Hexa X Crash, RCInput shows full yoke with no input!

Hi all,

I am assuming I have an interference issue can anyone confirm my assumptions. I was in hover, holding level when the hex rolled hard and subsequently failed to maintain clearance from terrain…

The plot shows to me what looks like some kind of failure at the tx/rx . The large step inputs are obviously not from me, and the hex did the best she could to follow the erroneous inputs.

Does any have experience with behavior like this from a X9D+ and R9 receiver?

Is this another case of incorrectly configured receiver failsafe ?
Switch off your RC radio and check RCIN then, I suspect it will go to those positions.

Will check tonight. Thanks.

OK I see it now. I had checked the failsafe in the failsafe tab of initial configuration, but I notice now in the fine print that the value there is only for the throttle failsafe. The FS_OPTIONS was still set to 0, I believe a value of 1, Continue if in auto is the one that I want.

I turned on disarm logging so I could check RC input on the bench. I tried intentionally setting the failsafe mode to NOT-SET on the X9D and NO PULSES. The NOT-SET setting did not produce the full input that I saw. They were both holding right at 1500us.

Looking at the logs, it is odd that the throttle channel didn’t experience this step input. If it was EMI related I would have expected it across all the channels. I’d like to return to flight, but I really want to replicate and mitigate this issue first.

I noticed something odd today while preparing for flight. My artificial horizon was rolling 180deg out from actual. This didn’t make sense as the hex wasn’t flying as if the controls were inverted. I couldn’t think of what could have caused this, so I set my ahrs orientation to a different value, rebooted and set it back to standard. My GPS and FC are both oriented forward. After setting it back to the value it already was, the horizon matched. I conducted a cautious test flight after verifying all failsafes and the log shows a clean Des Roll unlike my previous error. Do you think there’s a chance that the horizon inversion and the crashes were related? Its just odd because it was flying totally normally even prior to my adjustment. Log attached to show how well she’s doing now.

@Andre-K thanks for taking a look and helping out

Always best to set up your receiver failsafe as “output absolutely nothing” and ensure the throttle failsafe works before flying.
Then if there is a RC/throttle failsafe the craft will enter RTL mode and use WPNAV (or RTL) params to get home. Arducopter wont change flight modes or pitch/roll/throttle if the RC channels are beyond limits (no output).

Good point. I double checked that my Rx was set to “no pulses” for failsafe. I also flashed the latest to my 2018 R9 units. The frsky accst/access situation is a hot mess right now. Time will tell but for now, all systems go!

I do not think the AHI strangeness was related, the later sounds like one i an million crazy init error that would be caught by arming checks anyway.
When you switch off RC, you should see a clear indication of RC failsafe, not only some specifig RC input.
Also, with the speaker connected, there should be tones on RC loss.