By Ted Sander

Making those floppy things on your plane move - Pull-pull can be the most solid and precise way to move your controls possible, but it is not without it’s own set of headaches that can make it a particular challenge to install. First, though, we need to do a review of servos, conventional pushrods, and attachments.

The overall goal of activating your control surfaces is to convert your servo’s rotary motion into a linear forward and back, that then moves a control surface in it’s own rotation around its’ hinge line. This should ideally be done by conveying maximum power, the perfect amount of movement, and by not inducing any slop into the system. Sometimes this can be a tall order in any airplane.

Now that we know what we want, lets focus first on precision. You have no doubt seen either reviews or ads claiming a particular servo’s centering ability - the precision by which it returns to the exact same neutral point. A corollary of this is its ability to go to exactly the same point when you input a given amount of control. The better a servo does this, and holds it, the more accurately you can repeat a given maneuver. It can be very frustrating to chase a plane around the sky because it refuses to stay straight and level, or to respond to a stick movement exactly the same way twice. In the servo, this precision is governed by the quality of the electronics, the quality of the shaft bearings, and the quality of the gears transferring the rotation from the servo motor to the shaft. Between

brands and types, the differences can be striking.

While your trainer may not care that you have the cheapest kinds of servos in it, because it takes a fairly large amount of control change to get it to respond, a hot fun-fly or other aerobatic plane will show every flaw in it’s system. Gear slop in the servo, or flexing of the output shaft due to poor bearings or normal wear, can produce a significant amount of change at the control surface. Your ability to accurately fly your plane is directly related to the precision of the servos that that plane needs. You may not ever need the digital coreless ball bearing variety, but you will suffer if you move up the performance scale and always stick with the $10 servo.

Mounting the servo can make a difference. Servos must flex in their mounts slightly to allow the grommets to do their job in protecting from vibration. Cranking down on those mounting screws to keep it tight defeats that, and can lead to very premature servo failure.

Think about the loads the flight surfaces place on the servo. The pushrods (or cables) are either pushing forward, or pulling back all the time. The servo will naturally rock forward and back slightly in its grommets with this movement. Always mount your servos in line with the direction of force, i.e.-the long way. A sideways servo will twist in its mount much more; destroying all of that precision you paid so much for. Give it a longer base to resist pushing and pulling from! More will be said about servo arms later, but in this section we need to focus on properly attaching the pushrod to the servo.

There is a huge variety of ways to do this, more than can be touched on here. The critical part is to guarantee that there is no slop, with no friction, in the connection. In that regard, the worst is the common “Z ” bend through the hole of the servo arm. Improperly done, it can by itself override all the high-buck precision you paid for in the servo, and allow flutter to be induced on the flight surface. Yes, the holes in the arm never fit the wire you are using. That has been an issue since the dawn of RC. Most of us will grab the 3/32 ” drill and go for it. That’s a guaranteed sloppy connection. The hole is bigger than the wire. Some will take a micrometer to their wire, and number drill the hole to match. Better, but normal wear will elongate the hole, causing slop at a future date. Much better is to buy one of the many connectors sold. The current standard is the bolt-on ball link.

Whatever style of connector you choose, pay attention to how the connector fits. Check to see if the connector itself can rock back and forth while on the servo arm. Many with a snap-on retainer either have shafts that again do not fit the arm hole, or their retainers do not hold them tightly on the arm, or both, allowing slop to again hold sway. A bolt-on kind allows you to snug that thing down tight so it cannot wiggle on the arm. Finally, judge the strength of the servo arm against the loads you are going to put on the plane. In big plane/high load situation, with a connector sitting above the arm, the forces can actually cause the arm to twist and flex. Several companies sell high strength aftermarket arms, either made from composite or aluminum, just to prevent the problem. In most planes .60 and below, this is a lower order of issue that the others mentioned, but it is something to think about in a fun fly machine, a pattern/IMAC ship, or in anything getting up there in the size of 1/5 -1/4 scale and above.

Moving to the other end of the plane, the same concerns for connecting to the control surface are present, primarily “How much slop is in the hole of the control horn?” Again, bolt-on ball links present the current best solution in preventing slop, but you have the trade off of them being off side to the horn, again inducing flex, and their being more difficult to take off and on for trimming changes.

If you use any threaded connector at both ends, make sure to use jam nuts and thread lock to prevent pushrod rotation (from vibration) from unscrewing the pushrod from the connector, or changing the length of the pushrod! It happens! Some of the standard metal clevises, being made from rolled material, have a significant amount of play in their threads. Beyond the obvious fear of their pulling loose in a high-g maneuver, this wiggle of the clevis on the rod can again add in to any other slop you have, defeating your quest for perfection. Jam nuts and thread lock will help, at the loss of some ease of adjustability. At risk of alienating manufactures, I’ll go out on a limb and state that the Kraft/Hayes black plastic clevises, with the embedded metal pins, are by far superior for this kind of application. They stay tight, and their pins fit standard nylon control horn holes.

Watch for wear on the control horn, and replace any that begin to show signs of slop in their holes due to wear.

I’ll gloss over more on control horns, except with the notation that whatever you use must be solidly mounted. The biggest offenders are poorly installed torque rods used for ailerons. If they don’t have a perfect bearing surface to ride in, or they are not firmly in the hole in the aileron, the precision is again destroyed. I prefer to avoid this kind of setup, and go the two servos in the wing path, if the plane design allows. Even worse - the old fashioned bellcrank in the wing method. You’ve added two more pushrod to bellcrank connections, plus the bearing of the bellcrank itself to induce slop. These are almost impossible to set up perfectly tight.

Ah pushrods! - the meat of the matter. There are again dozens of types of pushrods to choose from - nylon tube within a tube, graphite and titanium kits, fiberglass arrow shafts, balsa square stock, solid wire, and many others. All have inherent pros and cons. The key in every one is to prevent the rod from bowing under a load. Horror is the model in a power dive that can’t pull out, because the forces on the elevator are so strong they force the pushrod to bend, preventing the servo from moving the surface. Slight bowing eliminates much of the precision you are striving to obtain. Lots of bowing results in a new kit with very funny shaped pieces. Tip: on your elevator, always install the pushrod so “up ” is a pull, and not a push! A push will bow, a pull cannot!

Primary rules for pushrods: First, they must be perfectly straight from the servo to the control horn. Second, they either cannot bend, or must be prevented from bending. As to rule one, the worst trend over the years in trainers has been the insistence by manufactures of routing the rudder pushrod out of the fuselage by using a large “Z ” bend. This leaves in the newcomers an impression that this installation is “OK ” in other circumstances. No matter how stiff the main body of the pushrod is, grab a rudder of one of these setups and notice how easy it is to force it against the torque of the servo. That bend guarantees flexing. And flexing is loss of control. Pushrods must be always dead on straight to convey maximum power to the surface. “Y ” pushrods, to drive split elevators from one servo, have again exactly the same problem. Loads tend to be better balanced, and the flex is less, but it is still there.

Tube-in-a-tube pushrods are by nature very bendy, and are good for going around modest curves. They must be anchored solidly along their whole length. Don’t assume that anchoring at the front and back is good enough. A minimum is to have them pass through a mount or brace in the fuselage every six inches and be glued in. Many people make the minimum unsupported length three inches. The “pushtubes ” therefore have to be installed during the construction process, and not after the plane is covered or the fuse is sheeted over. The inner tube should poke out of the outer tube the minimum amount possible before it transitions to the treaded rod going to the clevis/connector. Unsupported inner rod, even an inch, flexes like crazy. Best is to have your setup such that the metal rod actually fully goes inside of the outer sheath (which should be cut off slush with the surface) before screwing into the inner rod. Note that bends in tube push rods inherently induce slop proportional to the radius. Take a section of inner and outer tubes out of the package and lay them straight on your workbench. Move the inner rod back and forth. There is no play,correct? Take that same rod and bend it in a big circle. Move the inner rod again. Wow! You have to move the inner rod a noticeably large amount before the opposite end starts to move, and vice versa, pull on it a fair bit before the other end starts to reverse direction! Talk about slop! The rods have to have some clearance between the inner and outer diameters, and bending them exacerbates that. Same happens in your plane.

Solid pushrods run the gamut in stiffness. I have not tested the new graphite/titanium push rods yet. They promise to be the stiffest to date, thereby reducing the flexing problem. All others have to be “caged ” to a greater or lesser amount depending on their material.

In your fuselage, install strips of wood running across the fuse, going over, under, and on both sides of the pushrod, boxing it in. They should not normally touch the pushrod (friction would rob your servo of the power it is trying to get to the control), but should be as close to it as possible, so that they stop the pushrod from bowing under load. Small planes have the advantage here - lower flight loads and shorter lengths of pushrods mean less possibility of bowing. The bigger the flight loads, the more you have to guard against it. A .25 trainer doesn’t have to worry about it; a 1.20 IMAC plane certainly does have to worry.

The transition from the main body of the pushrod material to the wire connections is the second place to watch for bowing. Best is to have the wire directly exit from the center of the rod. Keep the wire ends as short as possible, again -the wire can flex and bend easier than the pushrod itself, so keeping it short reduces the problem.

Phew! Now that we have the preface and overview out of the way, let’s get on to the real reason for this lead-in to the meeting - cable systems!

The advantages of cables? Many! They only move the surface in a “pull ” direction, thereby eliminating the bowing issue. Hole fit of the ends on the servo side or the horn connector don’t matter - the pins of the clevises or the loop of cable are pulled snug against the sides of the holes, so no slop.

Thread fit on the clevises is tight, again due to the constant tension in the rigging. They don’t need to be run perfectly straight to the surface (but the amount of corner they can make is limited). If rigged direct to the servo, they exert a constant pull on it, eliminating the slop due to servo rock (but more on the downside of that at the meeting).

Temperature sensitivity is greatly reduced - nylon tube pushrods will noticeably change length due to temperature. Cables vary little to none, depending on material.

The draw backs? Setting up proper geometry is a must, and can be a big challenge! Some installations require going around a corner greater than they can accommodate. Over tensioning the cables can greatly increase the wear on the servos. Changing the amount of throw by repositioning in arm/horn holes is not the no-brainer that pushrods are. Initial installation can be more time consuming. Not possible for use in most aileron configurations.

Now that I’ve whetted your appetite for more, come to the meeting and hear all about it!

Welcome to the St. Paul Model Radio Controllers Club

By Ted Sander