Going to LOSE my mind: X8

ArduCopter Copter 3.1
1

Looking for some help here please. I have a custom built X8 (using the 3DR X8 frame, but with 750Kv motors and a 4S). It is ALL OVER the place. I have been working at this and working at this, and I just cannot figure out what is wrong.

In stabilize, it is just very awkward to fly. I don’t know how else to describe it. It overcorrects for everything, and just kind of wobbles around. Sometimes it will drift away, and bringing it back is near impossible - full rear pitch will just BARELY make it actually pitch it back, if at all. Loiter is an even bigger mess - it kind of toilet bowls and wobbles around all at the same time.

I originally suspected mag problems, so I put the 3DR gps on a 6" pole way above the props. Same issue.

I am including logs from a flight tonight. This starts with me trying to do an auto tune (hoping it would help), but it quickly gets too far away and I can barely bring it back. Then I do a Stabilize, reel it in, and then back off and switch to loiter.

In loiter, HANDS OFF (video below) you can see it just wobble all over the place. It is a dead calm day, with not even a breeze. I have no idea what it’s doing.

I would really appreciate if you experts can take a look at the logs (attached) and see if you see anything. Please ignore the battery warning, I ran it low since I only had one with me and wanted to get as much data as possible.

Video:

youtube.com/watch?v=By-OvsN … LvsdPX23-Q
^ This is in the log, attached, the loiter starts at about 65.

youtube.com/watch?v=Z9y634G … LvsdPX23-Q

Log: dropbox.com/s/k35iaa0rpvlks … 0.log?dl=0

2

Have you done any tuning at all? Just running through the basics on the three main sliders should solve almost any issue:

copter.ardupilot.com/wiki/basic-tuning/

3

Hi,
Analyzing your log, it detects still a compass/mag issue :
Test: Compass = PASS - mag_field interference within limits (20.44%)
Max mag field length (573.69) > recommended (550.00)

and then also a stability issue:
Test: Pitch/Roll = FAIL - Pitch (-45.10, line 101159) > maximum lean angle (30.00)

So I looked at your vibrations level and you indeed have way too much vibrations for a stable copter:
see picture

You should correct your mag issue, maybe redoing a compassmot and moving away ESCs, antennas and GPS/mag furthest as possible.

You should then also review how you fixed your fllight controller on the frame to reduce Zacc vibrations within acceptable levels : -+2 graduation maximum, around -10

4

[quote=“Hugues”]Hi,
Analyzing your log, it detects still a compass/mag issue :
Test: Compass = PASS - mag_field interference within limits (20.44%)
Max mag field length (573.69) > recommended (550.00)

and then also a stability issue:
Test: Pitch/Roll = FAIL - Pitch (-45.10, line 101159) > maximum lean angle (30.00)

So I looked at your vibrations level and you indeed have way too much vibrations for a stable copter:
see picture

You should correct your mag issue, maybe redoing a compassmot and moving away ESCs, antennas and GPS/mag furthest as possible.

You should then also review how you fixed your fllight controller on the frame to reduce Zacc vibrations within acceptable levels : -+2 graduation maximum, around -10[/quote]

Thank you! A couple things:

Regarding the compass/mag issues, I’m really not sure what to do here, the GPS is already at least 6" above the highest point on the copter. Is there any way to confirm that it is using the GPS mag and not the internal one? I am not sure which is which, MAG and MAG2.

As for the stability issue, that line is towards the very end, it caught a concrete lip while landing and tipped over.

I will definitely check into the vibrations - I am using the standard 3DR foam. I’ll see if I can find something that works better.

The other thing I don’t get, is how it would get to the point where I can’t even bring it back in Stabilize - as I mentioned in my original post, sometimes even with full back pitch it will barely start coming back. It’s almost as if the controller was completely trimmed in the opposite direction.

Thank you again for taking a look - hope someone has more info on the compass.

Dan

5

Another thing to verify/do: restore the default PID values from mission planner. I think there is a button for that in the full parameters list (you choose a config in a drop down list that corresponds to your config)

6

Thanks, that was actually one of the first things I did. I started with the 3DR X8 config and tried tweaking a bit, then restored back to it to see if it helped. Doesn’t seem to have.

7

Ok, so I raised the compass even more, did another compass cal, and replaced the foam with a floating rubber mount that I have successfully used on a number of other UAVs, and I am STILL having the same problem. I am at a complete loss.

It looks like my Z axis improved a bit, but it is still doing exactly the same thing.

Datalog attached.

Anyone?

8

Bumping this as I am at a loss. Thanks in advance.

9

Anyone?

10

I’ve had two quads run really funny as well. Seem to lose control and when you tried to correct it didn’t really go in the right direction. Turns out when initially setting up the autopilot from the wizard was causing the issue. Even though I select a X configuration, in the parameters it was actually setup as a I so all the controls were off by 45 degrees lol! Worth taking a look.

11

Thanks for the thought, I just double checked and the frame type is correct (X8). Still baffled!

12

Have you done anything to rule out issues with the motors?

13

If you have checked out all the electronics possibilities then you need to do a complete check of every motor/esc/prop to make sure they are plugged into the correct port on the APM for an X8 configuration, ALSO make sure all props are spinning the proper direction for top & bottom with the X8 configuration.

You can have esc/motors in the wrong ports and still have an X8 fly, about what like you have in the youtube and have little or very confusing control issues.

I would unplug and “rewire” from scratch again with the chart diagram in front of me, and double check the prop rotation. ALL props should be facing with the printed side “upward” as well, do not have the lower props facing down.

Hope this helps.

14

[quote=“jarlt”]If you have checked out all the electronics possibilities then you need to do a complete check of every motor/esc/prop to make sure they are plugged into the correct port on the APM for an X8 configuration, ALSO make sure all props are spinning the proper direction for top & bottom with the X8 configuration.

You can have esc/motors in the wrong ports and still have an X8 fly, about what like you have in the youtube and have little or very confusing control issues.

I would unplug and “rewire” from scratch again with the chart diagram in front of me, and double check the prop rotation. ALL props should be facing with the printed side “upward” as well, do not have the lower props facing down.

Hope this helps.[/quote]

Thanks for your response - I have done many, many rewires just to make sure, run the motor test, taken the props off to confirm direction, talked to a psychiatrist…etc.

I honestly got so fed up I just left the damn thing to sit, and I ended up stealing the GPS off it for a plane. Now, the plane is doing funny things related to compass that I’ve never seen before. Thinking I might have a bad GPS - going to swap out and see what happens.

Thanks!
Dan