Introduction & contents

Introduction Contents
With one of the most comprehensive product ranges ever Technology Overview ....................................... 3
offered, Parker’s Electromechanical Division provides
solutions for a wide variety of applications in the field of PART I - BASIC PRINCIPLES
programmable motion control. As well as drives, motors
and controls from Digiplan, Compumotor and Hauser, there Stepper motors ................................................. 5
is a choice of mechanical positioning systems and DC brush motors ............................................ 13
supporting products from Daedal and Automation Actuator
Divisions. This full range of hardware is complemented by Brushless motors ........................................... 14
sophisticated front-end software to simplify the process of
system analysis, motor-drive selection, program generation Stepping motor drives .................................... 19
and system setup. All products are supplied and supported
by Automation Technology Centres who offer expert DC brush motor drives ................................... 24
technical advice from initial concept right through to Brushless motor drives ................................... 26
commissioning and after-sales backup.
This handbook is intended as a companion to Parker Feedback devices .......................................... 28
Electromechanical’s product catalogues and covers basic
product technology as well as applications and engineering A comparison of stepper and
reference information. At the end of the book is a short servo motor technologies .......................... 33
section reviewing Parker’s range of electronic motion
control equipment.
PART II - APPLICATIONS
Torque and inertia calculations ....................... 35
Sizing mains transformers for
positioning systems ................................... 39
Dynamic braking of servo motors ................... 40
An introduction to X-Code programming ........ 42
RS232 communication troubleshooting ......... 45
EMC installation principles ............................. 46
Motors, electrical installation .......................... 46
Motors, mechanical installation ...................... 47
Stopping in an emergency ............................. 52
Safety earth wiring ......................................... 53
Limit switches in safety-critical applications ... 54
PARKER PRODUCT RANGE SUMMARY .... 55

2
Automation
Technology overview

function may be distributed between a host controller, such

Technology Overview as an industrial computer, and a slave unit that accepts
Motion control, in its widest sense, could relate to anything high-level commands. One controller may operate in
from a welding robot to the hydraulic system in a mobile conjunction with several drives and motors in a multi-axis
crane. In the field of Electronic Motion Control, we are system.
primarily concerned with systems falling within a limited We’ll be looking at each of these system elements as well
power range, typically up to about 10KW, and requiring as their relationship to each other.
precision in one or more aspects. This may involve
accurate control of distance or speed, very often both, and Choice of motor technology
sometimes other parameters such as torque or acceleration The ‘Technology Comparison’ section of this handbook
rate. In the case of the two examples given, the welding outlines the relative merits of each motor technology and
robot requires precise control of both speed and distance; discusses the type of application appropriate to each. The
the crane hydraulic system uses the driver as the feedback chart on the following page gives an overview of the
system so its accuracy varies with the skill of the operator. selection considerations and may be used as a general
This wouldn’t be considered a motion control system in this guide to the most appropriate motor type. Bear in mind
context. that many applications could be equally well served by
Our standard electronic motion control system consists of more than one technology, and there may be other
three basic elements: considerations which influence the choice, such as
compatibility with existing equipment or customer
preference. However, in general terms the stepper system
High-Level Command
Commands Signals offers the lowest-cost solution while the brushless servo
provides the highest overall performance.
Host Indexer/
Computer Drive Motor
or PLC Controller
Hybrid Stepper
DC Servo
Brushless Servo
Linear Stepper

Fig. 1.1 Elements of a motion control system

The motor
This may be a stepper motor (either rotary or linear), a DC
brush motor or a brushless servo motor. The motor needs
to be fitted with some kind of feedback device unless it is a
stepper motor.
Fig. 1.2 shows a system complete with feedback to control
motor speed and position. Such a system is known as a
closed-loop velocity servo system.

Encoder

Controller Drive Motor

Feedback loop

Fig. 1.2 Closed loop servo system

The drive
This is an electronic power amplifier that delivers the power
to operate the motor in response to low-level control
signals. In general, the drive will be specifically designed to
operate with a particular motor type – you can’t use a
stepper drive to operate a DC brush motor, for instance.

The control system

The actual task performed by the motor is determined by
the indexer/controller; it sets parameters like speed,
distance, direction and acceleration rate. The control

3
Automation
Technology selection chart

Start Here
No Is the No
Will a stepper meet load-to-motor
the torque/speed Use a brushless
inertia ratio servo system
requirements? greater than 10:1?

Yes Yes

Does the load Yes Do you need

change rapidly to run continuously
during operation? over 120 rpm?

No Yes
No

Do you need to run Yes Consider a

continuously at Use a brushless
direct-drive servo with a
speeds above servo system
2000 rpm? gearbox

No

Yes
Do you need to
control torque?

No

Is low-speed No Consider a
smoothness ministepping
important? drive

Do you need to
Yes detect position
loss OR measure
actual load
Is rapid No position to correct
Try a for backlash?
settling time microstepping
important? drive Yes

Yes
Add encoder
feedback
Use a microstepping
drive with electronic
viscosity & active
damping

Fig 1.3 Technology selection chart

4
Automation
Stepper motors

Stepper motors Variable reluctance (V.R.) motors

There is no permanent magnet in a V.R. motor, so the rotor
There are three main stepper motor types:
spins freely without “detent” torque. Torque output for a
■ Permanent Magnet (P.M.) Motors given frame size is restricted, although the torque-to-inertia
■ Variable Reluctance (V.R.) Motors ratio is good, and this type of motor is frequently used in
■ Hybrid Motors small sizes for applications such as micro-positioning
tables. VR motors are seldom used in industrial
applications. Having no permanent magnet, they are not
sensitive to current polarity and require a different driving
arrangement from the other motor types.

N N S S
S S N N
NS N S N S N S
S N
S N
N S

Fig. 2.3 Variable reluctance motor

Fig. 2.1 “Canstack” or permanent magnet motor
Hybrid motors
Permanent magnet (P.M.) motors The hybrid motor shown in Fig. 2.4 is by far the most
The tin-can or “canstack” motor shown in Figs. 2.1 and 2.2 widely-used stepper motor in industrial applications. The
is perhaps the most widely-used type in non-industrial name is derived from the fact that it combines the
applications. It is essentially a low-cost, low-torque, low- operating principles of the other two motor types (PM &
speed device ideally suited to applications in fields such as VR). Most hybrid motors are 2-phase, although 3- and 5-
computer peripherals. The motor construction results in phase versions are used. A recent development is the
relatively large step angles, but the overall simplicity lends “enhanced hybrid” motor, which uses flux-focusing
itself to economic high-volume production at very low cost. magnets to give a significant improvement in performance,
The axial-air gap or disc motor is a variant of the albeit at extra cost.
permanent magnet design which achieves higher
performance, largely because of its very low rotor inertia.
However this does restrict the applications of the motor to
Non-magnetic
those involving low inertia if the full performance of the Prelubricated Stainless
Bearing Steel Shaft
motor is to be realised.

Rotor

Stator cup A

Coil A

Coil B

Housing Rotor Stator

Stator cup B

Output shaft
Fig. 2.4 Hybrid stepper motor
The operation of the hybrid motor is most easily understood
Diagram courtesy of Airpax Corp., USA by looking at a very simple model that will produce 12 steps
per rev. (Fig. 2.5).
Fig. 2.2 Cutaway view of permanent magnet motor

5
Automation
Stepper motors

1A
By changing the current flow from the first to the second set
of stator windings (b), the stator field rotates through 90°
N and attracts a new pair of rotor poles. This results in the
rotor turning through 30°, corresponding to one full step.
S
2A N N 2B Reverting to the first set of stator windings but energising
them in the opposite direction, we rotate the stator field
S S
N through another 90° and the rotor takes another 30° step
(c). Finally, the second set of windings are energised in the
S opposite direction (d) to give a third step position. We can
now go back to the first condition (a), and after these four
steps the rotor will have moved through one tooth pitch.
1B
This simple motor therefore performs 12 steps per rev.
Obviously, if the coils are energised in the reverse
Fig. 2.5 Simple 12 step/rev hybrid motor
sequence, the motor will go round the other way.
The rotor of this machine consists of two pole pieces with If two coils are energised simultaneously (Fig. 2.7), the
three teeth on each. In between the pole pieces is a rotor takes up an intermediate position since it is equally
permanent magnet that is magnetised along the axis of the attracted to two stator poles. Greater torque is produced
rotor, making one end a north pole and the other a south under these conditions because all the stator poles are
pole. The teeth are offset at the north and south ends as influencing the rotor. The motor can be made to take a full
shown in the diagram. step simply by reversing the current in one set of windings;
The stator consists of a shell having four teeth that run the this causes a 90° rotation of the stator field as before. In
full length of the rotor. Coils are wound on the stator teeth fact, this would be the normal way of driving the motor in
and are connected together in pairs. the full-step mode, always keeping two windings energised
With no current flowing in any of the motor windings, the and reversing the current in each winding alternately.
rotor will tend to take up one of the positions shown in the
Fig 2.6. This is because the permanent magnet in the rotor
is trying to minimise the reluctance (or “magnetic
resistance”) of the flux path from one end to the other. This N S
will occur when a pair of north and south pole rotor teeth
S N
N S
are aligned with two of the stator poles. The torque tending N S
N S S N N S
to hold the rotor in one of these positions is usually small S N
N S
and is called the “detent torque”. The motor shown will
have 12 possible detent positions. S N

If current is now passed through one pair of stator

windings, as shown in Fig. 2.6(a), the resulting north and
south stator poles will attract teeth of the opposite polarity
S N
on each end of the rotor. There are now only three stable
N S
positions for the rotor, the same as the number of rotor S N
S N
teeth. The torque required to deflect the rotor from its S N
N
N S
S
S N
S N
stable position is now much greater, and is referred to as
the “holding torque”. N S

(a) (b)
N

S N S
Fig. 2.7 Full stepping, two phases on
N N
S
S S
N N S By alternately energising one winding and then two (Fig.
N
N S
2.8), the rotor moves through only 15° at each stage and
S the number of steps per rev will be doubled. This is called
half stepping, and the majority of industrial applications
make use of this stepping mode. Although there is
(c) (d) sometimes a slight loss of torque, this mode results in
S much better smoothness at low speeds and less overshoot
N
and ringing at the end of each step.
S N
S S
S N S N
N N
S S N

Fig. 2.6 Full stepping, one phase on

6
Automation
Stepper motors

1 2 3 4 5 6 7 8

S S

N N N
S S
S
S S Phase 1
S S N S N
N N
N N N
S N S S

N N

Phase 2

Fig. 2.8 Half stepping Fig. 2.10 Half step current

Current patterns in the motor windings
When the motor is driven in its full-step mode, energizing
two windings or “phases” at a time (see Fig. 2.9), the torque 1 2 3 4 5 6 7 8
available on each step will be the same (subject to very
small variations in the motor and drive characteristics). In
the half-step mode, we are alternately energising two Phase 1
phases and then only one as shown in Fig. 2.10.
Assuming the drive delivers the same winding current in
each case, this will cause greater torque to be produced
when there are two windings energised. In other words, Phase 2
alternate steps will be strong and weak. This does not
represent a major limitation on motor performance—the
available torque is obviously limited by the weaker step, but
there will be a significant improvement in low-speed
smoothness over the full-step mode. Fig. 2.11 Half step current, profiled
Clearly, we would like to produce approximately equal
torque on every step, and this torque should be at the level Microstepping
of the stronger step. We can achieve this by using a higher We have seen that energising both phases with equal
current level when there is only one winding energised. currents produces an intermediate step position half-way
This does not over-dissipate the motor because the between the one-phase-on positions. If the two phase
manufacturer’s current rating assumes two phases to be currents are unequal, the rotor position will be shifted
energised (the current rating is based on the allowable case towards the stronger pole. This effect is utilised in the
temperature). With only one phase energised, the same microstepping drive, which subdivides the basic motor step
total power will be dissipated if the current is increased by by proportioning the current in the two windings. In this
40%. Using this higher current in the one-phase-on state way, the step size is reduced and the low-speed
produces approximately equal torque on alternate steps smoothness is dramatically improved. High-resolution
(see Fig. 2.11). microstep drives divide the full motor step into as many as
500 microsteps, giving 100,000 steps per revolution. In this
situation, the current pattern in the windings closely
1 2 3 4 resembles two sine waves with a 90° phase shift between
them (see Fig. 2.12). The motor is now being driven very
much as though it is a conventional AC synchronous motor.
In fact, the stepper motor can be driven in this way from a
Phase 1 50Hz sine wave source by including a capacitor in series
with one phase. It will rotate at 60 rpm.

Phase 2 +
Phase 1 Current:
-

+
Phase 2 Current:
-
Fig. 2.9 Full step current, two phases on

Fig. 2.12 Phase currents in microstep mode

7
Automation
Stepper motors

The standard 200-step hybrid motor Bifilar windings

The standard stepper motor operates in the same way as Most motors are described as being “bifilar wound”, which
our simple model, but has a greater number of teeth on the means there are two identical sets of windings on each
rotor and stator, giving a smaller basic step size. The rotor pole. Two lengths of wire are wound together as though
is in two sections as before, but has 50 teeth on each they were a single coil. This produces two windings that
section. The half-tooth displacement between the two are electrically and magnetically almost identical – if one
sections is retained. The stator has 8 poles each with 5 coil were to be wound on top of the other, even with the
teeth, making a total of 40 teeth (see Fig. 2.13). same number of turns, the magnetic characteristics would
be different. In simple terms, whereas almost all the flux
from the inner coil would flow through the iron core, some
of the flux from the outer coil would flow through the
windings of the coil underneath.
The origins of the bifilar winding go back to the unipolar
drive (see Drive Technologies section). Rather than have
to reverse the current in one winding, the field may be
reversed by transferring current to a second coil wound in
the opposite direction. (Although the two coils are wound
the same way, interchanging the ends has the same
effect.) So with a bifilar-wound motor, the drive can be kept
simple. However, this requirement has now all but
disappeared with the widespread availability of the more-
efficient bipolar drive. Nevertheless, the two sets of
Fig. 2.13 200-step hybrid motor windings do give us additional flexibility, and we shall see
that different connection methods can be used to give
If we imagine that a tooth is placed in each of the gaps alternative torque-speed characteristics.
between the stator poles, there would be a total of 48 teeth,
two less than the number of rotor teeth. So if rotor and If all the coils in a bifilar-wound motor are brought out
stator teeth are aligned at 12 o’clock, they will also be separately, there will be a total of 8 leads (see Fig. 2.14).
aligned at 6 o’clock. At 3 o’clock and 9 o’clock the teeth will This is the most common configuration since it gives the
be misaligned. However, due to the displacement between greatest flexibility.
the sets of rotor teeth, alignment will occur at 3 o’clock and However, there are still a few motors produced with only 6
9 o’clock at the other end of the rotor. leads, one lead serving as a common connection to each
winding in a bifilar pair. This arrangement limits flexibility
The windings are arranged in sets of four, and wound such since the windings cannot be connected in parallel. Some
that diametrically-opposite poles are the same. So referring motors are made with only 4 leads, these are not bifilar-
to Fig. 2.13, north poles at 12 and 6 o’clock attract the wound and cannot be used with a unipolar drive. There is
south-pole teeth at the front of the rotor; the south poles at obviously no alternative connection method with a 4-lead
3 and 9 o’clock attract the north-pole teeth at the back. By motor, but in many applications this is not a drawback and
switching current to the second set of coils, the stator field the problem of insulating unused leads is avoided.
pattern rotates through 45°. However, to align with this new
field, the rotor only has to turn through 1.8°. This is
equivalent to one quarter of a tooth pitch on the rotor, giving
200 full steps per revolution.
Note that there are as many detent positions as there are
full steps per rev, normally 200. The detent positions
correspond with rotor teeth being fully aligned with stator
teeth. When power is applied to a stepper drive, it is usual
for it to energise in the “zero phase” state in which there is 4-lead 5-lead 6-lead 8-lead
current in both sets of windings. The resulting rotor position
does not correspond with a natural detent position, so an
unloaded motor will move by at least one half step at
Fig. 2.14 Motor lead configurations
power-on. Of course, if the system was turned off other Occasionally a 5-lead motor may be encountered. These
than in the zero phase state, or the motor is moved in the are not recommended since they cannot be used with
meantime, a greater movement may be seen at power-up. conventional bipolar drives requiring electrical isolation
Another point to remember is that for a given current between the phases.
pattern in the windings, there are as many stable positions Looking at the motor longitudinal section (Fig. 2.15), we can
as there are rotor teeth (50 for a 200-step motor). If a see the permanent magnet in the rotor and the path of the
motor is de-synchronised, the resulting positional error will flux through the pole pieces and the stator. The alternating
always be a whole number of rotor teeth or a multiple of flux produced by the stator windings flows in a plane at right
7.2°. The motor itself cannot “lose” individual steps – angles to the page, therefore the two flux paths are at right
position errors of one or two steps must be due to noise, angles to each other and only interact in the rotor pole
spurious or missed step pulses or a controller fault. pieces. This is an important feature of the hybrid motor – it

8
Automation
Stepper motors

means that the permanent magnet in the rotor does not

“see” the alternating field from the windings, hence it does
not produce a demagnetising effect. Unlike the DC servo
motor, it is almost impossible to de-magnetise a stepper
motor by applying excess current. However, too much
current will damage the motor in other ways. Excessive
heating may melt the insulation or the winding formers, and
may soften the bonding material holding the rotor
laminations. If this happens and the laminations are
displaced, the symptoms can be the same as if the rotor had
been de-magnetised.

Fig. 2.16 Three-stack hybrid stepping motor

As a rule of thumb, the torque-to-inertia ratio reduces by a
factor of two with each increase in frame diameter. So an
unloaded 34-size motor can accelerate twice as rapidly as a
42-size, regardless of the number of stacks.

Forcer

Permanent
Magnet
Phase A Phase B
Electromagnet Electromagnet Field Windings

Fig. 2.15 Longitudinal section through single stack Platen Teeth

motor Air
Gap N S
Fig. 2.15 also shows that the rotor flux only has to cross a
very small air gap (typically less than 0.1mm) when the rotor Platen
{
{

{
{
is in position. By magnetising the rotor after assembly, a A1 A2 B1 B2
Pole Faces
high flux density is obtained that will be largely destroyed if
the rotor is removed. Stepper motors should therefore
never be dismantled purely to satisfy curiosity, since the Fig. 2.17 Linear stepping motor
useful life of the motor will come to an end.
Linear stepping motors
Because the shaft of the motor passes through the centre of
the permanent magnet, a non-magnetic material must be The linear stepper is essentially a conventional rotary
used to avoid a magnetic short-circuit. Stepper motor shafts stepper that has been “unwrapped” so that it operates in a
are therefore made of stainless steel, and should be straight line. The moving component is referred to as the
handled with care. Small-diameter motors are particularly forcer and it travels along a fixed element or platen. For
vulnerable if they are dropped on the shaft end, as this will operational purposes, the platen is equivalent to the rotor in
invariably bend the shaft. a normal stepper, although it is an entirely passive device
and has no permanent magnet. The magnet is
Multi-stack motors incorporated in the moving forcer together with the coils
To produce a motor with a higher torque output, (see Fig. 2.17).
we need to increase the strength of both the permanent The forcer is equipped with 4 pole pieces each having 3
magnet in the rotor and the field produced by the stator. A teeth. The teeth are staggered in pitch with respect to
stronger rotor magnet can be obtained by increasing the those on the platen, so that switching the current in the
diameter, giving us a larger cross-sectional area. However, coils will bring the next set of teeth into alignment. A
increasing the diameter will degrade the acceleration complete switching cycle (4 full steps) is equivalent to one
performance of the motor because tooth pitch on the platen. Like the rotary stepper, the linear
the torque-to-inertia ratio gets worse (to a first motor can be driven from a microstep drive. In this case, a
approximation, torque increases with diameter squared but typical linear resolution will be around 500 steps per mm or
inertia goes up by the fourth power). 2 microns per step.
Nevertheless, we can increase torque output without The linear motor is best suited to applications that require a
degrading acceleration performance by adding further low mass to be moved at high speed. In a leadscrew-
magnet sections or “stacks” to the same shaft (Fig. 2.16). A driven system, the predominant inertia is usually the
second stack will enable twice the torque to be produced leadscrew rather than the load to be moved. Hence most
and will double the inertia, so the torque-to-inertia ratio of the motor torque goes to accelerate the leadscrew, and
remains the same. Hence, stepper motors are normally this problem becomes more severe the longer the travel
produced in single-, two- and three-stack versions in each required. Using a linear motor, all the developed force is
frame size. applied directly to the load and the performance achieved is

9
Automation
Stepper motors

independent of the length of the move. A screw-driven The magnetic flux passing between the forcer and the
system can develop greater linear force and better stiffness, platen gives rise to a very strong force of attraction. The
however the maximum speed may be up to ten times higher attractive force can be up to 10 times the peak holding
with an equivalent linear motor. For example, a typical force of the motor, requiring a bearing arrangement to
maximum speed for a linear motor is 2.5m/sec. To achieve maintain precise clearance between the pole faces and
this with a 10mm-pitch ballscrew would require a rotary platen teeth. Either mechanical roller bearings or air
speed of 15,000 rpm. In addition, the linear motor can bearings are used to maintain the required clearance.
travel up to 2.4 metres using a standard platen. When current is established in one of the windings, the
resulting magnetic field tends to reinforce permanent
How the linear motor works
magnetic flux at one pole face and cancel it at the other.
The forcer consists of two electromagnets (A and B) and a By reversing the current, the reinforcement and cancellation
strong rare earth permanent magnet. The two pole faces of are exchanged. Removing current divides the permanent
each electromagnet are toothed, the four sets of teeth being magnetic flux equally between the pole faces. By
spaced in quadrature so that only one set at a time can be selectively applying current to phase A and B, it is possible
aligned with the platen teeth. to concentrate flux at any of the forcer’s four pole faces.
Phase B
The face receiving the highest flux concentration will
Phase A
attempt to align its teeth with the platen. Fig. 2.18 shows
N S the four primary states or full step positions of the forcer.
The four steps result in motion of one tooth pitch to the
right. Reversing the sequence moves the forcer to the left.

Step motor characteristics

There are several stepper performance characteristics that
we could discuss, however to keep things simple we’ll
Flux Lines confine ourselves to those with the greatest practical
A1 Aligned Direction of MMF due
to electromagnet
significance.
Fig. 2.19 illustrates the static torque curve of the motor.
This relates to a motor that is energised but stationary, and
shows how the restoring torque varies with rotor position as
N S
it is deflected from its stable point. We’re assuming that
there are no frictional or other static loads on the motor. As
the rotor moves away from the stable position, the torque
steadily increases until it reaches a maximum after one full
step (1.8°). This maximum value is called the holding
torque and it represents the largest static load that can be
applied to the shaft without causing continuous rotation.
B2 Aligned However, it doesn’t tell us the maximum running torque of
the motor – this is always less than the holding torque
(typically about 70%).

N S
Counter Clockwise

4 Motor Steps

Max
Torque
Torque

Angle
Stable Unstable Stable
Clockwise

A2 Aligned

N S

Fig. 2.19 Static torque-displacement characteristic

As the shaft is deflected beyond one full step, the torque will
fall until it is again at zero after two full steps. However, this
zero point is unstable and the torque reverses immediately
beyond it. The next stable point is found four full steps
away from the first, equivalent to one tooth pitch on the rotor
B1 Aligned
or 1/50 of a revolution.
Fig. 2.18 The four primary states of the forcer

10
Automation
Stepper motors

Although this static torque characteristic isn’t a great deal of Under full dynamic conditions, the performance of the
use on its own, it does help explain some of the effects we motor is described by the torque-speed curve as shown in
observe. For example, it indicates the static stiffness of the Fig. 2.21. There are two operating ranges, the start/stop
system, in other words how the shaft position changes when (or pull in) range and the slew (or pull out) range. Within
a torque load is applied to a stationary motor. Clearly the the start/stop range, the motor can be started or stopped by
shaft must deflect until the generated torque matches the applying step pulses at constant frequency to the drive. At
applied load, so if the load varies, so too will the static speeds within this range, the motor has sufficient torque to
position. Non-cumulative position errors will therefore result accelerate its own inertia up to synchronous speed without
from effects such as friction or out-of-balance torque loads. the position lag exceeding 3.6°. Clearly if an inertial load is
It is important to remember that the static stiffness is not added, this speed range is reduced. So the start/stop
improved by using a microstepping drive—a given load on speed range depends on the load inertia. The upper limit to
the shaft will produce the same angular deflection. So while the start/stop range is typically between 200 and 500 full
microstepping increases resolution and smoothness, it may steps/sec (1-2.5 revs/sec).
not necessarily improve positioning accuracy.
Under dynamic conditions with the motor running, the rotor Holding
Torque Start/Stop Curve
must be lagging behind the stator field if it is producing
torque. Similarly there will be a lead situation when the Slew Curve
torque reverses during deceleration. Note that the lag and
lead relate only to position and not to speed. From the

Torque
static torque curve (Fig. 2.19), clearly this lag or lead cannot
exceed two full steps (3.6°) if the motor is to retain Slew
Start/ Range
synchronism. This limit to the position error can make the Stop
stepper an attractive option in systems where dynamic Range
position accuracy is important. Steps per second

When the stepper performs a single step, the nature of the

response is oscillatory as shown in Fig. 2.20. The system Fig. 2.21 Start/stop and slew curves
can be likened to a mass that is located by a “magnetic To operate the motor at faster speeds, it is necessary to
spring”, so the behaviour resembles the classic mass-spring start at a speed within the start/stop range and then
characteristic. Looking at it in simple terms, the static accelerate the motor into the slew region. Similarly, when
torque curve indicates that during the step, the torque is stopping the motor, it must be decelerated back into the
positive during the full forward movement and so is start/stop range before the step pulses are terminated.
accelerating the rotor until the new stable point is reached. Using acceleration and deceleration or “ramping” allows
By this time, the momentum carries the rotor past the stable much higher speeds to be achieved, and in industrial
position and the torque now reverses, slowing the rotor applications the useful speed range extends to about 3000
down and bringing it back in the opposite direction. The rpm (10,000 full steps/sec). Note that continuous operation
amplitude, frequency and decay rate of this oscillation will at high speeds is not normally recommended with a stepper
depend on the friction and inertia in the system as well as due to rotor heating, but high speeds can be used
the electrical characteristics of the motor and drive. The successfully in positioning applications.
initial overshoot also depends on step amplitude, so half- The torque available in the slew range does not depend on
stepping produces less overshoot than full stepping and load inertia. The torque-speed curve is normally measured
microstepping will be better still. by accelerating the motor up to speed and then increasing
the load until the motor stalls. With a higher load inertia, a
lower acceleration rate must be used but the available
torque at the final speed is unaffected.

Common questions and answers about

step motors
Angle

1 Why do step motors run hot?

Two reasons: 1. Current flows through the motor
Time
windings even at standstill. 2. PWM drive designs tend
to make the motor run hotter. Motor construction, such
as lamination material and riveted rotors, will also affect
Fig. 2.20 Single step response heating.
Attempting to step the motor at its natural oscillation
2 What is a safe operating temperature?
frequency can cause an exaggerated response known as
resonance. In severe cases, this can lead to the motor Most motors have class B insulation, which is rated at
desynchronising or “stalling”. It is seldom a problem with 130°C. Motor case temperatures of 90°C should not
half-step drives and even less so with a microstepper. The cause thermal problems. However, motors should not
natural resonant speed is typically in the range 100-200 full be mounted where operators can come into contact with
steps/sec (0.5-1 rev/sec). the motor case.

11
Automation
Stepper motors

3 What can be done to reduce motor heating? 12 Does the motor itself change if a
Many drives feature an automatic reduction of current at microstepping drive is used?
standstill set by a command or jumper link. This No, the motor is still a standard 1.8° stepper.
reduces current when the motor is at rest without Microstepping is accomplished by proportioning the
positional loss. current in the drive to give a higher resolution.
4 What does the absolute accuracy specification 13 A move is made in one direction, and then the
mean? motor is commanded to move the same
This refers to non-cumulative errors encountered in distance but in the opposite direction. The
machining the motor. move ends up short, why?
5 How can the repeatability specification be Several factors could be influencing the [Link]
better than the accuracy? motor has magnetic hysteresis that is seen on direction
changes, this is of the order of 0.03°. Any mechanical
Repeatability indicates how precisely a previous backlash in the system will create positional errors. A
position can be re-established. There are very few step pulse occurring before the direction signal is
errors in the system that cause a given position to vary, established will cause an error.
so basically the same inaccuracy is encountered on
returning to that position. 14 Why are some motors constructed as eight-
lead motors?
6 Will motor accuracy increase proportionately
with the resolution? This allows greater flexibility. The motor can be run as
a six-lead motor with unipolar drives. With bipolar
No. The basic absolute accuracy and hysteresis of the drives, the windings can then be connected in either
motor remain unchanged. series or parallel to give different torque-speed
7 Can I use a small motor on a large load if the characteristics.
torque requirement is low? 15 What is the difference between series and
Yes, but if the load inertia is more than ten times the parallel connection?
rotor inertia, extended ringing at the end of the move With the windings in series, low-speed torque is
may be experienced. maximised. But this also gives the highest inductance,
8 How can end of move “ringing” be reduced? so performance at high speed is lower than if the
windings are connected in parallel.
Friction in the system will help damp this oscillation. If a
start/stop velocity is used, lowering it will help. The 16 Can a flat be machined on the motor shaft?
electronic viscosity feature in the ZETA drive minimises Yes, but care must be taken to not damage the
this problem. bearings (see the Motor Installation section). The motor
9 Why can a motor stall during no load testing? must not be dismantled.

An unloaded motor develops a high torque in relation to 17 How long can the motor leads be?
its own inertia. Steppers are optimised for loads of the Under normal conditions, 15m for unipolar drives and
same order as the rotor inertia. 30m for bipolar drives should be OK. Shielded cables
10 How can you reduce the chance of are recommended.
resonance?
This is most likely to happen with full step systems.
Adding inertia lowers the resonant frequency, and
friction tends to damp out resonance. Start/stop
velocities higher than the resonant speed will help.
Changing to a half step, ministep or microstep drive will
also minimise any resonance. Viscous inertial dampers
are sometimes used.
11 Why does the motor jump at times when it's
turned on?
The motor has 200 natural detent positions, but the
drive powers up in a defined state corresponding to one
of only 50 positions. Movement can therefore be up to
3.6° in either direction.

12
Automation
Servo motors

DC brush motors Commutation

The force that rotates the motor armature is the result of
Principles interaction between two magnetic fields (the stator field and
The DC brush motor makes use of the force experienced by the armature field). To produce a constant torque from the
a current-carrying conductor when placed in a magnetic field motor, these two fields must remain constant both in
(Fig. 3.1). magnitude and in relative orientation.
Magnetic Field (B)
2

1 3

Conductor Current
In Out
Carrying
Current (I)
(Into Page) 6 4

Force (F) 5

Force on Conductor F = I x B
Fig. 3.3 Electrical arrangement of the armature
Fig. 3.1 Force on a conductor in a magnetic field This is achieved by constructing the armature as a series of
small sections connected in sequence to the segments of a
The force acting on the conductor is the product of current
commutator (Fig 3.3). Electrical connection is made to the
and flux density:
commutator by means of two brushes. If the armature
F=IxB rotates through 1/6 of a revolution clockwise, the current in
where B = magnetic flux density and I = current coils 3 and 6 will have changed direction. As successive
The available force can be increased by using more commutator segments pass the brushes, the current in the
conductors, or by winding a length of wire into a coil. This is coils connected to those segments changes direction. This
the basis of a DC brush motor. commutation or switching effect results in a current flow in
the armature that occupies a fixed position in space,
Practical considerations independent of armature position. This will result in
The problem now is how to use this force to produce the production of a constant torque output from the motor shaft.
continuous torque required in a practical motor. The axis of magnetisation is determined by the position of
To obtain the greatest possible force, a large number of the brushes. If the motor is to have similar characteristics
conductors must be placed in the magnetic field. In practice in both directions of rotation, the brush axis must be
this produces a cylinder of wire, with the windings running positioned to produce an axis of magnetisation at 90° to the
parallel to the axis of the cylinder. A shaft running through stator field.
the axis acts as a pivot, and this arrangement is called the Brush motor types
motor armature (Fig. 3.2).
Two different types of DC brush motor are currently in use.
Resultant
Field Due to Iron cored motor
Armature Shaft
Current This is the most common type of motor used in DC servo
systems (see Fig 3.4). It is made up of two main parts - a
housing containing the field magnets, and a rotor with coils
of wire wound in slots in an iron core and connected to a
Armature
commutator. Brushes in contact with the commutator carry
the current to the coils.

Stator Magnets
Direction
of Current
Stator Field Into Page

Rotor Winding

Fig. 3.2 DC motor armature

As the armature rotates, so does the resulting magnetic
field. The armature will come to rest with its magnetic field Commutator
Brushes
aligned with stator field, unless some provision is made to
constantly change the direction of the current in the
individual armature coils. Fig. 3.4 Iron-cored motor

13
Automation
Servo motors

Moving coil motor Motor equations

There are two common forms of this motor - the “printed” Unlike a step motor, the DC brush motor exhibits simple
motor (Fig. 3.5), using a disc armature, and the “shell” type relationships between current, voltage, torque and speed.
armature (Fig. 3.6). It is useful to look at these relationships as an aid to the
Since these motors have no moving iron in their magnetic application of brush motors.
fields, they don’t suffer from iron losses. This means that
The brush motor equivalent circuit
higher rotational speeds can be obtained with low power
inputs. However their thermal capacity is very small so they If you apply a constant voltage to the terminals of a brush
are more easily damaged by excessive current. motor, it will accelerate up to a steady final speed (n).
Under these conditions, the voltage (V) applied to the
Diagrams by courtesy of Electro-Craft Ltd. motor is opposed by the back emf (nKE); the resulting
Motion difference voltage drives the motor current (I) through the
motor armature and brush resistance (Rs).
The equivalent circuit of a DC motor is shown in Fig. 3.8.

I Rs

Permanent magnet v RL Vg
(8 pole)

Fig. 3.5 Disc-armature “printed” motor

Fig 3.8 motor equivalent circuit
Magnet pole Rs = motor resistance
Air gap S
Motion L = winding inductance
Vg = back emf and
RL represents magnetic losses.
Flux path
Armature Core S The value of RL is usually large and so can be ignored, as
(Hollow cup, shaped can the inductance L, which is generally small.
conductor array)
If we apply a voltage V to the motor and a current I flows,
then:
Fig. 3.6 Shell-armature motor V = IRs + Vg
Back emf but Vg = nKE
When driven mechanically, a permanent magnet DC motor so V = IRs + nKE (1)
will act as a generator. As the shaft rotates, a voltage will
appear across the brush terminals. This voltage is called This is the basic electrical equation of the motor.
the back emf and is generated even when the motor is If KT is the torque constant of the motor (in Nm per amp),
driven by an applied voltage. The output voltage is then the torque generated by the motor is given by:
essentially linear with motor speed and has a slope that is
T = IKT (2)
defined as the motor voltage constant, KE (Fig. 3.7). KE is
usually quoted in volts per 1000 rpm. There is an opposing torque due to friction and internal
damping, but this is usually small and may be ignored for
the purposes of initial calculation.
Equations (1) and (2) are all we need to work out the
required current and drive voltage to meet given torque and
speed requirements. The values of KT, KE, and Rs will be
found in the motor manufacturer’s data.
Output
volts Brushless motors
Before we talk about brushless motors in detail, let’s clear
up a point about terminology. The term “brushless” is now
accepted as referring to a particular variety of servo motor.
Shaft speed Clearly a step motor is a brushless device, as is an AC
induction motor. However, the so-called “brushless” motor
has been designed to have a similar performance to the
Fig. 3.7 Back-emf characteristic
DC brush servo without its limitations.

14
Automation
Servo motors

The chief limiting factor in the performance of iron-cored through 180°). By including the commutator and brushes,
motors is internal heating. This heat must escape either the reversal of current is made automatically and the rotor
through the shaft and bearings to the outer casing, or continues to turn in the same direction.
through the airgap between the armature and field magnets
and from there to the casing. Neither of these routes is
thermally efficient, so cooling of the motor armature is very
poor.
Backiron
Return
Path N S

Commutator
+ -
N Magnets N

Fig. 3.10 Conventional DC brush motor

To turn this motor into a brushless design, we must start by
S eliminating the windings on the rotor. This can be achieved
Winding by turning the motor inside out. In other words, we make
Stator
Lam Teeth the permanent magnet the rotating part and put the
windings on the stator poles. We still need some means of
reversing the current automatically – a cam-operated
Fig. 3.9 Brushless motor reversing switch could be made to do this job (Fig. 3.11).
In the brushless motor, the construction of the iron cored
motor is turned inside out so that the rotor becomes a
permanent magnet and the stator becomes a wound iron
core.
The current-carrying coils are now located in the housing,
providing a short and efficient thermal path to the outside
N
air. Cooling can be further improved by finning the outer
casing and blowing air over it if necessary (to effectively S
cool an iron-cored brush motor, air must be blown through
it.) The ease of cooling the brushless motor allows it to
produce a much higher power in relation to its size.
The other major advantage of the brushless motor is its lack
of a conventional commutator and brush gear. These items
are a source of wear and potential trouble and may require Reversing Switch
frequent maintenance. By not having these components,
the brushless motor is inherently more reliable and can be
used in adverse environmental conditions. + -
To achieve high torque and low inertia, brushless motors
Fig. 3.11 “Inside out” DC motor
use rare earth magnets that are much more expensive than
conventional ceramic magnets. The electronic drive Obviously such an arrangement with a mechanical switch is
needed for a brushless motor is also more complex than for not very satisfactory, but the switching capability of non-
a brush motor. contacting devices tends to be very limited. However, in a
Within the brushless category are two basic motor types: servo application, we will use an electronic amplifier or drive
trapezoidal and sine wave. Although the construction is which can also be used to perform the commutation in
virtually identical, the way in which the two types are driven response to low-level signals from an optical or Hall-effect
differs significantly. The trapezoidal motor is often referred sensor (see Fig. 3.12). This component is referred to as the
to as a brushless DC servo, whereas the sine wave motor is commutation encoder. So unlike the DC brush motor, the
sometimes called an AC brushless motor as it bears a close brushless version cannot be driven by simply connecting it
resemblance to the AC synchronous motor. To fully explain to a source of direct current. The current in the external
the difference between these motors, we must review the circuit must be reversed at defined rotor positions. Hence
evolution of the brushless motor. the motor is actually being driven by an alternating current.
A simple conventional DC brush motor (Fig. 3.10) consists
of a wound rotor that turns within a magnetic field provided
by the stator. If the coil connections were made through
slip rings, this motor would behave like a step motor
(reversing the current in the rotor would cause it to flip

15
Automation
Servo motors

A1
A1

B2 C2

C2 A2
N
C1 B1
S B2
A2 C1 B1

Fig. 3.14 Three-phase brushless motor

Commutation
Encoder
A1 A1

B2 C2 I
N
Drive N C2 A2
S B2
C1 S C1 B1
B1

A2 Stator
Field
Fig. 3.12 Brushless motor
Going back to the conventional brush motor, a rotor A1 A1

consisting of only one coil will exhibit a large torque B2 C2

N
variation as it rotates. In fact the characteristic will be N
I
C2 A2
sinusoidal, with maximum torque produced when the rotor S B2
C1 S B1
field is at right angles to the stator field and zero torque at C1 B1

the commutation point (see Fig. 3.13). A practical DC motor Stator A2

has a large number of coils on the rotor, each one Field

connected not only to its own pair of commutator segments

A1
but to the other coils as well. In this way, the chief A1

contribution to torque is made by a coil operating close to its Stator B2 C2

peak-torque position; there is also an averaging effect Field S N C2 A2
produced by current flowing in all the other coils, so the S N B2
C1 B1 C1 I B1
resulting torque ripple is very small.
A2

Fig. 3.15 Stator field positions for different phase

+ currents
Torque Direction of Ro
- 0° 90° 180° Field Relative field in increments of 60° (Fig. 3.15 shows the stator field
to Stator Field
positions for three of the six possible current paths). In
other words, the stator field stays in the same place during
60° of shaft rotation. So we can’t maintain a constant 90°
torque angle, but we can maintain an average of 90° by
Fig. 3.13 Position-torque characteristic working between 60° and 120°. Fig. 3.16 shows the rotor
position at a commutation point; when the torque angle has
We would like to reproduce a similar situation in the fallen to 60°, the stator field is advanced from 2 to 3 so that
brushless motor, however this would require a large number the angle now increases to 120°, and it stays here during
of coils distributed around the stator. This may be feasible, the next 60° of rotation.
but each coil would require its own individual drive circuit.
This is clearly prohibitive, so a compromise is made and a
typical brushless motor has either two or three sets of coils
or “phases” (see Fig. 3.14). The motor shown is a two-pole, Stator S
three-phase design; usually the motor has four or six rotor 2 Field
Rotation
poles, with a corresponding increase in the number of stator N
poles. This doesn’t increase the number of phases—each
60°
phase has its turns distributed between several stator poles.
Stator 120°
The torque characteristic in Fig. 3.13 indicates that Field
maximum torque is produced when the rotor and stator Rotor
fields are at 90° to each other. Therefore to generate 3 Field
Average Lag = 90°
constant torque we would need to keep the stator field a
constant 90° ahead of the rotor. Limiting the number of
phases to three means that we can only advance the stator Fig. 3.16 Position of rotor at commutation point

16
Automation
Servo motors

The trapezoidal motor angle very accurately, resulting in very smooth low-speed
rotation and negligible torque ripple.
With a fixed current level in the windings, the use of this
extended portion of the sinusoidal torque characteristic The drive for a sine wave motor is more complex than for
gives rise to a large degree of torque ripple. We can the trapezoidal version. We need a reference table from
minimize the effect by manipulation of the motor design to which to generate the sinusoidal currents, and these must
“flatten out” the characteristic – to make it trapezoidal, (Fig. be multiplied by the torque demand signal to determine
3.17). In practice this is not very easy to do, so some their absolute amplitude. With a star-connected three-
degree of non-linearity will remain. The effect of this tends phase motor, it is sufficient to determine the currents in two
to be a slight “kick” at the commutation points, which can be of the windings—this will automatically determine what
noticeable when the motor is running slowly. happens in the third. As previously mentioned, the sine
wave motor needs a high-resolution feedback device,
however this can also serve to provide position and velocity
information for the controller.
60°
To understand why a sine wave motor produces constant
torque, it’s easier to think in terms of a two-phase motor.
This has just two sets of windings that are fed with
sinusoidal currents at 90° to each other. If we represent
60° shaft position by an angle φ, then the currents in the two
windings are of the form Isinφ and Icosφ.
Going back to our original motor model, you’ll remember
Fig. 3.17 Trapezoidal motor characteristic that the fundamental torque characteristic of the motor is
also sinusoidal. So for a given current I, the instantaneous
Torque ripple resulting from non-linearity in the torque torque value looks like:
characteristic will tend to produce a velocity modulation in
T = I KT sinφ
the load. However, in a system using velocity feedback the
velocity loop will generally have a high gain. This means where KT is the motor torque constant. By making the
that a very small increase in velocity will generate a large motor current sinusoidal as well, and in phase with the
error signal, reducing the torque demand to correct the motor torque characteristic, the torque generated by one
velocity change. So in practice, the output current from the phase becomes:
amplifier tends to “mirror” the torque characteristic (Fig. T1 = (I sinφ) KT sinφ
3.18) so that the resulting velocity modulation is extremely = I KT sin2φ
small.
Similarly, the torque produced by the other phase is:
(Torque) T2 = I KT cos2φ
The total torque is:
T1 + T2 = I KT (sin2φ + cos2φ)
but: sin2φ + cos2φ = 1 for any value of φ
therefore: T1 + T2 = IKT
So for sinusoidal phase currents with a constant amplitude,
the resultant torque is also constant and independent of
Current shaft position.
60° For this condition to remain true, the drive currents must
accurately follow a sine-cosine relationship. This can only
occur with a sufficiently high resolution in the encoder or
Fig. 3.18 Current profile in velocity-controlled servo resolver used for commutation.

The sine wave motor Direct drive motors

In the sine wave motor, no attempt is made to modify the In some situations, a conventional servo motor may lack
basic sinusoidal torque characteristic. Such a motor can be adequate torque or resolution to satisfy the needs of the
driven in the same way as an AC synchronous motor by application. A mechanical reduction system is then used to
applying sinusoidal currents to the motor windings. These increase torque and/or resolution. Direct drive systems
currents must have the appropriate phase displacement, couple the load directly to the motor without the use of belts
120° in the case of the three-phase motor. We now need a or gears. The Dynaserv motor provides very high torque
much higher resolution device to control the commutation if directly and solves many of the performance issues of the
we want smooth rotation at low speeds. The drive needs to gear reducer.
generate 3 currents that are in the correct relationship to Fig. 3.19 shows the construction of the Dynaserv direct
each other at every rotor position. So rather than the simple drive motor compared with a conventional motor and gear
commutation encoder generating a handful of switching reducer. The gear reducer introduces significant frictional
points, we now need a resolver or high-resolution optical losses between the motor and the load, and accuracy may
encoder. In this way it’s possible to maintain a 90° torque be sacrificed due to mechanical errors and backlash. The

17
Automation
Servo motors

direct drive motor is brushless and gearless so it eliminates

friction in the transmission. Since the feedback element is Rotor Φm
coupled directly to the load, system accuracy and Φm
Φc
repeatability are maintained and backlash is eliminated.

Conventional Motor Direct Drive Motor

Excitation
Gear DC/AC Rotating Coil T
Reducer Motor Bearing
Element Permanent
Encoder PCB Magnet
Stator A
Slit Plate Stator B
Stator
Element Fig. 3.21 Dynaserv magnetic circuit

Direct drive motor advantages

High precision
Direct drive motors eliminate the backlash or hysteresis
Encoder Rotor Core Stator Core inevitable in using any speed reducer. Absolute
positioning of 30 arc-sec is typical with a repeatability of ±2
arc-sec.
Fig. 3.19 Conventional vs. direct drive comparison
Faster settling time
The motor contains precision bearings, magnetic
components and an integral feedback device within a Due to the gearless design, direct drive improves machine
compact package (see Fig. 3.20). The motor is an outer- throughput by decreasing settling times.
rotor type, providing direct motion of the outer housing High torque at high speed
which is attached to the load. The cross roller bearings that
support the rotor have very high stiffness to allow the motor The torque/speed curve of the direct drive servo is very
to support the load directly. In most cases it is unnecessary flat. This results in high acceleration at speeds which
to use additional bearings or connecting shafts. would be difficult to achieve with a conventional rotary
table (up to 4 revs/sec).
Smooth rotation
Encoder Plate PDA Kit
LED Kit The direct drive servo has very low velocity and torque
Clamp Ring ripple, contributing to its excellent controllability over a
Housing Kit speed range in excess of 1000:1.
Retaining Clean operation
Ring
Stator Core Encoder The brushless and gearless design results in maintenance-
Rotor Core free operation. With suitable preparation, the Dynaserv
Housing can operate in class 10 environments.
Hub

Speed Ripple Torque Ripple

(DM1150A) (DM1015A)
Core 20

Conditions
• Load 30 x Rotor Inertia
• Rotation: CW
15 • Speed Mode
15.3

Fig. 3.20 Exploded view of Dynaserv 15

Ripple (%)

14.7

The torque is proportional to the square of the sum of the 10 5%

Torque (N • m)

magnetic flux of the permanent-magnet rotor (Øm) and the 5

magnetic flux of the stator windings (Øc) - see Fig. 3.21. 5

High torque is generated due to the following factors: 3

1. The motor diameter is large. The tangential forces

0
between rotor and stator act at a large radius, resulting in 0.2 0.4 0.6 0.8 1.0 1.2 0° 90° 180° 270° 360°

high torque. Revolution (rps) Rotational Angle (degrees)

2. A large number of rotor and stator teeth means that the

motor has a high pole count, in other words many magnetic
cycles per revolution. This high pole count results in high Fig. 3.22 Dynaserv velocity/torque ripple
torque output.

18
Automation