What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

RV-14 Garmin GSA 28 Over Center

ShortSnorter

Well Known Member
I had an interesting experience this afternoon when I was doing some ground testing of a standard install G3X AP system. After I got home I did some "VAF'ing" to see what others experience has been and ran across this thread. Post #33 was exactly the scenario I was testing.

To be as brief as possible what I was doing was verifying that the torque on the GSA 28 was low enough I could over power. Previously I had done the directional control tests ensuring the wiring was correct, but had never done the "over power test." To my surprise, when I tested the right aileron up movement force (trying to over power with left stick), I heard a THUNK. At that point I had no left aileron (jammed controls) but could move center to right freely. I knew the only thing it could be was some sort of linkage issue so I pulled the inspection panel and found the first picture below. I couldn't for the life of me figure out how it went over center as I had check control stops and verified with the servo arm travel positions.

I was able to move the arm back from the over center position and verify the mechanical control stops and the the corresponding position of the servo arm (2nd picture). I'm not sure how, but my theory of what happened was as it reached it maximum limit, the servo arm was JUST far enough to allow it it pivot over center with enough force. I verified this by manipulating the servo arm by hand and pivoted past the mechanical stop to the over center position (1st picture).

(3rd Picture) I moved the push rod linkage to the center hole (from the hole nearest the shaft) and due to the changed geometry, was unable to replicate the over center position by hand or with AP servos active. I am definitely planning on installing the mechanical stops that came with the servo! The only reason I did not install was the instructions say as long as the geometry does not allow over center then the stops were optional.

Please go check your AP Servos!!! Even though though they check out visually through full range of motion you could have a situation like mine. I doubt there would ever be a real world scenario of full AP commanded aileron deflection that could potentially cause this scenario, but its worth a piece of mind to check it out.

**Not sure why the pics are turned?!
 

Attachments

  • IMG-4598.jpg
    IMG-4598.jpg
    248.4 KB · Views: 691
  • IMG-4599.jpg
    IMG-4599.jpg
    244.2 KB · Views: 637
  • IMG-4600.jpg
    IMG-4600.jpg
    186.2 KB · Views: 644
Last edited:
Plans call for an AN3 bolt for the rod end attachment, why are you using a screw?

Great question! I used a screw because that's what the plans call for ;) However, it appears I connected the screw from the wrong side.


Appreciate the feedback guys - My point with this post was that I installed per plans and tested range of motion with no obvious issues until the servo was powered and it was able to pivot within the rod end past over center. It took quite a bit of force to replicate by hand which is why I didn't catch it when testing normally.
 

Attachments

  • connection.JPG
    connection.JPG
    35 KB · Views: 537
Last edited:
Great question! I used a screw because that's what the plans call for ;) However, it appears I connected the screw from the wrong side.

.

The plans also appear to call for the attachment in the lowest hole, not the top hole. They probably specified this for the reason you experienced.
 
The plans also appear to call for the attachment in the lowest hole, not the top hole. They probably specified this for the reason you experienced.

It's hard to tell in the pictures, but the first two photos are the original install with using the lowest hole in the arm.

The final photo is where I moved it to the middle hole in the servo arm which limits the travel of the arm significantly.
 
Missing stop

I thought the Garmin servo came with a stop/ bracket that attaches to the servo and prevents this? That bracket/ stop appears to be missing in your photos. I think the missing bracket is a safety of flight issue and must be corrected.
 

Attachments

  • 327D5992-A129-4930-85AE-9C1CB1B1BE6C.jpg
    327D5992-A129-4930-85AE-9C1CB1B1BE6C.jpg
    211.6 KB · Views: 496
Last edited:
It's hard to tell in the pictures, but the first two photos are the original install with using the lowest hole in the arm.

The final photo is where I moved it to the middle hole in the servo arm which limits the travel of the arm significantly.

See that now. Misunderstood the pics.
 
Admittedly I am paranoid, but even when using mechanical stops I like to keep away from being close to center at either end of the rotational arc, preferably about the same number of degrees away from center on each end. If I can't manage that I do what you did - move the linkage out another hole on the lever.

The thought of a control linkage going over center while in flight gives me nightmares.
 
You should not move the arm to another hole in the servo arm or you will change the torque, it was spec'd there for a reason.
Adjust the length of the pushrod if you need to change the position of the servo arm. The arm should be approx. perpendicular to the servo when the controls are in the neutral position and swing approx. 45deg each way. I install the Garmin supplied servo stops on all my installs for safety (but have not done a 14 yet so maybe something prevents that?).
 
Last edited:
Garmin Limit stops

OP. Thanks for sharing this issue.


I’m interested if any RV-14 builders have been able to use the Garmin supplied limit stops. I’ve not been able to see a way in utilizing them due to how the brackets that support the servo prevent the limit stops from attaching to the servo. This is an issue on both the roll and pitch.

If you have been able to modify or make something work for stop limits please post some pic’s.
 
You should not move the arm to another hole in the servo arm or you will change the torque, it was spec'd there for a reason.
Adjust the length of the pushrod if you need to change the position of the servo arm. The arm should be approx. perpendicular to the servo when the controls are in the neutral position and swing approx. 45deg each way. I install the Garmin supplied servo stops on all my installs for safety (but have not done a 14 yet so maybe something prevents that?).

I may be missing something because when the controls are neutral and the servo is perpendicular there is a disproportionate amount of throw on the servo with up aileron vs down aileron. When I switched to the center position it limited the servo throw as seen in the picture. I suppose I could use the innermost hold but to do so I would have to adjust the linkage quite a bit beyond a perpendicular starting position. Since the opposite direction throw doesn't go as far (down aileron), the may be an option?

The fact that the servo stops were not installed - The fact that the only way this whole thing can happen is if you let the servo work against the linkage when the aileron stop is hit makes me wonder if there are others out there set up the same way. I guess my point is, full range of motion with the control stick on the ground cannot make this event occur. I cannot see this happening in flight because the autopilot would never be engage with full aileron up deflection.
 
... I guess my point is, full range of motion with the control stick on the ground cannot make this event occur. I cannot see this happening in flight because the autopilot would never be engage with full aileron up deflection.

But if it ever did happen in flight.....
 
I may be missing something because when the controls are neutral and the servo is perpendicular there is a disproportionate amount of throw on the servo with up aileron vs down aileron. When I switched to the center position it limited the servo throw as seen in the picture. I suppose I could use the innermost hold but to do so I would have to adjust the linkage quite a bit beyond a perpendicular starting position. Since the opposite direction throw doesn't go as far (down aileron), the may be an option?

The fact that the servo stops were not installed - The fact that the only way this whole thing can happen is if you let the servo work against the linkage when the aileron stop is hit makes me wonder if there are others out there set up the same way. I guess my point is, full range of motion with the control stick on the ground cannot make this event occur. I cannot see this happening in flight because the autopilot would never be engage with full aileron up deflection.


Differential ailerons. The up aileron travel is greater than the down aileron travel by design. Don’t center the arm based on equal travel.
Like Walt said, the servo arm must be perpendicular at neutral controls.
I have a 10 which is the same bits here. There’s no way it can over center if installed right.
That said I also have the servo stop brackets installed.
It appears that the -14 bracket has provision for the servo wiring molex connector which stops you installing the servo stops.
If I had to choose between servo stops and a plug holder I know which I’d be choosing. I don’t get why vans has over complicated this.
 
I was able to install in roll but could only use two of four mounting screws. The internal stops are hit well before an over center position but I installed anyway. The mounting bracket for the pitch servo did not have the required holes to pass the screws from the over center bracket into the servo. Since the internal stops are also hit well before an over center condition could be reached I didn’t install the Garmin bracket on the pitch servo.

Tom thanks for the photos - Is there a reason you chose the middle hole on the roll servo? Initially I started with the innermost hole but moved it for reasons stated in previous posts.
 
You do want maximum servo arm throw, as close to the garmin stop on each side as practical normally. The less throw the servo arm has the more sensitive/jittery the the AP will become as it take less movement for the same amount of deflection.
 
Last edited:
where did you get the drawing that shows a screw to be used instead of an an3 bolt? Garmin g3x manual shows a an3-7a bolt is to be used there, torqued to 34-46 in lbs. A screw with an angled head just seems wrong in this application, and extremely unsafe.
 
Hi Tom - I had a similar over center issue when setting up my pitch servo. I wound up shortening the push pull tube which connects the GSA 28 to the elevator bell crank. I was scratching my head and wound up calling vendor and manufacturers before doing so. Later, after I had riveted the turtledeck I had to reinstall everything aft and was very careful with all the washer, nuts and bolts called out in the G3X manual and in plans. And torques. Then I discovered I had inverted the bell crank on initial install. Because the hole into which hardware attaches on the bell crank is off center, my original error of inverting bell crank meant I had an over center condition with the proper length push pull tube. But after discovering the error the tube I had cut was now too short. So I had to cut and retap another one.
This is a long way of saying check the G3X and aircraft plans meticulously. And you will probably find your error like I did. I think the arrangement is well designed for the popular Vans models.

Warm regards,

Drum
 
where did you get the drawing that shows a screw to be used instead of an an3 bolt? Garmin g3x manual shows a an3-7a bolt is to be used there, torqued to 34-46 in lbs. A screw with an angled head just seems wrong in this application, and extremely unsafe.

It’s in the RV14 plans

4FF18CB4-A222-431D-B9C9-448AE3DA74C5.jpg

Maybe the factory can chime in with why. I’m sure there’s a reason.
 
I'm quite certain they spec'd the fastener to gain a little more clearance.
Personally I've always used the AN3 as called out in the Garmin documents with no issue.
 
Last edited:
I'm not an engineer, but it would seem to me a screw designed to be in a countersink should not be used in a flat washer, where only a small surface of the angle on the screw head is actually touching the washer.
 
Deviation from plans...

Understanding that clearance between the bearing attach and the servo body was a concern, I used a AN525-10R16 instead of the AN509.
 
As a non-RV builder/ home built owner (but still find endless help, information, and enjoyment in this forum) who had more sketchy construction manuals to work with, this discussion points out an issue that I've seen here several times. That is, just because the Van's manuals and instructions are so good and comprehensive, you should not blindly follow them. Remember that at the end of the day you are the manufacturer, not Vans (not to malign the company, but face it, there are so many places already where builders deviate sensibly). So, a situation like the flat head screw and washer in these AP instructions should raise a flag. Go ahead and call Van's and ask what they were thinking, but clearly using a threaded screw here in sheer is a mistake. This is an application that calls for a close tolerance fastener, non-threaded in the area of load, such that repetitive and reversing loads do not hammer it to looseness. So, keep your eyes open, question when necessary, and do what makes aviation engineering and safety sense.
 
To answer the question why the screw is used in lieu of a bolt: Clearance. An AN3 will not clear the servo when mounted per the plans.

Also, I did install the over center bracket as Tom showed in post #15.
 
I'm also curious about the stop brackets. The plans do NOT call for their use. I've looked all over section 56 for reference. If installed it looks like the mounting bracket would be cause issues and need washers to level the stop bracket.

I also show them using the middle hole on the throw arm on the roll servo, but the farthest hole on the Pitch Servo.

Thoughts? What have people done?

Bungalow
 
I'm also curious about the stop brackets. The plans do NOT call for their use. I've looked all over section 56 for reference. If installed it looks like the mounting bracket would be cause issues and need washers to level the stop bracket.

I also show them using the middle hole on the throw arm on the roll servo, but the farthest hole on the Pitch Servo.

Thoughts? What have people done?

Bungalow

Check out the pictures in post #15 - This worked out fine for me as well.
 
Back
Top