April 2003 75
■ Mike Hurley
Controlling
Systems
Volume IV:
IN PROJECT EXTRA Volume IV, it’s time to power up the beast.
We’ll install the components that will move the control surfaces,
power the aircraft, and power the electronics in the airplane. For
those of you not building the model, there is a great deal of useful
information about servo and control setup and geometry that can
help any modeler build a better control system.
Installing critical flight-control components, hardware, and
power systems will give your aircraft life. What kind of life will
depend on the components you choose and the care with which you
install them. This is a place where many modelers try to skimp to
save a few coins, but this is no place to go cheap! On an aircraft of
this size and power, I don’t believe there is a place to try to be
frugal. Costs for this project can add up fast. This airplane needs to
have a level of hardware and equipment that makes it reliable and
safe. A model this size can be dangerous, so I don’t recommend
building it on a budget. Servos, linkages, hardware, propeller,
spinner, etc. have to be of the best quality. Going with anything less
is irresponsible.
I like to think of my airplanes as having a generic setup that does
not involve a lot of complicated connections or procedures. But
understand, to properly set up an aircraft of this size and type is not
a simple matter and will be much more involved than your basic
sport airplane.
Electronics: The electronics lineup for Project Extra will be as
follows: one receiver and two receiver batteries running through two
switches. You can plug the second battery/switch into any open
channel on the receiver. There will be a total of nine servos; two in
Project Extra is a large model with a 106-inch wingspan and is built for performance aerobatics. Ed Alt photo.
04sig3.QXD 1.23.03 2:45 pm Page 75
each wing, two for the rudder, one in each elevator half, and one
throttle servo. The wing servos will be mixed like any other twochannel
wing so that differential is adjustable. The servos in each
wing half will be matched with JR MatchBoxes. The rudder servos
will be mixed through the radio with a multipoint mix.
Okay, I’m gonna take back that frugal comment. Here’s where
you can save a few bucks or, better, redirect a few coins toward
getting the right servos and hardware. Exotic electronics, multiple
receivers, optical isolators, regulated batteries, power distribution
systems, etc. are not needed in this airplane. They all do a job and
they do that job just fine, but in my experience complex electronics
are not necessary for you to have a successful 35% competition
aircraft that is safe and reliable.
For some of the larger models with more servos and bigger
control surfaces, sophisticated electronics can become a must, and I
have some of those systems in my own 40% aircraft, but for this
project I’d like to keep it simple and concentrate on making the
setup secure. It may be a letdown for some of you that we aren’t
With practice, maneuvers such as the elevator or harrier are easy to perform with the
Extra 300LX. Michael Schauer photo.
Slow-flight ability and positive control characteristics make this a model that will instill confidence. Schauer photo.
Right: Because of its neutral characteristics, the Extra is stable in difficult maneuvers
such as the torque roll or hover. Schauer photo.
76 MODEL AVIATION
04sig3.QXD 1.23.03 2:45 pm Page 76
With finished elevator in foam-core shuck, use drill press and 1⁄2-
inch sharpened brass tube to cut holes for servo horn dowels.
With aileron you’ll need to measure center of front, rear trailing
edges and use piece of foam to keep them level on drill-press
table.
Erik Richard used a router attachment on a common Dremel tool
to cut the servo bays in the wings. Two cuts are necessary.
The first cut will be at a depth for locating the servo rails, and
the second cut will be full depth for the servo body.
Measure and cut 1⁄2-inch hardwood dowels to be used as controlhorn
locations that will be tapped for control-horn screws.
going to discuss those systems, so here’s a bone: I’ll highlight
sophisticated electronics in an upcoming Scale Aerobatics column.
And although two receivers are not needed for this model, you can
learn more about multireceiver systems in the May 2002 Model
Aviation Radio Control Scale Aerobatics column.
One thing that separates the big airplanes from the smaller ones
is the amount of vibration that the model and all of its components
will experience. No matter what engine you use for your Giant
Scale airplane, it will have harder vibration pulses and all of the
electronics need to be isolated as much as possible. On the subject
of redundancy, most receiver failures are caused by vibration, so
proper isolation mounting should help protect your receiver. Even
the material you use to fasten your components to the aircraft is
important; a heavy nylon tie will transfer vibration more readily
than will a soft Velcro strap.
Erik Richards and I like to use the Du-Bro foam rubber sold in
hobby stores. It’s exactly the right density for protection from
vibration. A piece of 1⁄2-inch Du-Bro foam under your receiver,
battery, or ignition module fastened with a Velcro strap will work
fine. If you are going to use nylon ties, it’s a good idea to wrap the
entire component before it is tied down. I’ve seen many airplanes
out there with electronics Velcro-fastened directly to the model’s
wooden structure. That’s taking a risk in my opinion.
Let’s discuss servos and control linkage hardware, and why
we’ve chosen the parts that we’ll use. In doing research for the
Scale Aerobatics column, I’ve had the chance to test and evaluate
control rods, horns, connections, systems, and servos. Servo choice
is an interesting topic because there are so many brands and so
many opinions. It’s important for builders to stick with name brands
that have been proven to perform safely and accurately on large
Scale performance airplanes.
Futaba, Airtronics, and JR make excellent servos for Giant
04sig3.QXD 1.23.03 2:46 pm Page 77
Use templates you made during foam-preparation stage of building
process to locate servo rails, control-horn dowel locations.
Scale. We chose JR DS8411 digital servos for this project. They
have a rating of 155 inch/ounce of torque and a speed of 0.16
second/60° at 4.8 volts.
Servo Arm Geometry: Geometry and force/connection
relationships are going to be a big part of this phase of the project.
It’s important that you understand how your mechanical connections
will interact with the parts they link.
For the control rod ends we will utilize ball links that are bolted
to the servo arms. We use the ball links because they are simple to
use and reliable. But because the connection is offset to the
rotational center of the servo arm, any force will tend to create a
twisting motion on the servo arm. If the servo arm were to twist, that
in turn would put a side load on the control rod. So in order to tame
the twist and ensure a solid connection, aluminum servo arms are a
must when using ball links.
The heavy-duty plastic servo arms sold by some of the
aftermarket companies will work fine when used with a clevis that is
supported on both sides of the arm, but they will twist when used in
an offset environment such as a ball link bolted to one side of the
arm. I used SWB arms on this project. The arms come pretapped to
accept 4-40 bolts.
The wings and horizontal stabilizers will have the servos
mounted in the bottom of them, vertically, nearly flush with the
outer skin. There will be a short control rod linked from the servo
arm to a mild steel bolt that will act as the control horn.
As a basic starting point, the longitudinal centerline of the servo
should be 90° to the hinge line—not parallel to the aircraft’s
centerline (for the stabilizers they will be both). We will use SWB
aluminum arms that are 1.25 inches (L in Diagram 2) to achieve 45°
of elevator surface deflection at 100% travel. The idea is to strive
for a control-horn length of 1.25 inches measured from the center of
the hinge line (the beveled point) to the center of the control rod
78 MODEL AVIATION
A pad of 1⁄2-inch closed-cell foam was used to insulate receiver
from vibration. Put a piece of tape over crystal to keep it in place.
Diagram 1
Courtesy of Erik Richard
Diagram 2
04sig3.QXD 1.23.03 2:46 pm Page 78
April 2003 79
Parts labeled A through F come in Nelson 10-32 swivel
link kits. A is 10-32 mild steel machine screw that will be
used for control horn. It is threaded into 1⁄2-inch-diameter
birch dowel embedded in control surface. Clevis (B)
threads onto A. F is special link that fits into B with
bushing installed instead of steel ball. The 4-40 machine
screw (C) fits through B and F and is secured with
threaded insert (D) and locking nut (E). G is Hangar 9
titanium Pro-Link. H is Nelson gray (now dimpled black)
left-hand threaded ball link. Conical standoff (J) comes in
package with ball links (H), as does slotted 4-40 machine
screw and lock nut (K). Mike replaced slotted machine
screws for much more user-friendly Allen-head cap screw
(I). L is SWB 1.25-inch aluminum servo arm. L is attached
to servo (O) using metric 3 x 6mm cap screw (N). Washer
(M) comes with servo and acts as locking device, but
before model is flown you should lock all servo arm
screws in place with Loctite.
connection point (Diagram 1). We want to create a 1:1 ratio so that
we get all the deflection we need without compromising (or
reducing) the applied force (mechanical advantage) generated by the
servo.
You can increase the mechanical advantage from your servo by
utilizing a control horn that is longer than the servo arm, but you
will lose deflection degrees. You can also increase the surface throw
by using a servo arm that is longer than the control horn, but this
ratio decreases the servo’s mechanical advantage so it is not
recommended. The whole thing works kind of like gears on a bike. I
find that a 1:1 ratio is just right when 45° of deflection is desired.
The servo rails mounted in the wings are approximately 2.5
inches, so positioning the servo in relation to the control horn is just
a matter of mapping them out before the servo bays are cut. Since
the movement of the servo arm is on a different plane from the
control horn, let’s take a look at how we can arrange them to get the
best end result.
You can see in Diagram 2, Case A, that when the servo arm is
centered at 90° or parallel to the hinge line, the control rod is
positioned 90° to the hinge line. As the servo arm travels in an arc
the x displacement decreases, causing the control rod to change its
angle with respect to the hinge line. At 100% travel the arm has
moved approximately 45°. The movement up until this point is
fairly linear along the y axis, but past 45° the slope degenerates in a
nonlinear manner.
Mechanical force also decreases as the x axis distance decreases
and the control rod moves farther from 90°. We have a situation
where the mechanical force from the servo arm actually decreases as
the arm travels to full deflection, but the required applied force is
increasing from flight loads as the surface is deflected farther into
the air stream.
In Case B the control rod is 90° to the hinge line at the point
where the servo arm has reached full deflection of 45°. In this case
the force is greatest at full deflection, where flight loads are likely to
be the greatest and the deflection travel is closer to a fully linear
motion. None of this is truly critical, but it makes sense to arrange
the positions of the components to get the best advantage possible.
In Case B with a 1.25-inch servo arm, we found that the best
location for the control horn is 0.89 inch from the centerline of the
servo (d in Diagram 2).
For a sophisticated software program that will allow you to
design your own linkage systems, take a look at the Linkage Design
program from Envision Design at http://members.cox.net/evdesign/.
Find the locations of the stress-bearing plates using the template
you made for the wing cores, and if you embedded servo rails under
the skins as we suggested during the sheeting portion of the
construction phase, you’ll need to locate them with the original
templates as well. When determining a location, the dowel should
just touch the beveled leading-edge stock. Find the location for the
bay in reference to the dowel position as described, and mark it all
out on the wing panels.
We used a Dremel tool with a small router attachment for
cutting the servo bays. Mask off the area around the servo bay to
protect the wood. We cut the bays freehand, but if you want to be
more accurate with the edges of the bays you can pin some 1⁄4
square balsa sticks in the appropriate positions to act as a cutting
fence.
Decide how far you would like your servos to be recessed into
the wing panel. We recessed the JR DS8411 servos to 3⁄8-inch deep.
Make the initial cut to the depth of the desired servo recess
according to the outside dimension of your servo. The remaining
depth should be cut only between the servo rails to finalize the
servo bay. If all was done correctly, the servo lead tunnels in the
FlyingFoam.com wings should be accessible.
Installing the Dowels: Now that we know where everything goes,
let’s install the dowels we’ll use for mounting the control-horn bolts
into the ailerons and elevators. The rudder will use a special horn
manufactured by Jerry Nelson for his pull-pull system. With the
template that you made when you prepared your foam cores, find
the stress-plate locations for each control surface. We marked the
locations on the wood. Tape the control surface into its original
shuck, and check to ensure that the center points (leading and
After control-horn dowels are in place and sanded, drill hole for
tapping roughly an inch deep. Drill from the bottom!
04sig3.QXD 1.23.03 2:46 pm Page 79
As shown on the plans, aluminum tubing available from K&S is
suitable for tapping and makes a great control rod.
Bellcrank is used for rudder pull-pull system and is connected
in line with servos under hatch. Note use of light-plywood
spacer to align bellcrank to servos.
Make all control linkages, but don’t cut control-horn screws to
length until after they are epoxied in, which will be done after
the model is covered.
trailing edges) are equal all around (level to the work table) so that
the dowels will be aligned correctly.
Because the CNC-cut wings from FlyingFoam.com are cut with
dihedral built into the core (cut flat instead of on center), cutting the
aileron dowel holes is a bit more complicated. We measured front to
back and at the ends and learned that it worked to simply prop up
80 MODEL AVIATION
the trailing edge to match the height of the centerline of the leading
edge.
I’ve used reamers for cutting the holes but have found that a
sharpened 1⁄2-inch brass tube gives a smoother cut. The only way to
get a truly straight hole is to use a drill press. Be sure to set your
drill press to the slowest setting and work quickly so that the cutting
tube does not heat up and melt the foam. If you do melt some foam,
don’t sweat it; make a thick paste of epoxy with microballoons for
setting the dowels. Clean the cutting tube between each cut.
For dowels you need to find good-quality wood—preferably
maple, but a hard birch would also work fine. Avoid the pine or
poplar dowels found at hardware stores. Insert a length of 1⁄2-inch
dowel into the hole and mark it for cutting. Notice that it will follow
the contour of the control surface. Try to avoid finish-sanding as
much as possible because this wood is hard to sand. When satisfied
with the fit of the dowels, epoxy them in place.
Once the dowels are glued in place, you can drill the dowels for
your control-horn bolt. The bolt will be tapped into the dowel, and
the hole should not go completely through the dowel and exit the
top of the control surface. Make sure that you are drilling the dowel
on the bottom side of the control surface. Be sure to use the proper
drill sized to tap for the appropriate thread. Drill and tap into the
dowel to a depth of approximately an inch.
Control Hardware: You can see that we have paid a lot of
attention to geometric relationships. Here’s where the difference
between a 60-size sport airplane and a Giant Scale airplane gets
really important from a precision and reliability standpoint.
On the plans there is an isometric drawing depicting the servo
linkage system. The control rod shown is a thick-walled aluminum
tube that has been cut to length and tapped at each end to accept a 4-
40 stud backed up with an aluminum lock nut. (K&S manufactures
the tubing; ask for part number 6030 from your local hobby store.)
The aluminum-tubing system is reliable as a control rod, and it
looks great if you take the time to polish it. But for the sake of not
having to build each rod, we went with Hangar 9’s new titanium
control rods called Pro-Links.
Pro-Links are threaded opposite directions on each end,
turnbuckle style, so that you can perform adjustments while the
servo rod assemblies are installed in the aircraft. Easy maintenance
is one of my top priorities when building a model. Nelson Hobby
Specialties sells 4-40 ball links tapped both directions to work with
the Pro-Links. Black plastic ends have the normal right-hand
threads, and the black end with a machined dimple (formerly gray
plastic as labeled in the photo as “H”) ball links are tapped left-
Diagram 3
04sig3.QXD 1.23.03 2:46 pm Page 80
April 2003 81
The Nelson control horns were cut so that the author wouldn’t
need to remove a portion of the leading edge of the rudder.
The author used aluminum control horns specifically made for
Giant Scale rudder systems by Nelson Hobby Specialties.
The control horn was fitted into wooden sandwich with dowel
pins holding everything in place and glued into the rudder.
Balsa spacers were used fore and aft to get the positioning of
the rudder horn just right and to ease assembly.
handed. For additional information about linkages, see the January
2002 Model Aviation Scale Aerobatics column.
The point is to have a solid, firmly bolted, secure linkage
system that does not flex or vibrate. For control hardware we used
Nelson (formerly Rocket City) 10-32 swivel link kits, item
RCL71A. The 10-32 is slight overkill, so if you’d like to go a little
lighter you can use the 8-32 kit item RCL70A with no problem.
The labeled photo shows an exploded view of the parts that we
used to make up the control linkages. It is essential to build
linkages to this level to maintain the aircraft’s integrity in
operation.
The parts labeled A through F come in the Nelson 10-32 swivel
link kits. The 10-32 mild steel machine screw (A) will be used for
the control horn. It is threaded into a 1⁄2-inch-diameter birch dowel
embedded in the control surface. The clevis (B) threads onto the
machine screw (A). A special link (F) fits into the clevis (B) with a
bushing installed instead of a steel ball. The 4-40 machine screw
(C) fits through the clevis (B) and the special link (F), and it is
secured with a threaded insert (D) and locking nut (E).
G is a Hangar 9 titanium Pro-Link; they are sold in various
lengths in a package of two. H is the Nelson gray (now dimpled
black) left-hand threaded ball link. The conical standoff (J) comes
in the package with the ball links (H), as does a slotted 4-40
machine screw and a lock nut (K). I replace the slotted machine
screws for a much more user-friendly Allen-head cap screw (I).
L is an SWB 1.25-inch aluminum servo arm. The arm (L) is
attached to the servo (O) using a metric 3mm x 6mm cap screw
(N). I replace the factory Phillips screw for the much easier-to-use
cap screws. The washer (M) comes with the servo and acts as a
locking device, but before the airplane is flown you should lock all
of the servo arm screws in place with Loctite.
We prefitted everything in the wings and made up all of the
control linkages and servo extensions. The control-horn bolts will
not be glued in until after covering the airplane. At that time we
will epoxy them into the dowels and use a Dremel cutoff wheel to
remove the head of the bolt and trim to the appropriate length.
Pull-Pull: For the rudder, Erik wanted to use a slightly different
control-horn system from what I had on my airplane. To his credit,
the rudder horn on my prototype Extra (and the one shown on the
plans) is rather outdated. We are going to use the pull-pull rudder
control outlined on the plans with slight variations to accommodate
the new-style rudder horn.
The rudder will utilize two JR DS8411 servos ganged together
and attached to a bellcrank. The bellcrank will have two Kevlar
“Kev-cord” cables that attach to the rudder horn. Kev-cord and the
end fasteners (Kev-cord connectors) are available from Aerospace
Composite Products. The rudder horn and bellcrank we used are
04sig3.QXD 1.23.03 2:47 pm Page 81
Before rudder horn is glued into position, make sure
everything fits just right. Center of control-horn eyelets should
be lined up with hinge line.
If you’re using a pull-pull system for the rudder, you’ll need to
add some structure to bottom of fuselage floor FL3.
Two servos were ganged for maximum rudder authority. Servos
are linked with Hangar 9 titanium Pro-Links for easy adjustment.
After everything is in place and dry, sand whole assembly flush
to rudder. Use a couple pieces of masking tape to prevent
gouging soft balsa rudder skin.
82 MODEL AVIATION
from Nelson Hobby Specialties and are made to match each
other for this type of setup.
It’s important that the geometry for the pull-pull system be
exact; if it is not, the cables may droop when the surface is
deflected. Diagram 3 shows two examples of how you can set
your system up and ensure tight cables throughout the entire range
of motion.
In Case B (our prototype Extra), notice that the distance of
offset of the control-horn connection from the hinge line (B) at the
rudder needs to be duplicated at the bellcrank (A). In Case A, the
control-horn connection is in line with the hinge line and lined up
with the pivot point. The bellcrank should also have the
connection points in line with the pivot point (like our new Extra).
It is important that the width of the bellcrank be the same as the
width of the control horn for both systems. Do not cross the
cables.
We chose a 4-inch Nelson rudder horn and bellcrank. To
mount the rudder horn, Erik sandwiched the two plates in wood to
be epoxied into the rudder. Since much of the rudder’s strength is
dependent on the leading-edge hinge cap (rudder post), we did not
want to cut it when installing the horn assembly. So before the
plates were assembled, Erik cut the hinge-beveled shape from the
rudder-horn plates with a Dremel and a cutoff wheel.
For the horn assembly we used 1⁄4 balsa on top and bottom and
two 1⁄8-inch pieces of light plywood between the plates. Find the
best position for your rudder horn, and cut the balsa and foam
away to fit the horn when fitted with the wood sandwich; in our
case, it was 21⁄8 x 7⁄8 inches.
Erik fitted the plate separation to the steel ball of the ball ends
by sanding the center light-plywood section to the thickness of the
ball. Mark all of the pieces to fit the shape of the opening, and cut
it to leave a bit of overhang that will be sanded flush after
everything is glued in place. We used a piece of plywood at the
front of the system that we could sand and adjust to get the control
connection points to align with the hinge line. Aft of the sandwich
is a gap just less than 1⁄4 inch, so that installation would be easy
and a 1⁄4-inch piece of balsa could be wedged in to hold the whole
assembly tight.
Once the parts have been fitted and cut to size, drill six holes
through the sandwiched assembly for dowels. Erik used 1⁄8-inch
dowels at the front and 1⁄4-inch dowels for the rear four. Epoxy the
sandwiched parts together and sand the dowels flush. Epoxy the
whole assembly in place, making sure to align the horns 90° to the
centerline of the rudder. Once dry, sand everything flush with a
sanding block.
The business end of the pull-pull system uses two servos
ganged together in line connected to a bellcrank. We used SWB
2.5 full servo arms with a Nelson bellcrank and Pro-Link
control rods. From inside the fuselage we made a light-plywood
04sig3.QXD 1.23.03 2:47 pm Page 82
platform to raise the bellcrank up to line up
with the servo arm.
We use a bellcrank rather than
connecting the cables directly to the servo
for several reasons; it takes the load from
the tightly stretched cables rather than the
servo grommets, output shaft, and
bearings, and it enables the correct
geometry.
At this point I hope you have a better
understanding of Giant Scale performance
control systems. It would be impossible to
fully document each step of the building
process in the pages of this magazine, so
the basics are covered here but there is
much more waiting for you on the AMA
Web site. Go to www.modelaircraft.org/
mag/index.htm for further details on
control systems and loads of pictures in an
easy-to-download and -print PDF format.
In addition to more detail about what
we’ve discussed here, you can learn about
the fuel system, mounting the fiberglass,
and installing the engine.
Now that you’ve done all the work to
set up your airplane for its engine and
flight controls, rip it all back out and grab
some sandpaper. It’s time to start the
covering and painting process! That’s what
we’re gonna do in the next issue. See you
then. MA
Mike Hurley
11542 Decatur Ct.
Westminster CO 80234
[email protected]
