Matek H743-Mini Quad won't arm in 4.5.6

We can’t get our Matek H743-Mini build to arm since we upgraded ArduCopter to 4.5.6. We didn’t expect any issues so we never made note of the firmware version that was previously working fine.

Here’s the thing, we discovered by accident that our build will arm if the build is MavLink connected to Mission Planner running on our PC with USB cable. We do not have telemetry radios on this build or would have tried a wireless MavLink connection too.

Based on a thread in an old forum we already know that if we put the following firmware on the build we can get it to arm and fly but we don’t understand why the latest firmware is a bust:
https://firmware.ardupilot.org/Copter/stable-4.1.1/MatekH743/

Any suggestions greatly appreciated.

Can you post your param file?

Correct me if I’m wrong but it’s my understanding that an upgrade has no affect on parameters so here’s a link for the parameters for the 4.1.1 version that we got to work when we downgraded:

If you want the parameters listed for the 4.5.6 version I can reproduces those too just let me know. Hmmm?, sort of like maybe there’s new parameters in 4.5.6 that we are not aware of that could explain our issue(?)…

Thanks for you help : )

Here’s a link to the parameter list for the 4.5.6 version which doesn’t allow us to arm our model:

What messages does Mission Planner give when arming is attempted?

Maybe, and that is certainly developer intent.

However, parameters can be added, removed, or renamed over time. Upgrading from one version to the next (like 4.5 to 4.6) should be seamless. It’s hard to guarantee that upgrading from years old firmware to current stable will be as successful.

@Dave we don’t have telemetry on this build so there is no wireless communication on MP that we can check for messages.

Our build will arm if we are MavLinked to MP with a USB cable. We discovered that by accident.

So what messages are presented? None with USB?

Disable LOG_DISARMED. Who knows how many huge log files have been taken.

@Dave Here are the messages listed for a sequence that successfully armed our build; MavLink connected with USB cable. The parameter LOG_DISARMED = 0 (default) was already set.

9/17/2024 8:55:42 PM : EKF3 IMU0 is using GPS
9/17/2024 8:55:26 PM : u-blox 1 HW: 00080000 SW: ROM CORE 3.01 (107888)
9/17/2024 8:55:24 PM : GPS 1: u-blox saving config
9/17/2024 8:55:20 PM : Field Elevation Set: 143m
9/17/2024 8:55:20 PM : EKF3 IMU1 origin set
9/17/2024 8:55:20 PM : EKF3 IMU0 origin set
9/17/2024 8:55:13 PM : Frame: QUAD/H
9/17/2024 8:55:13 PM : IMU0: fast sampling enabled 2.0kHz
9/17/2024 8:55:13 PM : RCOut: DS600:1-6 PWM:7-13
9/17/2024 8:55:13 PM : MatekH743 00300034 32305113 34353331
9/17/2024 8:55:13 PM : ChibiOS: 6a85082c
9/17/2024 8:55:13 PM : ArduCopter V4.5.4 (fd1bcc61)
9/17/2024 8:55:13 PM : Frame: QUAD/H
9/17/2024 8:55:13 PM : IMU0: fast sampling enabled 2.0kHz
9/17/2024 8:55:13 PM : RCOut: DS600:1-6 PWM:7-13
9/17/2024 8:55:13 PM : MatekH743 00300034 32305113 34353331
9/17/2024 8:55:13 PM : ChibiOS: 6a85082c
9/17/2024 8:55:13 PM : ArduCopter V4.5.4 (fd1bcc61)
9/17/2024 8:55:12 PM : Frame: QUAD/H
9/17/2024 8:55:12 PM : IMU0: fast sampling enabled 2.0kHz
9/17/2024 8:55:12 PM : RCOut: DS600:1-6 PWM:7-13
9/17/2024 8:55:12 PM : MatekH743 00300034 32305113 34353331
9/17/2024 8:55:12 PM : ChibiOS: 6a85082c
9/17/2024 8:55:12 PM : ArduCopter V4.5.4 (fd1bcc61)
9/17/2024 8:55:10 PM : GPS 1: detected as u-blox at 230400 baud

That doesn’t show an attempt at arming.

The HUD said Armed and all four motors spun after I held the throttle stick down and to the right for about 5 seconds.

If it only not arm if USB is not connected and you don’t have a mavlink connection without USB I think best is to set for a really short test LOG_DISARMED to 1. Than disconnect USB. Powercycle the FC (Quad) and try for a short time to arm (only a few seconds) Powercycle again the FC. Download the log and present it here than don’t to forget to change LOG_DISARMED back to 0

Links to bin files for J-F’s suggested sequence:

2024-09-18 20-59-22.bin - Google Drive, 2024-09-18 20-58-19.bin - Google Drive, 2024-09-18 20-57-45.bin - Google Drive, 2024-09-18 20-56-41.bin - Google Drive

Here are messages for last step of J-F’s suggested sequence:
9/18/2024 9:00:11 PM : PreArm: Battery 1 below minimum arming voltage
9/18/2024 8:59:53 PM : EKF3 IMU0 is using GPS
9/18/2024 8:59:53 PM : EKF3 IMU1 is using GPS
9/18/2024 8:59:41 PM : PreArm: Battery 1 below minimum arming voltage
9/18/2024 8:59:38 PM : u-blox 1 HW: 00080000 SW: ROM CORE 3.01 (107888)
9/18/2024 8:59:36 PM : GPS 1: u-blox saving config
9/18/2024 8:59:32 PM : EKF3 IMU0 origin set
9/18/2024 8:59:32 PM : Field Elevation Set: 138m
9/18/2024 8:59:32 PM : EKF3 IMU1 origin set
9/18/2024 8:59:22 PM : GPS 1: detected as u-blox at 230400 baud
9/18/2024 8:59:18 PM : EKF3 IMU0 MAG0 initial yaw alignment complete
9/18/2024 8:59:18 PM : EKF3 IMU1 MAG0 initial yaw alignment complete
9/18/2024 8:59:18 PM : EKF3 IMU0 tilt alignment complete
9/18/2024 8:59:18 PM : EKF3 IMU1 tilt alignment complete
9/18/2024 8:59:16 PM : AHRS: EKF3 active
9/18/2024 8:59:16 PM : EKF3 IMU1 initialised
9/18/2024 8:59:16 PM : EKF3 IMU0 initialised
9/18/2024 8:59:15 PM : Frame: QUAD/H
9/18/2024 8:59:15 PM : IMU0: fast sampling enabled 2.0kHz
9/18/2024 8:59:15 PM : RCOut: DS600:1-6 PWM:7-13
9/18/2024 8:59:15 PM : MatekH743 00300034 32305113 34353331
9/18/2024 8:59:15 PM : ChibiOS: 6a85082c
9/18/2024 8:59:15 PM : ArduCopter V4.5.4 (fd1bcc61)
9/18/2024 8:59:15 PM : Frame: QUAD/H
9/18/2024 8:59:15 PM : IMU0: fast sampling enabled 2.0kHz
9/18/2024 8:59:15 PM : RCOut: DS600:1-6 PWM:7-13
9/18/2024 8:59:15 PM : MatekH743 00300034 32305113 34353331
9/18/2024 8:59:15 PM : ChibiOS: 6a85082c
9/18/2024 8:59:15 PM : ArduCopter V4.5.4 (fd1bcc61)
9/18/2024 8:59:15 PM : Frame: QUAD/H
9/18/2024 8:59:15 PM : IMU0: fast sampling enabled 2.0kHz
9/18/2024 8:59:15 PM : RCOut: DS600:1-6 PWM:7-13
9/18/2024 8:59:15 PM : MatekH743 00300034 32305113 34353331
9/18/2024 8:59:15 PM : ChibiOS: 6a85082c
9/18/2024 8:59:15 PM : ArduCopter V4.5.4 (fd1bcc61)
9/18/2024 8:59:14 PM : RCOut: DS600:1-6 PWM:7-13
9/18/2024 8:59:14 PM : AHRS: DCM active
9/18/2024 8:59:14 PM : ArduPilot Ready
9/18/2024 8:59:13 PM : Barometer 1 calibration complete

Pretty sure this message is all you need for root cause analysis…

PreArm: Battery 1 below minimum arming voltage

That’s the point where I had unplugged the battery according to J-F’s instructions.

So where is the arming attempt prior to that? Something like this:

message

That’s the problem, we can’t get our build to arm unless our model is MavLinked to MP with a cable. As soon as we disconnect the MavLink USB cable the model refuses to arm.
All I can tell you guys is that when the cable is connected between the model and MP the HUD indicates armed and the motors spin when I move the throttle stick on my transmitter down and to the right. When I disconnect the cable holding the throttle stick down and to the right the model sits there and does not respond.

Maybe GCS failsafe. I’m not in a position to review the logs right now, but that’s the logical place to start looking.

I replaced the value for FS_GCS_ENABLE = 1 (default) with a value of 0 instead and now the model arms and the motors spin without the USB cable.

Is there any risk to disabling the GCS failsafe? We don’t have telemetry on this build and it seems to make sense that the failsafe was invoked because there was no connection to the MP.

There’s no danger in disabling a failsafe that can’t possibly help you.