RV7A, Dynon SV-32, Pitch Servo install issue

[wirejock]

RV7A, Dynon SV-32 AP Servos.
I have a problem rigging the pitch AP servo. I rigged the elevators and controls. Everything moves from full up to full down smoothly to the stops and the sticks are positioned correctly to plans. Bellcrank is vertical with the elevators nuetral in trail. Deflection degrees up and down are close to max spec. I think it's correct.
Dynon servo bracket is installed per plan. I checked. Hole in the bellcrank is correct. I checked. The drawing shows a dimension of 6" from centers for the control rod and bearings.
The AP pitch servo arm is dangerously close to locking up over center at full down pitch. Photo is actually adjusted a bit. At 6", the servo arm is almost at 90 degrees. Second photo is full up. The anti rotation bracket should stop it, but I would feel better if it were no so close. The servo arm needs to be further aft at center. Dynon says it should be vertical at neutral. I can adjust the rod. That works and will move the arm position where it can't go over center either way. However, the threaded tube is too short. If I rotate the tube all the way it falls off. So if the jam nuts got loose the tube can fall out. I also tried it by screwing the rod together 1/2 way, plus one turn. That works but if the jam nuts get loose, the rod is only held by one thread turn.
I ordered two Vans hex flap pushrod tubes to try a longer pushrod tube.
I can shorten the pushrod distance from the bellcrank to the elevators. That will work, but it's not how Vans says it should be rigged.

Did anyone notice anything like this?
Have I got something rigged wrong?

 

[N890GF]

I experienced something sort of similar, with full forward feeling "squishy" because the servo was nearly reaching its limit. I ended up adjusting the main push-tubes to slightly shift the bell crank aft while maintaining sufficient threads in the AP pushrod. My justification for this is that the original design of the RV didn't have Dynon AP servos, and was probably sufficient for the time, but now that we've got additional hardware in play, its okay to deviate slightly without compromising the performance. I still get full up/down on the elevators and am no longer at risk of over-centering the AP servo or backing out the threads. at the end of the day, my goal is to be safe through all flight envelopes, and I don't really plan on going full elevator down during normal flight (but thats just me).