Potential FATAL ROLL due to Roll Servo bracket install error

[jjoffe@me.com]

Do not forget to install the rivet that holds the roll servo bracket on the torque tube - If it slips it can get to a position where it is directly in line with the servo arm and will jam full left with no chance of recovery from the roll.

The manual shows that rivet but it is easily overlooked - Be Aware.

I would recommend checking the position at least every 100 hours and possibly whenever you have a situation where you are forcing the stick in a roll against the force of the autopilot.
ps : when drilling that rivet hole be sure to use a drill stop.

[Andy]

Can you provide a bit more detail of your hazard analysis.

As I understand it, the purpose of the rivet is to prevent the roll servo torque link slipping on the roll torque tube. If the rivet was not fitted, or has sheared, how can the servo torque link jam the roll torque tube? Doesn't the torque tube slip inside the torque link clamp allowing full roll control?

I am a bit sensitive to autopilot servo linkage failure as I did experience a disconnect of the pitch servo bridle cable in my FX-3. The disconnected bridle cable wadded up round the pitch servo capstan. No control jam just a marked loss of AP pitch authority. Fixed with the new metal clamp plates and new bridle cable.

If you have the original nylon pitch bridle cable clamps probably should fit the newer aluminum ones.

[jjoffe@me.com]

Andy hi
The amount of slippage was minimal and the clamp was quite firmly attached - I am not sure how it rotated into the position it reached but I think the pressure of the elbow joint that was then 180 degrees forced the clamp
against the torque tube adding to the friction. I never really gave it much thought at the time but would have crashed if I was airborne when that happened. Since I was concerned about hurting something I never used that much force to attempt to get the stick back to level - if I was flying i would definitely have used a lot of force.

[Andy]

Sill trying to understand this and not doubting that you had cause for concern.

In a normal roll servo installation all the links are at right angles to each other when the ailerons are neutral (torque tube link vertical, servo arm vertical, link rod horizontal) at least that's how I remember it and what I see in the roll servo installation drawing.

I don't have a drawing or photo of the link geometry with the stick on the roll stops but I doubt any of those linkages reaches 45 degrees. Can you please provide a rough sketch of the linkages when they reached that locked 180 degree position.

When I got my FX-3 home I evaluated the stick force required to overpower the AP servos with the stick driven to the stops. I didn't like the force I had to apply to override the servos and I reduced the AFCS pitch and roll authority from the CubCrafters settings. The override force was reasonable with stick centered but, I thought, far too high with stick at or near the stops. No, that wasn't paranoia. I had experienced an autopilot nose down pitch hardover during phase 1 flight test.

I can't see the link rod in any of my photos but I can see that the anti-slip rivet is installed. One less thing to worry about.

[Andy]

Was the roll servo stop bracket installed?

[jjoffe@me.com]

Sill trying to understand this and not doubting that you had cause for concern.

In a normal roll servo installation all the links are at right angles to each other when the ailerons are neutral (torque tube link vertical, servo arm vertical, link rod horizontal) at least that's how I remember it and what I see in the roll servo installation drawing.

I don't have a drawing or photo of the link geometry with the stick on the roll stops but I doubt any of those linkages reaches 45 degrees. Can you please provide a rough sketch of the linkages when they reached that locked 180 degree position.

When I got my FX-3 home I evaluated the stick force required to overpower the AP servos with the stick driven to the stops. I didn't like the force I had to apply to override the servos and I reduced the AFCS pitch and roll authority from the CubCrafters settings. The override force was reasonable with stick centered but, I thought, far too high with stick at or near the stops. No, that wasn't paranoia. I had experienced an autopilot nose down pitch hardover during phase 1 flight test.

I can't see the link rod in any of my photos but I can see that the anti-slip rivet is installed. One less thing to worry about.

Hi these pics are from original install date about 2 years ago

[Andy]

I do not see the servo stop bracket in your installation photo. That could be far more critical than the missing ant-slip rivet.

This is a snip from the CubCrafters roll servo installation drawing -

[jjoffe@me.com]

I do not see the servo stop bracket in your installation photo. That could be far more critical than the missing ant-slip rivet.

This is a snip from the CubCrafters roll servo installation drawing -

Andy - does it really make a difference with stops on the torque tub roll side? - what could be an issue - I left it off on purpose thinking the stops on the torque tube rotation would suffice??

[Andy]

Andy - does it really make a difference with stops on the torque tub roll side? - what could be an issue - I left it off on purpose thinking the stops on the torque tube rotation would suffice??

Well they clearly didn't suffice as you seem to have experienced the over-center locking that Garmin cautions about. Could that link misalignment have been possible with the servo stop bracket installed?

It seems unwise to reject the instructions of the servo manufacturer and the kit designer unless you have a very good understanding of the possible failure modes.

[jjoffe@me.com]

Well they clearly didn't suffice as you seem to have experienced the over-center locking that Garmin cautions about. Could that link misalignment have been possible with the servo stop bracket installed?

It seems unwise to reject the instructions of the servo manufacturer and the kit designer unless you have a very good understanding of the possible failure modes.

Good points made - I will add it