Page 1 of 1

Autopilot servo reading

Posted: Mon May 26, 2025 5:17 pm
by pavel569
Is it possible to reset zero servo position? In my capstan servo pitch setting, the pitch down position is about 3500, when moving elevator up the servo numbers go 4095 to zero up to about 2500. But there is no plus or minus numbers. My numbers are as following - (down) 3500 - 4095 - 0 - 2500 (up).

Re: Autopilot servo reading

Posted: Tue May 27, 2025 8:41 am
by rainier
Sure - just remove the shear screw and rotate arm by 90 or 180 degrees - the boss has 4 positions for the screw.
Fix screw in place with blue loctite please. Also the boss itself can be rotated - there is a big grub screw on its side - again please ensure you use a drop of blue loctite here.
Use of the loctite is crucial as otherwise these fastners will work themselves loose introducing play into the coupling.

Re: Autopilot servo reading

Posted: Tue May 27, 2025 3:13 pm
by pavel569
Thanks Rainier, will do as you suggest - rotating 90 deg should do it. I didn't think of checking the servo position reading before attaching the capstan drum.

Re: Autopilot servo reading

Posted: Sun Jun 15, 2025 3:08 pm
by pavel569
rainier wrote: Tue May 27, 2025 8:41 am Sure - just remove the shear screw and rotate arm by 90 or 180 degrees - the boss has 4 positions for the screw.
Fix screw in place with blue loctite please. Also the boss itself can be rotated - there is a big grub screw on its side - again please ensure you use a drop of blue loctite here.
Use of the loctite is crucial as otherwise these fastners will work themselves loose introducing play into the coupling.
Hi Rainier, I've rotated the drum to a position where my pitch up/down limits are between 304 and 3644 (travel about 290deg) but the servo is still not working as it should be. When doing the preflight check the bank servo moves correctly left/right to its limits and then center, but pitch - when flight stick is in slightly up initial position the servo will start moving up (screen reads "pitch up") then continue moving up (screen reads "pitch down") and then move down (screen reads "pitch zero") so there is no change in servo move direction when "Pitch down" is on the screen. When flight stick is initially in pitch down position - the same happens in opposite direction but the movement never changes - "Pitch up" "Pitch down" are in the same direction, "Pitch zero" moves in the opposite direction. What am I missing here?

Re: Autopilot servo reading

Posted: Wed Jun 18, 2025 7:04 am
by rainier
The pitch or bank servo must stay under 180 degrees of rotation. Only the Yaw servo may go to under 360 - this is why it has three calibration points instead of two.
The calibration points are used to decide on servo direction and are used during the autopilot test.

You can do a "hack" - when you set the left/right and up/down positions make sure you do not exceed 180 degrees and the arm does not go through zero.

The autopilot does not use the limits other than to decide which direction the servo must go (the circle is split into two and the section that is less than 180 degrees decides the direction).

Do not engage the autopilot when the stick is outside of your set limits as this could lead to the initial correction going the wrong way but it would be unlikely that you would engage when at extreme maneuvers.

As a side note I am aware of a case during a round the World flight the aircraft fitted with our AP encountered storms over the ocean and flew into IMC with the pilots loosing control and orientation in the severe turbulence. They rapidly lost altitude and where unsure of attitude. At the last opportunity they engaged the AP - the AP recovered the aircraft with not a lot of room to spare just as they broke through cloud bottom over the ocean.
Just mentioning this as the "hack" described above may not do this if the AP is engaged at extreme attitudes with the controls not within the 180 degree servo limit.

Re: Autopilot servo reading

Posted: Sat Jun 21, 2025 4:12 pm
by pavel569
Thanks for the explanation Rainier. I have designed and made a custom Capstan (easier to install and adjust than the original) and even though I don't have the exact dimension of the original drum, I think mine is a bit bigger. But still, with the 170deg limit, the active travel of the pitch A/P servo is only 55mm which is even less than using the standard servo arm. Why to limit the mechanical advantage of Capstan spool with software? It is possible for the yaw, why not for pitch? Is it possible to add option in the selection menu for Capstan servo with larger range as for the yaw? I could make a larger Capstan drum which would cover my cable range even with the 170deg limit but then the movement precision and torque might create another problem.