Loss of RC control when switching modes

Ziblube · February 22, 2022, 2:28am

Hi everyone,

I recently had my plane crash after switching out of RTL mode. The plane stopped responding to control inputs. It seemed as if the control surfaces were locked in one position and were not moving. This resulted in a gradual roll to the left with some down elevator.

The plane had returned to launch and was halfway through its first circle when I switched from RTL to FBWA. In FBWA the plane did not level itself or respond to control inputs. As it would not respond, I flicked RTL on and off again, but still had no control. I then tried MANUAL and FBWA, still no control. The result was an inverted 60deg nose down impact at about 95km/h. It was quite an impressive crash to see really

I am not sure of my abilities to interpret the log, however, to me it seems that my RC inputs were recognized, it also seems that there were servo outputs but I am not sure if they match the RC inputs. I reasonably sure that the servo outputs don’t match my inputs when in MANUAL mode though. Although there were servo outputs, there did not seem to be roll and pitch changes corresponding to the servo outputs. But as I said I don’t think I am very good at interpreting the log so I am hoping someone might be able to help track down the issue.

After the crash I had a look at mechanical issues, like servo to elevon connections or unplugged servos but did not find anything. Besides, I don’t think it was a hardware failure because when one control surface stops working on a flying wing, pitch inputs cause the plane to roll. Which did not happen. The crash was a continuous gentle roll.

The FC is a Matek Sys H743 v2 and it’s on a flying wing.

RCIN C1 = Roll
RCIN C2 = Pitch
RCIN C3 =Throttle

RCOU C1 = Throttle
RCOU C2 = Elevon Left
RCOU C3 =Elevon Right

Things to note about the log:
You will see that at about 12:40:31 the plane is in a nose dive and there is a large pitch down RC input. This was deliberate, the plane had rolled inverted at this point and down elevator would have been the fastest way to avoid the ground… if it was responding to control inputs

I have my modes setup on two switches on my radio. Both switches are three position switches.
Switch A positions are: MANUAL - FBWA - AUTOTUNE.
Switch B positions are: OFF - AUTO - RTL.
When switch B is set to OFF, switch A is active. But when switch B is on AUTO or RTL it overrides what ever switch A is set to. I am only explaining this so the mode changes in the datalog might make sense. You will see in the log that AUTO is on for a brief period when switching in and out of RTL.

If anyone has an ideas or could point me in a direction it would be greatly appreciated.

I could not attach the log file as it was too large so I have put the log and param files in the OneDrive folder below. Let me know if you cant open it or want the files by another means.

Cheers

Allister · February 23, 2022, 10:47am

Have you been using the trim on your radios? What appears to be your resting stick positions for roll is about 1459, however RC2_TRIM is set to 1504. This means that you are commanding a roll to the FC at all times, even when you think you have set the stick in neutral.

I suggest you reset the trims on your radio to neutral. (and if possible on your radio disable the radio trims) Re-do the radio calibration, and then set SERVO_AUTO_TRIM to 1.

Ziblube · February 24, 2022, 1:04am

Hi Allister, thank you for having a look at the data.

I did use the radio trims for this flight. I had just changed to a bigger diameter prop and it was causing a larger torque reaction so I trimmed the roll out on the radio and kept flying.
I think I can disable trims on my radio, so I’ll do that and set SERVO_AUTO_TRIM too, thanks for the pointer.

Could this have caused the plane to stop responding to control inputs? Even when in manual mode?

Allister · February 27, 2022, 10:53pm

Sorry for taking so long to get back to you.

The log file shows the outputs were still sending signals servos, even during manual mode. However there doesn’t seem to be a lot of reaction to the outputs. To me this suggests some kind of mechanical failure. Either with the servos them selves (loss of power maybe?), issue with the airframe, or significant shift in the c of g (battery came free).

Ziblube · March 2, 2022, 1:56am

No worries, take as long as you want I am appreciative for any help I can get.
Sorry for this large response too, I have not mastered the art of explaining things simply yet.

Though the datalog shows outputs to the servos, I am wondering if that output actually got to the servo pin. Do you know at what point or how does the data logger gets the PWM values to save? For example, does the logger receive a calculated PWM value before it is sent to the output pin? Or is the PWM value read from the output pin?

I just feel like the control surfaces were not moving. The link bellow is a video of the last few moments before the crash. The footage shows the roll and pitch rate to be fairly constant before it hits the ground. If the control surfaces were moving, I would have expected it not to be a constant(ish) rate.

I am leaning away from a mechanical failure as the servos still work and are connected to the control surfaces, and the way the batteries are put in the plane does not allow them to move during flight. I can’t say whether there was a loss of power to servos though. However, the servos are powered from the flight controller and so is the camera, as the camera and controller (somewhat the controller) were still working, I am assuming that the servos were receiving power.

I have found two things going through the data that I am unsure about. The first is the current vs voltage vs throttle out. For the whole flight you can see that the current follows the same trend as the throttle out, and that the voltage dips as current increases. Both of these are logical behaviors. Though at the moment RTL is switched on, the throttle out drops but the current increases for some reason. I don’t think the current actually increased here though, because you can see that the voltage increased as the throttle decreased, which is what would happen with a reduced current.

The other thing I found is that at the moment RTL is turned on, AETR.ail drops to -4500.

I am not sure what either of these plots signify, but looking back at previous log files, I have not seen either of these behaviors before.

Allister · March 3, 2022, 5:32am

I think you may be on to something regarding the voltage. Even in the video you can see the current draw continuing to climb, in spite of the lower throttle setting. I think you should look into what caused that higher current.

zeta696 · June 19, 2024, 5:25pm

Hi there - did you ever figure this out? I had the exact same thing happen to me this morning. I was flying my nano talon and having a great time - all was well. Was testing three different modes as this is a new plane/setup. Did an auto takeoff, switched to FBWA, and then eventually RTL. Once in the air and high enough I just played around by switching between FBWA and RTL a few times. Then, for whatever reason, the nano talon simply stopped responding and the motor cutoff and it flew straight ahead in a slow descending form eventually landing (thankfully w/o damage) in a small tree about 100 yards away. Craziest thing I’ve seen in awhile. Anyhow, when I recovered the plane, the battery was still at 56%. All I know is that this happened when switching b/w RTL and FBWA. Radio is ELRS/Radiomaster pocket.