Pitch on the controller seems to be broken?

[k4kfh]

I just had my first (small) crash with the Solo yesterday. I went to take off, and it shot off in one direction and caught the grass and flipped over. No huge damage, but I'm waiting on a new set of props in the mail.

I've been trying to figure out what exactly caused the crash (so I can avoid it in the future) and I've checked all sorts of things. If I put the logs through the auto-analysis tool in Mission Planner, the only thing that's a little wonky is the compass, but that's only slightly out of normal spec and probably has a lot to do with me taking off next to a big iron grill.

The desired pitch value and the actual pitch value in the logs don't seem to sync up either. Desired pitch is 0 degrees and actual pitch is all over the place. I thought it was a motor failure but all motors seem to work just fine.

Now I am thinking it may be the controller actually commanding it to pitch so wildly. I looked at the "radio calibration" screen in ArduPilot and when I wiggle the sticks, all the channels on my controller work fine EXCEPT pitch. Pitch stays at 2000 (about 85%) regardless of what I do to any sticks on the controller.

Is this a normal bug with reading data from the Solo controller, or am I correct in assuming this is the problem? Certainly seems out of the ordinary.

Here is a screenshot to clarify. When I took this screenshot, I had my hands off the Solo controller, and all sticks were in the center.

 

[Maddog]

Welcome to the forum- you can learn a lot here!

I'm not sure about your analysis, but most of us take off and land in "FLY-MANUAL" just to avoid any wonky behavior.

Any problems once you are airborne?

If you haven't already, we highly recommend you map that to your A or B button in case you ever need to take over manual control quickly.

 

[DaveB]

It looks to me that the pitch rheostat is stuck at 2000. If you can find some contact cleaner, you could open the controller and spray some in it and then move it around and then retest it. IF you can find someone to loan you one for a quick check

 

[XevetS]

Pull the battery on the controller give it a few minutes and reinstall, also you can try a stick calibration,

see this>> https://3drpilots.com/threads/recalibrate-sticks-how-to.364/

 

[k4kfh]

Hey guys. I ended up (partly out of curiosity, and partly out of utter panic that my controller might be toast) opening the controller and checking all the connections. All was well in there. Then just for good measure I performed a factory reset (back to fresh Open Solo 3.0.0). Right now everything seems to be working alright according to reading stick values in Mission Planner, although we won't know for sure till my props come in the mail tomorrow.

NEW SAFETY MEASURE:
As you can imagine I am still a little gunshy after this, especially since I never really identified exactly what caused the pitch stick to do this. And the scary thing is that there is no plausible way I could've detected this issue before arming the Solo and prevented the crash. So I decided to use DroneKit and make a small Python script to sanity check the sticks on the Solo.

It checks for each stick to exhibit a certain amount of change over a period of 10 seconds. So you run the script, then move the sticks in all directions wildly. If the script detects sufficient change in all the axes, it tells you you're safe. If any one channel doesn't show sufficient change, however, it will display a very obvious "DO NOT ARM!" message and tell you which channel(s) are being problematic.

This eliminates the need for a PC with Mission Planner to perform what I am now considering an essential pre-flight sanity check. I can put the script directly on the Solo's IMX6, then set it up as a remote command within Solex. Then all I have to do is tap the "Stick Sanity Check" button in Solex, and wiggle the sticks around. In 10 seconds I will get a pop-up that either says "Sticks passed all sanity checks. Safe to arm." or "DO NOT ARM VEHICLE!"

If anyone wants to implement this on their own Solo, please reply and let me know, and I will make a tutorial.

I also talked to the Open Solo guys, who passed along this issue to the ArduCopter team in the hopes that "centered sticks" will become a new arming check (that would have prevented this crash without a doubt). Although that will probably take a while to become a usable feature, so in the meantime I will be making use of the script I wrote.