Technical References

System Elements

Overview The focus is on the most common type of control system, digital
The motion controller is part of a complete closed-loop system position systems controlling DC motors utilizing incremental
which also includes the power amplifier, motor, and an encoder. encoders. However, most of the discussion also applies to systems
Such systems, also known as servo systems, sense the motor posi- utilizing AC or hydraulic motors with resolvers or absolute
tion and feed the position signal back to the controller. A typical encoders, etc.
servo system is shown in the figure below.
DC Motors
Elements of a Servo System
DC motors convert electrical energy to mechanical energy, or more
HOST specifically, convert current into rotational [Link] key parame-
COMPUTER
ters of a DC motor are the torque constant Kt, the armature resist-
POSITION ance r, the moment of inertia Jm, and the maximum torque level.
COMMAND
MOTION The torque constant is expressed in units of Nm/A or oz-in/A and
COMMAND CURRENT
MOTION
CONTROLLER
POWER
AMPLIFIER
MOTOR indicates the amount of torque that the motor generates for a unit
of current. For example, a DC motor with a torque constant of 0.1
POSITION
FEEDBACK
Nm/A converts a current of 2 A to a torque of 0.2 [Link] armature
POSITION
SENSOR resistance is the total resistance of the armature winding and the
POSITION
brushes; it is expressed in [Link] moment of inertia Jm is the
sum of the moments of inertia of the rotating parts of the motor; it
The operation of the various system elements may be is expressed in units of kg-m2 or oz-in-s2.
described by comparing the servo system with the human body. A motor is also characterized by the level of torque it can pro-
The combination of the motor and power amplifier is analogous [Link] motor torque capability is expressed by two parameters:
to the muscle that moves the human [Link] motor is the device the continuous value and the peak [Link] continuous torque is
that generates the motion; the power amplifier generates the what the motor can produce continuously,often at any speed,with-
current required to drive the motor. For example, the amplifier out [Link] peak torque is the maximum that can be gener-
takes a low-current signal and amplifies it to a higher level of ated for short periods of time without causing mechanical damage
current of 10 A. or [Link] peak torque is several times the continuous
The controller is the intelligent element that commands the [Link] motor can generate any level of torque below the peak
motion. As such, it operates as the brain of the system. It gener- torque so long as the root-mean-square (RMS) value of the torque is
ates a signal, referred to as the motion command, which is within the continuous torque level.
applied to the power [Link] function of the position sen- The current that drives the motor is generated by a power ampli-
sor is analogous to human eyes. It senses the position of the [Link] operation of such amplifiers is described in the next section.
motor and reports the result to the controller, i.e. closes the loop.
A closed-loop system receives its command from an outside
source, often a host computer. Continuing our human society
analogy, the command source may be seen as the Boss, generat-
ing commands and often requesting status [Link] com-
mands can also be generated by other sources such as a pro-
grammable controller, terminal, or a set of switches.
The following discussion briefly describes the system elements.

[Link] / Galil Motion Control, Inc. 1

 Technical References
System Elements

Power Amplifiers Now that we have discussed the operation of motors and
Power amplifiers receive a command signal, typically an analog amplifiers, we will proceed to describe the operation of the posi-
signal within the range of ±10 V, and amplify it to the required tion [Link] most common sensor is the incremental
level of current. Such amplifiers can be configured in the current encoder.
or velocity [Link] velocity mode is preferred when velocity
feedback is utilized; otherwise, a power amplifier typically is con- Incremental Encoders
figured in the current mode. Incremental encoders generate pulses that represent the shaft
In the current mode, the amplifier produces a current that is [Link] encoder output includes two signals, commonly
directly proportional to the input [Link] is achieved by a called Channel A and Channel B, which generate N pulses per revo-
current feedback loop which monitors the current and assures [Link] two signals are shifted by a quarter of a cycle, as shown
that it is proportional to the command signal. Current amplifiers in the accompanying [Link] shift between the two signals
are characterized by the current gain Ka which indicates the enables the controller to determine the direction of rotation,
amplification per 1 V of command signal. depending upon whether Channel A leads Channel B or vice-versa.
Amplifiers can also be configured in the velocity mode. Here The shift of the two signals also increases the sensor resolution by
the amplifier includes a voltage amplification stage which com- dividing each encoder cycle into four quarters, or quadrature
pares the applied voltage with the motor velocity and amplifies [Link], an encoder with N cycles per revolution produces 4N
the difference before it is applied to the current loop. Such an quadrature counts per revolution.
amplifier is shown in the accompanying figure. Most encoders produce square wave signals with TTL levels.
Other forms include sinusoidal signals or square waves at higher
Elements of an Amplifier in Velocity Mode
voltages. Industrial systems often use encoders with differential
VELOCITY POWER
AMPLIFIER AMPLIFIER signals, i.e. channels A and B with their [Link]
VELOCITY CURRENT
COMMAND COMMAND devices reduce system sensitivity to noise. Incremental encoders
∑ + ∑
+ may also produce a third signal known as the index or marker
- - MOTOR [Link] signal appears once per revolution and can be used for
CURRENT
FEEDBACK
M initialization purposes.
VELOCITY
FEEDBACK
Output Signals of an Incremental Encoder

T CHANNEL A
TACHOMETER

CHANNEL B

Amplifiers use one of two methods to generate the required

voltage or current: linear amplifiers that produce constant output POSITION
voltage or pulse-width-modulated (PWM) amplifiers that gener-
ate a voltage switching between the high and low levels. Most
amplifiers today, especially those with power ratings above 100
watts, are switching amplifiers using the PWM method to mini-
mize power losses. Linear amplifiers are more common when low
power is required.

2 [Link] / Galil Motion Control, Inc.

Technical References
System Elements

Motion Controllers The most common type of stabilizing filter is proportional-inte-

The motion controller performs the intelligent tasks of the system as gral-derivative or [Link] proportional term is for speed of response.
shown in the block diagram [Link] basic tasks of the motion The derivative term provides stability and [Link] integral
controller include: term is for system [Link] tuning a servo system for opti-
mum performance involves adjusting the proportional,integral,and
■ Decoding position feedback
derivative coefficients or KP,KI,and KD.
■ Generating the desired position (profiling) The motion controller also performs “profiling” functions, gen-
■ Closing the position loop erating a time-dependent position function R(t), which corre-
■ Stability compensation sponds to the required velocity profile. In the velocity profile
shown below, the motion time of 150 msec is divided equally
The most fundamental function of the controller is to decode the between the acceleration,slew,and [Link] slew velocity is
motor position and to close the [Link] motor position C is deter- 100,000 count/sec and the total displacement is 10,000 counts.
mined from the feedback signal (often an incremental encoder) and Typically, the motion requirements are specified by the host in
compared with the desired or reference position [Link] difference of terms of the total distance, slew speed, and acceleration. It
R-C is known as E,the position error. remains for the controller to generate the position profile R(t)
The objective of the controller is to reduce the value of the posi- shown as a time-dependent function. Since the motor position C
tion error E to a minimum without causing system [Link] follows R, the generation of the profile R controls the motion path
achieve this,the controller often includes a stabilizing filter whose and rate. In addition to the basic tasks described above, an
output is applied via the digital-to-analog converter (DAC) to the advanced motion controller may perform high-level functions such
amplifier and the motor. as processing commands from a host computer, program sequenc-

Elements of a Motion Controller Desired Velocity Profile (a) and Corresponding Position (b)
MOTION CONTROLLER VELOCITY
(counts/sec)
100000 (a)

COMMUNICATION

MOTION R E X
PROFILE
∑ FILTER DAC 0 50 100 150 TIME (ms)
+
-
POSITION
C AMP (counts)
(b)
POSITION 10000
DECODER

7500

E M
ENCODER MOTOR 5000

2500

0
0 50 100 150 TIME (ms)

[Link] / Galil Motion Control, Inc. 3

 Technical References
System Elements

ing, I/O processing, and error handling. It is these high-level func- The motor may approach the required position in several [Link]
tions that allow the controller to operate as a complete stand- the motor approaches the required position from one direction and
alone machine controller. stops,we say that the system response is [Link] the motor
To describe the operation of the system,consider an example position overshoots the target several times before settling,the
with the following parameters: response is [Link] some conditions the motor
PARAMETERS DEFINITION response may oscillate and never stop; such a system is called unsta-
[Link] system response is a function of the system hardware and
Kt = 0.1 Nm/A Motor torque constant
PID filter [Link] ideal system response has minimum rise-
Jm = 10 -4 kg m2 Moment of inertia
time,minimum overshoot,and quick settling to the commanded
r = 2Ω Motor resistance [Link] tools,such as Galil’s WSDK servo design kit software,
N = 500 lines/rev Encoder line density are available to tune the PID filter for the optimum response.
Ka = 3 amps/V Amplifier current gain
Suppose the system is required to be at position zero (R=0)
but the actual motor position is -20 (C = -20).The controller deter-
mines the position error according to the equation E = R - C,result-
ing in an error of 20 [Link] error signal E is then processed by
the [Link] the combined gain of the filter and the DAC is 0.1 V
per count:this implies that an error signal of 20 counts produces a
motion command signal X of 2 V.
The signal X is applied to the amplifier with the given gain of
3 amps/V, resulting in a current of 6 [Link] the current is
applied to the motor, it produces a torque Tg which equals the prod-
uct of the current and the torque constant [Link] this case,the posi-
tive torque of .6 Nm is generated which drives the motor forward,
reducing the position [Link] generated torque accelerates the
motor at the rate a where:
a = Tg / Jm
for the given system parameters,a = 6000 rad/sec2.
As the motor moves closer to the desired position,the error E
decreases and,along with it,the drive signal to the [Link] the
motor reaches the desired position exactly,the position error drops
to zero,as does the motion command [Link] current is applied to
the motor while it remains in the commanded [Link],
whenever the motor moves away,a correction torque is applied to
force it back to the required position.

4 [Link] / Galil Motion Control, Inc.