] Preflight Fail: height estimate not stable ERROR [uavcan] couldn't start parameter count: -2

Add screenshots to help explain your problem.

Drone:

-Mini TestCopter -Matek H743-Slim V1 / V1.5 with Matek M9N GPS (Tested both HW-Versions) -FrSky SBus Receiver on RX6

Additional context

You can also notice, that the FC-PCB Heats up Significantly. Further more the Altitude Estimate Becomes unavailable after some time.

[taileron]

it seems that the voltage divider is wrong or missing you can try BAT1_V_DIV=11.05 instead of -1

I´m using the mini variant that heats up extremely, cause the small footprint, especially when becs are used for cams or vtx it urgently needs to be cooled in the air

In this case it´s recommended to do the sensor calibration after the system has reached final operating temperature otherwise it gets high acc bias error

[PrcsnFlyer]

Thanks for replying! The Voltage divider is already set. Voltage and Current Measurement works just fine. It seems like the error is related to CBRK_SUPPLY_CHK Circuit breaker. Disabling this one with value 894281 "solved" the issue somehow. Now the drone can be armed. But now I can notice a huge drift on the yaw axis during movement. I think this is related to the accelerometer and messes up the yaw Fusion with the compass. @taileron as you recommended I did all the calibrations on operating temperature. But still there is a lot of noise and drift.

For reference I created a log file on the bench. After reaching the operating temperature I rotated the drone 90° Counter Clockwise - Back - 90° Clockwise - Back, around it's yaw axis. You can observe, that the Estimated Yaw angle drifts way further. Also the vibration metrics do not match with the more or less careful movements that i did with the drone. I also disconnected the ESC to reduce overall EMI for the test.

Could this be related to the high clock rate of the CPU? Maybe lowering the clock rate would help with the overall Board Temperature and EMI of the PCB....

Flight log: https://logs.px4.io/plot_app?log=29970e0b-f221-4e13-99a6-c2ce60ec4759

[taileron]

You could test with EKF2_MAG_TYPE=none if the drift comes only from the compass (the estimator then starts without physical compass). If you do tests indoors and loudspeakers are in the same room, you need at least 2.5m distance. If the system is not powered by lipo but by a power supply, there may be some interference from there as well. At the moment I build the H743 mini in a small 2S Whoop platform, then I will test more precisely which interference factors have the strongest effect or whether a change of the CPU clock multiplier (board.h) would be useful. (currently a F405 fc is in use, mag and GPS are separated to place GPS on top and mag as far as possible from interference sources. The strongest disturbance there is temperature drift in the z-axis of the ACC and requires EKF2_ABL_LIM=0.8)