Volume 5, Issue 12, December – 2020 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Microcomputer Application in Motion Control

Muhtar H. Alhassan
Department of Computer Science
National Open University of Nigeria
Abuja

Abstract:- This paper briefly examines the microcomputer This paper focuses on the roles of the microcomputer
both as a control element and as a feedback-signal both as a control element and as a feedback-signal synthesizer
processor in modern closed-loop motion-control systems. in a closed-loop electric drive system.
Problems of interfacing, effective data-acquisition and
real-time control are briefly highlighted. II. INTERFACING

I. INTRODUCTION Interfacing is the means by which the microcomputer

bus can communicate with any peripheral device. It is the
The gradual replacement of analogue controllers for means by which a computer is connected to external
speed, position, torque, etc, with digital ones has facilitated equipment for data acquisition, monitoring or control. Signals
the evolution of the digital computer into one of the major coming from the external device must be transformed into a
components of modern motion control systems. Generally, form suitable for access by the computer. Similarly, any
the microcomputer when properly interfaced is suitable for signals generated by the computer for transmission to the
implementing several types of functions in a power electronic external device must be appropriately transformed before
based motion control system. These include: transmission to the equipment. The set-up is as shown in Fig.
1.
 Gate-firing control of power converters,
 PWM and step-wave signal generation, Thus, interfacing usually involves one or more of the
 Feedback control loops following:
 Signal Processing  Electrical buffering
 Sequencing control  Data transmission control
 Monitoring and warning  Code conversion
 Diagnostics  Interrupt handling
 Data conversion
 Timing
 Multiplexing

Fig 1:- Interfacing the computer to a control system

The IBM PCs and compatibles can only function as the computer’s expansion slots. A typical board, like the
effective tools in a data-acquisition / control context if they PC30A, includes 16 analogue input lines, 2 analogue output
are extended by the addition of appropriate expansion lines, 24 programmable I/O lines and usually a counter timer.
boards. Such boards are designed to plug directly into any of

IJISRT20DEC208 www.ijisrt.com 590

Volume 5, Issue 12, December – 2020 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Some non-IBM compatible machines, like the BBC In memory mapped I/O, the I/O lines are grouped in
microcomputer, come equipped with on-board A/D sets of 8, and each set is called a port and treated as a single
conversion capabilities. The 12-BIT A/D converter on the memory location. The ports are accessed using memory
BBC, which is usually the integrating type, can accept transfer instructions such as PEEK and POKE in BASIC, or
analogue inputs in the range between 0 and 1.8 volts. Here, LDA and STA in assembly language. Since memory
the upper range (1.8 volts) will produce the number 65520, addresses are used up for this type of I/O, communication is
which represents a 12-bit increment of 16. Thus in order to relatively slow.
obtain the normal 12-bit range of 0 to 4095, a division by 16
is necessary. Although there may be just one A/D converter In dedicated port addressed I/O, ports and channels
on-board a machine like the BBC, several analogue inputs are assigned unique addresses on a dedicated I/O bus. They
can be accommodated by multiplexing. Software access to are therefore accessed using additional software instructions
the A/D converter is provided by the BBC BASIC keyword like IN (INP) and OUT in low- or high-level programming
ADVAL (n) where n refers to the channel number (1, 2, 3 or languages.
4 in this case).
Direct memory access (DMA) is usually adopted when
III. SIGNAL CONDITIONING high sampling rates are required. Since non-DMA data-
acquisition techniques use program control for data transfer to
Signals generated from sensors in motion control memory, they have frequency limitations of about 100 kHz
systems are sometimes analogue in nature and generally must on the sampling rate [2]. DMA, on the other hand, is
be appropriately treated before they become suitable for hardware based, and with the help of tri-state buffers, causes
sampling by a computer. This is referred to as signal the microprocessor to momentarily be isolated from the
conditioning and can take any of the following forms: system buses so that the DMA device can directly access
 Voltage Amplification: This helps to match the full-scale memory, thereby increasing the data transfer rate up to a
voltage change of a sensor output to the full-scale voltage possible 1 MHz.
range of the input interface. Thus, a sensor might produce
a voltage from 0V to 1mV full-scale whereas the input V. COMPUTER CONTROL
interface may linearly convert voltages in the range of 0 to
1.023 V to a digital number between 0 and 1023. In such a Microcomputer /digital control of electric drive systems
case, it becomes necessary to amplify the input signal by a has several advantages over analogue types, notable among
factor of 1023 to utilise the full digital range.Units which are:
 Level Shifting: This refers to the alignment of the 1. Significant reduction in controller-hardware cost,
minimum sensor output voltage with the minimum voltage especially in systems using powerful VLSI
that can be converted by the input interface. For example, microcomputers with high speed and greater functional
if a sensor produces a voltage between –0.5 and 0.5 V, integration.
and the input interface converts a voltage between 0 to 1 2. Improved reliability leading to higher MTBF (mean
V to a digital number, a voltage of 0.5 V must be added to time-between-failures), due to the replacement of
the sensor output voltage to match the interface multiple electronic components by LSI and VLSI chips.
requirement. 3. Drift-free operation with minimal parameter variation.
 Frequency-Range Limiting: This is effectively the 4. Possibility of implementing sophisticated control
filtering out of unwanted high-frequency components functions involving complex computations as for
from the input signal. Filtering can also be used to block a example in variable speed ac drives.
dc voltage component using a high-pass filter, or by ac 5. Easy implementation of such functions as temperature
coupling using a simple capacitor in series with the input monitoring, over current warning, etc.
signal.
 Electrical Isolation: It is often necessary to electrically There are two notable limitations to the use of a
isolate signal generators from the computer to reduce the microcomputer in the control of an electric drive system:
effect of power-line voltage difference and noise that are 1. The introduction of quantisation error due to D/A and
usually generated when there is a common electric circuit A/D conversions. Increasing both the bit-size and the
connecting two physically separate systems. sampling rate can usually reduce this type of error.
2. A relatively sluggish response due to the serial way
IV. DATA ACQUISITION signals are processed in the microcomputer. Sluggishness
is most noticeable in centralised control of multi-drive
There are three main ways that the PC acquires data systems involving multitasking operations that tend to
from the outside world, namely. slow down conventional microprocessors. This situation
 Memory Mapped I/O has improved tremendously with the development of
 Dedicated Port Addressed I/O faster RISC type processors that can be used to produce
 Direct Memory Access microcontrollers with excellent dynamic responses.

IJISRT20DEC208 www.ijisrt.com 591

 Volume 5, Issue 12, December – 2020 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
VI. TYPES OF COMPUTER CONTROL VIII. THE PID CONTROLLER

The digital computer can play a variety of roles in The output equation for a PID controller is given by
motion control systems based on the strategy adopted. Thus
in Computer Supervisory Control, the controllers are de
essentially analogue with the computer accessing the drive v o  K p e  K i  edt  K d (1)
variables and generating driving signals for establishing the dt
desired set point.
where e is the error, vois the output and Kp, Ki and Kd
In Direct Digital Control, the computer directly are the proportional, integral and derivative coefficients of the
controls all the loops of the system. Thus the computer controller. The precise value of constants Kp, Ki and Kd
directly sets the driving power of each loop-actuator, loop by determine the quality of control of the system. Implementing
loop. A serious setback in this type of system is that the PID controller in a microcomputer requires the conversion
whenever the computer ‘goes down’ or ‘crashes’, all the of Equation 1 into a corresponding difference equation. First,
loops become open thereby making the system ‘runaway’. An both sides of the equation are differentiated with respect to
effective but rather expensive remedy is to have a stand-by time to obtain
computer acting as a monitor ready to take-over control when
things go wrong with the main machine.
(2)
Microprocessor control is a distributed type computer
control strategy, in which each loop of the system has its own Since computer implementation is discrete in nature, the
microprocessor or single-chip microcomputer. This method equation must now be transformed into the discrete mode
guarantees stability, self-tuning, remote re-programmability byreplacing continuous time with discrete time as follows:
and ease of implementation of complex algorithms for
compensating dead time delays and enhances the application
of feedforward control. (3)

In Integrated Computer Control, several Here, Ts is the cycle time, vois the change in output
microcomputer-based single-loop controllers are individually and e is error change. Thus
connected to a central computer referred to as a cell (4)
supervisor. The cell supervisor provides displays, keeps
records and issues commands to the individual loop (5)
controllers. In some cases, the supervising computer itself
reports to a higher-level management computer system.
The difference equation for the PID controller implies
Compatibility of physical, electrical and data structures in
that we compute the present value of the output based on the
such a system must be ensured by the adoption of standards
previous output, the present error, previous errors, the
such as the Manufacturing Automation Protocol (MAP)[3].
sampling time and the controller constants.
VII. IMPLEMENTING DIGITAL COMPENSATORS
IX. SOFTWARE CONSIDERATIONS
The block diagram of Fig.2 represents a
The PID equation can easily be implemented using a
microcomputer-based position control system. The
high-level language like C, Pascal, FORTRAN PL/M, and so
discontinuous nature of the power converter that usually feeds
on. Alternatively, an assembly language can be used for this
the motor is often neglected for the purpose of analysis. The
purpose. High-level implementation has the following
system is characterized by Multirate sampling, i.e. the various
advantages:
loops have different sampling times. Thus, the speed loop,
which usually has a large bandwidth, is characterized by a  Possibility of using floating point arithmetic
smaller sampling time than that of the outer position loop.  Automatic handling of negative numbers
 Possibility of utilizing a math co-processor
In the position loop, the microcomputer samples the
command position , feedback position and updates the Assembly language programming would require the
separation of the PID equation into several smaller steps that
speed command * through the PI position controller all at are definable in the assembly language of the host processor.
a sampling rate of 1 . In the speed loop, the
T s1
Computer implementation of PID controllers, whether
software- or hardware-based, can lead to unstable response
microcomputer samples the speed command *, the for some specific combinations of the proportional gain Kp
feedback speed , and generates the q-axis current and the sampling time Ts especially in second-order systems.
Another setback is the fact, that derivative computations in a
I 1
*
command at a sampling rate of
sq
T s2
sampled-data system can produce unpredictable output jerks
due to the introduction of DAC related delays. This

IJISRT20DEC208 www.ijisrt.com 592

Volume 5, Issue 12, December – 2020 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
phenomenon makes the use of simple derivative controllers signals are not directly measurable, they have to be computed
usually undesirable. More sophisticated derivative from any that have been measured. This is an important area
computations are usually resorted to, if stability is to be of application for the microcomputer in feedback control. For
achieved. Alternatively, derivative action can be entirely example, in a PWM-inverter based AC drive system, it is
dropped so that a PI controller results. possible to synthesise such feedback signals as flux, torque,
and the like, from measured voltages and currents.
X. THE PI CONTROLLER
Another example of feedback signal synthesis is the
The transfer function of a PI controller has the form digital measurement of rotor speed from incremental position
signals generated by a simple incremental encoder. The pulse
C( s ) Ki train from the encoder is accumulated in a counter, and the
= Kp + (6) speed in rpm computed as
R( s ) s
(11)
In finite difference form, the transfer function becomes
where Ts is the sampling time in seconds, m is the
C( k + 1) − C( k ) R( k +1) − R( k ) number of encoder pulses in one sampling interval, and M is
= KiR( k ) + Kp[ ](7) the number of encoder pulses per revolution. The larger the
Ts Ts sampling period, the more accurate the speed measurement.
Furthermore, the pulses per revolution should be as high as
The smaller is Ts , the sampling time, the more accurate possible for a given sampling period. The sampling period is
is the representation of equation 6 by equation 7. Further limited by the bandwidth of the speed-control loop, whereas
manipulation of equation 8 yields the pulses per revolution are speed-sensor dependent. Thus in
the final analysis, a compromise has to be made between the
C( k + 1) = C( k ) + KpR( k + 1) + ( KiTs − Kp ) R( k )(8) accuracy with which the speed is estimated, and the
bandwidth of the speed loop. It should be noted, that at lower
For computer implementation, the difference equation is speeds, due to the fall in frequency of the encoder pulse train,
transformed into state variable form as follows: the method of pulse-interval measurement is usually adopted.
In this case the speed in rpm is obtained as
X ( k + 1) = AX ( k ) + BR( k ) (9)
(12)
C( k ) = MX ( k ) + NR( k ) (10)
where fcis the clock frequency in Hz, M is the encoder
Here X(k) is the state variable, A = 1, B = KiTs, M = 1 pulses per revolution, and n is the number of clock pulses in
and N = Kp. These equations can be easily implemented the interval Tw of the encoder pulse.
digitally. The gain coefficients are adjustable software
parameters that can be tuned experimentally using, say, the Tests and Diagnostics
Ziegler-Nicol’s method. Computer based tests and diagnostics of motor drive
systems are easily and quickly carried out either on-line or
XI. FEEDBACK SIGNAL SYNTHESIS off-line. In the first case, the tests are carried when the system
is in operation, with the microcomputer acquiring data and
Most modern strategies applied to the control of servo executing the necessary application in real time. In the second
drives rely on complete state feedback. Unfortunately, case, data obtained from the system while operating is
however, it is often impossible or very costly to monitor some analyzed separately on the microcomputer.
of the state variables. In such cases when some state feedback

Fig 2:- Multirate position control

IJISRT20DEC208 www.ijisrt.com 593

Volume 5, Issue 12, December – 2020 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
The level of skill needed to carry out even the most
complicated tests is considerably lowered. Furthermore,
because the microcomputer-based procedure is methodical,
there is little chance of faulty diagnosis. Also, the need for
sophisticated purpose-built test equipment is eliminated and
safety problems are minimal during the test.

Another advantage here is the possibility of making the

test highly automated without any need for user intervention.

XII. CONCLUSION

Some of the functions that can be implemented on a

digital computer in a closed loop motion control system have
been briefly described. The need for proper interfacing of the
microcomputer to the target control system has been
highlighted.

REFERENCES

[1]. BOSE, B.K.: ‘Technology Trends in Microcomputer

Control of Electrical Machines’,
IEEE Transactions on Industrial Electronics, 1989,35,
(1), pp 160 - 177.
[2]. BRICKWEDDE, A.: ‘Microprocessor-based adaptive
speed and position control for electrical drives’, IEEE
Transaction on Industry Applications, 1985, IA-21,
pp1154 - 1161.
[3]. FRASER, C.J and MILNE, J.S.: ‘Microcomputer
Application in Measurement Systems’,
Macmillan Education Ltd, Hounmills, 1990.

IJISRT20DEC208 www.ijisrt.com 594