NASA Engineer Develops Servo-Tab Control
September/October 1965 American
couple years ago I experimented with adding rudder control to my
via a radio control system salvaged from an Estes Sky Ranger model.
It used a solenoid actuator mounted on the vertical fin, and the
receiver mounted in the cabin area of the fuselage. A single cell
LiPo battery provided power. To save weight I stripped the components
that came with the receiver for motor control. It worked kind of
OK, but the actuator really wasn't powerful enough to more the rather
large rudder with any authority. I vowed to make it better, but
have not had the opportunity.
This article from the September/October
1965 edition of American modeler proves that the idea has been around
for a long time. Even in 1965 people were lamenting the disappearance
of wide-open spaces for flying free flight models (unless you live
in the Midwest or Southwest deserts). Mr. Phillips' airborne system
boasted a mere 1.3 ounces - a ton by today's standards, but quite
an accomplishment 50 years ago. He even included plans to a custom
airplane model to host the equipment.
NASA Engineer Develops
Servo-Tab Control for Rubber-Powered R/C
By W. Hewitt Phillips
The author's son brings model down by gliding it in circles
overhead. No chase!
Solo launching technique demonstrated by Mr. Phillips. Let the
model lift off and then you can take up the transmitter. All
radio flying should be like this!
Many of the old-timers agree that the greatest thrill in modeling
is to have a rubber-powered plane spiral up, hook a thermal and
soar off, leading the builder on a merry cross-country chase, unfortunately,
this kind of flying requires the wide, open spaces, which are rapidly
disappearing. Even the use of dethermalizers and the 3-minute or
5-minute max has not helped the situation much. In a 10-mph wind,
a model flying in neat circles will drift one-half mile in three
minutes. How many city dwellers have a half-mile stretch of field
available? And then, there are always times when the model will
head straight down wind.
The author has always dreamed of
being able to control one of these high-performance rubber-powered
jobs to keep it flying overhead. With an air-speed of 12 to 15-mph,
the average Wakefield-type or unlimited category rubber-powered
model will make headway into the breeze on any reasonably calm day.
Recently, developments in radio control have progressed to a point
where this type of control is practical. As a result, the local
playground can become a flying field.
All that is needed
is a radio installation, including batteries, receiver and actuator,
weighing less than about 1 1/2 oz. Fortunately, miniature transistorized
receivers weighing 3/4-oz or less are readily available in inexpensive
kit form. A pair of the smallest size pen cells (Eveready No. 912
or equivalent) weighs just under 1/2-oz. These weights leave just
1/4-oz for the actuator. In this article, an actuator weighing less
than .07-oz is described!
How can an actuator weighing .07-oz
control a model of 3 to 4-ft span? The secret is the use of an old
idea from full-scale aircraft, the servo-tab control. This device
was patented in 1922 by
Anton Flettner, the German, who also invented the rotorship.
For this reason, the device is sometimes called a Flettner tab.
As early as 1925, this tab control was used on airplanes such as
a huge old Handley-Page bomber when pilots found that the barn-door
sized rudders were too heavy to move by human muscle power alone.
Today, airplanes such as the DC-6, the Boeing 707 and the DC-8 use
servo-tab controls so that the pilot can move the controls with
comfortable forces when flying along at 300 to 600-mph.
The principle of the servo-tab control is that the pilot (or in
the case of a model, the actuator) is connected to a small tab on
the trailing edge of the control surface. As the tab moves, air
forces developed on the tab push the control surface in the opposite
direction. Because the tab is so small, it can be moved with extremely
To control a rubber-powered model a small magnetic
actuator, similar in principle to a relay, is connected directly
to the tab control horn by a pushrod. Normally, the tab is held
full left by a small spring, producing right rudder in flight. On
transmission of a solid signal, the actuator pulls the tab full
right, producing left rudder. For neutral, a pulsed signal with
equal on and off periods may be used. In practice, however, it is
found that full control in one direction or the other is used most
of the time anyway, so that quite satisfactory control can be obtained
with a simple on-off pushbutton on the transmitter.
rubber-powered model used by the author for the radio-control installation
is fairly conventional. Its design, therefore, is not emphasized
in this article. Instead, the control installation is described
in detail. A three-view drawing of the model and a table of weights
is shown. Any conventional rubber-powered model of this size could
readily be converted to R/C by the method described.
radio receiver and batteries are mounted on top of the fuselage
near the nose in order to move the center of gravity as far forward
as possible. To save the weight of a battery box, the batteries
are attached with Scotch Tape and wired by soldering hookup wire
to make the connections. A flea clip is used as a switch. The magnetic
actuator is so light that it may be mounted back in the tail, eliminating
the need for long control linkages which would add friction and
interfere with the rubber motor. To get the power to the actuator,
a pair of strands of No. 40 magnet wire are run from the back along
the length of the fuselage. These wires are preferably glued to
the longerons before the fuselage is covered, but they may be attached
on the outside with a coat of dope if it is desired to convert a
completed model to R/C. A pair of light contacts made from bits
of brass are soldered to the wires and glued to the ends of the
longerons to make contact with a similar pair of contacts soldered
to the wires in the removable tail section. The No. 40 magnet wire,
.003 inches in diameter (almost as thin as a human hair), adds no
appreciable weight to the model, yet has very low resistance compared
to the actuator coil itself. Ordinary hookup wire would be prohibitively
heavy for this installation.
(above and below) Practical, inexpensive
radio control of rubber-powered planes is now possible by a system
consisting of batteries, receiver and actuator which weight less
than 1.3 oz. Enjoy sport flying from small fields. Interesting new
contest events are proposed.
Details of construction of the actuator, tab and linkage are now
described. The small magnetic actuator used to operate the tab is
shown. The small electromagnet may be made by winding wire on a
nail of 3/32 inch diameter. The nail should first be heated to red
heat and allowed to cool slowly to soften the iron. Fit celluloid
or cardboard end plates and wind with No. 40 magnet wire, using
a hand drill, to a depth of about 1/16" along a length of 3/8".
About 35-ft of the fine wire will. be required. Finally, bend the
back part of the nail around to a U-shape and file the ends square.
The coil should have a resistance of 30 to 35-ohms.
U-shaped magnet, when energized, attracts a small piece of iron,
or armature, in a manner similar to a relay. Instead of using this
armature to operate contacts, however, it is attached to a light
aluminum lever about 1 1/4" long which serves to magnify the motion.
The armature may be made from a piece of soft iron 1/32" x 1/8"
x 3/8". If sheet iron cannot be obtained, a piece of a nail filed
flat on one side may be used. The aluminum extension is attached
by binding with nylon thread and gluing with model cement. Soft
.012" sheet aluminum suitable for this and other aluminum parts
in the control system may be obtained from hardware stores in the
form of aluminum flashing. A small clip of .016" music wire keeps
the armature in place as shown. The armature should be a loose fit
in the clip. The armature may be inserted or removed by turning
it 90° inside the clip. A small spring, made from .006" music wire,
tends to keep the armature against the magnet. This spring is mounted
oppositely from the spring found in an ordinary relay, which tends
to keep the armature away from magnet. In the present arrangement,
the armature is held away from the magnet by the tab return spring.
This return spring overpowers the spring on the armature. The two
springs working against each other serve to take up any backlash
in the tab linkage and hold the armature in the correct position.
If .006 music wire is hard to obtain a piece of 1/32" rubber may
be substituted for the spring.
The method of hooking up
the actuator to the tab and rudder is shown. A most important detail
is the location of the end of the actuator arm which moves the tab
pushrod. With the tab in neutral, this arm must be directly below
the rudder hinge line. With this setup, movement of the rudder will
not affect the tab angle. Full travel of the actuator arm is about
1/16 inch. This motion is shown exaggerated for clarity. In order
to get adequate tab travel of plus and minus 30°, the tab control
horn must be quite short, about 1/16".
Though the design
of a model for R/C conversion is not critical, some attention must
be paid to the details of the vertical tail and rudder design. The
tail size can be quite small because the radio receiver installation
in the nose moves the center of gravity farther forward than usual.
The rudder should be aerodynamically balanced. In other words, some
rudder area should be located ahead of the rudder hinge line to
reduce the moment required to deflect the rudder. In this way, the
tab size required to deflect the rudder may be reduced. Experiments
have shown that the location of the hinge line should be at 1/3
the total rudder chord. A tab chord of 15% of the total rudder chord
and a tab span of 2/3 of the rudder span have proved satisfactory.
Stops should be placed on the fuselage to limit the rudder motion
to plus and minus 25°.
The rudder, tab and all linkages
must be perfectly free from friction. They should be so free that
when the model is tilted from side to side, with the pushrod disconnected,
the tab and rudder will readily flop back and forth due to their
own weight. Once the importance of this low friction is realized,
it is not so hard to obtain as might be thought. Rather than use
thread or cloth hinges, common in most R/C model practice, hinges
should be made with .010" music wire shafts fitting in holes punched
in .012" sheet aluminum. Drilling the holes is quite unnecessary.
The holes for the bearings may be punched in with common pins. Small
bushings or washers to separate the moving parts may be made by
stripping a short length of plastic insulation from hookup wire
and slicing it into 1/32" lengths with a razor blade. Care must
be taken when covering and doping the model to avoid getting any
dope, paper fibers or glue strings into the "works" to add friction.
The tab is made of 1/32" sheet balsa. Small pieces of .012"
sheet aluminum, with holes punched in them with a common pin, are
used as bearings at the top and bottom. The lower bearing is made
integral with the tab horn. The two holes, one for the tab pushrod
and the other for the lower tab bearing, should be 1/16" apart.
A small hook is attached to the tab even with the lower rudder rib
to hold the tab return spring. This spring may be made of .006 music
wire or a 1/32" rubber strand may be substituted.
is of conventional built-up construction. The tab and rudder hinges
should be installed before covering the rudder.
The pushrod to operate the tab is made from .010" music
wire. This size may sound pretty small to conventional R/C builders,
but it is plenty strong for this application and the small diameter
contributes to low friction. Pinholes punched in the tab horn and
in the actuator lever to act as bearings can be a sloppy fit because
the springs on the tab and actuator keep the linkage preloaded,
thereby taking up the backlash. The pushrod is simply slipped in
place and retained by a light balsa block glued to the bottom of
the rudder. A U-shaped bend in the tab pushrod will allow adjustments
in its length to get exactly the tab motion desired.
limit of tab motion when the actuator is energized is provided by
the armature hitting the actuator magnet. A thin coat of glue or
dope should be applied to the armature to prevent actual metal-to-metal
contact between the two parts, to prevent them from sticking due
to residual magnetism. The limit of tab motion in the opposite direction
is provided by a stop which bears against the hook holding the tab
An additional wire link, the lateral support
link, is shown. The purpose of this link is to keep the top of the
actuator lever from moving sideways. This link is desirable because
the clip at the bottom of the relay armature does not restrain this
lever very effectively against lateral motion.
adjustment and checkout of the completed installation should be
made before attempting a flight. First, the tab pushrod should be
adjusted in length so that, when the magnet coil is energized, the
tab deflection is 30° left. No other travel adjustments are necessary
since a stop on the tab limits the deflection to 30° right when
the magnet is not energized. The linkage design determines the separation
of the armature from the upper magnet pole when the magnet is not
energized. This separation is about 1/64 inch. This value should
not be exceeded, as the ability of the magnet to "pull in" the armature
decreases rapidly as the separation increases.
spring tension on the tab return spring should be adjusted. As a
rough guide or as a starting point for the adjustment, the spring
tension may be balanced against a small weight as follows: Temporarily
glue a piece of 1/32" square balsa, 2 inches long, to the tab to
act as a lever for hanging a small weight. Cut out a piece of the
.012" aluminum 1/8" wide and 1/2" long. Lay the tail on its side
and hang this piece of aluminum (about .002 ounces) over the lever.
With the weight 1" from the tab hinge line, the tab should just
move to its neutral position against the tension provided by the
combination of armature spring and tab return spring.
connect a pair of flashlight batteries producing 3 volts to the
magnet coil. The pull of the magnet should snap the tab to the full-left
position. For a final adjustment of the spring tension, reduce the
voltage gradually to 2 volts. The tab should not cease to operate
until the voltage drops below about 2
volts. With this setting,
the control will continue to function as long as the battery voltage
is sufficient to operate the receiver.
A final check of
the completed installation can be performed by holding the rudder
in neutral with a piece of Scotch Tape and keying the transmitter
on and off. The tab should snap smartly from side to side. With
the rudder free, the rudder motion will be random, but in flight
the air forces will cause the rudder to flip over instantly when
the tab moves.
The procedure for flight adjustment of the
model is very similar to that for any conventional rubber-powered
job. Particular care should be taken to adjust the model to fly
straight with rudder in neutral, both power on and in the glide.
In this way, equal control effectiveness for left and right turns
will be available at all times. First, tape the rudder and tab in
neutral and check the glide. If any turn is noted, correct it by
warping the wing slightly. Then adjust for straight flight under
power by varying the amount of side thrust.
a rubber-powered model requires both hands, solo launches present
a problem. The author has found that a satisfactory procedure is
to set the transmitter on the ground, launch the model, then quickly
pick up the transmitter and take control as required. Control of
the model in flight comes naturally with a little practice. A flyer
accustomed to gas-powered radio-control jobs will consider the model
docile because of its low speed. To the rubber-powered fan, however,
this means of control offers a world of exciting possibilities.
The model may be steered into thermals, allowed to hang in them
or drop out as desired, and be brought down to spot landings. All
this can be done in a field no larger than the average playground.
Because the current drain of the actuator is very small
(about 100 milliamps for a 30-ohm coil) a pair of the smallest pen
cells will last for a whole day's flying. To get a little extra
boost in model performance, these pen cells may be sawed in half,
using a jeweler's saw, and the ends covered with a double layer
of Saran Wrap bound on with nylon thread to keep the moisture from
evaporating. These half-size cells weigh only 1/4-oz per pair and
still give enough energy for about five flights.
the more interesting aspects of the rubber-powered R/C model is
its contest possibilities. These models should not be flown in competition
with uncontrolled models. Such a procedure is certainly contrary
to the spirit of the present AMA rules, even if it is not specifically
prohibited. Instead, the rubber-powered R/C model offers the opportunity
for a special type of contest in which all artificial design restrictions
and limitations are removed, and in which the element of luck is
Consider, for example, a contest for rubber-powered
R/C models in which the rules are as follows: (A) Winner determined
by the maximum endurance (best of three attempts); (B) Flight disqualified
if the model fails to land in a specified 100-ft radius circle.
The usual artificial restrictions such as weight rules or
limitations on maximum endurance are completely absent. A review
of the history of outdoor rubber-powered model development shows
that in the early days no restrictions of any kind were imposed,
but primitive design and construction techniques prevented unduly
long flights or loss of many models. As models improved, however,
weight rules were imposed, first 1-oz per 50 square inches and then
2-oz per fifty square inches. This restriction failed to hold endurance
in bounds for very long, but it did end the incentive to make structurally
light models. Finally, as dethermalizers were perfected, the max
concept was introduced. This rule has the objective of preventing
out-of-sight flights. As the model designs improved, the ability
of more and more modelers to make maxes as a matter of course has
resulted in all experienced builders being placed on more or less
equal terms. The natural result is more rounds and more flyoffs,
until the contest is decided partly by physical endurance of the
modeler and, in some cases, by luck.
How does radio control
change this situation? With contest rules of the type proposed previously,
there is no need for the max restriction. The flyer must bring the
model in to a landing near the launch point and is thereby prevented
from losing it. In addition, there is no need for any weight rules,
because the wing loading must be kept high enough for the model
to make headway into the wind and because the radio gear provides
an unavoidable "payload" weight. Skill in flying becomes important,
because the model can be allowed to circle in a thermal for a while,
but can never be allowed to drift downwind so far that it cannot
glide back to the launch point. Since thermals do drift with the
wind, very long flights are unlikely.
If more emphasis on
precision of flying is desired, the rules suggested previously might
be modified such that the endurance in seconds is decreased by the
distance in feet that the model lands from the launch point. With
rules such as these, the modeler is faced with a true test of skill
in designing, building and flying. He may gain an advantage by better
aerodynamic or structural design, more sophisticated or lighter
control equipment, and by skill in riding the air currents. He is
no longer placed on a common level with other builders by artificially
Plans for Rubber-Powered R/C Model
for larger version>
Posted November 13, 2011