I have been using Pull Pull for the Rudder and Elevator controls on most of my aerobatic planes. I find them rather easy to install, light weight and very good performance.

However.... I have used tiller arms. pulleys, direct connects etc... but have never been really satisfied with the non-linear motion of the cable pull from a straight arm. That is to say, when the servo or tiller arm turns, the cable does not get consistent, linear motion. They slack at one point.

I am wondering how some of you have addressed this.

If you are using tiller arms it's important to have the servo arm exactly the same width as the rudder horns. If not one side will get loose or the system will bind. I use this system with a single servo set-up, with multiple servos I like a bearing bell crank, or wheel depending on the application.

The way to avoid assymetrical slack in a closed-loop system is to ensure that the control linkage is a perfect parallelogram; IOW, the distance between the connectiing points on the servo wheel/tiller bar/whatever are the same distance apart (from center) as the connecting points on the control surface horns; and the cables or pushrods are exactly the same length.

I dunno how to post multiple pics on here, so the pic here will show the tiller bar bellcrank I'm using, and the next post will show the rudder horn set up.

The idea is that the links on the bellcrank are exactly the same distance from each other (and the center point) as are the links on the rudder horns. That is the critical part. It is no big deal to adjust the cable tension so they are of equal lengths, if you are using the proper hardware.

Here's the rudder horn set-up:

H9 Taylorcraft. The width of the holes in the servo arm and control horn are the same size. The set up works great and it hardly cost anything to set up.

Du-Bro servo horns I see. Why does Hi-Tech supply such poor horns?

Dunno know about Hi-Tec horns, but neither JR nor Futaba makes horns big enough for what I needed; hence the large Dubros.

Steve

Cool, I didn't know that the distance between the connection points on the servo arm and the control contact points needed to be the same.

I started to fabricate two wheels for the ele. and rudder last night, providing equal spacing on the exit points of the wheel to the control horn spacing on the control surface.

I have read a few threads about cable length. Still a bit confused as some stated the cable lengths needed to be equal and others stated they made no difference. I dont know how differences in length can introduce problems

Otter, if non-linear motion is your focus, the only way to fix that is to use a wheel at the servo and another at the control surface. This way, the cable travel doesn't decrease at the extremes of servo throw, rather remains constant.

Cables getting slack at one point is another issue altogether, and is caused when the clevis holes on the control surface horn aren't exactly in line with the hinge line. For example, on a rudder pull-pull if a line is drawn from the clevis hole on the right control horn to the clevis hole on the left control horn, ideally it should pass exactly through the rudder's hinge line. Most of the time it doesn't, and this line passes behind the hinge line, causing slack in the cables as the servo moves to the end of it's travel. If this line passed in front of the hinge line, the cables would actually tighten up as the servo moves to the end of it's travel.

As long as the cables are tight and the rudder has zero play at center, it's really not a problem having slight slack as the servo moves, because if, for example, you give right rudder command, the air pressure on the right side of the rudder keeps the right cable tight, and prevents flutter, so the left cable being slightly loose doesn't matter. After all, BOTH cables work when the surface is centered, but only ONE cable is working when the surface is deflected, right?

Tim

Dubro heavy duty servo arms...........

On a 1/4 scale or larger airplane, standard size servo arms are an accident waiting to happen. I have seen standard arms fracture on 2 high dollar big planes. The planes were lost because they didn't want to spend the extra $$ on better servo horns. The two futaba servos in the wings also have the same servo arms.

I know, I'm a little paranoid. I don't use balsa pushrods anymore after seeing one break in flight. That is the reason for the Dave Brown pushrod system in everything I fly. I'll never use a clevis unless it has a metal pin. I've seen a couple of planes go in because the little plastic pin shears off. I double up servos on aileron and elevator on my planes larger than .40 size for redundancy. (saved one because of it) I also check the battery before every flight. I just really hate crashing planes. If I see something that causes a crash, I'll make sure I don't do it.

[QUOTE]Originally posted by Steve Campbell
[B]The way to avoid assymetrical slack in a closed-loop system is to ensure that the control linkage is a perfect parallelogram; IOW, the distance between the connectiing points on the servo wheel/tiller bar/whatever are the same distance apart (from center) as the connecting points on the control surface horns; and the cables or pushrods are exactly the same length."

This is exactly correct and the only way to make the system work properly.:thumbup:

[QUOTE]Originally posted by OUTCAST
[B]Outcast, consider a 6 inch pulley driving a 3 inch pulley via a fan belt. This isn't a perfect parallelogram, but there are zero slack issues, right? The reason is that the pivot points for both pulleys are centered. If one used a 3" wheel at the servo (or a 3" sheel for a tiller bar) and a 1 1/2" wheel at the control surface (exagerrated example, yes) the wheel movement would be multiplied by a factor of two (every 5 degrees of servo travel yields 10 degrees of control surface travel) but, as long as the control surface wheel's center point was exactly on the control surface hinge line, there would be zero slackening or tightening, just like the above pulley example, because essentially they're are the same. Even if they were a perfect parallelogram, there would STILL be cable slackening or tightening UNLESS the control surface horn's pivot point was exactly on the hinge line of the surface.

Ideally, one would use equal length servo arms and control surface wheels, but, most guys use arms rather than wheels, and arms don't give consistent cable movement. At center, when the arm is at 90 degrees to the cable, moving the servo 10 degrees may yield 1/4" pull cable travel (an example, because travel depends on the length of the arm, right?). As the servo arm continues to rotate to 40 degrees from center (and the arm is almost 45 degrees to the cable) , moving the servo 10 degrees now only yields 1/8" pull cable travel, because the some of the movement at the servo horn clevis is sideways, and doesn't pull the cable. If you could continue servo rotation to 85 degrees so that the arm is almost in line with the cable, 10 degrees of movement would yield almost zero pull cable movement. With wheels on both servo and control surface, the pull cable travel would remain constant for each degrees of servo movement, and allow 45 degrees of servo movement to yield 45 degrees of control surface movement, eliminating the need for, say, 3" servo arms/tiller bars and 1" rudder horns. to get that BIG travel. Matter of fact, as the planes get larger and use tiller bar arrangements like yours, the cables would be attached to the OUTER holes on the tiller bar, and the servo would drive the tiller using one of the inner holes. Your arrangement wouldn't allow sufficient travel for my plane, for example. I used a pair of JR 8411 servos (360 oz in torque total) with 3" servo arms on a pull-pull to my rudder. I also used floating bellcranks to eliminate the problem of my digital servos "fighting" each other and excessively draining the batteries. The rudder horns are maybe 2" across to give me the travel I need. Pics of my setup are here: http://www.********.net/pages/tmills/page15.html . Cables are tight at center, but do slacken slightly as servo moves, but, I've found that to be a non-issue, because as I said in my earlier post, two cables are working when the surface is centered, but only one cable does the work when the surface is deflected, the other cable does nothing. But.....you knew all that, I'm sure.

Quote:

This is exactly correct and the only way to make the system work properly.

Yeah, well, maybe you're right, and your way actually *IS* the only way. (JUST KIDDING!!)

Tim I agree with what you are saying as long as you use Pulley's not arms. With a pulley the belt always leaves the pulley 90 degrees to the center line, not so with arms. As a control arm travels through its ark the amount the push rod {or cable} travels is NOT linear. The rod will travel more per degree of rotation at 90 degrees to the center line than it will at 45 degrees as you stated, if the rudder arms are wider than the servo arms the distance between the cable pivot points will not remain constant and one side will go slack, if the servo arm is larger than the rudder the cables will turn into guitar strings. I agree that I would not be concerned about a loose cable, but a tight one......I gotta think about that when I'm awake. :^}
I really like your set-up :thumbup: