This is only a preview of the December 1995 issue of Silicon Chip. You can view 26 of the 96 pages in the full issue, including the advertisments. For full access, purchase the issue for $10.00 or subscribe for access to the latest issues. Items relevant to "Build An Engine Immobiliser For Your Car":
Items relevant to "Five Band Equaliser Uses Two Low-Cost ICs":
Articles in this series:
Articles in this series:
Articles in this series:
|
REMOTE CONTROL
BY BOB YOUNG
The mysteries of mixing
This article has nothing to do with the
making of alcoholic drinks, about how to
behave at parties or even the design of radio
receivers. It is about mixing the control
signals to servos in models. Mixing makes
difficult models easier to fly.
Now that my Mk.22 transmitter
design is close to realisa
tion, it is
appropriate to consider the mysteries
of mixing. Why? – because one of the
most powerful features of the proposed
Mk.22 transmitter is the provision for
mixing any or all channels from 1 to
24. Now mixing is a little understood
subject and so we will spend some time
examining the interaction between elec
tronic theory and practical application.
It is the ability to mix controls on
the modern transmitter that has contributed greatly to the vastly improved
standards of performance and skills of
the operators. Without mixing, some
flying manoeuvres would be virtually
impossible, particularly in helicopters
and high performance gliders.
Mixing is best defined as the modification of one or more control positions
by inputs from one or more different
control channels. In its simplest form,
it consists merely of a small shift in
neutral on one channel controlled
by the full excursion of another
channel. In its most complex form, it
may require inputs from three or four
channels, some with add-subtract (dif
ferential or inversion) inputs.
There are many practical reasons
for using mixing and mostly they fall
into the category of making it easier on
the driver. A good example is the tail
rotor control on a model helicopter.
The prime function of the tail rotor is
to hold the tail boom in the desired
location against the torque of the main
rotor blades.
If the throttle/collective pitch control (these are usually coupled on
model helicopters) is increased, there
will be more torque and the tail rotor
will therefore require more pitch to
Fig.1: a simple mixing
circuit which could be
useful for easier control
of a helicopter. Some of
the throttle control input
(CH1) is fed to the tail
rotor control channel
(CH2) to introduce
automatic compensation
for torque changes in the
main rotor. VR1 is used
to set the mix ratio of the
feedback voltage.
compensate. Likewise, if the throttle
is reduced, the tail rotor will require
a reduction in pitch. Now flying helicopters is a real handful at any time
because all four primary controls are
constantly in motion and the level of
manual dexterity required from the
pilot to co-ordinate all four controls
simultaneously is very high.
Here then is a prime application for
mixing. If we take some of the throttle
control input and feed it across to the
tail rotor control channel, then we can
effectively introduce automatic compensation for torque changes in the
main rotor. Fig.1 shows a representative circuit for a mixer of this style.
It is the most simple of the mixing
circuits in that a small percentage of
the main control channel is used modify the neutral of the second channel.
The direction of the feedback remains
constant with no inversions required.
I must point out that the circuits
presented here are representative of
the type of mixer for use on voltage
driven encoders (they will not work
on the old 1/2 shot encoders). These
encoders use a reference which is
1/2 of the regulated supply rails. In
this manner, the control pots can be
inverted for servo reversing without
any neutral shift in the servos. Thus,
the REF input is connected to the 1/2
regulated supply rail of the transmitter.
Referring to Fig.1 the main control
pot of Channel 1 (CP1) supplies a
feedback voltage to the control input
of Channel 2 (CP2) via the gain set
control pot VR1. This pot is used to set
the mix ratio of the feedback voltage.
The values will vary depending on
which encoder you are hooking the
mixer into. Typically, pots CP1 and
CP2 are 5kΩ, VR1 is 50kΩ and all fixed
resistors are about 100kΩ.
In the practical example of our
helicopter model, CP1 is the control
December 1995 81
Fig.2(a): mixed elevators/flaps are used for aerobatics or com
pensation for trim shift induced by large angles of takeoff/landing
flaps. It is desirable to arrange for the mixing to be switched in or
out very quickly and easily during normal flight. Fig.2(b) shows
elevator trim compensation for the pitch change that takes place
when the takeoff or landing flaps are selected. The direction of
compensation will depend on the configuration of the aircraft
Fig.3: the plan and end elevation of a typical glider wing. The
outboard trailing edge panels are the ailerons and perform some
unusual functions. The inboard panels are the variable camber
panels and they also perform multiple tasks.
pot (stick) for the Throttle/Collective
pitch and is thus the primary control.
CP2 is the stick control pot for the tail
rotor. To set the system up, you would
place the Throttle and Tail Rotor control sticks in neutral and set VR1 for
an approximation of the desired mix
ratio. Moving the Throttle stick will
now induce a neutral shift on the Tail
Rotor pitch.
The amount of Tail Rotor pitch
change is adjustable via potentiometer
VR1 and is found by experimentation.
This will vary from model to model
82 Silicon Chip
due to aerodynamic influences.
Model aircraft
Another application is the mixing of
flaps and elevators in a model aircraft.
There are two basic scenarios here: (1)
the use of mixed elevators/flaps for
aerobatics; and (2) compensation for
trim shift induced by large angles of
takeoff/landing flaps. In both cases,
unlike the helicopter scenario, it may
be desirable to arrange for the mixing
to be quickly switched in or out during
flight. To do this, a switch inserted in
the feedback line from VR1 is all that is
required. This switch is best mounted
on the front of the Tx case.
In this case, the flaps work in reverse
to the elevators but deflect equally
about neutral (Fig.2a). The ratio of
elevator movement to flap movement
is again set via VR1. This is an old
control-line trick and the effect of this
arrangement is to tighten the radius of
inside and outside loops to the point
where square loops are possible.
A further extension of this circuit is
used for elevator trim compensation of
the pitch change that takes place when
the takeoff or landing flaps are selected
(Fig.2b). Putting the flaps down can
result in violent trim changes on full
size and model aircraft. This is brought
about by the large change in angle of
attack on the wing and the sudden shift
in the centre of drag in relation to the
thrust line of the aircraft. As a result,
large control inputs may be required
on the elevators.
The direction of compensation will
depend on the configuration of the
aircraft. As a general rule, high wing
aircraft will require down elevator trim
and low wing aircraft, up elevator trim.
Further variations are possible in
that the flaps may be proportional or
switched. In the first case, a further
complica
tion is introduced in that
there will be a full excursion of the flap
channel from the up or closed position
which will be the neutral position for
the elevator feedback, hence the flaps
only supply a one-way correction. A
more simple system is the fitting of a
3-position switch as the flap control
instead of the pot. This would provide
closed (0°), takeoff (15-20°) and landing (60-90°) flap positions.
The same circuit could be used to
control the cavitation plates on high
speed model boats. Here, they could
be coupled to the throttle and possibly
even with some rudder mixed in to
help control the turns. All of the above
come under the heading of operator
aids – nice touches, designed to make
life easier for the driver.
Glider controls
A more complex situation arises in
the class F5B and F5J gliders. These are
required to perform a variety of tasks
which include endurance, distance
and pure speed runs. These tasks virtually call for three separate airframes
and the design of a single airframe to
achieve the best compromise is a very
flap movement. Also, during the speed
run, a small amount of up flap deflection may improve the aerofoil, again
depending on the aerofoil selected
for the model. All of the above only
requires a simple mixer.
Getting complicated
Fig.4: this mixer provides add-subtract outputs. Thus, the two channels
controlling the aileron servos are coupled together, with a reversal on one
channel for normal aileron control. It may also be desirable to mix some aileron
control into the flap panels to help improve turns.
demanding exercise indeed. To get the
results they require, the glider operators make extensive use of mixing.
Here we find mixing being used
to actually reconfigure the physical
properties of the entire wing and this
application falls well and truly outside
the bounds of mere operator comfort.
For the competition glider pilot, this
is life and death stuff.
One of the big problems they face is
getting the model back on the ground
due to the cleanness of the airframe.
These models are capable of very high
speeds and most enter the speed trap
at speeds around 220km/h. (Yes I did
say they were gliders. You know, no
motor). Once these models hit ground
effect, they can glide on forever and
so very effective spoilers are a must.
In addition, the endurance run requires a different camber on the wing
aerofoil to that required for the speed
run. Thus, the entire trailing edge
of the wing is given over to variable
camber devices which are required to
carry out a variety of functions. Fig.3
shows the plan outline of a typical
glider wing. The outboard trailing edge
panels are the ailerons and perform
some unusual functions. The inboard
panels are the variable camber panels
and they also perform multiple tasks.
In addition to the complex wing
functions, these models need aileron/
rudder coupling for the entire flight.
This is largely due to the reduction in
drag on the inboard wing tip and the
increase in drag on the outboard wing
tip screwing the aircraft in the opposite
direction to the turn. The long, high
aspect ratio wing (typically 13-15:1)
makes this effect more pronounced on
gliders, particularly during the slow
speed endurance flights.
To discuss the mixing required for
contest gliders, we need to understand
that each control surface on the wing
requires a separate servo and thus four
servos and four separate channels are
used, all with mixing applied. In addition, there is the usual configuration of
a separate elevator and rudder servos,
the only unusual feature being that
the rudder and elevator servos may
be buried in the fin or rear fuselage for
balance. So we are talking about a very
sophisticated little aeroplane capable
of a wide range of tasks.
To begin, let’s put the simple mixer
of Fig.1 in place for a coupled aileron/
rudder. This is usually switched out
during the speed run. During the high
speed runs, very snappy turns are
required and here the old control line
trick discussed previously is of great
benefit. Thus, we must add another
mixer for coupled flaps/elevators, only
this time we mix in both flap servos.
So, when the elevators go up both
flap servos go down, the mix again
being determined by experimentation.
This must be capable of being switched
in and out, as it is not desirable to use
this feature in the endurance run, for
example. The typical maximum deflection of the flap is about 5°.
It is also desirable to use variable
camber on the trailing edge of the wing
to provide the best lift/drag ratio on the
aerofoil for each task, so we must have
normal flap control. Hence, we select a
bit of flap to increase the camber during
the endurance run, to improve the lift/
drag ratio of the wing. Thus, both flaps
need to be able to be moved down as
a normal flap, the angle of deflection
depending on the aerofoil section used.
In addition, during winch launch,
the wing camber is increased for maximum lift and thus line tension. This
calls for approximately 20° of down
Now we get to the really complicated bit. The ailerons which control the
roll axis require opposite rotation from
each servo, thus any mixing applied to
these controls will require an inverter
with a gain of -1. The mixer in Fig.4
is typical and provides Add-Subtract
outputs. Thus, the two channels controlling the aileron servos are coupled
together with a reversal on one channel
for normal aileron control.
It may also be desirable to mix some
aileron control into the flap panels to
help improve the turns.
The landing configuration calls for
the lift to be dumped and the drag to be
increased as much as possible. Here we
see a remarkable configuration used on
the wing which is known as “crow”. In
this configuration, the ailerons which
usually work in opposition are both
raised up 20°. This reduces the lift
across this portion of the wing and also
ensures that the wing tips do not stall
before the centre section. Conversely,
the centre section flaps are deflected
down by approximately 60° to provide the drag necessary to slow these
missiles down for landing. All of this
requires very complex mixing facilities and a great deal of experience on
the pilot’s behalf to set up and master.
All of the above combinations must
be capable of being switched in and
out instantly and in the heat of a turn
at 220km/h, initiated up to 1km from
the operator and sometimes close to
the ground. This is definitely not for
the fainthearted.
So there you have it. It only takes
a moment’s reflection to see that the
development of a commercial computerised transmitter with the flexibility
to handle all of the above scenarios is
a serious undertaking. You can also see
why the modern computer radio has
become so complex and why in many
instances it has outgrown the requirements of the average club modeller.
The proposed Mk.22 transmitter will
have a simple system which can be
tailored to your own requirements. You
add only the features you need. Only a
handful of people require a system as
SC
complex as described earlier.
December 1995 83
|