Selecting programmable-logic controllers.
PLCs were first introduced to industrial control in the early 1970s as replacements for electromechanical control relays that were slow, hardwired, at times unreliable, inflexible, and of limited utility.
As acceptance of PLCs grew, so did the demands for increased function, more memory, and larger input/output (I/O) capacities. Manufacturers responded by introducing models of PLCs covering small (50 to 150 relays), medium (150 to 500 relays), and large (500 to 3000 relays) applications.
Computer capability was also added to the PLC so it could be applied to more complex applications. This was accomplished by expanding the basic programming language to include the power of advanced functions.
Relay ladder-diagram programming was especially well-suited to solving simple functions such as relays, latches, timers and counters. But advanced function programming, which uses alphanumeric functions in combination with input and output addresses, allowed the operator to go beyond the limitations of ladder diagraming and use many logical and arithmetic functions otherwise not available.
Advanced function language permits writing programs such as Proportional, Integral, and Derivative Control (PID), machine diagnostics, and sequencing. This expanded memory made the PLC an ideal electronic administrator for material-handling and tracking systems, parts programs and recipe storage, and extended the PLC's control capability in both process and discrete parts manufacturing applications.
Advanced function power added new dimensions to programmable control. For example, material-handling applications require MOVE functions, automotive users insist on MATRIX for machine-fault diagnostics, and the primary metals industry requires MATRIX, MOVE, and PID.
Processes such as injection molding, synthetic fuel production, soaking pits, and heat-treating ovens require MOVE for recipe storage. Sophisticated discrete parts manufacturing processes use MATRIX to drive many different sequencers.
Today's PLC is high technology. It is capable of controlling PID loops at remote locations, detecting and annunciating alarm limits on hundreds of analog values, doing position loops for controlling multiaxis robots and transfer lines, and acquiring and analyzing torque, pressure, or temperature values from test stands operating at high speed. How to tap that talent
PLCs can be employed to handle a multitude of control assignments. The challenge is to tailor the electronic package so it is cost-effective and contains the level of intelligence necessary to master the application.
The following evaluation should be made to implement a successful PC system:
1) Describe the process. Prepare a complete description of the functional requirements of the control system. This consists of block diagrams and words that relate the PLC and its I/O to the controlled process and to other controllers and computers (Figure 1).
The sequence of operations or machine states must also be determined because it contributes to the efficient functioning of the machine or process. The various states and conditions of the process have unique input/output states that relate directly to the process step.
Equally important in today's modern automation picture is the ability to monitor for faults in the physical input/output pilot devices. State-of-the-art PLCs are capable, with a minimum of user programming, of comparing the real-time input/output status to a previously "learned" correct sequence. Any miscompares can be highlighted for rapid diagnosis and repair to minimize process downtime.
Operating environment is another factor that should be considered when reviewing the process. Today's PLCs are capable of functioning in a high range of temperatures and humidities: generally 0 C to 160 C and 5 to 95 percent relative humidity (noncondensing). Vibration and foreign particle atmosphere are other environmental considerations that will determine mounting methods and enclosure needs.
Because of the broader scope that PLCs are being applied in today, the latest innovations are distributing the functionality and processing capabilities to different parts of the system. This is accomplished by parallel processing of data and I/O separate from the main Central Processing Unit (CPU) scan. This is done either in the input/output chain (smart I/O), or by extending the processing power through a data interface to a smart coprocessor.
2) Define number and type of inputs and outputs to the PLC. All I/O must be sorted by type (analog/discrete), function (input/output), and field signal level. Group the inputs according to function, because doing so reduces programming effort and significantly improves ease of documenting and understanding of programs.
Some of the smaller PLCs on the market today have dedicated or "fixed" I/O structures with perhaps 12 inputs and 8 outputs, all 115 VAC. These are on occasion cost-effective, but have the limitations of not being expandable to larger capacity I/O systems, or of not being adaptable to a variety of voltage levels.
A more cost-effective alternative is to install a modular system that allows the control engineer to tailor the input/output types to the task at hand. This building-block approach has a twofold payback. First, it can successfully satisfy current control needs. And second, it allows you to make modifications to adjust to changing process requirements.
3) Estimate logic memory requirements. PLCs are usually segmented into various sizes for different application needs. This relates to the different memory sizes and the ability to control different ranges of inputs and outputs. Sizes range from 1/4K words to greater than 96K words of memory, and I/O control ranges from 4 to 8000 inputs and outputs in one system.
Typical relay replacement applications have I/O in the ratio of 6:4, and the rules of thumb used to estimate logic memory requirements are: one control relay equals one I/O point, and one control relay represents eight contact references. Since one contact reference requires about one word of logic memory, it requires eight words and one I/O point per relay replaced.
Exceptions to this rule of thumb generally involve applications where other than relay replacement tasks are required. This might involve the ability to manipulate large quantities of data in logical or arithmetic functions as part of the recipe/parts program storage operation, or as part of PID loop solution. This would require the Pc to have added capacity of variable memory over and above the relay replacement requirements.
Special PLC functions such as intelligent ASCII interface, analog motion control, and machine diagnostics can be estimated separately by considering the availability and functionality of special-purpose "smart" I/O modules. These units allow specific tasks to be performed asynchronously to the main ladder logic solution in an optimum manner.
Examples of this are closed-loop, single-axis motion control, basic/ASCII programming interface, high-speed pulse counters, and intelligent analog and thermocouple inputs/outputs.
The primary benefit of using any of these "smart" I/O solutions is the opportunity to eliminate the external "black boxes" normally required to perform these special functions. For example, using a single-board motion-control module allows all of the intelligence associated with motion control to be integrated intelligently into the I/O structure of the PLc without encumbering the scan time for data processing.
Previously, this would have required a stand-alone motion control system to be connected to the PLC with numerous discrete I/O points individually wired. Integrating the motion control function into the I/O structure reduces both the cost and complexity of implementation.
4) Determine which functions are required. Applications requiring just relaying, timing, and counting can be handled with conventional relay logic. Advanced functions reduce scan time and enable the electronic brain to perform machine diagnostic, matrix, list, and table functions that are not readily implemented with basic functions.
Some PLCs can be programmed to "learn" the correct chain of events in a normal machine sequence. Then, during real-time operations, the actual input/output status can be compared to the learned sequence, and the miscompares can be loaded to a chronological list, including the sequence of fault occurrences.
This programming uses advanced functions to manipulate data and matrix operations. A source-to-table instruction is used to load the learned table with the actual machine sequence; a masked/compare instruction is used to compare actual to learned sequence, loading miscompares to a fault register while masking out any "don't care" situations.
An add-to-top-of-list instruction is used to create the chronological list of faults from the fault register, and can also key diagnostic messages to be displayed on a CRT.
All of these instructions can be written in traditional ladder logic, using contact power flow/coil conventions.
5) Determine response time requirements. The time it takes for the PLC to process an input signal and produce an output is known as "throughput time." With a scan time of 20 milliseconds, throughput time for an AC input is 60 milliseconds, which is sufficient for most industrial-control applications.
Larger and more complex systems can be expected to take longer in execution and result in higher scan rates. Total scan rates may exceed 150 milliseconds, and could provide unpredictable operating results in some machinery and process applications. In addition, remote (serial) I/O is also going to be limited to a lower maximum update time, so special consideration must be given here.
Interrupt inputs, however, can reduce throughput time to 1.2 milliseconds; this could be significant in applications where timing is critical. To ensure fastest response to interrupts, use functions with consistent maximum executing times, and keep subroutines brief.
It's important to know the scan and update speed required for your application. Most controllers have a maximum speed limit, and some can cause problems if not properly applied.
6) Estimate register or variable data requirements. The most important principle in the selection of registers is structure. It is easier to write a program with structured data, because data entry from thumbwheels and display of data on a seven-segment LED becomes easy with the use of source-to-source table and table-to-destination advance function. Also, programs are shorter and require less memory with structured data, because subroutines are easier to use.
In the final analysis, a control system is only as intelligent as the windows provided into it. This includes peripheral devices used for programming, real-time monitoring, and documenting the process.
In summary, today's PLC plays host to an ever-increasing array of intelligent peripheral devices including CRT-operator interfaces (color and monochromatic), printers for program documentation and report generation, and, of course, the programming device itself. Some industry trends are demonstrating the cost-effectiveness of combining one or more of these functions. A typical machine control application
A hypothetical machine makes five types of widgets. The requirement is to monitor production and provide the operator with control and feedback on the type of widget being produced (see Figure 3).
The application utilized 22 limit switches for detection of mechanical positions, plus six auxiliary contacts on motor contactors, to provide confirmation of the on/off status of these devices. A total of six contractors and nine solenoids must be controlled by the PLC, in addition to four electronic drives.
The operator's panel included a single-digit thumbwheel for type selection, one for parameter selection, and four for numerical entry. Operator push-button controls totaled five for start, stop, new data entry, increment display, and halt-after-next operation.
A similar quantity of devices was also required to feed back to the operator--LED displays and indicator lamps--the status of machine (type of widgets currently being produced, parameter being examined, go/stop state, halt, pending, etc).
To estimate the control hardware required, the I/O requirements are first examined. All similar devices (inputs or outputs of same voltages) are counted.
In this example, there are:
* 28 inputs (24 VDC) for limit switches and contactors
* 24 inputs for 6 digits of BCD thumbwheels (4 per digit)
* 5 inputs for push buttons
* 15 outputs (115 VAC) for contactors and solenoids
* 4 outputs (analog) for electronic drives
* 24 outputs for 6 digits of LED display (4 per digit)
* 5 outputs for indicators
The number of modules required is determined by dividing I/O counts of similar devices by the quantity of circuits available per module (rounding high).
Since 10 to 50 VDC and 5 VTTL inputs are available with the PLC on modules of 32 circuits each, one module can accommodate the 24 VDC inputs and another 5 VTTL. The driving outputs for contactors and solenoids (up to 2 A) are provided on modules of eight circuits, therefore two are required.
One analog output module provides the necessary four circuits to control the drives. Finally, one 32-circuit output module will supply more than sufficient outputs for the displays and indicators.
The size of the CPU is based upon the controlled discrete outputs, which in this example are the solenoids and contactors. A minimum of 10 words per output is required.
To this product (150 words), allowances for the drives (50 words per drive) and handling the information flow to the operator's panel (estimated 200 words) are added. This totals up to a memory requirement of 550 words.
Since functions beyond relay replacement such as numerical-data display and entry, control of drives, and analog outputs are anticipated, the power of the extended functions is also selected. The application calls for a total of six I/O modules, which can be accommodated without adding additional racks.
The complete bill of materials for this application is:
* One Model 60 CPU, with extended functions and 2K of CMOS memory.
* Two 5 VTTL/10 to 50 VDC input modules (32 circuits).
* Two 115-VAC output modules (8 circuits).
* One 5 VTTL output module (32 circuits).
* One -10 to +10 VDC analog output module (4 circuits).
|Printer friendly Cite/link Email Feedback|
|Author:||Johnson, David G.|
|Publication:||Tooling & Production|
|Date:||Jul 1, 1984|
|Previous Article:||Automation versus employment.|
|Next Article:||Built-in workhanding options.|