controls and sensors - uvajesman/bigseti/ftp/actuadores/...1998 pt design a159t ogether, controls...

1998 PT Design A159

T ogether, controlsand sensors regu-late the transmis-

sion of power. Throughswitching devices, powercontrols regulate the energythat motors use to turnshafts, heaters use to heat,etc. Through sensors, pro-cess controls fine tune oradjust performance param-eters to regulate machine and processoperations and are part of a systemthat is either open or closed loop.

Open loop systemsThe main characteristic of an open

loop system is that there is no feed-back of a process control variable. Ma-chines and processes that work wellwith open loop control include thosethat:

•Have predictable behavior andare self-regulating through the use oftimers, cams, or switches.

•Are insensitive to outside distur-bances.

•Do not cause harm if the desiredresult deviates widely from command.

•May be simple, slow in operation,do not affect other process sections, orare under the supervision of an operator.

Step motors, Figure 1A, are com-monly used in open loop control sys-tems. An indexer receives a commandsignal from an operator, pro-grammable controller, or computer.The indexer then provides the driverwith acceleration, deceleration, veloc-ity, and position information. Signalsto the driver continue unaltered untilthe indexer’s program is changed.

push-button or a computerkeyboard command. Push-buttons come in many con-figurations, from hard con-tact systems that operateby closing one contactagainst another, to unitswith Hall-effect devices andsaturable ferrite cores.

SwitchesPushbuttons

a r e a v a i l a b l ewith normal lyopen and nor-mally closed con-tacts, and with acombination of

both. Some selector switches, for in-stance, let the operator select a partic-ular operation or sequence of opera-tions with over 30 combinations.

For digitally controlled units,thumb-wheel switches and operatorkeyboards are available, which let op-erating personnel give specific digitalcommands to the control system.

Other operator controls include po-tentiometers that provide an analogoutput voltage, determined by the po-sition of a contact wiper attached tothe shaft. These units come in single-turn units that operate over a 270 degarc, to multiple-turn units that re-quire 100 turns to go from zero to fullvoltage.

DisplaysTwo basic types of displays are

used to keep operating personnel in-formed of machine conditions: analogand digital.

Analog — These devices are fre-quently voltage or current sensingunits that consist of a dial or metercalibrated in the desired units of mea-sure (volts, amperes, rpm, etc.). Theymay offer a charting or graphing func-tion, as well as a visual display. Com-plex models also incorporate a control-ling function that converts a specificlow level signal — such as thermocou-ple input — to a 4 to 20-mA output sig-

Closed loop systems

Closed loop systems, Figure 1B, usefeedback from the work process tocontrol an operation. Servosystemsare typical of closed loop control. Thedigital controller in this system com-pares position feedback from an en-coder to a programmed motion profile.The error between the two signals isused to alter the velocity signal.

OPERATOR CONTROLS ANDDISPLAYS

With the increased use of automaticdevices and theneed for personalsafety, an almostinfinite number ofoperator’s controlsand displays areavailable.

In most cases, before any operationstarts, the operator must initiate astart command — usually through a

CONTROLS AND SENSORSOPEN-LOOP SYSTEMS ...........................................A159CLOSED-LOOP SYSTEMS .......................................A159LOGIC AND PROCESS CONTROLS..........................A160SENSORS AND TRANSDUCERS..............................A165POWER CONTROLS ................................................A171MOTOR CONTROLS ................................................A172MOTION CONTROL SOFTWARE..............................A173CONTROLS AND SENSORS ADVERTISING.............A175

Step and direction signals

Controller Drive Motor

Figure 1A — An open-loop system doesnot use a tachometer, encoder, or resolverfor feedback.

Figure 1B — Closed-loop systems usefeedback for velocity and position control.A traditional servo system may use atachometer for velocity control and anencoder for position information. Currentsystems typically use an encoder toprovide velocity and position feedback.Some systems use a resolver in place of an encoder.

Controller Drive

10V signal or digital signal Encoder

Servo motor

Position and velocity feedback

_ __

nal for controlling process systems.Digital — Built with either LEDs,

or more recently LCDs, these displaysprovide digital or alphanumeric indi-cations. Some units with microproces-sors combine visual display with con-trol functions and are programmed inmuch the same manner as a pro-grammable controller. These displaysare usable as stand-alone controllersfor simple operations, or to augment aprogrammable controller function.

Cathode ray tubes — Common totelevision and computer terminal use,CRTs are rapidly expanding into theindustrial environment and are fre-quently included with equipment con-taining microprocessors. The systemsprovide a graphic display of machineoperating parameters.

Liquid crystal displays — A re-cent development, Figure 2, the mainbenefit is a depth of only a few inches,particularly useful for applicationswhere space is a premium. They arelight weight and not subject to con-ducted EMI and radio interference.Monochrome gas plasma and electro-luminescent (EL) displays are themost common.

LOGIC AND PROCESSCONTROLS

Process controls make shafts turnsmarter. They use reference signalsprovided by sensors to adjust or regu-

A164 1997 Power Transmission Design

Figure 2 — Liquid crystal displays are arecent development. With a depth of onlya few inches, they are particularly usefulfor applications where space is apremium, and they are light weight andnot subject to conducted EMI and radiointerference.

Always a PLC?PLCs aren’t the only solution for motion control, particularly those appli-

cations that need few I/O points. The industrial personal computer (PC) isoften a competitor with the PLC.

When choosing between a PC and PLC, keep in mind the benefits and lim-itations of each.

Despite the limitations, PC benefits have made it a strong competitor inmany motion applications. To compete, PLC vendors often embed single-board PCs in their products. Software is loaded on these PC boards throughmodules that plug into a PLC’s VME backplane. This gives users a productthat combines PC flexibility in a PLC environment, and eliminates the needto integrate one to the other. Information courtesy of Aromat Corp.

For the PC, features and limita-tions include:

• An open architecture. Add-inboards let engineers add hardware fea-tures such as sensor interfaces, specialmath capabilities, sound, and others.

• Communications. Ability to net-work with other PCs, PLCs, and othercomputer products.

• Large availability of software.Programs exist for analysis, program-ming, business functions, and somecontrol functions. Also, large avail-ability of programming languages.

• Flexibility. Users can alter PCfunctions in software or add-in hard-ware boards.

• Easy-to-use operator interface.The PC set the standard for windows-based, pull-down menus, and icon-graphics style operator interface.

• Single-tasking operating system.DOS and Windows are not real-time oper-ating systems. They are oriented towardfile handling and operator interface.

• Does not fail in a predictablemanner. When a PC fails, it “crashes”regardless of the step in the controlprocess. The PC does not leave cluesto possible causes of failure and thus,is not easily, or quickly, diagnosedand repaired.

• I/O limitations. PCs lack I/Opoints and are limited in their abilityto operate with a large number of I/O.Some motion-control applications,such as driving labeling machines,may require at least 64 I/O points,which are not readily available on anindustrial PC. The cost of I/O for in-dustrial PCs can be high. An indus-trial PC, excluding software, can costup to $1,000/point.

For the PLC, features and lim-itations include:

• Semi-open architecture. Open-ness in hardware and software isgrowing in PLCs, although it is notyet the same as PC openness. MostPLCs, though — even micro and nanoversions — can interface with wide-bit, high-speed, single-board comput-ers that use standard software andgraphical user interfaces. Motor pa-rameters can be set up on a controlpanel, giving an operator the abilityto adjust parameters “on-the-fly.”

• Communications. Ability to net-work to many manufacturing opera-tion and control devices from sensorsto motors to mainframe computers.

• Limited software. Tradition-ally, only ladder logic was available.The IEC-1131 standard, however,increases the number of program-ming languages to five. Software isgeared to controlling a processrather than displaying status.

• Limited flexibility. PLCs areoptimized for factory control.

• Operator interface is usually a PC.• Multitasking and interrupt

handling capabilities. • Predictable failure. PLCs shut

down in a logical manner that safe-guards the process. Specific memoryregisters record the last functions exe-cuted for easier diagnosis of why thesystem shut down. This feature makesit easy to bring a PLC back on line.

• I/O. A small 14 I/O PLC, can bebought for about $16/point. Theprice advantage between PCs andPLCs changes with mid-size PLCs.A complete mid-size PLC can cost$4,000 to $5,000, while a 486 indus-trial PC can cost less than $2,000.

1997 Power Transmission Design A165

late machine operations. Control-level signals are low-voltage, low-cur-rent signals often measured in mil-liamperes. Logic and process controlsrange from simple relays that controlone function to sophisticated pro-grammable logic controllers (PLCs)and personal computers (PCs) thatcontrol and monitor every facet of amanufacturing process. The abbrevia-tion “PLC” is often used for pro-grammable (logic) controllers to dif-ferentiate from “PC” for personalcomputers.

RelaysRelays are used to control other de-

vices in the same or another circuit,and are either electromechanical orsolid-state units.

Magnetic relays — Many types ofrelays exist, each having unique me-chanical and electrical construction.The most common magnetically oper-ated types are covered here.

Reed relays — are basically an as-sembly of reed switches within an op-erating coil. The reeds can be any typeor configuration, but the quantity islimited by the coil size. Most manu-facturers limit the size of the coil tohandle a maximum of 12 switches. Toobtain additional contacts, relays areconnected in parallel.

Mercury-wetted contact relays —consist of one or more glass switchcapsules surrounded by a coil. Whentwo contacts wetted with mercury areseparated, the mercury stretches andthen breaks at two points, leaving athin rod of mercury in the middle thatdrops to the bottom of the switch. Theloss of mercury from the contacts dis-turbs the system equilibrium, andmore mercury is fed up the armaturefrom the pool. Thus, in effect, the re-lay provides a new contact surface forevery operation.

Armature relays — are a large classof relays with armatures that carry oractuate electrical contacts in responseto small control signals.

AC relays are the most readilyavailable, but ac is the least flexiblepower source for relay operations.However, most ac relays designed for120 V line operation tolerate line fluc-tuations from 102 to 132 V.

DC relays have inherently greatermechanical life expectancy than ac re-lays. Of the many sources of dc, themost frequent is probably rectified ac.

be located near the equipment or pro-cesses they control.

A typical PLC consists of a centralprocessing unit (CPU), a pro-grammable memory, input/output(I/O) interfaces, and a power supply,Figure 3.

The CPU is the “brain” of the PLC,where all logic operations are per-formed. The CPU compares inputs toa program stored in memory and al-ters the output signals in accordancewith instructions in the program.

Often ac ripple influences relay oper-ation. Some dc relays can tolerate rip-ple; others need filtering.

Solid-state relays (SSRs) —These devices control load currentsthrough solid-state switches such astriacs, silicon-controlled rectifiers(SCRs), or power transistors. Theseswitch elements are controlled by in-put signals coupled to the switchingdevices through isolation circuitssuch as transformers and reed relays.Since the semiconductor switch candissipate significant amounts ofpower, solid-state relays are generallyheat-sinked to minimize the operat-ing temperature.

Solid-state relays are used in appli-cations where rapid on/off cyclingwould quickly wear out conventionalelectromechanical relays. General-pur-pose SSRs have on/off cycle lifetimes ashigh as 100,000 actuations. SSRs canbe actuated with conventional CMOSand TTL logic level voltages.

Transistor-transistor logic — (TTL)devices operate with 5-V power sup-plies, and recognize voltages between2 and 5 V as logical 1 level, and volt-ages between 0 and 0.8 V as logical 0level. When these devices are con-nected to form digital circuits, the twomost important characteristics arecalled fan-out and propagation delay.Fan-out is the maximum number oflogic gate inputs that can be fed fromany given logic gate output. Propaga-tion delay is the amount of time a gatetakes to react to a change in state atits input. This delay time is causedprimarily by capacitive delays inducedby components making up the gate.

Complementary metal-oxide semi-conductors — (CMOS) can operatefrom power supply voltages rangingfrom 3 to 15 V. These devices are gen-erally slower than equivalent TTLs,however, they dissipate much lesspower. Propagation delay times areset primarily by supply voltage; thehigher the supply voltage, the longerthe propagation delay. Since these de-vices dissipate so little power, fan-outis essentially unlimited.

Programmable controllersA programmable logic controller

(PLC) provides many of the advan-tages of computer control in indus-trial applications. PLCs are solid-state control devices that withstandtough factory environments and can

Digital signalprocessors

Motion control products previ-ously used ordinary microproces-sors to calculate and execute mo-tion algorithms. Today, most use a“screamingly fast” new micropro-cessor — the digital signal proces-sor (DSP). This computer-on-a-chip executes motion algorithmsup to ten times faster than previ-ous controls.

It consists of an arithmetic andlogic unit, address generation andmanagement units, and se-quencers to control the flow ofdata and instructions. It has twodistinct sets of memory — one fordata, one for instructions — andtwo busses. One bus grabs dataoperands while the other grabs in-structions. This ability to do twothings at once is part of what givesa DSP its speed.

All microprocessors execute in-structions in a measure of timeknown as a cycle. In a DSP, most in-structions take one cycle. Depend-ing on how multiplication is imple-mented in the motion algorithms,execution can take one or two cycles,including floating point numbers.

Microprocessors such as thePentium family or the Power PCchips may compete with DSPs inmotion control. Both types of mi-croprocessor chips are moving to adesign where they have one largememory outside and two memorychannels inside, similar to DSPs.But unlike these microprocessors,DSPs can directly process sensorinputs because they can operate inthe same microsecond time frame,and they have a software programthat users can access and alter.

Most CPUs perform mathematicaloperations on the input data or uselookup tables in order to alter the out-put signal. Other capabilities includeremote input/output, run-time diag-nostics, and high-speed communica-tion networks that link one PLC toother PLCs, supervisory computers,and sensing and monitoring devices.

During operation, the CPU contin-ually scans all I/O circuits, comparesinput to memory, performs logic oper-ations, and directs output. Scan timesrun from less than 1 to 200 millisec-onds, depending on the size of mem-ory and the number of I/Os that mustbe read.

Memory in a PLC is divided be-tween executive and user. The execu-tive memory is usually read onlymemory (ROM) and often pro-grammed by the PLC manufacturer.This memory is permanent. It cannotbe altered nor will it be lost if power isinterrupted or removed. Its program

outages by lithium batteries havingservice lives of up to several years.Non-volatile random access memory(NOVRAM) acts like RAM but storesdata like EEPROM. No battery back-up is required.

The amount of memory that a PLCcontains determines its size. Gener-ally, PLCs are available with 1K to 8or more Megabytes of memory. Manyare expandable so that users can in-crease memory size.

Input/Output (I/O) devices are ei-ther digital or analog types. Digitalinputs are typically supplied by on/offdevices such as switches, analog-to-digital converters, and counters. Ana-log signals are provided by process in-strumentation and transducers.Generally, PLCs handle digital sig-nals as high as 125 Vdc and 230 Vac.Analog signals are converted to digi-tal signals by input modules. Stan-dard analog inputs are 4-20 mA, 10-50 mA, 0-5 Vdc, and 0-10 Vac.

contains instructions and algorithmsfor its specific operation.

Programmable read only memory(PROM) is more versatile in that itcan be programmed (once) by theuser. Thereafter, it effectively be-comes ROM. Often, users purchaseseveral PROMs and program each fora different operation.

Erasable programmable read onlymemory (EPROM) permits data era-sure when it is exposed to an intensesource of ultraviolet light. Electricallyerasable programmable read onlymemory (EEPROM) is another non-volatile memory that can be erasedelectrically and reprogrammed. BothEPROM and EEPROM provide long-term program retention without bat-tery backup.

Random access memory (RAM) isuser memory that is easily accessedand altered. Should power be re-moved, the contents in RAM will belost. It can be protected from power


Figure 3 — Programmablecontroller based system.


I/O systems can be rack mounted orremote (also termed distributed) andcontain from 4 to 8,000 I/O points.Maximum remote mountings typi-cally run to 2 miles away from theCPU.

Power supplies provide all of thevoltages required for a PLC’s internaloperations. These supplies can bemounted directly inside the PLC orremotely mounted and connected tothe PLC by cable. The power supplyconverts 120 or 240 Vac line power toregulated dc power required by theCPU and I/O modules.

Programming devices range fromsmall hand-held keypads to personalor mainframe computers. In the lattercase, the PLC is connected to a largerPLC, a PC, or a mainframe computerthrough a data bus. Programs andlookup tables can be downloaded intothe PLC from the larger computer.Thus, one PLC can control hundredsof types of operations.

Programming languages for PLCsare moving towards true portability— enabling engineers to use variouslanguages on any PLC regardless ofmanufacturer. The IEC 1131 pro-gramming standard is responsible forthis increased level of portability.

IEC 1131 has been in developmentsince 1981. It consists of five parts:General information, specificationsfor equipment and test requirements,specifications for programming lan-guages, user guidelines, and commu-nications. The third part on program-ming languages, 1131-3, receivedinternational approval in 1993.

The standard defines two graphicallanguages — ladder diagram andfunction block diagram — and twotext-based languages — instructionlist and structured text. Each of theselanguages also uses an organizationlanguage commonly known as se-quential function chart. This lan-guage is similar to the European stan-dard graphical language Grafcet.Sequential function chart helps orga-nize the sequences of functions in aPLC program.

Ladder diagram logic is the oldestPLC programming language, and theone most electricians and mainte-nance people know. Instruction list isa low-level language, similar to an as-sembler language for computers. Forsimple applications, it executes onlyone operation per line. Structured listis a high-level language that offers

one to four bits. They send on/off,there/not there signals, or simplecounting data. They transmit at thefastest rate — usually less than 5msec/bit — because they don’t sendseveral bytes of format coding withthe data.

Byte-type networks send bytes ofdata. Data transmission speeds typi-cally range from 125 to 500 Kbaud. Afew can go higher. These buses typi-cally handle mid-level automationfunctions such as process or cell-levelcontrol. Byte-type buses includeCAN-based buses such as DeviceNetand SDS, and low-level versions ofProfibus and Interbus-S.

The packet buses include Ethernetand TCP/IP, which send largeamounts of data, called packets, tovarious computing systems.

Selection tips:• Select a bus that connects with

the sensors you use, or plan to use. • Be sure the bus meets your speed

and data throughput requirements. • Select a bus that communicates

over the distances you require. • If you have other buses or net-

works, be sure the topologies, i.e., aring configuration vs. a trunk-dropconfiguration, are compatible.

Personal computersPersonal computers (PCs) are fre-

quently found with programmablelogic controllers (PLCs) on the factoryfloor. They often work side-by-side tooptimize automation and motion con-trol solutions, generally with thePLCs controlling production pro-cesses, and the PC collecting andmanaging information. More re-cently, PCs are used in direct motion-control applications in addition todata acquisition and manipulation,data trending for production sched-ules and quality control, logic solving,handling of screen and operator func-tions, and interfacing with other de-vices such as printers and graphicsterminals. Through I/O modules andPLCs, they control and monitor in-dustrial devices such as motors andsensors. In addition, PCs are used tocreate ladder-diagrams off-line, thenprogram PLCs and debug programson-line.

Compared to PLCs, PCs generallyhave more memory and greater calcu-lation capabilities.

Programming languages for PCs in-

Boolean and arithmetic statements as well as statements usingIF...THEN...ELSE...and WHILE. Itcan be used to represent analog anddigital values. Function block lan-guage defines code that is common toseveral applications or repeated inseveral segments of a program, e.g.,PID loop control, counters, timers,and similar algorithms.

A programmer can use any one or allfive languages in one PLC program.The benefit: programmers will havethe freedom to program in the mostsuitable language for each function orapplication segment, eliminating manyof the limitations typical of using onlyone language for PLC programming.

This standard also forces vendorsto use a standard “look,” or commonoperator interface for each language,which will lessen the amount of train-ing required to use PLCs.

With this common interface, allladder programs will use the samegraphic objects for rungs, open andclosed symbols, the same names forfunctions, and so on. All Structuredlists will have the same Boolean andarithmetic statements and program-ming constructs. Once an engineerknows a language, he or she can go toany vendor’s PLC and program it.

Communications. In 1994, controlmanufacturers introduced a new typeof communications product for con-necting switches, sensors, small mo-tor starters, and other such devices toPLCs or personal computers (PCs).These buses cut wiring, cabling costs,and device installation time in half,while they increase uptime and relia-bility of automated systems. They of-fer an alternative to discrete, propri-etary I/O solutions and fieldbuses.

One four-wire cable connects theabove-mentioned devices to PLCs.

These buses take advantage of ad-vances in microprocessor technology,especially application specific inte-grated circuits (ASICs). A bus’s proto-col, or the rules governing data for-mat, transmission format, andnetwork housekeeping details, arecoded onto an ASIC chip or chips.These chips are small enough thatsensor and switch manufacturers canembed them in their 18-mm diamproducts.

Device-level buses exist in threeversions: bit, byte, and packet. Bit-type buses, such as Seriplex and ASI,transmit a few bits at a time, typically

clude conversational languages suchas Windows, BASIC, PASCAL, and C,plus various assembly languages.

Industrial controlsThe PLC and the PC have reduced

cycle times, improved productivity, andmet the information gathering needs ofproduction and manufacturing pro-cesses. But as each goal is reached, cus-tomer requirements change.

In 1996, the requests were for con-trols that offer open, modular archi-tectures that enable users to gain life-cycle economies. The goal of thisopen, modular industrial control is toenable users to:

• Run control software on anyhardware platform.

• Purchase peripheral equipmentthat will plug and play into any hard-ware platform.

• Reduce the amount of time ittakes to change a control’s capabilities

• Reduce the costs of retraining, in-stallation, and spare parts.

In 1994, engineers at General Mo-tors reviewed their manufacturingoperations and came up with strate-gies to cut their development andmanufacturing time. But to imple-ment these strategies, the engineersneeded a control that they could up-grade with minimal additional invest-ment. They developed a specificationfor this control, which ignited devel-opment of the open architecture, mod-ular, industrial control. Eventually,such a control would facilitate thesame rapid revolution of capabilitiesfor the shop floor that has occurred inoffice computing applications.

While GM is one of the primarydrivers for this open architecture con-cept, other drivers include relativelysmall firms:

• Searching for ways to meet cus-tomer needs of mass production whilesimultaneously manufacturing ahighly custom product.

• Searching for the best-of-breedamong computer-control components.Users want to choose from the bestI/O modules, the best peripheral in-terfaces and the best software pro-grams — not just those componentsthat work with a specific control.

• Demanding solutions that seam-lessly integrate control and informa-tion throughout a company.

• Distributing microprocessor con-trol and I/O to other devices.


on ofand I/O

Control and I/O migrating into other devices

nce

Increased importance on product packaging

Figure 4— Part of the reason open, modular controls are in demand now is becausecustomers are seeing the benefits of distributing microprocessor control and I/O to otherdevices and of connecting these components through innovative network strategies suchas device-level networks.

Deg

rees

of

au

tom

atio

n

Time1960 1970 1980 1990 2000

Relay-electromechanical control

Centralized (rack-based) control

Distributed processing & I/O

Distributed control

Fixed control

Programmable control

Networked manufacturing systems

Networked business & manufacturing systems

Hardware interfaces

I/O systems

Hardware platforms

Information & display systems

Programming software

Programming interfaceI/O interface

Information interface

Control system

Figure 5 — Several manufacturers are moving from a present control strategy ofdistributed processing to a new strategy of distributed control.

Figure 6 — These modular control components will be connected through applicationinterfaces (APIs), software or hardware that will send data back and forth yet it will be aseasy to install as a new program on an office PC.

The drive toward distributed control

Evolution of automation

System interfaces


• Developing innovative networkstrategies such as device-level net-works, Figure 4.

• Seeking to leverage the technologyshifts found in electronic component, soft-ware, and network technologies, Figure 5.

Because there is such a huge in-stalled control base, the opportunityfor open, modular systems revolvesaround application interfaces (APIs).These consist of software or hardwarethat will let logic engine, I/O, and hu-man-interface components send databack and forth, Figure 6.

Control manufacturers are investi-gating several approaches:

• A control that incorporates anoperator display with a look and feelcommon with other controls. Yet, op-erators can tailor the display to theirpreferences, eliminating retraining.It also uses a common set of core fea-tures and interfaces to enable load-and-run software. This software linksWindows-based programs to the con-trol’s real-time operating system. Anelectronic message system shuttlescommunications among the modulesand add-on software and hardware.

• A control with a PC compatiblesoftware-hardware system. Cost re-ductions will result from easier sys-tem design, less installation, and lessdowntime when compared to tradi-tional control systems.

• A control that consists of openmodules that can be packaged to-gether or sold as individual pieces.These modules offer communications,I/O drivers, a real-time control en-gine, human-machine interface, andinterfaces to other PCs, networks,and information systems.

SENSORS AND TRANSDUCERSSensors and transducers can be clas-

sified by the machine or process variabledetected (such as temperature), or bythe method used to detect a machine orprocess variable (such as photoelectric).

Temperature sensorsTemperature sensors range from

simple conventional thermometers tosophisticated solid-state devices.Some are used merely to sense andprovide a visual indication; others,combine sensing with control capabilities.

Mercury thermometers — Amercury in glass thermometer is a

Position sensors

Position sensors may be simple me-chanical devices like switches that re-quire physical contact with the objectwhose position they are sensing, orelectronic devices like photoelectricscanners that determine positionthrough the use of a light beam.

Limit switches — These devicestranslate mechanical position into anelectrical indication of position. Severaltypes are available and usually referredto by their physical configuration.

Push rollers are actuated by directthrust or the vertical component of alinear or rotary cam. Side forces cancause binding and should be mini-mized, Figure 7.

Rotary switches are one of the mostcommonly used due to the variety ofheads available, Figure 8.

Cam limit switches consist of one ormore rotating cams shaped to actuatelimit switches during part of thecam’s rotation in order to initiate aspecific action. Several switches canbe arranged in parallel to actuate asequence of actions, Figure 9. Bothmechanical and electronic pro-grammable types are available. Theelectronic type, Figure 24, can providemultiple functions with digital and

simple industrial sensor used only toindicate temperature. Mercury risesand falls within a capillary tube witha calibrated scale in response to tem-perature changes.

Bimetallic thermometers —These devices have a sensing elementthat consists of two strips of differentmetals bonded together. Each metalhas a different thermal expansion co-efficient. Consequently, the metalsexpand at different rates as tempera-ture rises, and bend in the direction ofthe metal that expands least. Somebimetallic sensors are formed into ahelix with one end attached to a shaft.A pointer is attached to the shaft and,as temperature changes, the pointermoves along a calibrated scale. Theshaft could be attached to a positionsensor (such as a potentiometer) toproduce an output signal that can beused for process control.

Thermocouples — Two wires,such as iron and constantan, are joinedto form a measuring junction and a ref-erence temperature junction. A ther-mocouple produces a voltage propor-tional to the temperature difference ofthe two junctions. This voltage differ-ence is only a few millivolts, but is am-plified for control applications.

Thermistors — A semiconductorthat provides high sensitivity andquick response to temperaturechanges. Thermistor resistance de-creases with temperature increases.Because a large change in resistanceoccurs per degree of temperaturechange, thermistors provide verysensitive indications of temperaturewhen used in a bridge circuit. Often,bridge output is fed to an amplifier,and then to a control circuit, such asa relay.

Radiation pyrometers — A non-contact sensor used to measure ex-tremely hot (700 to 3,500°C) tempera-tures. Because they do not have tocontact an object to measure its tem-perature, the object can be in a corro-sive atmosphere, in a vacuum, or mov-ing. The radiation pyrometer sensestemperature by measuring an object’sradiation of thermal energy. Opticalor brightness pyrometers compare theradiation from the surface of an objectwith the radiation emitted from a fila-ment of a calibrated lamp. The bright-ness of the object is matched to the ref-erence bulb. Then the reference bulbcurrent is measured and translated totemperature.

Figure 7 —Typical push rollerlimit switch.

Figure 8 — Typicalrotary limit switch.

Figure 9 — Typical rotating cam limitswitch. Cam switches are also availablein solid-state versions, Figure 19.

analog outputs. These electronic unitscan be reprogrammed to change func-tions without stopping the machine,and have diagnostic capabilities. (Formore on electronic programmablecams, see Smart Sensors later in thissection.)

Lead screw types use an actuatornut to trip stop contacts at each end oftravel, Figure 10. Intermediateswitches can be located between stopsfor position indication or circuitswitching.

Potentiometers — Available inrotary and linear configurations, po-tentiometers change the value of thevoltage output. By moving a slideacross a resistor material, outputvoltage changes and indicates themeasured object’s position. A typicaluse for potentiometers is in measur-ing fluid levels. A float is attached tothe potentiometer shaft and, as fluidlevel changes, voltage output variesproportionately.

Proximity sensors — Actuatedby the presence of objects withoutphysical contact, proximity sensorscan be used as limit switches, posi-tioning devices, speed controls, andcounting devices, Figure 11. Theirspeed generally varies from 5 to 1,000pulses per second. Different types ofsensors detect either metallic or non-metallic materials.

Inductive proximity sensors detectmetallic objects by means of changesin the characteristic of an inductiveelectrical circuit. These sensors gen-

target breaking a beam of light be-tween the scanner and a retroreflec-tor. This method is usually the leastexpensive of the five methods. Surfaceproperties and color of the object arenot critical, but it must be largeenough and sufficiently opaque tobreak the beam, Figure 13.

Through-beam is used to indicatethat an object broke a beam of lightbetween a sender and receiver, Figure

14. Restrictions are similar to retrore-flective types, and precise alignmentof sender and receiver is necessary.Through-beam sensors offer thelongest range (about 300 ft), but havethe highest cost.

Diffuse reflective/proximity typessense an object when the object com-

pletes a beam by reflect-ing any element of thebeam back to the sen-sor, Figure 15. Thismethod requires a dis-tinct difference in re-flectivity between theobject and backgroundarea that is in visiblerange. If this is not pos-sible, a retroreflective or

through-beam type should be used.Specular reflective/proximity types

are similar to diffuse types, but re-quire that a highly focused beam bereturned to the sensor. This type isusually used for sensing glossy andother highly reflective surfaces.

Color mark sensing types differen-tiate between colors and are used fordetecting registration marks, partsinspection, and color sorting. Extremesensitivity lets them sense ripe/un-

erally have a sensingrange of a few thou-sandths of an inch toseveral inches.

Other types of prox-imity sensors include

capacitive and sonic designs, whichsense nonmetallic materials.

Some of the advantages of proxim-ity sensors include high speed actua-tion, solid-state reliability, ability tosense motion fromany direction in anysequence, and a ca-pacity to sense smallparts without hinder-ing flow.

Ultrasonic sen-sors — Able to detectferrous and non-fer-rous objects without physical contact,these sensors use sound waves to de-tect the presence, distance, and orien-tation of objects. Operating at ultra-sonic frequencies (typically 140 kHz),the sensor emits pulses in a concen-trated beam from a cylindricallyshaped housing, Figure 12. A re-ceiver, within the same housing,senses returning (reflected) pulses.

Photoelectric scanners— These noncontacting sen-sors detect nonferrous ob-jects. They operate by sens-ing the completion orinterruption of a light beam.Photoelectric scanners canbe used tofine tune

automated machineprocesses by manipu-lating the timing andduration of their out-put signal. Five typesof sensing methodsare available.

Retroreflective types depend on a


Figure 11 — Typical proximity sensor.

Figure 12 — Typical ultrasonic sensormeasures about 41/2 in. long and 13/4 in. indiameter.

Figure 13 — Retroreflective principles.

Figure 14 — Through-beam.

Figure 10 — Typical lead screw limitswitch.


Improve encoder performanceDuring the selection process, motion system designers gener-

ally focus on encoder resolution, overlooking the electronics asso-ciated with an encoder. But the operating bandwidth, the type ofoutput driver, and the length and type of cable have an equal af-fect on system performance.

Resolution. Resolution is the number of up-down cycles producedon one channel within one revolution of the encoder shaft, referred toas counts per turn or pulses per revolution (ppr). The number and ac-curacy of the patterns or slots on the code disc and the rigidity andstability of the mechanical assembly affect encoder accuracy.

Discs with the highest accuracy have photolithographicallyproduced patterns on glass. Common accuracy is a few ppm, or 5arc seconds. Less accurate are chemically milled holes in metaldiscs. Errors can be as high as 100 ppm, or 2 arc minutes.

The direct-read (native) resolution value is dictated by thenumber of optical slots on the disc. Too many lines or slots de-crease the percentage of light that can pass, which can producefringing effects and crosstalk.

The bearing and spindle design determine the mechanical sta-bility of any industrial encoder. Resolutions higher than 20 ppm(50,000 counts per turn) will reflect bearing noise in the form ofposition error. Thermal mismatches between mating componentsin the bearing assembly can also degrade performance over tem-perature, and reduce encoder life.

Techniques to obtain better resolutions than through the direct-read value include quadrature decoding and electronic interpolation.

Using a 2.5-in. diameter encoder with 2,500 slots per turn as anexample, the quadrature technique improves resolution because ituses both up and down transitions of a channel to get up to 5,000counts per turn, or both up and down transitions of both channelsto get up to 10,000 counts per turn. Most common controller chipshave these functions built in, so it is a matter of setting a few hard-ware or software switches to activate these higher resolutions.

The choice of direct-read resolution or electrical interpolation isa design decision made by the manufacturer. A common methodfor electronic interpolation is to use a voltage divider circuit to sub-divide the raw analog signal into the desired number of interpola-tion steps. Interpolations as high as 20 times are possible. This isusually done internally to the encoder and is transparent to themotion system designer. It pushes the practical upper limit toabout 200,000 counts per turn (10,000 direct × 20 = 200,000) whichwill handle most common industrial applications.

When specifying encoder resolution for any system, look at theerror analysis for the system and choose an encoder that will readtwo to four times better resolution than the maximum errorsource. In the above example on the encoder, the error sourcesare typically about 20 ppm (26 arc seconds) and the maximumresolution available is about 5 ppm (6.5 arc seconds), or fourtimes better.

Bandwidth. It is not possible to entirely separate the designparameters of resolution and bandwidth. They are related,through operating speed, by the formula:

where: F = Frequency, HzN = speed, rpmR = Resolution, counts per turnThe bandwidth is dictated by the type of output driver (line

driver vs. open collector), length and type of cable, and the type ofterminations used at the controller end.

Line drivers. Consisting of two transistors with collectors con-nected, line drivers ensure a low impedance output, typically less

than 50 ohms. They can be used with cable lengths to 1,000 ft andoperating frequencies to 1 MHz. However, their high switchingspeeds mean that they are prone to ringing while their low im-pedance makes them more noise immune.

For highest performance, the signals for these outputs shouldbe specified to include the channel complements and be carried inshielded, twisted pairs. These signals should be fed to a high im-pedance differential-line receiver, or opto-isolator, which offerscommon mode noise rejection. Cable termination resistors matchthe receiver to the cable impedance to minimize signal ringing.This is usually done through trial and error.

Open collectors. Due to their higher impedance, open collectorsare more bandwidth limited. A good rule of thumb is to use themat a maximum of 50 kHz with a maximum of 50 feet of cable.Their advantage is that they are the least expensive type of out-put. The pull-up resistors can be user supplied or factory in-stalled within the encoder, or both. The arrangement is often dic-tated by the controller or the need to operate the encoder at adifferent voltage than the supply voltage.

Encoder bearing assemblies and load. To ensure a highquality, reliable output, code disc and spindle assemblies for en-coders must be mounted in a bearing assembly. There is a commonmisunderstanding regarding encoders that since they have a bear-ing structure, they can carry significant loads such as those thatresult from installation misalignment. They can carry some loads,however, even a good installation can have a 0.003-in. parallel mis-alignment. The relationship between bearing loading and life is:

where:n = bearing life, revolutionsC = dynamic capacity (from manufacturer’s data)P = bearing load, lbLH = design life of bearing, hrIf the encoder is hard mounted to the motor shaft and the en-

coder housing is hard-mounted to a base plate, the bearing as-semblies could experience side loads of several hundred pounds.Bearing life is approximately inversely proportional to the thirdpower of the load. Therefore, doubling the side loads results in1/8th the useful life.

The most common encoder mechanical failure is caused by ex-cessive bearing load.

Standard, high quality bearing assemblies operate well atspeeds to 10,000 rpm. At higher speeds, bearing operation is lim-ited by a combination of heat generation in the races and surfacemicro-cracking caused by the high relative speeds between theraces and the balls.

Bearing assemblies can be built with tolerances that allowtheir use at higher speeds, to 30,000 rpm. However, temperaturecan still be a problem. Most industrial encoders are built withshaft seals to protect the internal assemblies from dirt, oil, metalchips, etc. As spindle speeds increase, these seals can be a sourceof heat generation. Manufacturers take different approaches toshaft sealing techniques, so be sure to check with the particularmanufacturer to ensure that they have addressed this concern.

Installation. All encoders must have a means to couple di-rectly to the drive shaft to be controlled or measured. Because thecenter of rotation of the encoder is never exactly co-aligned withthe center of rotation of the driving shaft, use a coupling or flexi-ble mounting scheme to resolve the misalignment.

Excerpted from an article by the Industrial Encoder Div., BEISensors & Systems Co., in the September, 1996 issue of PTD.

FN R= ×

60

nCP LH

=

×

31665

ripe fruit, and quality of miniatureelectronic components.

Vision systems — Industrial vi-sion systems aid manufacturers inidentifying, measuring, and inspect-ing parts for defects. Each systemconsists of a camera, lights, processor,and computer. In operation, the videocamera scans the object line-by-line,sending video signals for each line tothe processor for conversion into aform understood by the computer.The processor digitizes these signalsin the form of a grid where eachsquare in the grid is a picture elementcalled a pixel. Light intensity is deter-mined for each pixel and expressed asa number from zero to 256 in a systemcalled gray scale.

The gray scale data is sent to thecomputer for analysis of the image,using one of several techniques:

•Feature analysis — the determi-nation of characteristics such as size,shape, and location of holes.

•Edge detection — finding the

the probe. The probe is a differentialmagneto-resistive element attached toa powerful permanent magnet. Itsoutput is connected to an amplifierand drive transistor. The imbalance ofthe magnetic field caused by the geartooth passing under the legs of thesensor produces an output signal. Ad-vantages and disadvantages of thistype are similar to that of magnetictypes; however, they can operate tozero rpm.

Resolvers look like small motorsand are, in some respects, similar inconstruction. A resolver contains sin-gle winding rotors that revolve insidefixed stators with two windingsmounted 90 deg apart. The rotorwinding is typically driven by a refer-ence voltage at a frequency rangingfrom 400 Hz to several kHz. The re-solver circuitry contains a resolver-to-digital converter that compares out-put to a fixed reference frequency todetermine shaft speed. In most cases,an analog output proportional to theshaft velocity eliminates the need fora separate tachometer feed. Resolversare rugged, wide-temperature-rangedevices that operate reliably to zerorpm. Disadvantages of resolvers arethe relatively high moment of inertiaand the complexity of the electroniccircuitry.

Recent design innovations, how-ever, may reduce or eliminate thesedisadvantages.

A new design uses a solid rotorwithout windings, which makes theresolver less complex and less costlyto manufacture. The primary andsecondary windings are in the stator,therefore it does not need a rotatingmagnetic coupling, Figure 17. The de-sign is inherently brushless.

The rotor contains a diagonal sec-tion (not perpendicular to the shaft) ofhigh permeability material thatvaries the magnetic field across thestator as the rotor turns. Transferredenergy remains magnetic from theprimary coil through the air gap tothe sinusoidally shaped poles of thesolid rotor. The total flux through thegap is constant—the rotor determinesthe angular position within the statorbore where the coupling occurs, andthus, the relative amplitudes of theoutput signals.

The primary coil is wound circum-ferentially between the two stators.The secondary windings are wound inthe stator slots in space quadrature,

edge of an objectbased on a change ingray scale level.Edges can be used toderive other informa-tion such as part di-mensions.

•Template match-ing — comparing features (pixel bypixel) of an object with a standard image.

Rotary encoders and resolvers— Sometimes referred to as incre-mental encoders, these devices pro-duce discrete electrical pulses dur-ing each increment of shaft rotation,and provide high-resolution feed-back data on shaft position, velocity,and direction. Available in a range ofresolution capabilities from 1 to30,000 pulses/rev, four types arecommonly used.

Optical encoders use one or morelight sources and a set of photo detec-tors, separated by a rotating disk anda stationary reticle. Advantages to op-tical encoders are: operation down tozero speed, low cost, high resolution(2,500 pulses/rev), excellent immu-nity to shock and vibration, low mo-ment of inertia, high reliability, andthe capability to operate over a widetemperature range.

Magnetic encoders consist of avariable reluctance probe, an irongear, and housing, Figure 16. The

probe consists of astrong permanent mag-net wrapped with elec-trical wire to form amagnetic coil. The re-luctance path of themagnet is varied by thepresence or absence ofteeth. The continuouslychanging magnetic fieldinduces a voltage in thecoil that produces a sinewave output from theencoder. Advantages tothis type of encoder are:lowest in cost, no exter-nal power require-ments, highest reliabil-ity, and widestoperating range. Disad-vantages are: resolutionis l imited to 360counts/rev, relatively

high moment of inertia, and cannotbe operated at zero or low speed.

Magneto-resistive encoders are sim-ilar to magnetic encoders except for


Figure 16 — Magnetic rotary encoderconsists of variable reluctance probe,iron gear, and housing. Output is a sinewave.

Figure 15 — Diffuse reflective.


sim-ilar to atraditionalresolver. The induced voltageamplitudes correspond to the sine andcosine of the rotor angle as in a tradi-tional resolver.

This design is mechanically andelectrically compatible with tradi-tional resolvers, Figure 18. Becausethere are no coils in the solid rotor, it israted to 30,000 rpm. Higher speeds area matter of mechanical balance. Thetop speed (100,000 rpm) is limited bythe excitation frequency, and the burststrength of the rotor.

The single-stage magnetic designmeans low source impedance, so theresolver is less susceptible to noisepickup and tolerant of long cableruns. It can be excited at frequenciesto 40 kHz.

Linear variable differentialtransformers — LVDTs are fre-quently used as position sensors be-cause they provide excellent linearity,infinite resolution, and high sensitiv-ity, Figure 19. The oscillator convertsdc input to ac, exciting the primarywinding of the differential trans-former. Voltage is induced in the sec-ondary windings by the axial core po-

rated voltage over a 20:1 range and of-fers repeatability of 60.1%, of maxi-mum rating. Response is adjustablefrom 600 ms to 250 ms through the op-eration of a damper. Load cells withdual LVDTs offer ranges to 125:1.

Strain gages — These devices un-dergo a change in electrical resistancewhen subjected to strain—a material’schange in length due to applied force.Strain gages are also used to measurephysical quantities such as load,torque, pressure, tension, and weight.In industrial process control, straingages are probably the most commondevices used to measure force.

They are also used for static anddynamic measurements. They can re-spond quickly to minute displace-ments. However, they may lose accu-racy due to hysteresis, when the gage

sition. The two sec-ondary circuits con-sist of a winding, afull-wave bridge, andan RC filter. The cir-cuits are in series op-position so that theresultant output is adc voltage propor-

tional to core dis-placementfrom the elec-trical center.Voltage po-

larity is afunction of the

direction of coredisplacement from

the electrical center.

Today, LVDT circuits built on a sin-gle chip greatly simplify interfacing.These chips contain a programmablefrequency oscillator, a synchronousdemodulator, and an amplifier thatproduces a buffered voltage output.

Force transducersA force transducer produces an

output signal proportional to appliedforce. LVDTs and strain gages aremulti-purpose electrical devices thatcan measure a variety of physicalquantities, including force. In force

measurement applications,these devices convert me-chanical displacement due toapplied force into an electri-cal signal.

LVDTs are often used inload tables containing an inter-

nal spring and flexure, Figure 20.The load moves the load table, which

moves the core in the LVDT. DC out-put signal is linear within 0.2% of

Figure 17 — A new resolver design hasprimary and secondary windings in thestator and a solid rotor.

Figure 18 — In traditionalbrushless resolvers, a rotarycoupling transformer transfers energyfrom the stator to the rotor.

Figure 19 — Principle of a dc-to-dc LVDT.

Figure 20 — Typical LVDT-type forcetransducer.

must measure the same displacementfor many cycles, or due to overload.Strain gages are available in three ba-sic types: bonded, unbonded, andsemiconductor.

Bonded gages are generally avail-able in two types. One type, Figure21, uses thin metallic filamentsbonded to a thin paper or plastic base.Another type, Figure 22, uses a foil(thin film) design. These gages haveresistive elements that are vacuumdeposited on a ceramic film. Pressureapplied to the pressure summing di-aphragm deflects the diaphragm andchanges the resistance of the foilstrain gages that are wired into aWheatstone bridge.

Unbonded gages are generallysmaller than bonded types, but arenot as rugged, nor as stable as thin-film pressure types. They differ frombonded types in that the filaments areattached between movable and fixedsupports. Unbonded gages can bemade to measure very low and veryhigh pressures — from less than 0.05to 20,000 lb/in.2

Semiconductor gages includebonded bar and diffused types that of-fer important advantages such assmall size and high output signal(about 100 mV).

established between the stationaryand rotating components of the trans-ducer, usually by means of slip ringsor rotary transformers. Where appli-cation constraints such as space, con-figuration, or speed prohibit the use ofthese types, other designs may beused, such as the torsional variable-differential transformer (TVDT),phase-shift transducer, or reactiontransducer. Torque transducers areused in quality control to check com-pleted motors for rated torque, and forprocess control to maintain constantmotor torque on production lines.

Smart sensorsIn distributed control systems, op-

eration logic is often placed close tothe application being monitored bycombining a sensor and microproces-sor into one unit, commonly called asmart sensor. This reduces responsetime and interference from noise, andallows a smaller, less expensive PLCto be used because routine, repetitiveoperations are controlled by the sen-sor’s own microprocessor. Functionsperformed by smart sensors includecorrecting for environmental condi-tions, diagnosing operation, and mak-ing decisions.

One type of smart sensor, an elec-tronic programmable cam switch in-cludes a position sensor and a micro-processor, sometimes in oneenclosure, Figure 24. An encodingdisc converts shaft position data intobinary coded decimal data (BCD). Amicroprocessor compares the BCDdata to user-defined dwell settingsstored in an EEPROM chip, and then

determines the on/offstatus of several out-puts. The outputs areprogrammed into theswitch with a built inkeypad or a hand-held programmer.Dwell settings can bemaintained to withinplus or minus 1 deg at2,000 rpm.

Because it is pro-grammable, theswitch can be inte-grated into an auto-mated factory systemthrough an intelligent

I/O module. For example, a host com-puter can direct a controller (PC orPLC) to adjust machine parameters

Bonded bar gages are similar tobonded foil gages in that they arebonded with epoxy to the mechanicalstructure (cantilever beam or di-aphragm). The bonded bar gage is apiece of pure single-crystal siliconinto which an impurity such as boronis introduced during the growth of thesilicon crystal. The entire piece of sili-con serves as the strain gage.

Diffused gage transducers areformed from a piece of pure siliconwith boron diffused areas that arestrain sensitive. Because the straingages are an integral part of the sili-con wafer, which also serves as the di-aphragm, no bonding is required. Theresult is a structure whose stability issuperior to that of unbonded wire,bonded foil, and bonded semiconduc-tor strain gage transducers.

Torque transducersA torque transducer consists of a

mechanical structure that distortswhen subjected to a torsional force.Strain gages, which are bonded to thestructure, deform with the structureand produce corresponding stress orstrain values. The two most widelyused structures are hollow-cruciformand square-section, Figure 23. Hol-low-cruciform types produce highstrain values at low torque and areused extensively in low-capacity (be-low 500 lb-in.) applications. Squaresections are used for high capacity(500 to 500,000 lb-in.). Torque trans-ducers are accurate to 0.1% or betterand operate up to 35,000 rpm.

When used in rotating applica-tions, a signal transmission path is


Figure 23 — Typical torque transducersconsist of a physical structure and straingages.

Figure 21 — Bonded resistance-typestrain gage

Figure 22 — Foil strain gage.


for a product changeover. The con-troller downloads instructionsthrough the intelligent I/O modulewhich receives position feedback datafrom the switch to fine tune machineoperations.

Flow sensorsFlow rate is a common process vari-

able that must be controlled in thepetrochemical, food-processing, andsimilar industries. Many flow rate sen-sors are actually pressure sensitive de-vices. They respond to changing fluidpressure. Generally, flowmeters areclassified as head flowmeters, positivedisplacement flowmeters, and variablearea flowmeters.

Head flowmeters measure flow bypassing the fluid through a restriction.Pressure on both sides of the restrictionis measured, and flow rate is directlyproportional to the square root of thedifference of the static head (high pres-sure before the restriction) and dy-namic head (low pressure after the re-striction). Commonly, orifice plates;flow nozzles, and venturi tubes com-prise head flowmeters.

Positive displacement flowmeters andmetering pumps measure fluid flow di-rectly, not indirectly like head flowme-ters. The positive displacementflowmeter has a measuring chamberbetween inlet and outlet ports. Avane, disc, or piston in the measuringchamber displaces a fixed fluid vol-ume with each revolution. The fluidvolume passing through the flowme-ter is a function of the meter’s rota-tional speed. Fluid flow is a product of

clude the capillary and rotating discviscometers.

Capillary types force fluid througha small diameter tube at a constantflow rate. The pressure drop acrossthe capillary tube is proportional toviscosity. This viscometer requiresclose temperature regulation. It can-not be used with fluids that have non-Newtonian viscosity characteristics.With a non-Newtonian fluid, the ratioof shear rate (velocity per unit thick-ness) to shear stress (force per unitarea) is not constant.

Rotating disc viscometers measurethe torque required to rotate the discin a fluid at constant speed. With onetype, a strain gage measures the forceexerted by a spring loaded wormgearthat drives the rotating disc spindle.The signal generated by the straingage can be fed to a recorder to moni-tor viscosity.

POWER CONTROLSPower controls direct energy to

power consuming devices, such asheaters, lights, and other similaritems that can be associated withpower transmission products. Three ofthe most common elements used incontrolling electrical energy to theseitems include switches, circuit break-ers, and contactors. Motor controls arecovered separately in the next section.

SwitchesUsually operated manually,

switches provide one of the simplestmeans of turning electric power onand off, and are rated from 10 throughseveral thousand amperes. Having aninternal design similar to early knifeswitches, modern switches are fre-quently enclosed in a protective hous-ing and operate in a snap-acting man-ner to reduce sparking. Enclosuresare available in general purposeNEMA 1 through NEMA 7 for explo-sive atmospheres, and NEMA 12 forcommon oil and dust environments.Frequently termed disconnectswitches, these units come with orwithout fuses depending on applica-tion requirements. They are alsoavailable as transfer switches, fortransfering the source of power from amain power system to an emergencysystem. Typically, these units aresimply power rated multiple-pole,double-throw devices. In addition to

displacement and chamber cross-sec-tional area. Typical applications in-clude water, gas, and gasoline flow.

Variable area flowmeters includerotameters and piston-type meters.Variable area flowmeters operate onthe same principle as other differen-tial headmeters. They have a variableorifice and relatively constant pres-sure drop.

Rotameters are the most commontype of variable area flowmeter. Theyconsist of an upright tapered glasstube containing a float, usually madeof metal. When fluid flows throughthe rotameter, the float’s weight op-poses the upward force exerted byfluid flow. You can determine flowrate by reading the float position ofthe rotameter’s calibrated scale. Somerotameters are designed for directviewing; others are coupled magneti-cally or hydraulically to allow auto-matic recording of flow rate.

Piston types consist of a fitted pis-ton in a sleeve or cylinder, held in acast body. The cylinder has inlet andoutlet orifices cast into it. The pistonindicates the flow rate. These flowme-ters can be used to measure highlyviscous or corrosive fluids. However,they should not be used with fluidscontaining sediment.

Viscosity sensorsViscosity is a measure of a fluid’s

resistance to flow. It should be consid-ered in industries where fluid must bepumped, such as petrochemical, phar-maceutical, and food processing appli-cations. Typical viscosity sensors in-

Figure 24 — Electronic programmablecam switch senses position and controlsseveral outputs.

being manually operated, they cancome with different types of operatingschemes — such as magnetic unitsthat will automatically switch to astandby power system should themain supply fail.

Switches are available for switch-ing ac and dc power. Typically, the dcswitch will include blow-out coils (typ-ically permanent magnets), whichhelp extinguish the arc of high cur-rent dc to increase contact life.

Circuit breakersCircuit breakers are usually con-

tained in a single enclosure and areavailable in fractional through sev-eral thousand amp ratings. Thermaland compensated-thermal breakersdepend on heat for tripping. The ther-mal tripping range is usually 101 to120% of their current rating and of-fers an inverse time delay character-istic. Their characteristics differ onlyslightly when used on dc or 60 Hz ac.

Circuit breakers with a magnetictrip feature offer instantaneous re-sponse, tripping whenever a predeter-mined value of overload occurs. Somebreakers contain both thermal andmagnetic trip functions. Typically,the magnetic portion is set to tripwhen the overload exceeds 600% ofrating. The relationship betweenmagnetic and thermal trip character-istics can be varied, usually throughan adjustment screw.

Circuit breakers can also come withground fault interrupters that causethe breaker to open whenever a highamount of current flow to ground isdetected. This feature is especially im-portant when the circuit breaker con-trols devices operated by personnel.Many types of ground fault detectorsare available to meet a range of appli-cations. Consult manufacturers withthe specific application requirementsto assure proper personnel protection.

Circuit breakers are not designedfor a high number of operations. If itis desirable to open or close a powercircuit several hundred-thousandtimes, use a magnetic contactor.

ContactorsThese are magnetically controlled

devices for controlling electricheaters, large banks of lights, andother power consuming devices. Un-like magnetically operated circuit

(across-the-line starters), and softstart (reduced voltage starters). Bothoffer two functions — turn the motoron and off, and provide overload pro-tection, as required by the NationalElectrical Code. This code also re-quires that almost all motor branchcircuits contain a disconnectingmeans and short circuit protection.These functions are provided by cir-cuit breakers and fused disconnects.If a single enclosure contains the mo-tor starter plus the disconnect andshort circuit protection, the unit is acombination starter.

Full-voltage startersThese starters apply full line volt-

age directly to the motor terminalsand come in two basic types: manualand magnetic. Other types, such as

breakers, contactors are designed forhundreds of thousands of operations.

Contactors are available in manyenclosures, including explosion-proofand NEMA 12 industrial use. Avail-able for operation on ac or dc (whichusually include blow-out magneticsfor rapidly extinguishing the arc)these units should be specified basedupon the type of load they are control-ling. For example, units designed forcontrolling incandescent lights andhighly inductive loads are derated togive reliable operation for long periodsof time, even if the contacts must openor close under abnormally high cur-rents. Contactors are available withnormally open or normally closed con-tacts, or a combination of both.

The most common type of contactoris built to operate with ac poweredcoils for controlling ac power devices.Many manufactur-ers offer variouscombinations ofcoil voltages andtypes, including4,160 Vac and sev-eral hundred am-pere rating.

MOTORCONTROLS

Motor controlsare usual ly re-ferred to asstarters or drives.Motor startersstart and stop mo-tors by applying orremoving electric power. Drives aremore complex than simple starters, inthat they conditionthe electricalpower prior to de-livering it to themotor. Motorstarters and re-duced voltagestarters representthe simplest formof motor control.

AC motorstarters andcontrols

AC motorstarters and controls are grouped intotwo broad classifications: full voltage


M

M

OL

OL

OL

OL

Single phase

Three phase

Figure 25 — Across-the-line manualstarter.

L1

L2

L3

M

M

M

Motor

OL

OL

OL

OL

OL OL

Fusesecondary

StartStop

Figure 26 — Across-the-line magneticstarter.


reversing starters, are magneticstarters with added control functions.

Manual starters are hand operatedmechanisms that connect or discon-nect the motor circuit. A thermal pro-tective circuit in the starter guardsagainst excessive motor temperatureconditions. Manual starters are lim-ited to single-phase motors up to 5 hpat 230 V, and three-phase motors upto 15 hp at 600 V, Figure 25.

Magnetic starters are similar tomanual starters but the electricalconnection to the motor is madethrough a set of contacts, which arecontrolled by a coil. De-energizing thecoil opens the circuit, Figure 26.

Reversing starters switch two of thepower leads going to a three-phasesquirrel cage motor to reverse its di-rection of rotation. Reversing startersare usually magnetic and switch thetwo leads through another set of con-tacts, Figure 27.

Reduced voltage startersThese starters apply less than

rated voltage during the starting se-quence of ac motors. During starting,ac motors typically draw 600% of fullload current and develop 150 to 200%of rated torque. This high current cancause the power line voltage to drop,which can make lamps flicker, andcause sensitive instruments to givefalse indications. The high starting

does not exceed a selected value, on acurrent limit ramp, or, by usingtachometer feedback, on a linear mo-tor speed ramp. This solid-statemethod offers many advantages, including a high degree of controlla-bility and stepless acceleration, Figure 28.

Any application of reduced volt-age starters must be carefully engi-neered to assure proper coordina-tion of required starting torque,starting duration, starter capabili-ties (current and time), and motorheating characteristics.

MOTION CONTROL SOFTWARERecent developments in pro-

gramming software let engineersconcentrate on designing a motioncontrol solution rather than writ-ing l ines o f code. The most fre-quently executed programmingtask for the engineer is to point and

torque can breakbelts and damageother mechanicaldevices.

By reducing theapplied voltage tomotors duringstarting, the initialline currents andtorque surges arereduced. The linecurrent reductionis approximatelyproportional to thevoltage reduction.However, thetorque is approxi-mately propor-tional to thesquare of the volt-age change.

Three methodsare commonly used

to reduce the motor starting voltage:resistors, auto-transformers, andsolid-state con-trollers.

Resistors —Connected in se-ries with the mo-tor during start-ing, resistors areshorted out toapply full voltageduring running.A timer fre-quently deter-mines the dura-tion of thestarting phase.

Auto-trans-formers — Con-nected in the mo-tor circuit tosupply reducedvoltage duringstarting, full linevoltage is sup-plied to the mo-tor after thestarting se-quence.

Solid-state —Also termed softstart controls,these use SCRsto increase theapplied motorvoltage from zeroto rated. This in-crease can becurrent limitedso the motor

L1

L2

L3

R

F

Motor

Fusesecondary

Stop ReverseForward

R1

R2

F2

F1

R

F

OL

OL

OL

OL

OL

OL

Figure 27 — Across-the-line reversingstarter.

Figure 28 — Power circuit of typical solid-state, reduced-voltagestarter. SCRs control motor voltage during starting sequence only.After starting sequence, motor runs at line voltage.


click. Engineers can use prewrittenroutines, or fill in the blanks, orkeystroke parameter data such asdesired final position, speed, num-ber of axes, and position points; thesoftware does the rest of the pro-gramming.

The popular Windows operatingsystem is a large part of the reasonwhy motion control software hasbecome easier to use then earlierversions. Windows offers program-development tools and program-ming aids (such as Visual Basic,DDE, and others) that manufac-turers can use to create applica-tion programs that almost writethemselves.

Some motions must be pro-grammed with more sophisticatedsoftware. For 85 to 90% of pro-grammed motion, however, the Win-dows-based software works well.

Typical features of the easier-to-use software include:

Graphic icons. Engineers linkthese icons to describe the motionsystem in flow chart form. The soft-

ware converts the icon flowchartinto code and transmits that codeto the drive or motion system.

Integrated diagnostics. Theseprograms offer digital versions ofoscilloscopes, spreadsheets, andreal-time signal plots, which aidmotion system diagnostics. For ex-ample, using the oscilloscope ver-sion to set up servo drives can be aone key-stroke operation. Such afeature eliminates the time an en-gineer needs to f ind cables andwire up the scope, and reduces thetime needed to measure the cor-rect signals.

Dynamic Data Exchange (DDE).The Windows operating system istoo slow for motion control applica-t ions . DDE is an addit ion thathelps alleviate this limitation. Itcaptures and quickly transmitsdata from a device, such as a mo-tion controller, to software applica-tions such as spreadsheets, plot-t ing rout ines , and operatorinterfaces.

File management . Some pro-

grams will also manage file upload-ing and downloading, and archivethe application programs.

CAD interface. A few programscan receive files from a CAD sys-tem. These programs will take theCAD data and create lines of codeto manipulate the axes to trace thepattern.

In addition, most software offersautomatic tuning of drives to themotion system. You enter in a fewparameters and the software doesthe rest. Most programs also offerHelp functions for on-line explana-tion or instruction. ■

controls and sensors - uvajesman/bigseti/ftp/actuadores/...1998 pt design a159t ogether, controls...

Documents