Electronic Building Blocks
By Julian Edgar
Quick and easy construction
Great results on a low budget
Selecting and Using Actuators – Part 1
Turning electronic signals into physical movement – linear and rotary
O
ver recent issues we’ve covered simple and
quick ways of controlling linear actuators and stepper
motors and soon we will look at RC servos. But what
we haven’t covered is how to select these for different applications. Since that’s a critical part of building any project that
uses actuators, that’s what we’ll do in this article.
For our purposes, an actuator is any device that can turn
an electronic signal into a physical movement of sufficient
size and force to achieve the required mechanical task. That
task might be to open a door or a cooling vent, move a model
(or real) robot, oscillate a fairground attraction in a model
railway layout, steer a solar reflector or PV panel – or even
raise an aerodynamic air brake on an amateur racing car. In
addition to linear actuators, stepper motors and RC servos,
we can add to the list gearmotors and electrical solenoids.
Let’s take a look – it’s a fascinating subject!
Controlling DC motors
The great advantage of DC motors is that their speed is
easy to control by simply varying the supply voltage (and
so motor current). Pulse-width modulation (PWM), where
a varying duty cycle square wave is used (duty cycle is the
‘on’ versus ‘off’ percentage of time), is highly effective at
controlling motor speed. A typical PWM frequency for a
small DC motor is 20kHz – that is, the supply is switched on
and off 20,000 times per second. Of course, the motor cannot
respond at this rate and so behaves as if it is seeing a lower
average voltage – for example, a 50% duty cycle halves the
effective supply voltage. Motor speed controllers (Fig.3) are
cheap and widely available (they use a manual pot to control
speed) or a MOSFET module can be duty-cycle controlled
from a microcontroller to achieve automatic speed control.
DC brushed motors
Permanent magnet, brushed DC motors are widely used in
actuators. These motors are cheap and powerful for their
size. In the type of actuators we’re talking about, they vary
from being smaller than your little finger to being perhaps
as large as your fist (Fig.1).
DC motors of this type are relatively low in torque but
high in speed. (See the breakout for more on power and
torque.) Therefore, unless they are being used in low-torque
applications like driving fans or rapidly spinning an offcentre weight (so creating a vibration alert), this type of motor
is normally used in conjunction with gearing. Examples of
geared DC motors include those used in RC servos, linear
actuators and gearmotors (a gearmotor is a motor with a
gearbox built onto one end).
Not widely used in hobby pursuits, but well worth
considering if you want a very high-torque, low-speed
output are car windscreen wiper motors (Fig.2). These
use a worm-drive to give a huge speed reduction and so
produce immense torque. And while we’re on the topic of
salvaging parts from cars: if you want to be able to control
flow through a small duct, electronically controlled throttles
are now available cheaply second-hand. These devices use
geared DC motors and incorporate feedback pots (more on
feedback in a moment).
64
Fig.1. Small ‘hobby’ brushed DC motors are widely available and
cheap. They can also be salvaged from old tape recorders and toys.
These motors are low-torque, high-speed designs and so unless
they are driving a fan, they typically need to work through a reduction
geartrain to give a usable output.
Practical Electronics | October | 2022
Fig.2. Car windscreen wiper motors use a DC motor and high-reduction worm-gear drive
to give the output shaft immense torque. These motors are available cheaply both new
and second-hand.
Disadvantages of DC motors
And the downsides of DC brushed
motors? Most small DC motors tend
to have quite a limited life. They
generally used plain (bush) bearings
that are not easy to lubricate. Also,
the carbon brushes used to transfer
power to the commutator wear away
over time. (Precious metal leaves,
sometimes used to replace carbon
brushes, have even shorter lives). So,
if you are thinking of an application
where the motor is in constant
operation, steer away from actuators
that use brushed DC motors. On the
other hand, many actuators are used
Voltage
source
+
–
DPDT
switch
DC
motor
+
When the polarity of the power
supply is reversed, DC brushed
motors reverse in direction. Therefore,
directional control can be achieved
with a simple double pole, double
throw (DPDT) switch or relay (Fig.4), or
a microcontroller-controlled H-bridge.
Fig.3. Variable duty cycle motor speed control
module. This one is rated at 3A and costs
under £2. It will work with an input voltage of
6 – 28V with speed control via a manual knob.
It’s ideal for varying the speed of brushed DC
motors. These controllers are widely available
– check eBay and Banggood.
M
–
Fig.4. By using a double-pole, double-throw
(DPDT) switch or relay, the direction of a DC
motor is easily reversed.
Power and torque
L
et’s imagine that we want to open a heavy
farm gate. When doing this manually, we go to the
end of the gate furthest from the hinges and pull at
right angles to the gate. We then ‘walk’ the gate through 90
degrees until it is fully open. Easy, huh?
But let’s now replace the manual work of opening the
gate with an automatic actuator, for example a motordriven winch. We’re going use one winch to open the gate
and another to close it. Because we don’t want to obstruct
the road, the winch cables will connect nearer the hinge
of the gate. If you’ve ever tried to open a gate pulling or
pushing close to the hinge, you’ll know it takes a lot of
force, especially to get the gate moving.
So in this case, to pull in the cable, the winch will need
to develop a lot of twisting force, which is called ‘torque’.
But does the gate need to move fast? Not really – it’s not a
problem if it takes (say) 30 seconds to open. That’s good
because it will reduce the power we need.
Let’s look at that in more detail. To put it simply, power
is the product of torque and speed. Therefore, as in the gate
example, if we need high torque and low speed, the power
demand won’t be great. In this situation, we’d use a relatively
low power motor teamed with a geartrain to multiply the
torque (and so at the same time, slow the winch speed).
Another example of where we have a low power demand
is where we need a high speed but a low torque. This might
be the cooling fan in a piece of electrical equipment – the
Practical Electronics | October | 2022
speed is high, but the torque is low (it’s easy to spin the
fan), so requiring only low power – perhaps just a few watts
for a small cooling fan.
But what about where you need high torque and high
speed – and so a lot of power? For this very reason, you
don’t see this very often in actuators used in hobby electrical
and electronic systems. However, one example might be a
home-built electric bicycle climbing a hill. The speed needs
to be sufficiently high that the bike doesn’t fall over, and
yet the torque must be great enough to propel the weight
of the bike and rider up the hill. The resulting power may
well be 300 or 400 watts.
So, a fundamental decision when selecting an actuator
is to decide on the required power and torque:
Torque
requirement
High torque
Low torque
High torque
Speed
requirement
Low speed
High speed
High speed
Resulting
power requirement
Low power
Low power
High power
Note that you don’t need to quantify these values – that is,
to actually measure the required torque and speed. (Well,
you would if you were a professional engineer designing
a system.) In our case, just having a feel for the required
speed and torque – and so power – will stop us making
gross errors in selecting actuators.
65
Fig.6. Mounts for linear actuators need to cater for some rotational
movement. Light-duty mounts (left) are folded from sheet metal while
heavy duty mounts (right) are cast or machined from solid. The mount
pins fit through holes in the linear actuator’s body and extending rod.
Fig.5. A linear actuator – when power is applied, the DC motor rotates,
so turning a leadscrew and extending the actuator rod (top) from
the body. This linear actuator incorporates potentiometer position
feedback and costs about £75.
over their entire life for a total of only a few hours, so it
very much depends on the intended application.
If heavily loaded or stalled, the current demand of a DC motor
rises rapidly. This may overload the supply circuit or cause the
motor itself to overheat and potentially burn out. Therefore,
some type of overload safety cut-out is best used in situations
where the motor may become overloaded. This safety cut can
be as simple as a fuse or circuit breaker, or as sophisticated
as current monitoring by measuring the voltage drop across
a low value resistor in series with the actuator power supply.
Let’s now look in more detail at three of the most common
applications of DC motors – linear actuators, RC servos and
then (next month) gearmotors.
Linear actuators
Where movement in a straight line is required, linear
actuators (Fig.5) have major advantages over other actuators.
They are easy to control for speed and direction of movement,
are available in a wide range of sizes, and develop a lot of
force – albeit at low speed. This combination of force and
speed means that they are relatively low-power devices,
which is advantageous in terms of energy consumption and
required current supply.
Inside a linear actuator is a DC motor that drives a reduction
gear-train connected to a threaded rod – sometimes called a
lead-screw. The rod normally uses a square-cut thread (like
on a bench vice), and has an internally threaded nylon saddle
(a nut) riding on it. The saddle is prevented from turning
by a tab sliding within an enclosing tube and is connected
to the extension rod. Therefore, when the motor is turning,
the saddle slowly moves along the threaded rod, extending
(or retracting) the rod. With the current switched off, the
rod stays mechanically locked in position.
Limit switches are fitted at each end of the rod’s travel.
These automatically turn off the supply current when the rod
is either fully extended or retracted. Reversing the polarity of
the supply then causes the rod to move in the other direction.
A two-position DPDT (or sprung centre-off) switch can then
be used to control the extension of a linear actuator – the
two-position switch causes the actuator to either fully extend
or retract, and the sprung centre-off switch allowing the
actuator to be manually positioned at points of less than full
movement. (This is the approach already shown in Fig.4.)
Some linear actuators are fitted with a feedback pot. This type
of linear actuator can be used in conjunction with an electronic
controller, giving accurate positioning under different loads
or with different motor supply voltages. The Pololu Jrk 21v3
USB Motor Controller uses feedback control and works well
with linear actuators drawing up to 5A peak (higher currents
Position feedback
I
n many applications where we are using
actuators, we need to have position feedback. For example, take a linear actuator. For a given actuator, the
extension speed of the actuator’s rod will depend on the
supply voltage and load to which the actuator is subjected.
Therefore, powering the actuator for (say) 3 seconds will
not give a specified actuator position – the actuator position will vary over a range. Even a stepper motor, when
stepped by a given number of pulses, may not end up in
the designated position – it may have stalled under load,
for example. Position feedback tells the controller where
the actuator actually is, not where it is supposed to be.
This feedback information can be communicated to the
controller in several ways, but an analogue voltage signal
66
from a potentiometer is often used. For example, an RC servo
has an inbuilt pot-based position sensor and the servo’s
internal control electronics use its signal to help attain the
specified output shaft rotation. Some linear actuators use
an internal pot for position sensing, and this signal can be
used by an external controller.
Where position feedback is being used, the controller
typically makes use of a PID (proportional, integral,
differential) control software approach. This allows the
magnitude of the correction to be proportional to the size
of the error, in addition to taking into account the rate of
change of the error and when a small (but fixed) error exists.
In practical use, a well set-up PID system will react quickly
but without overshoot or jittering.
Practical Electronics | October | 2022
Speed (mm/sec, loaded and
unloaded)
Maximum duty cycle (eg, 10%,
meaning that the actuator can operate
for only 10% of the time)
Maximum extension
Physical dimensions
Fig.7. A demonstrator I built to show the action of a microcontroller (an eLabtronics Stemsel)
controlling air suspension height in a car. An air compressor (rear) provides pressurised air
via a solenoid valve to an air spring (the black bellows). The air spring supports a hinged
lever laden with paving stones (left). Position feedback is given by a pot working through
two linkages (the red arms). When a paving stone is removed, the load is reduced and the
arm rises, with the controller then venting air from the spring to maintain the same level.
Rotating the front control knob can vary the height of the arm by altering the air pressure in
the spring. In a system like this, position feedback is critical.
are available in other Pololu controllers
in the family).
Linear actuators – at least those
priced at hobby levels – are not
designed for constant movement. So,
for example, it’s better if controlling
the orientation of a solar reflector or PV
panel to move the actuator every hour,
rather than every few seconds.
If you are moving large loads with a
linear actuator, you will need to design
a mechanical system that can cope with
the high forces. For example, using a
linear actuator to change the height
of a desk – moving it from a sit-down
to a stand-up design – may involve
quite large loads – especially if a child
chooses to sit on the desk as it changes
height! In this situation a strong frame
will be needed.
Linear actuators have the following
specifications:
Low operating voltage (usually 12 or
24V DC)
Current draw (minimum and
maximum of around 1A and 7A)
Maximum load (usually expressed
in newtons – divide by 9.8 to get
kilograms force)
Fig.8. A micro RC servo. The transparent case of this unit allows you to see the DC motor,
the reduction geartrain, feedback position pot and control circuit board. (Courtesy Wikipedia)
Practical Electronics | October | 2022
Linear actuators usually need swivel
mounts at each end (Fig.6). This is
because in most applications, one end
of the actuator will move through an arc
as the actuator extends. Swivel mounts
are available in a variety of strengths,
with light-duty ones made from folded
sheet metal and heavy-duty units cast
or machined from solid.
Linear actuators are more expensive
than the other actuators described here,
but for slow, powerful movement in a
straight line, a linear actuator is the pick.
RC Servos
RC servos are especially suited for
relatively small, low-power movements
made with fine control. That makes
perfect sense, because of course they
were originally designed to move
control surfaces in radio-controlled
models boats and aircraft.
RC servos use an internal geareddown DC brushed motor. Most servos
use plastic gears, but some are available
with metal internal gears, with the metal
gears being stronger. The internal stepdown gear ratio is in the order of 100:1
(Fig.8). A splined output shaft is used
and various levers and discs (‘horns’) can
be fitted to this shaft. As quality (and so
cost) increases, servos are more likely to
be fitted with ball bearings (sometimes
only on the output shaft) and have metal
rather than plastic cases.
An RC servo is a ‘smart’ device –
it contains control electronics. An
internal pot is used as a position sensor,
and the motor rotates until the shaft
reaches the required angular position.
This position is then held – if the shaft
is mechanically rotated from its set
position, it actively resists.
Servos use three connections – power
(4.8-6V), ground and signal. The control
signal comprises a pulse train with
a varying pulse width ‘on’ time. The
pulse width determines the servo’s
position, with 1.5ms corresponding
to the servo’s neutral point. That is,
pulse widths shorter than 1.5ms cause
the servo to rotate one way, and a
pulse width longer than 1.5ms causes
the servo to rotate the other way. The
normal pulse width range is 1.0ms to
2.0ms. Servos can be controlled very
accurately to attain precise angles.
A servo cannot be used without a
specific controller – unlike DC motors
and linear actuators, a servo will not
67
Fig.9. Servos are available in three sizes – micro, standard and giant. Dimensions are in mm. (Courtesy Sparkfun)
turn when simply connected to power.
Dedicated servo controllers are available
that can interface with manual controls
like a pot or switch, or can work with
a microcontroller. These controllers
can also work many servos at once –
a good example is the Pololu Micro
Maestro controller and others in the
Maestro family. Microcontrollers can
also directly operate servos – remember,
the controller doesn’t have to handle the
current requirement of the servo because
the servo is fed power directly.
Servos are available in standard and
continuous rotation forms. A standard
servo has a movement range of 90° or
180°, while a continuous rotation servo,
as the name suggests, continuously
turns. In continuous rotation servos, the
control is of only speed and direction.
Servo specifications include:
Torque
Speed
Size and weight
Weatherproofness
Maximum torque and speed depend
on the voltage with which the servo is
being fed. As a result, these values are
often specified for both 4.8V and 6V
supply voltages.
Torque is specified in kg-cm (eg, 4kgcm). To understand this specification,
consider a servo with this torque rating
that has been fitted with a 1cm long lever.
If you connect the end of the lever to a
spring balance positioned at right-angles
to the lever, and operate the servo, the
maximum ‘pull’ value the spring balance
will show is 4kg. If you double the servo
lever length to 2cm, the maximum pull
value will drop to 2kg. (That is, the lever
length in cm multiplied by the force
value in kg will, for this servo, always
equal 4.) Large servos can have a torque
specification of over 35kg-cm.
Speed is normally specified in the time
taken for the servo to rotate 60°. A typical
68
value is about 0.15 seconds, with a fast
servo taking less than 0.1 seconds and a
slow servo needing at least 0.25 seconds.
Servos are available in three sizes –
called micro, standard and giant (Fig.9).
Servos are available at ratings up to
IP67 – that is, dustproof and protected
against being submerged in one metre
of water for 30 minutes. However, most
low-cost servos are not weatherproof.
Curiously, servo current is not often
specifi ed but typical stall currents
(that is, the maximum possible) are
1A for a standard servo, a few hundred
milliamps for micro servos and as much
as 10A for a giant servo.
Prices for RC servos vary over a very
wide range. You can pay as little as £8
for a standard servo and up to as much
as £120 for a giant servo. Whatever the
price, when using servos you need to
remember that even in upscaled form,
they are basically devices developed for
use in toys and models. So, for example,
even if the torque and speed figures of a
servo indicate that one could be used for
opening a gate or lifting a TV out of a piece
of furniture, the life of the servo in those
applications is likely to be quite short.
On the other hand, moving the limbs on a
small robot or animating items in a model
railway layout are ideal uses for servos.
Next month
In Part 2, we will look at gear and
stepper motors, in addition to exploring
a simple way of switching off actuators
at fully open and closed positions.
Fig.10. A servo with its normally supplied accessories. Shown are various ‘horns’ (attachments
for the shaft), mounting grommets and collars, and mounting and horn attachment screws.
To mount a servo, a rectangular hole is normally cut and the servo placed through the hole,
its mounting tabs are bolted to the material around the hole via the grommets.
Practical Electronics | October | 2022
