Autopilot swinging left/right

[FSGOM]

After update to 12.2 our autopilot in Mode TRK+ALT swings left/right (+- 5 degrees, ~ 8 seconds left ,then right and so on) to keep the direction. :-?

Any hints for that?

Thanks
Mike

 

[Schorsch]

Maybe try the relatively new magnetic compass inflight calibration?
I get large swings like you describe on the autopilot only in slow flight when not using my rudder (messy slips&skids).

 

[FSGOM]

We did this now but there was no change of this situation.

@Dynon support: Do you have any ideas?

 

[dynonsupport]

We sent you a PM on how to send us a Dynon Diagnostic File.

We did this now but there was no change of this situation.

@Dynon support: Do you have any ideas?

 

[FSGOM]

Thx, we will do this next week.

Could it also be that the servos shear screws are broken ??

 

[Dynon]

If the shear screw is broken, it would normally not have any authority at all with the servo. But it's worth investigating whether there's a mechanical issue, whether the shear screw or other issues that can affect performance like excess play.

 

[DY351]

We sent you a PM on how to send us a Dynon Diagnostic File.

We did this now but there was no change of this situation.

@Dynon support: Do you have any ideas?

Hi Dynon
On my plane WT9 the same thing than FSGOM
Il you give me the way for create a diagnostic file...
Thanks for your help
DY351

 

[dynonsupport]

Sent you a PM.

Hi Dynon
On my plane WT9 the same thing than FSGOM
Il you give me the way for create a diagnostic file...
Thanks for your help
DY351

 

[rfazio1951]

My plane was doing that also. All I did was to up the sensitivity on the roll axis, fixed. I noticed that if the aileron trim was off a little it would tend to do it more. Trimming the ailerons first and then engaging the auto pilot helps a lot.

 

[FSGOM]

Settings are and where ok , no change from our side.

In the picture you see the swing (violet line), a continuous sinus wave.

 

[FSGOM]

Solution: It was the shear bolt , replaced it and now everythinb works fine!
Thanks Dynon Support

 

[jakej]

thanks for being open about the result, I'm sure it could help others   
For me, a note to self - engage a/pilot & gently apply pressure on stick to move ailerons with someone to hold them, do not clamp them in place otherwise you can/will break the screw if it wasn't broken 

Jake J

 

[tedbain]

Michael, do you why the shear screw failed?

Ted

 

[swatson999]

How was the servo controlling the ailerons if the shear pin was, er, sheared? Shouldn't that effectively disconnect the servo arm from the actuator?

 

[swatson999]

This is also why I think there should be a TEST mode for the A/P which should be exercised during run-up (like most other autopilots I've used)...put it in TEST mode, dial the heading bug and left and right and confirm the stick moves left and right, etc., then A/P to SDBY or OFF.

 

[jakej]

When the shear screw breaks there can, & sometimes is, be some drag in the system caused by 1. the broken screw end 'dragging' against the servo arm & 2. the arm being held firmly by the arm screw (not shear screw) & nylon washer -with light ailerons the only way (at the moment) to confirm for sure is to test at the aileron after auto pilot engaged.

Just a couple of thoughts here - Maybe there could be 2 ways to 'fix' this 1. have the machined part of the shear screw moved slightly back towards the servo & 2. have slightly less tension on the arm screw.

Jake J

 

[bernardsv]

I frequently see the 'Roll Slip' message displayed with A/P engaged. Will it cause premature failure of the shear pin if I continue to leave the A/P engaged in this situation? Or is it advisable to disconnect it to spare stress on the system?

 

[dynonsupport]

This is also why I think there should be a TEST mode for the A/P which should be exercised during run-up (like most other autopilots I've used)

Steve,
You can engage the AP on the ground and do your own test as you described if you consider the AP to be a flight critical pre-flight item. I believe this is how you test something like an S-Tec as well, not a specific "Test" mode. We specifically allow engagement on the ground so you can test.

Not sure you need to do all 70 steps on each pre-flight like S-Tec wants you to do

What does the test modes do that you are familiar with that just engaging the AP wouldn't do?

We also have the deeper test in the setup menu as well.

 

[dynonsupport]

I frequently see the 'Roll Slip' message displayed with A/P engaged. Will it cause premature failure of the shear pin if I continue to leave the A/P engaged in this situation? Or is it advisable to disconnect it to spare stress on the system?

This is not an issue. It's fine to leave it engaged with "Roll Slip" occurring. Do make sure your roll torque is set to 100%.

 

[dynonsupport]

How was the servo controlling the ailerons if the shear pin was, er, sheared?  Shouldn't that effectively disconnect the servo arm from the actuator?

As Jake says, there's always some residual friction that allows the servo to impart some force on the controls even when the shear pin is broken. In planes with very light controls, this could be enough to sort-of fly the plane. Isn't the single finger on the stick control forces one of the things we love about RV's?

The point of the shear screw is to allow a human to fly the plane after it breaks if the whole servo seizes up. A little friction from the washer and other elements doesn't prevent this at all, so it probably isn't something that needs to be "fixed". It's just something that can make diagnosing a broken shear screw a bit harder for someone that isn't used to the way a properly functioning autopilot feels.