Pixhawk/APM - Unflyable/Spontaneous Movements

ArduCopter Copter 3.1
1

Hey Guys,

Long time lurker, first time poster.

I have a really weird, weird issue. It’s with a custom built Hex that I’ve had for 2.5 years and been flying on APM without an issue this whole time.

So, I will try to describe what is happening during flight. Generally speaking, there is very little total control over the Hex. I give it a bit of throttle, it goes into a very “unstable looking” hover then with spontaneously want to pitch or roll. The more input I give it on the pitch, roll and yaw, the quicker it does this. I can give it 100% throttle whilst holding the Hex on the ground and it will spin all motors hard and generate full thrust, when it is in the air it will give out a maximum of what seems like 50% power. Sometimes when it lands and I go to give it power, it just won’t take off and spins the motors at what looks like 20-30% power, even though I am giving it up to 100% throttle.

Now, that is flying in stabilise with “out of the box” tuning on both APM 2.6 and Pixhawk. When I fly Acro (with ACRO_TRAINER = 0) I get exactly the same issues.

This has 5 of us all left scratching our heads. We were flying this Hex on an APM 2.6 (normally updating firmware as available) for months without an issue, I then had a fly away whilst in Loiter that resulted in a heavy crash. We decided to tear it all down and rebuild and put a Pixhawk in, now we have these issues, no matter what we are trying.

Current Hardware:

  • Custom Hex Frame
  • T-Motor MT-3506-25 650KV
  • Pixhawk with 3.1.2
  • 40 AMP Plus ESCs with BLHeli
  • Satellite into Pixhawk & DX8
  • 3DR uBlox with Compass
  • 3DR Power Module
  • 2 x 4S 4000mah Nano-Tech Batteries
  • 13x4.7 Props

What we have tried:

  • APM 2.6 (firmware 3.1.2)
  • Pixhawk (firmware 3.1.2 then rolled back to 3.1)
  • Changed ESCs from old Multistars with SimonK to Plush 40amp with BLHeli
  • Run Pixhawk from its own Power Module instead of the power module that then goes to the distribution board and supplies power to ESCs/Motors
  • Changed from using Satellite plugged into Pixhawk & a DX8 to EzyUHF and a Taranis
  • Used 2 different GPS units
  • Removed all other hardware not needed (Vid TX/RX, Mavlink, OSD)
  • 2 different PCs running Windows 7 to load Firmware onto Flight Controllers. Cleared eeeprom before every firmware change.
  • Re-Checked every single wire and solder join. Re-soldered Distribution board

We are lost at to what might be causing this. I have uploaded 2 flight logs from the Pixhawk (transferred across from Micro SD card). I am very limited in my ability to read these, so maybe you guys can see something I can’t.

Any help would be greatly appreciated! We have a lot of hardware here, so can keep trying different combinations…which is weird, as we have changed just about every bit of hardware and am still having problems.

Cheers,
Ash

2

This is almost certainly a motor setup problem. Have you run the motors test in the CLI to check the order? I also recommend updating to 3.1.4.

3

Thanks jschall.

I have updated to 3.1.4. Run the motor test a few times.

You were onto something, discovered slight damage to 2 motors that seemed to have got worse. The windings were slightly damaged inside the can. Testing them on a servo tester showered up their erratic behaviour.

Terrific, I now think the problem is solved…go to fly and the Hex is still unstable (expected as it hasn’t been tuned) but when I give it more than 50% throttle, it flips and looses throttle.

I am stuck here guys, any suggestions?

Here is the flight log of the flight that just occurred…