Solis State Sensors

7/18/2019 Solis State Sensors

http://slidepdf.com/reader/full/solis-state-sensors 1/43

Basically, a sensor is any of various devices designed to detect, measure or record

physical phenomena — such as radiation, heat,pressure, or presence — and to respond bytransmitting information, initiating changes, oroperating controls (Figure 1).

PRESENCE DETECTION

Presence sensors detect the presence of anobject — a target material. In industry, thereare as many targets as there are products andcomponents of products. As you will see, thereare many problems unique to certain targets,but there are different sensing technologieswhich can be applied to solve the problem.

In the context of an automated manufactur-ing process, sensors are essential in providing positioning data so controllers, such as PLCs,can perform their tasks all along the produc-tion line. In the context of safety, sensors in

safety interlock doors and guards keep opera-tors safely guarded from hazardous processes.They also aid manual processes, such as posi-tioning of fork lifts for material handling andtrucks in shipping docks. They are used incounting and rotary speed monitoring. In anyautomated material handling application, sen-sors close the data loop between the operator,the controller, and the process. There are 2general types of sensing:

1. Discrete (on-off)2. Analog (continuous-proportional)

Analog gives a continuous response, answer-ing the question, “Where is it?” Analog re-sponds best to continuous processes, such aslevel control, positioning, etc.

The discrete type answers the question, “Isit here, or not?” Discrete sensors have a con-tact, and the ON-OFF position corresponds tothe discrete states. The discrete type is mostapplicable to manufacturing, since most sens-ing applications deal with processes whichdemand that a target be in a definite positionfor assembly, or at a specific level for filling. Insituations such as materials handling or theassembly of products, manufacturing systemsneed to know when the target is at a givenlocation. They don’t need to know when thetarget is almost there, or almost gone.

There are 2 general categories of presencesensors (Figure 2):

1. Physical contact (example: limit switch)

2. Noncontact (example: photoelectricsensor)

1

CHAPTER 1

Introduction to

Sensing Technologies



2

face translates the distance into an electricalsignal. A trigger, or threshold level detector, isused to trigger the output circuit when thesignal reaches a predetermined level. An out-put circuit interfaces with data acquisitionsystems (PLCs, dedicated controllers), or othercontrol circuits (relays, counters, timers, etc.).

DISCRETE SENSINGIN THE REAL WORLD

The wavy line in Figure 4 expresses the realvariation of the distance between the target andsensor. The line is wavy because we are dealing with nonuniform movement caused by vibrationor jerkiness in the mechanical systems whichare found in a real industrial environment. Thevariations can be a problem, especially in slow-moving applications. With only a single thresh-old, those small changes in distance cause achattering effect between the ON and OFFstates. This is noise that we need to eliminate.

If we introduce a second threshold for OFFabove the threshold for ON, the small wave-forms between the two are filtered-out and thechattering is suppressed. The distance be-tween the ON and OFF thresholds is thedifferential.

Differential, also called hysteresis, is thedistance between the operating point and the

As an example, you would want anoncontact sensor to detect the presence of freshly painted car doors as a means of count-ing how many had been sent on to the curing area.

NONCONTACT SENSORSVERSUS LIMIT SWITCHES

With limit switches, the target directly actu-ates the switch through physical contact.Noncontact sensors are, with few exceptions,solid-state electronic devices which offer thefollowing advantages:

• Physical contact is not required.• No moving parts to jam, fail, or be

broken by extreme physical contact.• Faster.• No bouncing at contact state change.

Advantages of limit switches include:

• Carry more current.• Are safer–when opened, there is an air

gap between the contact terminals.

COMPONENTS OF ASOLID-STATE SENSOR

In generic terms, solid-state sensors willhave an interface, generally an oscillating cir-cuit, the kind of which is the main distinctionbetween the technologies (Figure 3). The inter-

FIGURE 2. The 2 general categories of pres-ence sensors.

PHYSICALCONTACT

NONCONTACT

TARGET

E R

FIGURE 1. Components and systems forautomation.

DISTRIBUTION

POWERSWITCHING

ACTUATORS

DETECTION

MAN AND MACHINE DIALOGUE

PROCESSING

MACHINES ORINSTALLATION



3

release point as the target is moving away.Differential is necessary in order to make aprecise determination of target presence with-out factors of the environment intervening tocreate a noisy output signal. All discrete sens-ing technologies must have a differential. Insome technologies, differential occurs as a by-product of the basic laws of physics. However,in other technologies it must be manufacturedthrough additional circuitry.

MAJOR NONCONTACTSENSING TECHNOLOGIES

The major noncontact sensing technologiesare:

• Reed Relay (Magnetic Target)• Inductive (Metal Target)• Capacitive (All Materials Target)• Photoelectric (All Materials Target)

Inductive, capacitive and photoelectrictechnologies compose the majority of industry’s applications, but other technologiesare in the field and you should be aware of them. They are:

• Ultrasonic (All Materials Target)• Magnetic Reluctance (Magnetic Metals

Only)• Hall Effect (Magnetic Target)

Generally, the more control you have overwhat passes in front of the sensor, the better.The less control you have over what passes in

TABLE 1 — TERMINOLOGY

Nominal sensing distance: The rated operating distance for which the switch is designed. This valueshould only be taken as a guide, since no manufacturing tolerances or changes in external operating environment are taken into account. You should be aware that manufacturers arrive at performancestandards for their products using standardized criteria, so you can compare “apples to apples” indetermining which product is right for your application. These criteria reflect, to a reasonable extent,the real performance which can be expected in an “average” controlled environment.

Usable sensing distance: This sensing distance is used for specifying a specific sensor for an appli-cation. Relative to a standard target, the usable sensing distance takes manufacturing tolerances andexternal operating conditions into account. This specification gives confidence to the specifier — alldevices are guaranteed by the manufacturer to perform within the usable sensing distance.

Effective (or Real) sensing distance: This is the actual sensing distance realized by the actualapplication. The effective sensing distance will be no less than the usable, and is usually closer to thenominal in average circumstances.

Standard target: An object with standardized dimensions or characteristics–common among manufac-turers–used in the product laboratory to determine benchmark performance characteristics for a sensor(Figure 5).

Target: The real object to be sensed.

FIGURE 3. Components of a solid-statesensor.

TARGET

INTERFACE

ON

OFF

TRIGGER

OUTPUT



front of the sensor, the more selective thesensor must be for identifying the target. Reedrelay is among the best in this regard becauseit senses only magnetic targets, and magnetsare rare in industry. If you differentiate yourtarget by putting a magnet on it and use a reedrelay to detect its presence, it won’t be fooledby other objects that don’t have magnets at-tached.

Inductive sensors detect metal and wouldnot be fooled by non-metal objects. This is ahighly accurate technology, so if you are able tocontrol what metal objects pass in front of the

sensor, this technology is good at letting youknow precisely when they are within sensing (detection) range. Photoelectric can be fooledvery easily — if you break the beam, some-thing is sensed. That object could be the target,operator’s hand, or nothing but a cloud of dust!

REPEATABILITY

Repeatability is the ability of a specificsensor to repeatedly detect a target at agiven point. It is related to the precisionneeded for the application. For example, areyou filling a bottle that requires a positioning

TABLE 2 — NEMA ENCLOSURE RATINGS (AMERICAN)

Classifications

1 For indoor use — protection against contact with enclosed equipment.

3 For outdoor use — windblown dust, rain, sleet, external ice formation.

4 Windblown dust, rain, hose directed water.

4x Corrosion, splashing, hose directed water.

6 Short-term immersion in water, limited depth.

6p Long-term immersion (8 hours at 6 ft. depth).

6x Saline mist, withstanding UV.

11 Oil immersion, corrosive liquids and gases.

12 For indoor use — dust, falling dirt, dripping non-corrosive liquids.

13 For indoor use — lint, dust, seepage, external condensation and spraying of water, oil and

non-corrosive liquids.

4

TABLE 3 — IEC ENCLOSURE RATINGS (EUROPEAN)

1st Characteristic numeral

Protection against contact and penetration ofsolid bodies.Conforming to IEC, NFC, DIN.

0 Non-protected.

1 Protection against solid objects greater

than 50mm.


than 12mm.


than 2.5mm.


than 1mm.

5 Dust protected.

6 Dust tight.

2nd Characteristic numeral

Protection against contact and penetration of liq-uids.Conforming to IEC, NFC, DIN.

0 Non-protected.

1 Protection against dripping water.

2 Protection against dripping water when

tilted up to 15°.

3 Protection against rain.

4 Protection against splashing water.

5 Protection against water jets.

6 Protection against heavy seas.

7 Protection against the effects of immersion

(short term).

8 Protection against immersion in cutting

oils. (Telemecanique/Square D proposal)



5

accuracy of ±1/16 of an inch at the filling station, otherwise you have product spilling on the floor? Or are you interested in sensing

the presence of a car in an automatic carwash with an accuracy of ±1 foot?

RATING THE TECHNOLOGIESFOR REPEATABILITY

Different technologies offer varying levels of precision. Rating the technologies for repeat-ability in descending order:

1. Limit Switches2. Reed Relay (magnetic target)3. Inductive4. Photoelectric

5. Capacitive

ENCLOSURE RATINGS

An enclosure is a surrounding case con-structed to provide protection to personnelagainst incidental contact with the enclosedequipment, and also to protect the equipmentfrom environmental conditions: water splash-ing, short-time immersion, cutting oils, andsaline mist. The American standards for enclo-sures are provided by NEMA, and the Euro-pean standards are provided by IEC. The most

important NEMA classifications are listed inTable 2.

The IEC system has IP ratings (Table 3); 2digits are normally used. The first digit de-scribes the dimensions of the solid object thatcan penetrate the enclosure. Maximum protec-tion is 6, which is dustproof. The second digitrepresents the enclosure’s ability to withstandliquids, with 7 as the maximum, representing short-term immersion in water, similar to

FIGURE 4. Discrete sensing in the real world requiresdifferential.

SensingDistance

SensingDistance

Threshold

Time

Time

Time

OffThreshold

Time

OutputSignal

OutputSignal

HOnThreshold

FIGURE 5. General terminology for tar-

StandardTarget

RealTarget

NEMA 6. Therefore, an IP67 is an excellentrating common to most sensors.

The IEC standard allows an even higher

letter grade: IP68. The 8 for liquids is notdefined by IEC, rather, it is left to the usersand manufacturers to define for specializedcircumstances.



CHAPTER 2

Sensor Types

REED RELAY: FOR MAGNETIC TARGETS ONLY

Reed proximity switches are actuated bythe presence of a magnetic field caused by apermanent or electromagnet. Typical sens-ing range is 0 to 1.5".

The sensor is called a reed relay becauseit has two little ferromagnetic reeds her-metically sealed in a glass bulb such that

they overlap, but don’t touch. When a mag-net is brought close to the reed bulb, thefoil reeds take on opposite magnetic prop-erties and attract each other. The reedsmake contact, closing the circuit.

Referring to Figure 1, when the northend of a magnetic field is introduced to thereed head-on, the relay actuates in the areashown by the solid line. This distance willvary according to the strength of the mag-net. As the magnet is removed from thearea, the reed relay opens at the dotted line.The difference between the point of actua-tion as the magnetic target comes closerand the point of de-actuation as the targetmoves away is the differential. This is anaturally occurring phenomenon when us-ing this technology. Note that when themagnet approaches the reed in a planeperpendicular to the reed’s axis, the area of commutation is altered, and 2 areas of actuation are created. This is not desiredbecause it gives 2 signals for one target

pass.

MAGNETICALLY-ACTUATEDSWITCH APPLICATIONS

The more selective your sensor can be forthe target, the better, and the more controlyou have over what passes in front of thesensor, the better. Reed relay is among thebest in this regard because it senses onlymagnetic targets, and magnets are rare inindustry. Applications, advantages and dis-advantages of magnetically actuated

switches follow.

Applications:

• Security and safety — when non-targetmaterial can accidentally trip other de-tectors. Ideal for security applications,since magnets are not very common ma-terials. Because reed relay is anoncontact sensing technology, break-down from mechanical wear is virtuallyeliminated. Examples: Security door in-terlock; secondary switches for operatorsafety applications (redundant systemfor safety, in addition to a switch with adirect opening action contact).

• Sensing through metal walls (non-fer-rous). Example: Piston position in an airor hydraulic cylinder (the piston has amagnetic band).

• Identification of metal bins, containers.Example: Magnetic tape or small magnetsin preassigned slots can uniquely identify

a carrier on an automatic assembly line.



• Environments where sand, dirt, dust, oilor coolants are prevalent.

• Ideal for high-speed operation.

Advantages:

• No bouncing. Great for interfacing with

PLCs or other solid-state circuits.

• Excellent low-current characteristics.

• A low-cost alternative to other proximitysensing technologies.

Disadvantages:

• Welds easily, causing failure. As a pre-caution, use a resistor in series to limitcurrent surges for long cable runs.

• Sensitivity to welding fields; do not use inclose proximity.

• Needs magnet.

INDUCTIVE PRINCIPLES

An inductive detector senses the proximityof a metal object utilizing an oscillator prin-ciple. Surface currents, known as eddy cur-rents, are induced in the metal target by theelectromagnetic oscillating field. These targetcurrents react with the oscillator’s field — the

closer the target, the stronger the reaction.

When the metal target is outside theoscillator influence, the magnitude of theoscillations is not impeded. When the targetis inside the field, it attenuates the magni-tude, eventually stopping the oscillator.That’s why it’s named ECKO (Eddy CurrentKilled Oscillator).

Figure 2 shows how the proximity of ametal target affects the magnitude of oscil-lations, and therefore the output voltage.

SELECTING INDUCTIVE SENSORS

The usable sensing distance and repeat-ability are some of the most importantattributes in matching sensor technology toreal-world applications. Nominal sensing distances range from 0" to 2". Choose tubu-lar or rectangular sensors based on themounting requirements of the application.Tubular sensors install in holes, while rect-angular sensors mount to flat surfaces (Fig-

ure 3). Tubular devices allow for sensing distance adjustment when mounting, while

the rectangular do not.

SHIELDED AND NONSHIELDEDINDUCTIVE SENSORS

The nonshielded sensor has a fieldwhich extends around the front surface tothe sides of the unit (Figure 4). When ametal target enters this field, there is apredictable point at which the unit willsense its presence. The field must be loadedwith metal to a threshold point, and thenthe sensor will change states. When thetarget is closer to the sensor, less metal hasto be introduced to the field to make thesensor change states. Using a standardtarget, a predictable curve can be plotted todefine the sensing distance characteristics.

If a nonshielded inductive sensor is

mounted in metal such that the metal is inthe field (embedded), this will “pre-load”the field and the sensing pattern will nolonger be predictable.

The shielded sensor introduces metal intothe field within the switch itself. Metalmounting brackets or fixtures will not influ-ence the sensing pattern. The shielded sen-sor can even be flush-mounted in metal.Note that a shielded sensor of a given typewill have a shorter sensing distance than its

nonshielded counterpart. On the otherhand, it will be better protected mechani-cally.

CAPACITIVE PRINCIPLES

This is an “all materials” type of sensor.Capacitive proximity switches contain ahigh-frequency oscillator with one or bothof its capacitor plates built into the end of the sensor. Typical sensing ranges are from0 to 3/4 of an inch. When the target isoutside the oscillator field, there is a weak

capacitive coupling. When the target iswithin the oscillating field, a strong capaci-tive coupling develops. The sensing zone isinfluenced by the physical properties of themedium being sensed. All materials in thesensor’s proximity are sensed by the changein dielectric characteristics versus air.

Capacitive proximity switches lend them-selves to applications requiring the sensing of bulk and liquid level, and nonmetallicobjects such as glass and plastic. It is the last

choice for detecting metal targets because of



its non-selectivity (relative to inductive tech-nology).

A typical installation is in the side of anonmetallic vat (Figure 5). The capacitivesensor can detect the level of the liquid ordry material right through the wall, with-

out coming into contact with it.Weaknesses of capacitive sensors include:

• The actual sensing range of the switchvaries widely according to the mediumbeing sensed.

• Very sensitive to environmental factors.

• Not selective for its target — easilyfooled. Control of what comes close to thesensor is essential.

• If tank is grounded or too thick, will needa special “well bracket.”

PHOTOELECTRIC SENSING

The principles of light emission and recep-tion are used to sense the target. Figure 6shows the emitter and receiver. The emitteris a device that emits a beam of light (usuallyinfrared) when an electric current is applied.The receiver is a device that changes itsoutput state when light from the emitter is

received. There are three basic configura-tions for photoelectric sensing (Figure 7):

1. Thru-beam, where the target passesbetween an emitting unit and receiving unit, blocking the beam. These are sepa-rate, powered units.

2. Retroreflective, where the target passesbetween the sensor and a reflector. Theemitter and receiver are in the same hous-ing. The reflector establishes the beambetween emitter and receiver.

3. Proximity (diffuse), where the unit sensesthe light directly from the target. The emit-ter and receiver are in the same housing,the same as retroreflective, but the receiveris more sensitive to the weaker light whichis diffused by the surface of the target.

The types and use of photoelectrics areincreasing at a rapid rate because they areextremely versatile. They can be used todetect the presence or absence of nearly any

material or object without contacting itphysically, and can function at a consider-

FIGURE 2. A schematic showing the principles ofinductive sensing.

Metal Target

• Metal Objects

• Eddy Currents

• ECKO (EddyCurrent KilledOscillator)

Magnitude

ofOscillations

OutputVoltage ofInterfaceCircuit

Output

On

Operating Level Releasing Level

Off Off

FIGURE 1. Reed relays are used to sense magnetictargets. The graphic shows a sensing pattern — not

a magnetic field.

Output

Signal

SN

N

S

FIGURE 3. An example of tubular and rectangularinductive sensors.

Tubular Rectangular

• Sn = Nominal Sensing Distance = 0-2"

FIGURE 4. Shielded and nonshielded inductive sen-sors.

MetalShield

Metal

Field Field

OnThreshold

OnThreshold

OffThreshold

OffThreshold

Target Target

Metal

SHIELDEDNONSHIELDED



able distance from the target.Photoelectronics are reliable and have rapidresponse times because they employ a solid-state design and no-touch actuation. Thisresults in long, low-cost operation.

GENERAL PHOTOELECTRIC

TERMINOLOGIESRetroreflector: A corner cube reflector doessomething that a mirror cannot do — itreflects energy back to the source fromwhich it came when the source is not in aperpendicular direction (Figure 8). That’swhy similar kinds of reflectors are used oncars and road signs to provide a passivelylighted warning at night.

Compare the diagrams (Figure 8) andnotice how mirrors differ from reflectors.

The mirror reflects light such that theangle of reflection is equal to the angle of incidence. The reflector sends light backparallel to the rays of incidence.

You might wonder if a mirror could beused with an angle of incidence of 90°, butyou would find this only workable in thelaboratory, and only then after much care-ful alignment. A good reflector sends lightback to the sensor even at a 45° angle. Thecorner cube reflector makes retroreflective

photoelectric sensing practical.

Reflectors are available in different sizesand forms to do different jobs. For specialsituations, there are reflective tapes.

Sensing Range: With respect to the threebasic modes of photoelectric scanning, sens-ing range is the distance between an emit-ter and a receiver, an emitter and reflector,or emitter and an object in the path of thebeam within which nominal operation isachievable. Note that this does not take

into account environmental factors orvariations in the nature of the target.

When relating sensing distance to pho-toelectric proximity sensors, certain refer-ence standards for reflectivity are used,such as 90% white Kodak paper for thestandard target. A 3" reflector is used forretroreflective sensing mode.

Usable Sensing Distance: Taking the ac-tual operating conditions into account, the

usable sensing distance is the maximumreliable operating range for a given system.

FIGURE 5. Using a capacitive sensor in an isolatedand a grounded tank application.

• “All Materials” type of sensor• Typical sensing range: 0-3/4"• Typical Application: Level Sensing

Isolating CapacitiveSensor in GroundedTank Application

FIGURE 6. Photoelectric emitter and receiver.

Emitter

Receiver

E

R

FIGURE 7. Types of photoelectric sensing.

Target

Target

ReflectorTarget

Proximity (diffuse)

Retroreflective

Thru-Beam

ER

RE

ER

FIGURE 8. Retroreflector principle.

Reflector(corner cube)

Mirror



FIGURE 10. Retroreflective scanning.

FIGURE 9. Thru-beam scanning.

FIGURE 11. Retroreflective polarized scanning.

FIGURE 12. Proximity (diffuse) scanning.

Detects shiny objects

Constraints:• More expensive

• Shorter distance

Typical range: 0-24 feet

Typical range: 0-3 feet

Target

Most Suitable for:• Transparent or translucent objects• Short range

Constraints:• Range depends on the color and reflective nature of objects

E

R

Target

Reflector

Target

Most Suitable for:• Opaque or reflective objects• Dirty, polluted environment• Long range• Accuracy with small objects

Constraints:• Accurate alignment• Cannot be used to detect

transparent objects• Two powered units must be installed E R

Most Suitable for:• A power supply is available from only one side• Rapid, simple installation is required• A relatively clean environment

• Packaging, conveyor systems, etc.

Not recommended for:• Shiny or reflective objects• Dirty or polluted environments• The detection of very small objects

R

E

The most important factors to consider areatmospheric environment and the reflec-tive nature of the target.

Repeatability: Unlike inductive sensors,which have repeatability that can be ex-pressed as a percentage of their usablesensing distances, photoelectrics have in-definable repeatability because there areso many external factors that can affecttheir performance. The precision of photo-electric sensing resolution is dependent onnot only the size of the beam and the size of the target, but also where the target iswithin the sensing range, the humidity,and the amount of dust in the area.

THRU-BEAM SCANNING

In thru-beam scanning, the emitter and

receiver are in separate enclosures (Figure9). Of the photoelectric scanning types, thru-beam is the most predictable, with typicalranges from 0 to 200 feet.

The system is most suitable for:

• The detection of opaque or reflective objects.

• Use within a dirty, polluted environment(dust, rain, mist). At shorter distancesthan the nominal one, the excess powerhelps “burn through” pollutants.

• Applications requiring long-range detection.• Applications where accuracy is important,

where small objects are being detected, andfor counting systems.

The constraints of thru-beam scanning in-clude:

• Accurate alignment in setup is essential.

• The target should be larger than the beam.

• Cannot be used to detect transparent objects.

• Two powered units must be installed, sothis route is more costly and requires morewiring to install.

RETROREFLECTIVE SCANNING

In retroreflective scanning, transmitter andreceiver are contained within the same enclo-sure (Figure 10); the beam is closed by areflector. Typical ranges are from 0 to 30 feet.

This system is the most suitable for:

• Applications where a power supply is avail-

able from only one side of the target.


http://slidepdf.com/reader/full/solis-state-sensors 11/43FIGURE 15. Fiber-optic scanning.

FIGURE 14. Color registration scanning.

FIGURE 13. Proximity (diffuse) background sup-pression.

• Utilizes two receivers behind the receiving lens• Distinguish a target in front of a background• Sensor is not fooled by target color

patterns or texture

Applications:

• Material handling-conveying systems• Collision detection for AGVs• Car/truck wash• Level sensing

• Identifies index marks of various colors, or the presence of labels on shiny surfaces

• Fast response times

Constraints:• Fixed distance

MarkCutter

SensingArea

ER

ER

Fiber OpticThru-beam ScanningTypical range: 0-50"

Proximity (diffuse)using a bifurcated fi-

ber optic cableTypical range: 0-50"

• Applications where rapid, simple installa-tion is required. The reflector may bemounted at an angle of up to ±15° from theline perpendicular to the light beam.

• A relatively clean environment.

It is the most common system used on

packaging equipment, conveyors, etc.

This system is not recommended for:

• Applications where shiny or reflective objects are to be detected.

• Very dirty or polluted environments.

• The detection of very small objects. Usu-ally, the target should be larger than thereflector, and very small reflectors are im-practical.

RETROREFLECTIVE POLARIZED SCANNING

A polarized photosensor is actually aspecial retroreflex system which can detect,in addition to normal opaque objects, theshiny objects that fool a normal reflex sen-sor — mirrors, metal straps, foils, metalboxes, cans, shrink wrap, and mylar tape(Figure 11). Typical ranges are from 0 to 24feet.

This type of sensor is used extensively

by the canning industry to sense aluminumcans and bottles.

Constraints of this technology include:

• More expensive than standard retroreflec-tive.

• Shorter sensing distance.

PROXIMITY (DIFFUSE) SCANNING

In proximity (diffuse) scanning, thetransmitter and receiver are contained

within the same enclosure (Figure 12).Detection is based on light from the emitterbeing diffused from a target back to thereceiver. Typical ranges are from 0 to 3 feet.

Diffused light is scattered light. We areable to see objects because when lightstrikes them, the light scatters. Our receiv-ers, our eyes, really don’t see the objectsthemselves, but rather, we see the lightdiffused by the objects. Diffused light gen-erates a very small signal, compared with

retroreflective sensing. That’s why we needvery sensitive amplifiers to deal with theweaker signal.

This system is most suitable for:

• The detection of transparent or translucentobjects such as empty bottles on a conveyor.

• Short range applications.

• Detection of small objects.

Constraints include:

• Usable range depends largely on thecolor and reflective nature of objects to bedetected. Extremely vulnerable to environ-mental factors such as dust and humidity.

• The area behind the target should beclear of reflective surfaces.



• Their use is not recommended for dirtyenvironments.

PROXIMITY (DIFFUSE)BACKGROUND SUPPRESSION

Diffuse proximity sensors with back-

ground suppression were designed to distin-guish a target in front of a background, but aside benefit of the way background suppres-sion works is that the sensor is not fooled bytarget color patterns and texture (Figure 13).Typical ranges are from 0 to 6 feet.

One implementation of background sup-pression utilizes 2 receivers behind thereceiving lens. Just as our eyes provide uswith stereo vision by giving our brain 2images to compare, the background sup-

pression sensor relies on 2 receivers thatare adjusted so that they sense the pres-

Solid-State Sensor TechnologiesAdvantage Disadvantages Applications

Magnet Operated (reed relay)

Hall Effect

Ultrasonic

Inductive

Capacitive

Photoelectric

Security and safetyinterlocking.

Sensing throughmetal.

Keyboard.

Anti-collision on AGV (AutomatedGuided Vehicles).

Doors.

Presence detectionon all kinds of ma-chines and installa-tions.

Level sensing withliquids and non-me-tallic parts.

Parts detection. Material handling. Packaging.

Magnet required. Sensitive to welding fields.

Magnetic targetonly.

Extremely sensitiveto industrial envi-ronment.

Resolution. Repeatability; sen-

sitive to background

and environmentchanges.

Distance limitationto 60 mm.

Very sensitive to en-vironment changes.

Subject tocontamination.

ence of a target when the output of bothreceivers is equal.

Since equality of reflected light ampli-tude is the only characteristic sensed, itmakes no difference whether white, grey,or any other color is present — or whetherit fluctuates in amplitude as it would whenencountering a textured surface or letter-ing or graphics on a container.

Applications:

• Material handling-conveying systems.• Collision detection for AGVs (Automated

Guided Vehicles).• Car/truck wash.• Level sensing.

Inexpensive;very selective targetidentification.

Complete switching function is in a singleintegrated circuit.

Operates up to150 kHz.

High temperature(150 °C).

Good resolution.

Senses allmaterials.

Resistant to harshenvironments.

Easy to install;very predictable.

Senses allmaterials.

Dectects throughwalls.

Senses all materials. Longest sensing

range. Versatile.



COLOR REGISTRATION

Color registration is used in the packag-ing and material control industry to iden-tify index marks of various colors, or thepresence of labels on shiny surfaces (Figure14). These units are characterized by fast

response times and are capable of detecting small marks that pass the unit at extremespeeds. A demanding application might bethe detection of a 0.03" mark that passesthe unit at 9 feet per second.

The color registration sensor is a highlyspecialized diffuse proximity sensor whichhas the ability to detect fine changes incontrast on a surface. But unlike the standarddiffuse proximity sensor, this type of unit usesa powerful lens system and must be posi-tioned at a specific focal distance from thetarget.

Constraints of color registration scanning:

• Fixed distance.

FIBER-OPTIC SENSORS

What do you do when the physical con-straints of the application don’t allow forinstalling regular, self-contained sensors?Maybe the target is in a high temperature orchemically aggressive environment. Per-

haps the target is small or very fast-moving.Fiber-optics, applied to photoelectric scan-ning, solves these problems (Figure 15).

The principle of conducting light aroundcorners was discovered more than 100years ago. However, the practical use of thisprinciple is relatively new. It was acceler-ated by the development of communica-tions, especially the need for noise-free datatransmission. Medical imaging systemsalso played an important role in expanding

the technology to the point of making itaffordable for industrial and commercialuse.

A fiber-optic cable is made up of a bundleof hundreds, and sometimes thousands, of glass or plastic fibers which are protectedby a flexible armored sheathing. You canthink of these as being analogous to con-ducting electricity using electric cables,only fiber-optic cables conduct light. Allfiber-optic sensing modes are implemented

using one type of amplifier which containsboth emitter and receiver in one housing.

Using two opposed, individual fiber-op-tic cables, the object to be detected breaksthe beam. The target must be at least thesame dimension as the effective beam,which in this case is the bundle diameter.Because the beam is small, the detectioncan be precise. A typical application might

be edge detection for a web printing press.Needle tips reduce the beam dimension foruse with extremely small targets, typicalfor applications in semiconductors andpharmaceutical industries. Typical rangeof this type of sensor is 0 to 50 inches.

FIBER-OPTICS APPLICATIONS

When you mix the emitter bundle withthe receiver bundle at the sensing tip, theresult is a bifurcated cable. The setup israther straight forward: the sensing tipmust be placed close enough to the targetto directly sense diffuse light.

This mode of scanning is the least precisebecause the properties of the target surface(color, shining, texture, etc.) greatly influ-ence the quantity of light bounced back.However, it is the mode of choice for trans-parent targets. Typical range is 0 to 5 inches.

Typical applications:

• Small parts detection.

• Very fast moving targets.• Edge detection.• High temperature environment (600°F+).



14

FIGURE 1. Schematic showing the principles of inductivesensing.

Inductive sensors were developed as anoncontact alternative to the tried and truelimit switch. Because limit switches requirephysical contact, they are prone to failurethrough wear, fatigue and accidental jamming.They also have mechanical contacts which canexhibit bouncing effects that PLCs don’t like.

The fact that inductive sensors are solid-state,and have no moving parts or bouncing effects,make them especially compatible with PLCs.

Inductive sensing technology was developedin what used to be West Germany and intro-duced to the U.S. marketplace about 20 yearsago. To the credit of the technology’s develop-ers, inductive sensing is one of the best stan-

dardized technologies. This has helped themarketplace gain confidence in specifying thisalternative to the limit switch.

INDUCTIVE PRINCIPLES

An inductive detector senses the proximity of

a metal object using an oscillator principle.Surface currents, known as eddy currents, areinduced in the metal target, and they react withthe oscillator’s field — the closer the proximity,the stronger the reaction.

When the metal target is outside of theoscillator field, the magnitude of the oscilla-tions is not impeded. When the target is insidethe field, it attenuates the magnitude of theoscillations, eventually stopping the oscillator.That’s why it’s named ECKO (Eddy CurrentKilled Oscillator). Figure 1 shows how the

proximity of a metal target affects the magni-tude of oscillations, the output voltage of theinterface and of the output. The graphic showshow the threshold level detector (trigger cir-cuit) shapes the output signal of the sensorinto a clear-cut ON-OFF (here/not here) type of signal. If you look closely, you will notice thatthe ON point (operating level) is a differentlevel than the OFF point (releasing level). Thedifference between these points is called differ-ential or hysteresis.

Metal Target

Magnitude ofOscillations

Output Voltageof Interface Cir-cuit

Operating level Releasing level

On

OffOff

CHAPTER 3

Inductive Sensors



15

SENSING DISTANCE

Figure 2 shows a sensor and the field around it.This field is not an indication of sensing distance.The field is more intense around the sensor andless intense at its perimeter. (The perimeter P

indicates the field limits within which the target-sensor interaction has a measurable effect.) If thiswere an analog system, the presence of metalwould be detected at a weak level at the fringesand a strong level near the sensor.

If you introduce the standard target atdifferent distances from the sensor, you willfind that as the target gets farther away fromthe sensor, more metal has to enter the field totrigger the sensor ON. Finally, if the targetenters the field at a certain distance from thesensor, there isn’t enough metal in the field to

trigger the sensor (Figure 3). If you plot allthese points of entry that cause ON, you’ll geta sensing distance curve that looks like Figure4.

As we showed in Figure 1, a discrete systemhas a differential. That means that when wemove the target back from the ON position, thesensor remains ON for the distance of thedifferential. We plot the OFF curve by intro-ducing the target to the sensor until the ONpoint is triggered, then backing it away untilthe sensor triggers OFF (Figure 5).

The sensing pattern is 3-dimensional. Itmight help you to think of the shape as being like a brandy snifter turned upside-down. Thecone-shaped area enclosed by the intersectionof the sensing curves describes a sensing areathat is used most in practical applications(Figure 6).

Figure 7 shows the sensing pattern graph fortypical shielded and non-shielded inductive sen-sors. The horizontal axis gives the radius of thesensor from center, and the curves are eachmarked with the sensor diameter.

CALCULATING USABLESENSING DISTANCE

There are 4 factors that will affect theperformance of inductive sensors:

1. Supply Voltage2. Ambient Temperature

3. Target Material

4. Target Dimension

Sn = Nominal Sensing Distance

The rated operating distance for which the switchis designed, obtained using a standard target.This value should only be taken as a guide, sinceno manufacturing tolerances or changes in volt-age or temperature are taken into account. Itserves as a base for determining the usable rangeand for comparison between various designs andmakes.

S = Usable Sensing Distance(0.81 S

n < S < 1.21 S

n)

Sensing distance S is an application designparameter used for selecting a specific sensor foran application. Relative to a standard target, theusable sensing distance takes manufacturing tolerances and external operating conditions

into account. It is derived by testing with astandard metal target, during extremes of volt-age and temperature variations. Generally 80%of the nominal sensing distance.

Effective (or REAL) Sensing Distance

This is the actual sensing distance realized bythe actual sensor that you take out of the boxand install in the actual application. The effec-tive sensing distance will be no less than theusable, and is usually closer to the nominal inaverage circumstances.

Standard target

Used in the product laboratory to determinebenchmark performance characteristics for asensor. For inductive sensors, the standard tar-get is mild steel that is 1mm in thickness andeither:

1. A square with sides equal to the diameter of the sensing head for round sensors.

2. A square which is 3 times the nominal sensing distance for rectangular sensors.

Target

The actual object that needs to be detected by

the proximity sensor. In real life, targets maydiffer from the standard, both in dimensions andin material. As a general rule, you should selecta sensor with dimensions compatible with yourtarget or smaller. If your target is small or madeof a metal other than mild steel, you shouldcarefully calculate the usable sensing distance,rather than rely on the 80% of nominal rule of thumb.

GENERAL TERMINOLOGY



16

FIGURE 7. Sensing patterns for typical shielded/non-shielded inductive sensors.

FIGURE 6. The inductive sensing pattern is 3-dimensional.

FIGURE 4. Plot of asensing distance curve.

FIGURE 5. Plot of the“off” curve showing dif-ferential.

FIGURE 3. Notice thatthe top target does nottrigger the sensor, yet iscompletely in the field.

FIGURE 2. Inductivesensors create a field tosense a target.

Off

On

On

On

Off

Off

2-D Graph

— On

... Off

— On

... Off

30 mm dia. shielded

18 mm dia. non-shielded

12 mm dia.

12 mm dia. non-shielded

Drop-out points

S e n s i n g D i s t a n c e (

m m )

0 16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 160

2

4

8

10

16

1214

6

To determine the usable sensing distance,apply this formula:

S = Sn • K

t • K

q • K

m • K

d

where:

Sn = Nominal sensing distance. The rated oper-

ating distance for which the switch is de-signed, obtained using a standard target.

Kt = Supply voltage coefficient. Use 0.9, which

is the standard correction coefficient. Thistakes into consideration the maximum volt-age variation within the rated voltagerange.

Kq = Ambient temperature coefficient. The stan-

dard correction coefficient is 0.9, which isthe worst case within the whole rated tem-perature range. Figure 8 shows a curve ofK

q versus temperature. If your installation

will be indoors and you need to squeeze

maximum performance out of the sensor,the most conservative scenario might notbe necessary to achieve full product confi-dence. As an example, for a temperaturerange of between 0˚C and 50˚C, a variationof no more than 3% might be adequate (K

q

= 0.97).

Km = Target material coefficient. Use data from

Table 1 for mild steel, stainless, brass,aluminum, or copper. Mild steel is thestandard target material, which has a coef-ficient of 1. Mild steel has the best charac-

teristics for inductive sensing.

Kd = Target dimension coefficient. As a rule, you

should select a sensor having comparabledimensions to that of your target. If your realtarget is smaller than the standard target, usedata from Figure 9 to determine the correc-tion coefficient. If the target is greater than orequal to the standard target dimension, use acoefficient of 1.

Most of the typical applications use targetsthat can be considered standard. Therefore,

the usable sensing range (0.8 of nominal sens-ing distance) covers the maximum variationdue to temperature (10%) and power supply(10%).

Example: What if the real target is mild steelbut the dimensions are smaller than the stan-dard target?

For a 120 mm rectangular sensor, a stan-dard target is 120 x 120 mm. What if the real

Pick-up pointsPick-up points

Sensor Head Size (mm)Sensor Head Size (mm)

.

. . . . ... .

.. .. .

.

..

.... ...

..



17

Mild Stainless Brass Aluminum

Material steel steel UZ33 AU4G Coppe

Km

1 0.35-0.65 0.38 0.34 0.27

target is 60 x 60 mm? Following Figure 9,locate 60 mm for a 120 mm sensor, and you willfind that it is in column C. If you look up thecoefficient for target dimension on the graph,you will find the value is K

d = 0.9. That means

that the sensing distance for your real targetshould be 10% smaller than for the standardtarget. Therefore, from a nominal sensing dis-

tance of 40 mm, we find a usable sensing distance of 32 mm (.8 times the nominal sens-ing distance) and an effective sensing distanceof 0.9 x 32mm = 28.8 mm.

Example: We need to sense a large aluminumtarget at a distance of 4 mm. To arrive at thenominal, we divide 4 mm by .8 because theusable sensing distance is 80% of the nominal.This makes the nominal sensing distance (S

n)

equal to 5 mm. But, since the target materialcoefficient (K

m) equals .34 (see Table 1):

Sn = 5/.34 = 14.7

Because we need a sensor with an Sn that’s

greater than or equal to 14.7 mm, we choose asensor with the next standard nominal sensing distance. Conclusion: a 15 mm nominal sensing distance is required for sensing an aluminumtarget at a maximum of 4 mm.

INFLUENCE OF TARGET THICKNESS

Because the eddy currents are of the surfacetype (sometimes called “skin currents”), if thethickness of the target is lower than the cur-

rent penetration, an interesting phenomenonhappens — the sensing distance increasessharply — several times the norm. The param-eters depend on the nature of the material.

Aluminum foil 0.01 mm thick results in thesame sensing distance as mild steel. In applica-tions where inductive sensing would normally beimpossible due to the nature of the target, a thincoating of metal paint on the target can increasethe usable sensing distance.

Suppose you wanted to detect the fluid level

of the contents of a plastic vat. Mount thesensor to the wall, paint a ping-pong ball withaluminum paint or foil, mount a plastic con-duit to the side of the tank, and float the ping-pong ball in the pipe. The sensor will detect theping-pong ball through the wall as if the ballwere made of steel.

Surface treatments (platings) can improvethe properties of otherwise poor sensor materi-als. For example:

• A .2 mil thick plating of nickel on 1020 steelincreases sensing distance by 1.6%.

• A .25 mil thick plating of tin on 1020 steelincreases sensing distance by 4.2%.

REPEATABILITY

Repeatability is the precision of the sensor

in its ability to repeatedly detect a target at agiven point, taking into account a set of stan-dardized operating conditions as environmen-tal variables. Repeatability is expressed as apercentage of the effective sensing distance,such as: 8 hours, T = 20˚C, repeatability = ±5%of S

n, worst case.

Example: A 120 x 120 mm sensor has an Sn of

40 mm. Its usable sensing distance is 32 mm.Now let’s take a sensor out of the box and put itin the application, and we find that it has aneffective sensing distance of 32 mm. It would

have a repeatability of 5% of 32 mm = 1.6 mm,so every time the target approaches the sensorfrontally, the sensor triggers at a distance of anywhere between 30.4 and 33.6 mm, depend-ing on the real temperature and power supplyvalue.

A side approach closer to the sensing sur-face is recommended when more repeatabilityis desired. At 5 mm from the sensor moving sideways, the same target will be detectedwithin ±0.015 mm of the effective distance.

Not all applications require a high level of repeatability.

Example: You are filling an aluminum can thatmust have a positioning accuracy of ±1/16 inchat the filling station, otherwise you have prod-uct spilling on the floor. Does your choice of sensor have the repeatability to achieve this allday long, despite changes in supply voltage andtemperature?

TABLE 1 — TARGET MATERIAL COEFFICIENT Km



18

FIGURE 8. Graph of ambient temperature coefficient (Kq)

versus temperature.

Temperature °C

Kq

1.1

0.9

−25 0 20 50 70

SHIELDED AND NON-SHIELDEDINDUCTIVESENSORS

A sensor has a field which extends aroundthe front surface to the sides of the unit. If metal is brought into the field without trigger-ing the output, the sensing distance and patternof the sensor will not be predictable. You will

recall that the amount of metal in the field iswhat causes the sensor to trigger, and if there isalready metal in the field before the targetenters, the field is preloaded, causing the targetto be sensed prematurely. Common situationsfor having a preloaded field are when the sensoris mounted in metal or when metal is in thebackground.

Because most of the time the ability tomount a sensor in metal is desirable, we needto cancel the influence of the metal preloading the field. If we introduced metal into the field

in a controlled manner (by incorporating it intothe sensor itself), the sensing pattern becomespredictable. This sensor can now be flushmounted. This type of sensor is called shielded,as opposed to non-shielded. You should notethat the shielding that makes flush mounting

possible also reduces the sensing distance.Plastic sensors can also be shielded. As withsensors having metal bodies, the shielding ison the inside of the housing.

MOUNTING RECOMMENDATIONS — TUBU-LAR SENSORS

To install a non-shielded sensor in a metalenvironment, an area 3 times the diametermust be cleared and recessed to a depth of twice the nominal sensing distance (Figure10). An L-bracket is commonly used for mount-ing. However, it is recommended that me-chanical stops be installed to protect the unitfrom physical damage. Mounting the sensor inC-channel can also protect the sensor, but besure that field clearances are observed.

Compare the ease of installation for theshielded sensor (Figure 11) to the non-shielded.

Flush mounted sensors have a better chance of surviving mechanical shocks because the fix-ture protects the unit. If the usable sensing distance of the shielded sensor suits the appli-cation, it presents the least amount of mount-ing considerations.

MOUNTING RECOMMENDATIONS — RECT-ANGULAR SENSORS

Shielded models can be mounted on metalplates without any problem. The non-shieldedsensor (Figure 12) can be mounted directlyfrom the side to a nonmetallic surface, but it

must be kept at least 1.5 times its width fromother metal objects or fixtures on two of thethree other sides (Figure 12A). If it will besurrounded by metal on three sides, a distanceof twice the sensor’s width must be maintained

FIGURE 9. Determining target dimension coefficient.

A B C D0

0.4

0.6

0.8

0.97

0.2

1

Kd

Dia. (tubular)A B C D

4mm .8mm 1.3mm 1.8mm 2.8mm

5mm .9 1.6 2.2 3.4

6.5mm 1.2 2.0 2.8 4.7

12mm 3 5 7 10Size



19

MUTUAL INTERFERENCE

If 2 sensors are positioned too closely, thefrequency of one mixes into the detecting coilof the other causing output to latch in the ONposition without a target being present. Thiscan be overcome by either maintaining therecommended side-by-side spacing, or some

models have the ability to have their oscillat-ing frequency altered at the factory. The latterare referred to as “alternating frequency” mod-els and can be mounted side-by-side withoutany problem.

With regular sensors, the way to guard againstmutual interference for side-by-side mounting isto maintain an interval twice the nominal sensing

(Figure 12B).

The non-shielded sensor can be directlymounted from its back to a metal surface.Metal must be kept away from the front of thesensor a distance of three times the nominalsensing distance (Figure 12C). If the sensor isto be mounted near metal on one side, it

cannot be mounted any closer than 1.5 timesthe sensor’s width. If the sensor must berecess-mounted in metal, a clear area of twicethe width of the sensor must be maintained(Figure 12D).

FIGURE 10. Mounting recommendations for non-shielded sensors.

FIGURE 11. Mounting recommendations forshielded sensors.

FIGURE 12. Non-shielded rectangular sensor mounting.

Metal

Metal

Metal

Metal

Metal

Metal Metal

Sensing face

Sensing faceSensing face

Sensing face

1.5a

3Sn

3Sn

2a2a2a2aa

1.5a

1.5a

2a

Non-Shielded

Non-Metal

A

B

C

D

aa

Non-MetalNon-Metal



20

distance. If the fields oppose each other, theymust be kept apart by a factor of ten times theirnominal sensing distance.

WELD FIELD IMMUNE (WFI)

The electric welding process generates astrong electrical field near the welding points.

This field can interfere with the normal opera-tion of an inductive proximity sensor situatednearby, creating false output signals. Normalinductive proximity sensors should not beused for such applications. Specially designedsensors, called weld field immune, can be usedas close as 1 inch from the welding point.

RADIO FREQUENCY IMMUNE (RFI)

Radio frequency fields also can induce falsesignals. Typical examples are walkie-talkies —used quite frequently on the factory floor. Spe-cial techniques are used to create radio fre-quency immunity. Today RFI sensors can actu-ally be touched with the walkie-talkie antenna

while it transmits without changing the outputof the proximity switch.

VARIABLE SENSING DISTANCE

Few inductive sensors are available with anadjustable sensing range. This usually undesir-able feature is sometimes recommended forsituations where you have metal in the back-ground of long range sensors. The sensitivityadjustment should be decreased to reduce thebackground influence and make target detec-tion look “normal.”

APPLICATIONS

Figures 13-15 illustrate some common ap-plications of inductive sensors.

FIGURE 13. Monitoring the presence of bottle tops.

FIGURE 14. Speed monitoring.

FIGURE 15. Control of a hydraulic actuator.











25

CHAPTER 5

Interfacing Sensors

With Loads

The two most common types of loads thatwe will discuss are resistive (e.g., lamps,

heaters, solid-state relays and PLC inputmodules) and inductive (e.g., relays and sole-noid valves).

RESISTIVE LOADS

Common resistive loads do not pose anyproblems interfacing sensors as long as thecurrent is within the sensor specified range.

Resistive loads that warm up significantlywhen activated, such as lights and heaters,present a problem. Their cold-state resistancecan be 10 times lower than their hot-stateresistance, causing high current levels whenturning ON. So high, in fact, that they cancause sensor failure or, for sensors soequipped, short-circuit protection circuits toengage. A resistor wired in parallel with the

switch keeps the load warm, raising the re-sistance to levels that are well within thespecifications of the switch (Figure 1) but lowenough not to cause visible effects.

PLC INPUT MODULES -TWO-WIRE AC

PLC (Programmable Logic Controller) in-put modules are another example of a resis-tive load. In principle, they require little en-ergy; their internal requirements are in the

low milliamps. You can think of the role of PLCinput modules as "signal switching" rather than"power switching." The current for input mod-ules was made artificially higher than required,in the range of 12 to 16 milliamps, because themodules originally were designed for electrome-chanical switches with silver contacts. Thesecontacts develop an oxide film on their surface,

but arcing cleans this film away. To promotearcing, the current to be interrupted must be aminimum level - 15 mA at 24 V or higher. Thisis an established industry standard.

PLC input modules have been designed toprovide a wide latitude for signal switching.They are ideal loads for sensors because theydo not require much current from the sensor.This is why two-wire technology started to gainacceptance.

In order for a two-wire AC switch to be di-

rectly compatible with a PLC, two conditionsmust be met:

1. PLC Requirement: Leakage current must beless than 1.7 mA when the switch is off (rule of thumb). This level is below the standard PLCthreshold for OFF state. If the leakage currentof the sensor is greater than 1.7 mA, the PLCwill be ON all the time, regardless of the condi-tion of the switch.



FIGURE 1. Wiring resistive loads to correct for coldstate.

FIGURE 2. Wire bleeder resistor in parallel withload.

Vline

Rb R

in

PLCInputModule

Switch

orSensor

Load

2. Switch Requirement: Load current must begreater than the switch minimum rated valuewhen the switch is ON. Rule of thumb: if theswitch minimum current is above 20 mA, it willnot work with most PLCs.

A method to satisfy these requirements is touse a bleeder resistor in parallel with the load(Figure 2). You dimension the bleeder in such away that, when the switch circuit is closed, itwill make up the difference between the mini-

mum current required by the switch and thecurrent required by the PLC input module(switch requirement).

When the switch's minimum load current isgreater than the PLC's load current during theON state, the bleeder resistor is dimensionedusing the following formula:

Ibleeder min

= Iswitch min

- Iinput PLC

When the sensor's leakage current is greaterthan 1.7 mA during the OFF state (PLC require-ment), the bleeder resistor must be dimensionedusing the following formula:

Typical values for PLC input resistance arebetween 7.5 to 10 kΩ. NOTE: The value shouldbe kept close to the maximum value determinedby the formula in order to minimize heat dissi-pation. The smaller the resistor, the higher thedissipated power (power = heat).

Rbleeder max

=V

line

Ibleeder min

where:

Rin

= the input resistance of the PLC module orany input of a similar solid-state device.

Vo max

= the PLC maximum OFF voltage (usually 20 -40 V AC).

Vs

= the power supply voltage.

Pb

= the minimum power rating of the bleederresistor. It needs to be calculated in order toavoid overheating and self destruction.

Ioff

= the leakage current of the actual sensor.

PLC INPUT MODULES -THREE-WIRE DC

PLC input circuits are identical for sinking or sourcing input modules. The difference is thecommon wire. To reduce wiring and the num-ber of connectors:

• Sinking PLC input modules have all thenegative terminals internally bussed toone terminal.

Pb min

=(V

s)2

Rbleeder max

Rbleeder max

=R

in • V

o max

(Ioff

• Rn) - V

s

26



FIGURE 3. Connecting sensors to a PLC usingthree-wire DC.

FIGURE 4. Electromechanical switches create tran-sients.

SourcingSensors

SourcingPLC

SinkingPLC

+

−

SinkingSensors

• Sourcing PLC input modules have allpositive terminals internally bussed to oneterminal.

To complete the circuit, sensor must becomplementary to the PLC. By complementarywe mean that a sinking PLC input module musthave sourcing sensors and a sourcing PLC in-

put module must have sinking sensors. The in-dustry is pretty well established in the use of the terms sinking and sourcing, but there are

exception. Be aware of this and check the PLCwiring diagram. The industry is in the processof standardizing these concepts to avoid anyconfusion.

INDUCTIVE LOADS (AC)

When inductive loads conduct, they accumu-late energy. When these circuits are interrupted,the accumulated energy must be dissipated.Electromechanical switches typically arc as theyare opened with inductive loads; the arc dissi-pates the energy (Figure 4). When solid-state

devices are not used, the transients generatedby this process are not a problem, even thougharcing will degenerate the contacts of electro-mechanical switches. However, these transientscan destroy solid-state devices in an instant.

Because AC solid-state switches (TRIACs andSCRs) stop current at a zero crossing - when noenergy is present - there is no energy to dissi-pate when they interrupt an inductive circuit.Therefore, AC solid-state switches do not gen-

erate transients. This explains why there is noproblem using only solid-state switches for ACinductive loads.

The only problem is when both solid state andelectromechanical devices are used to switch theload (Figure 4). The solution is to dissipatesurges in the line caused by the electromechani-cal contact when interrupting the current. AnMOV (metal oxide varistor) or similar device isput in parallel with the load (Figure 5) to ab-

sorb the energy surge. The MOV is a variableresistor which changes its value in the presenceof increased voltage. The higher the voltage,the lower the MOV's resistance value. Whenthe MOV "sees" the increased voltage, it opensup and diverts the surge. This not only pro-tects the solid-state device, but substantiallyreduces arcing in the electromechanical device,which also prolongs its service life.

INDUCTIVE LOADS (DC)

In DC, when the inductive circuit is opened,

the polarity reverses until the energy is dissi-pated. This results in an arc across the electro-mechanical switch, or an overvoltage across thesolid-state ones. The overvoltage across ablocked (OFF) solid-state switch tends to per-manently damage the blocking capability of theswitch in the OFF state. Most solid-state sen-sors today have an internal mechanism in par-allel with the switch that opens only in thesecircumstances and directs the coil energy safelyto the power supply.

−

+

Coil

27



28

FIGURE 6. Use of a flywheel diode to protect solid-state switches.

FIGURE 5. MOV surge suppressor protects bothsolid-state and electromechanical switches.

Coil

FlywheelDiode

DiodeCharacteristics

I

V0.6V

+

+

−

−

To protect solid-state switches without thisprotection, a flywheel diode is wired betweenthe coil and the sensor (Figure 6). When thepolarity changes as the circuit is broken, theenergy is safely dissipated through the diodeback to the negative terminal. Fast diodes arerecommended for this application.

WIRING IN SERIES

Normally-open sensors in series with the load

perform the AND function, while normally-closed sensors perform the OR function.

The maximum number of sensors in series isrestricted by supply voltage, the voltage dropacross each sensor, and the minimum operat-ing voltage of the load. When you have a num-ber of switches wired in series, their combinedvoltage drop might reduce the available loadvoltage to a level that is too low to activate theload.

We need to provide enough voltage to the loadto energize it. To determine the maximum num-ber of sensors we can wire in series, use the for-mula:

For two-wire types of sensors another majorrequirement is to provide enough supply volt-age for each of the sensors wired in series. Theyneed to work properly when in the OFF state.Divide the minimum line voltage by the num-ber of sensors you intend to use. The resultshould be within the sensor supply operating range.

Figure 7 shows an example of two sensorsused in series. Suppose the line voltage is 110

V AC, and each sensor requires 93 to 134 V AC.If 100 V AC is the minimum line voltage, theavailable voltage to power each sensor (100 V / 2 = 50 V) is insufficient.

In the example of Figure 8, we use extendedrange sensors, each having an operating rangeof 20 - 264 V AC. Now 50 V AC is enough topower each of these two sensors. In fact, up tofour extended range sensors can be used in se-ries with this load.

WIRING IN SERIES - THREE-WIRE TYPE

Solid-state devices have consumption cur-rents that are a concern. In Figure 9, noticethat in addition to the load current, sensor num-ber one handles the consumption currents forthe rest of the sensors.

Notice that when sensor one is OFF, the othertwo sensors have no power and are not able todetect a target. When the first sensor comesON, the next sensor in line must first "warmup" before it can detect a target. Since it is also

Number or sensors = - 1Supply Voltage - Minimum Operating Voltage of Load

Voltage Drop Across Each Sensor

Coil



29

FIGURE 7. Narrow range sensor application inseries (OFF state).

FIGURE 8. Extended range sensors wired in series(OFF state).

110VAC

(100Vmin

)

50V is less than 93V;

sensors don’t work.

110VAC

(100Vmin

)

a normally-open switch, the third sensor has notreceived any power yet, and its power-up timemust be added in evaluating the total responsetime.

WIRING IN PARALLEL - TWO-WIRE TYPE

When sensors are wired in parallel, normally-open switches perform the OR logic function,and normally-closed switches perform the ANDfunction (Figure 10).

The maximum number of sensors is restrictedby the minimum ON current of the load - thesum of the leakage currents must be less thanthe minimum ON current of the load.

WIRING IN PARALLEL - THREE-WIRE TYPE

In the example (Figure 11) of three-wire DCtechnology, each of the sensors is now fully acti-vated without the need for power-up time.

There are practically no restrictions to thenumber of devices used in parallel because theleakage current is relatively small. However,when a large number of sensors is wired in par-allel, check the sum of the leakage currents

against the load's maximum OFF threshold -don't exceed it.

WIRING ADVICE AND SHORT-CIRCUIT PRO-TECTION

Don't try to economize by running powercables with control cables, just because it mightbe easier. Control cables have very low energysignals that should be shielded from the fieldsof power cables. Power cables can induce tran-sients just by being close to the control cablesbecause the long run of wire acts like a big an-tenna. Always use separate cable runs. Yourextra effort in setup will reward you with asmooth running system.

Do not connect an AC switch directly to asupply source. Doing so will destroy the switchunless a short circuit protection option is used.

A suitable load must be wired in series with theswitch - even for testing.

When one sensor is ON, it short circuits therest of the sensors. The response time of the

sensor two through n is affected by their power-up delay. For instance, assume sensor 1 is ON.Sensor 2 "sees" a target but is not working be-cause sensor 1 practically short circuits it. Whensensor 1 turns OFF, sensor 2 doesn't turn ONimmediately because it was just powered-up.Only after its power-up delay will it respond tothe target presence.

Number or sensors = - 1

Holding Current of the Load

Leakage Current

* Caution for multiple sensors applications: Whenever possible, use one sensor in load circuits. Load circuits with multiple sensors

are more difficult to design and troubleshoot. Try to avoid mixing solid-state and electromechanical switches in the same circuit.

Coil

50V50V

93-134VAC 93-134VAC

50V is greater than 20V;

sensors work.

Coil

20-264VAC 20-264VAC

50V50V



30

FIGURE 9. Three-wire sensors wired in series. FIGURE 10. Wiring sensors in parallel.

+

−

+ −

Probably the most common situation for ashort circuit is during the installation of thewires into screw-type terminal blocks. Straywire strands can touch other terminals, by-pass-ing the load, and then your sensor is destroyed- unless you have short-circuit protection.

Short circuit protection (SCP) is a practicalfeature for switches to have in industrial envi-ronments. Once triggered, it blocks the current,effectively protecting the switch from destruc-

tion.

Short circuit protection (SCP) should not beused with loads such as lamps and heaters.When these loads are cold, they will trigger theSCP electronics.

There are two basic types of short circuit pro-tection.

Automatic Reset: When the threshold isreached, the power is shut off, and the SCP elec-tronics sample the condition of the power sup-ply at short intervals to see if the short is stillthere. When it samples, it only closes the cir-cuit for 2 or 3 µs, which is too short a time tocause any damage to the solid-state circuitry.When the sample shows that the circuit is backto normal, the SCP electronics automaticallyrest, and the device operates normally. Auto-matic reset is only a DC feature because ACswitches cannot be closed by a control signal insuch a short time. Automatic reset is commonin DC devices. Many users consider it conve-

nient because it does not react to common tran-sients in a visible manner.

Manual Reset: When the threshold isreached, the power has to be shut down. Some-one must then correct the problem, power upand manually reset the device before it will op-erate normally again. For safety reasons, themanual reset is considered the best because itforces the operator to check the source of theproblem. It is primarily found in AC circuits,

but some situations warrant the added safetyof manual reset in DC applications.

DC POWER SUPPLIES

The mechanics of converting an AC power sup-ply to DC involves a few steps (Figure 12). First,a transformer steps 110 or 220 volts down totypical levels for DC applications (V

e). As a gen-

eral rule, a transformer is selected having alower secondary voltage than the required DCvoltage - i.e., 18 V transformer to obtain 24 V, or36 V transformer to obtain 48 V.

Next, a diode bridge rectifies the AC current. At this point, the power supply is DC, but watchthe instant values. At each half cycle the volt-age goes down to 0 V. A sensor rated as being able to operate on between 10 to 30 VDC willnot work properly because the supply voltagegoes under its absolute minimum: 10 V.

To correct this, a smoothing capacitor filter isadded to achieve V

s. During the peak, the ca-

Load

I leakage

I leakage 1S1

Sn

I leakage 2

Load



31

FIGURE 11. Three-wire sensors used in parallel. FIGURE 12. Steps involved in converting AC to DC.

Voltage

TransformerAC

50-60 Hz

C

Reg

pacitor stores energy like a battery. When thecurrent falls, the capacitor discharges and sus-tains the current flow. The result is an averagecurrent, well within the operating range of thesensor. The fluctuation in the curve Vs is called"ripple." The ripple must be within the sensor'spower supply range. Note that some sensorsare designed to work with unfiltered power sup-plies. These sensors use built-in smoothing ca-pacitors.

If you have to filter your DC power supply, youcan select the correct smoothing capacitor byfiguring 400µF minimum for one switch, or 2000µF per ampere of load current.

For example, if you have three switches andthree 200 mA loads:

Remember that for three-wire sensors you mustalso add the sensor's "no load" consumption cur-

rent which powers their electronics. If you have20 switches with power supply requirements of 12 mA each, that's a total of 240 mA. The powersupply should be dimensioned for at least0.25 A plus the total load current.

3 x 200 mA = 600 mA = 0.6 A.6 A x 2000 µF = 1200 µF for the capacitor

Vs

Vsr

Vr

Ve

− Filter

+

Vs

Vsr

Vr

Ve

SnS2

L

O

A

D

S1

+

−



32

CHAPTER 6

Target Speed and

Response Time

The factor of timing is inherent in sensorsand is measured in milliseconds. If your

application is sensing the presence of an autobody on an assembly line, chances are good thatyou won’t have to worry about missing the tar-get because it will be there for a duration of min-utes. On the other hand, if you are counting the pins on an electronic component which istravelling by the sensor at 3 feet per second,you are pushing the technology to its limits.

RESPONSE TIME: “INERTIA”Response time is defined as the duration of

time required for the interface to trigger an

output. The specifications that indicate a sen-sor’s response time are power-up delay, ON de-lay and OFF delay.

You can think of response time as “inertia”(the nemesis of perpetual motion inventions).

FIGURE 1. Parasitic inductance and capacitancegenerate transients when a switch closes. Thetransient passes through the sensor, which falselysignals “ON” to the load..

FIGURE 2. Inductive, photoelectric and capacitiveproximity switch power-up delays range from 50 msto 200 ms.

LoadSensor

Vss2 - 3 times

normal

PowerSupply

TargetPresence

Output

t

t

t



33

There is no such thing as a perfect sensor - onethat responds instantly. In fact, such a sensor

would be too sensitive to be practical in indus-trial environments. Because the sensor inter-face generates a low-energy signal, the deviceis extremely susceptible to external transients.Filters are imposed between the trigger and theoutput circuit to suppress transients. However,

filters also affect overall response time perform-ance.

Terms for response time - power-up delay, ONdelay and OFF delay - are conservative ratingswhich the manufacturer guarantees as being the maximum performance for that model of device.

POWER-UP DELAY PROTECTION

Power-up delay* prevents false signals from

triggering an ON condition to the load (Figure1). Parasitic inductance and capacitance buildsup in long cables when switches are open; then

when they close they dump a transient spikethat can be two or three times the normal volt-

age. If a power-up delay was not built in, thistransient would pass through the sensor switchand signal an ON condition to the load - even if there was no target present in the sensing area.This false signal might cause dire consequences.Because this unsafe condition cannot be toler-ated, industry has required that all devices notchange state during the power-up phase. Theoutput is intentionally turned off during this

period.

You can see in Figure 2 that even when atarget is present in the sensing area when poweris applied, the output state does not change. Forconfidence, the delay must last longer than theduration of any start-up transient. Normalpower-up delays range from 50 ms to 200 ms

for the inductive, photoelectric and capacitiveproximity switches.

* Some people call power-up delay “warm-up time.” This term is a holdover from vacuum tube technology. Solid-state technology

requires no warm-up, so “power-up delay” is a more accurate term.

FIGURE 3. Example of electromechanical switchesmixed with solid-state switches.

FIGURE 4. ON delay is the time it takes for anoutput change to be triggered when a target entersthe sensing area.

MotorM

StartStop Interlock

Sensor not powered

Resistor powers switch

MotorM

TargetPresence

Output

ONDelay

t

t



34

You should avoid using electromechanical

switches in series with solid-state sensors if you

require a fast response time. When the electro-

mechanical switch is open, no power reaches the

solid-state sensor. The solid-state sensor, there-

fore, has to power-up from a completely inac-

tive state every time the electromechanical con-

tact is closed.

Let’s say that your electromechanical switch

is a push button and the sensor serves as an

interlock (Figure 3). The button must be held

down long enough for the sensor to go through

its power-up delay every time the button is

pressed. It won’t sense a target, even if the tar-

get is present in the sensing area, until the

power-up delay has been completed. The solu-

tion for this problem is to wire a resistor with a

relatively large value in parallel with the start

push button to keep the sensor “warm.”

FIGURE 5. OFF delay is the time it takes for anoutput change to be triggered when a target leavesthe sensing area.

FIGURE 6. Response time includes the entire dataacquisition system.

FIGURE 7. Critical response time (ON delay). FIGURE 8. Critical response time (OFF delay).

TargetPresence

Output

OFFDelay

t

t

Sensor

PLC

InputFilter

PLC

ScanTime

20 ms 15 ms 15 ms

Target

Output

OFFDelay

OFFDelay

A B

TOT

O

Sensor misses target:T

O<T

OFF sensor

OK:T

O>T

OFF sensor

Target

Output

ONDelay

ONDelay

A B

TOT

O

Sensor misses target:T

O<T

ON sensor

OK:T

O>T

ON sensor



35

ON DELAY

Don’t confuse ON delay with power-up de-

lay. ON delay is the response time after the

unit is powered up. It’s the duration of time

required for the interface to trigger an output

change of state when a target is introduced to

the sensing area (Figure 4).

ON delay is usually much faster than power-

up delay. Typical values are 0.1 ms to 5 ms for

DC sensors and 5 ms to 10 ms for AC sensors.

OFF DELAY

OFF delay is the amount of time required

for the sensor to trigger a change of output state

when a target is removed from the sensing area

(Figure 5). Typical values are 0.1 ms to 5 ms

for DC outputs and 8 ms to 20 ms for AC out-

puts.

FIGURE 9. Definition of observable time for aninductive sensor.

FIGURE 10. Methods to increase observable time.

FIGURE 11. Determining observable time for aninductive sensor in a hole or keyway application.

FIGURE 12. Determining observable time for a thru-beam or retroreflective photoelectric sensor.

Target

Output

OFFDelay

OFFDelay

A B

TOT

O

Choose a fastersensor

Increase the duration of thetarget in the sensing area

ONDelay

ONDelay

TO>T

ON

TO = a + b

V

a b

1 2 SpeedV

TO>T

ON

TO = b - a

V

a

b

1 2

ab

Start TO

End TO

V

Dia TargetDia Beam

TO = b - a

V



36

RESPONSE TIME

We have been discussing the response time

of the sensor. However, your application de-

mands that you consider the response time of

the entire data acquisition system.

In the model shown in Figure 6, consider that

there is a 20 ms ON delay at the sensor, a 15

ms delay at the PLC input module, and the sig-

nal must be maintained long enough for the PLC

to pick up the change of state in its scan cycle.

If the signal does not last long enough, the PLC

won’t see it during the scan.

During a PLC scan, the PLC reads the sta-

tus of the inputs, solves the logic programmed,and updates the outputs. The duration of the

sensor signal must be longer than the PLC scan

time, plus the PLC input filter delay.

The overall system response time is the sum

of the ON or OFF delay of the sensor and the

PLC input filter, and the PLC scan time.

FILTERS SLOW RESPONSE TIME

Line transients generated in industrial en-

vironments need to be suppressed. Filters are

effective in doing this so that transients do not

trigger false output signals, but filters slow

down the output response time.

Larger sensors have larger sensing areas.

They have more power to detect, and so have

more sensitivity to noise; the more noise, the

greater the amount of filtering required. There-

fore, large sensors are generally slower than

small sensor.

In the real world, there is always a compro-

mise between speed and transient protection.

There is also a real-world advantage to using sensors with small sensing areas to detect small

or fast moving targets - fast response time.

WHEN DOES RESPONSE TIME BECOMECRITICAL?

Response time becomes critical when you

have targets passing the sensor at durations

FIGURE 13. Observable time for thru-beam orretroreflective photoelectric sensors used to detectholes.

FIGURE 14. Fiber-optic sensors are useful fordetecting targets at high speeds.

TO = b + a

V

ab

Start TO

End TO

ba

V

* Reducing the diameter of the beam increasesthe duration of beam obstruction.



37

that are close to either the ON delay or the OFF

delay of that sensor. Here’s what can happen.

If the observable time of the target (TO) is less

than the sensor’s ON delay, the sensor will not

change states (Figure 7A). If it is longer than

the sensor’s ON time (Figure 7B), then it is OK.

TARGET LENGTH RELATION

When designing an application, you wouldlike to “arrange” things to be like Figure 7B,

which has a decent margin of safety for reliableoperation. The observable time should be sub-stantially longer than the sensor’s responsetime. Remember to consider the system’s totalresponse time when specifying.

Figure 8 illustrates sensing holes (or the

space between targets) instead of targets. Inthis case, the OFF delay is important. In Fig-ure 8A, a target entered the sensing area be-fore the sensor OFF delay expired, so the out-

put did not change state - it remained ON. InFigure 8B, a target is absent from the sensing area for a time longer than the OFF delay, sothe output does change state - to OFF.

Again, you would like to “arrange” things tobe like Figure 8B, which has a conservative mar-gin of safety with respect to both the sensor’s

response time and to the entire system’s re-sponse time requirements.

OBSERVABLE TIME

Observable time (TO) is the real time that the

target can be observed by the sensor. This rep-

resents the time elapsed between position 1 andposition 2 in Figure 9. To solve for T

O, you can

use the formula:

curve (2). This represents a length of a+b. Di-vide it by the speed of the target and the result

is the maximum operating time delay, or thetotal observable time for the target. The ob-servable time of the target should be longer thanthe sensor reaction time (T

ON) in order for the

output to change state.

Applying the formula, if the sensing range is1 inch wide and the target is 1 inch wide and

the speed is 2 inches per second, the result is 1second. ON delay for the switch must be lessthan 1 second for the target to initiate an out-put change of state. For practical reasons, usu-ally the “a” term is neglected. If the results aretoo close for comfort, you have to take the widthof the sensing area into consideration. Moving the sensor closer to the target increases “a” and

accounts for a longer observable time.

When the observable time of the target isdangerously close to the sensor response time,we need to bring about an increase in targetobservable time, as well as the time betweentargets, to levels that give confidence. You’d liketo have a minimum of 2 times T

ON , rather than

a fraction over the minimum. How do we do

this?

1. As we see in Figure 10A, we can choose a

faster sensor. One of the largest improve-ments in response time occurs when youchange from AC to DC output.

2. We can also increase the amount of timethat the target is in the sensing area bychoosing a sensor or sensor technologythat has a smaller sensing pattern inrelation to the target (Figure 10B).

3. In the most critical cases, we have to doboth.

CALCULATING OBSERVABLE TIME

A keyway in a motor shaft is actually a holein a metal surface as far as an inductive sensoris concerned. If we have an application wherewe need to detect the keyway (or holes), we needto analyze how long the sensor will observe thehole (Figure 11).

Between position 1 and 2, the elapsed timeis calculated by dividing the length (b-a) by the

Target observable time (TO) is equal to the

sum of the width of the sensing range (a), plus

the width of the target (b), divided by the speedof the target (V). For inductive sensors, whenthe front of the target passes the ON triggering

curve (1), the switch turns ON and will remainON until the end of the target passes the OFF

TO =

a + b

V



38

A. Let’s say that we have a 30 mm

metal target that we want to sense

with a 30 mm inductive sensor. TON

for this sensor is 20 ms. If the target

is travelling at 1 mm per second, can

the target be reliably sensed?B. Now, let’s change the speed to one

meter per second. When we plug this

into our equation, we get 30 ms for

maximum observable time. When

we compare this to the response time

of the sensor (TON

= 20 ms), we

should conclude that this is too close

to be reliable since we would like to

have a factor of 2 for TO.

What can we do to improve reliabil-ity?

1. Select a DC sensor with TON

= 2

ms.

2. Get as close as possible to the

surface of the sensor to maximize

“a’.

TO = a + b

V

TO = 30 mm

1 mm/sec= 30 s

TO = 30 mm

1000 mm/sec= 30 ms

30 mm target30 mm sensor

A. Speed = 1 mm/sec

B. Speed = 1000 mm/sec

OK

Close

TON

=20 ms

Case Study #1

A. In sensing a 30 mm target using an AC photoelectric

sensor in thru-beam mode, ON delay is 20 ms and the beam

diameter is 10 mm. What is the observable time (TO) forthis application if the target is moving at 1 mm per sec-

ond?

B. How about if the speed of the target increases to 5 m

per second?

We will check the DC version of the standard thru-beam

sensor which has ON delay of only 4 ms. This is exactly

TO, so the tolerance is too small for reliability. If we apply

fiber optics, the beam diameter is now only 1 mm. This

increases the target observable time by about one third. If we apply a DC fiber-optic sensor, T

O will be 5.8 ms and

TON

will be 4 ms.

Success is marginal since we do not have a factor of 2 for

TO. We will recommend that the user slow down a little to

3600 mm/sec (C.), or a sensor with 2 ms or better TON

is

chosen.

b - a

V

Target = 30 mmBeam diameter = 10 mmb

a

V

TON

=

20 ms (AC)

4 ms (DC)

A. Speed 1 = 1 mm/sec

C. TOmin

= 2 x TON

= 8 ms

B. Speed 2 = 5 mm/sec

TO =

30 mm-10 mm

1 mm/sec= 20 s

TO =

30 mm-10 mm

5000 mm/sec= 4 ms

Vmax

= 30 mm-1 mm8 ms

= 3600 mm/s

OK

NO

Better

Case Study #3

Case Study #2

A. Let’s push fiber-optic technol-

ogy to the limit. We will be sens-

ing the 2 mm pin on an electronic

component in a counting applica-

tion. Can we apply a DC fiber-

optic sensor in thru-beam mode if

the target is moving at 1 mm per

second? (TON=4 ms)B. How about 1000 mm per sec-

ond?

C. What we need to do is increase

the sensing area, which will also

increase the target’s observable

time. We can do this with fiber

optics using proximity diffuse

mode. This makes a sensing pat-

tern of 20 mm, which is 20 times

larger than the thru-beam mode.

In proximity mode, sensing time

starts when light from the target

is bounced back to the receiver -

it is not based on beam obstruc-

tion. Therefore, the target is ob-

servable as long as it is in the

beam pattern.

Using proximity diffuse mode,we are now able to detect this

small, fast-moving target. The

observable time of the target is

approximately 20 ms, which is

well within the specifications of

the sensor (TON

=4 ms).

Since the beam pattern is substan-

tially larger than the target, the

latter can be neglected in the ob-

servable time formula.

A. Speed = 1 mm/sec

B. Speed = 1000 mm/sec

C. Speed = 1000 mm/sec

b - a

VFor thru-beam mode:

Beam diameter = 1 mmTarget = 2 mm

NO

OKTO =

2-1

1= 1 second

TO = 2-11000

= 1 ms

TO =

20 + 2

1000= 22 ms OK

Beam diameter = 20 mmTarget = 2 mm

For proximity diffuse mode:a + b

V

TON

=

4 ms

TON

=

4 ms

V

V

2 mm

2 mm

Beam Pattern =

20 mm

1 mm



39

speed (V). The observable time of the holeshould be longer than the response time T

OFF of

the sensor in order for the sensor to turn off during the passing hole. In looking at Figure11, what could we do about this situation - physi-cally - to increase the observable time? Assum-ing the keyhole width (b) cannot be changed,

we could lower the velocity (V), we could reducethe value of (a) by using a smaller sensor, orretract the sensor farther from the target.

Photoelectric sensors use a similar formulafor determining the correct response time (Fig-ure 12). It is based on the duration of total beamobstruction. Observable time is the differencebetween the target diameter and the beam di-

ameter, divided by the target speed. One can-not consider ON time to start with partial beamobstruction because of the many variables in-

volved with photoelectric technology (humidity,dusty atmosphere, distance, etc.). Total beamobstruction is the conservative approach fordesign purposes.

If, for instance, we are close to excess gain 1,very little beam obstruction is needed to turnthe switch OFF. On the other hand, if we arefunctioning at a higher excess gain (perhaps 50),we don’t really know how much beam obstruc-tion will actually turn the switch OFF until wetry it. Another conclusion is that the larger your

target is with respect to beam diameter, thebetter.

When using photoelectrics to sense holes

(Figure 13), the start and stop points for deter-mining T

O is again based on total beam obstruc-

tion. Therefore, when the first ray is revealedby the hole, this is the starting point. Whenthe last ray is totally obstructed, this is the endpoint.

For fiber-optic applications (Figure 14), thebeam diameter is extremely small. Reducing

the diameter of the beam in relation to targetsize increases the time of beam obstruction.This makes fiber optics an excellent choice forhigh-speed applications or for small targets.Rule of thumb: your target dimension divided

by speed should be greater than the rated ONdelay time of your sensor by at least a factor of 2. For all practical purposes, the beam diam-eter can be neglected as long as it is several

times smaller than the target.

In all cases (inductive, photoelectric, fiber op-tic), if the target (plain or hole) is substantiallylarger than the sensor’s sensing area, you canneglect the sensor’s sensing pattern character-istics and concentrate on the target. This makeslife easier because when you can neglect the “a”term in the formula, you can forget about the“+” or “-” signs.

Three case studies of typical applications arepresented.



40

What is standard operating frequency,and what is it good for? The standard

operating frequency specification gives you ageneral idea of how fast the sensor reacts.Because it is a standardized parameter, youcan compare the products of various manufac-turers using this figure, so you can be sureyou are comparing “apples to apples.” You canalso use it as a guide for choosing sensorseries or types from a single manufacturer.For instance, smaller sensors and DC sensorstypically are faster than larger sensors and

AC sensors, respectively.

The maximum operating frequency is anindustry standard just like the nominal sens-ing distance. The method of arriving at thisparameter is to place standard targets on a

rotating disk. The targets are spaced on the diskin a well-defined geometrical configuration (Fig-ure 1).

For inductive, capacitive and magnetic technol-ogy, the sensor is mounted perpendicular to theshaft at a distance of one half its nominal sensing distance from the target as illustrated in Figure 1.Photoelectric and ultrasonic sensors are mountedparallel to the shaft facing the targets. The disk isrotated and the output of the sensor is displayedon an oscilloscope. (Photoelectric sensors need anadditional fast sensor to precisely trigger theoscilloscope.) The speed of the disk is increaseduntil reaching a point where detection practicallystops. The frequency of targets passing the sensorat this point is the maximum operating frequency.

T

On

T

Off

2M

M

1/2 Sn

Frequency =1

(Ton

+ Toff

)

CHAPTER 7

Standard Operating

Frequency

Target

Output

Present

Absent

ON

OFF

FIGURE 1. Standard operating frequency setup. FIGURE 2. ON delay.

ON Delay ON Delay



41

Caution: Operating frequency is a useful selec-tion guide, but do not use it to solve yourapplication problems. Operating frequency re-lates to a standard target in a standardizedlaboratory setup. Always use actual sensor re-sponse time(s) — ON/OFF delay — to analyzeyour application.

USING TIMING FUNCTIONS

Up to now we have analyzed the applicationto see if the sensor can detect progressivelyfaster and smaller targets. Now we are going totalk about adjusting the response times to solve

some common applications. We will introducean adjustable time delay between the triggerand output circuits in the sensor. You have threebasic tools for this:

1. ON delay which is used to prolong the OFFcondition after the target is detected.

2. OFF delay which is used to prolong theON-condition after the target is no longerdetected.

3. Monostable (one shot or pulse stretcher)

which is used to provide a measured ONtime “window” regardless of the presence of the target. It can be triggered either bytarget or non-target detection.

Note: Once any of these types of time delayshas been triggered, the output is controlled bythe timer regardless of the presence or absenceof the target.

ON DELAY

Figure 2 shows a target that comes into thesensing area for a long period of time, maybe 10seconds. We have adjusted the ON delay for 5seconds, so the target is present in the sensing area for 5 seconds before the output is energized.When the target leaves the sensing area, theoutput changes states without delay. When a

target comes into the sensing area for only 2seconds, it is neglected. The timer keeps theoutput ON for all its set 5 seconds.

Example: A car pulls into a drive-thru bankteller. A thru-beam photoelectric sensor detectsthe presence of the car and an alarm sounds tocall the bank teller to the window. What if a leaf momentarily blocks the beam? Or a dog runs by(Figure 3)? These things could cause the systemto be a nuisance. Controlling ON delay lets youdetect targets that enter the sensing area andstay there longer than a preset time (e.g., 5

seconds), thus reducing or eliminating nuisancetrips.

ON

Delay

ON

Delay

ON

DelayON

Delay

FIGURE 4. ON delay for underspeed detection.FIGURE 3. Cars are detected; interference is neglected.

OFF

Output

Target

Present

Absent

ON



42

Example: A thru-beam photoelectric system is

set up for detecting the level of fluid in a vatwhich has transparent walls. As fluid is added,the surface ripples. Ideally, the photoelectric sen-sor system would like a calm (no ripples) surfaceto detect. Because ripples break the beam, thesensor output will rapidly turn ON and OFF. This“chattering” effect can quickly damage a solenoidvalve or contactor used to control the liquid level.

You can use ON delay to neglect signals caused bythe waves and change output states only after acouple of seconds pass without ripple. This isrecommended when the normal hysteresis of thesensor cannot cope with the physical variations of the target. A HI-LO type of control would worksmoothly using the ON delay feature.

Example: You can use ON delay to detect

underspeed conditions. This is important inmany systems where an improper condition,such as a jam or a mechanical malfunction, iscausing a motor to slow down, and thereforecausing it to heat up. Before that motor heats upto the point where its own thermal protectionsaves it, but makes you wait many preciousminutes before you can restart, you’d like toknow that you’ve got a problem you can managewith minimum downtime. You set the sensor ONdelay such that it extends the output OFFcondition until just before the next target entersthe sensing area (Figure 4). The sensor staysOFF-at normal speed. When the target stays inthe sensing area longer than the ON delay, theoutput will change states, signaling theunderspeed condition.

FIGURE 5. OFF delay.

Present

Absent

ON

OFF

Target

OutputOFF Delay OFF Delay

FIGURE 7. One shot or pulse stretcher.

Present

Absent

ON

OFF

Target

Output

ONWindow

ONWindow

ONWindow

FIGURE 8. A thru-beam fiber-optic sensor detecting mis-placed labels with a pulse stretcher.

Target

Pulse Stretcher

Output

PLC Input

0 15

10

30

50 70

Target

Output

Present

Absent

ON

OFF

OFFDelay

OFFDelay

OFFDelay

OFFDelay

• Detecting Underspeed conditions: whenhole stays in the sensing area longer thanOFF Delay, output is triggered.

FIGURE 6. A thru-beam fiber-optic sensor detecting anunderspeed condition with OFF-delay.



OFF DELAY

OFF delay is basically a mirror image of ON delay (Figure 5). Just as ON delay pro-longs the OFF-condition, OFF delay prolongsthe ON condition. The idea with OFF-delay isto detect an irregularity in the rate of holes(spaces) between targets.

You can apply OFF delay in the same wayyou used ON delay — i.e., to detect underspeedand to eliminate “noise.” One of the mainreasons you use OFF-delay is when spaces areeasier to detect than targets.

Example: Cans of product are spaced evenlyon a conveyor system. An inductive sensor isinstalled to detect the spaces between thecans. Using OFF-delay, the output state re-mains ON as long as the metal cans keepentering the sensing area at regular intervals.

If a can is on its side, a space that is longerthan the OFF-delay will trigger an outputchange of state.

Example: Instead of detecting gear teethusing ON delay, we are now detecting holes ina flywheel using OFF delay (Figure 6). As long as the holes pass in a time less than the OFFdelay, the sensor does not change its ONoutput state. When the hole is in the sensing area longer than the set OFF delay, the out-put is triggered.

ONE SHOT OR PULSE STRETCHER

(MONOSTABLE OUTPUT)

A monostable sensor is commonly called a“one shot” timer or “pulse stretcher.” You pre-set a window of time that you want the outputto remain ON. Once the target is sensed, theinternal timer takes over, so that the outputstays ON, irrespective of the target status. It’scalled “one shot” because it turns ON only onetime per target. It takes another target totrigger another “shot.”

Figure 7 shows a one shot that is triggered

when energized (from absent to present). Itstays ON for its preset duration and does notcome ON again until the next time the sensorinterface changes states from absent topresent. Notice that if a transition from ab-sent to present happens within the ON timewindow, it is neglected.

you can usually preset. You must choose onetransition mode or the other, according to theneeds of your application (ON or OFF delay).

The pulse stretcher is useful if you have aslow system response time and a short targetobservable time. The event might be too fast forthe system to respond. That’s when you want to

use the ON “window” functions as a pulsestretcher — it lets you “stretch” the signalduration of a fast event to one that will be“seen” by the system.

Example: In detecting misplaced labels oncanned food, a thru-beam fiber-optic sensor is setfor just above the height of the can (Figure 8). If the label is misapplied, a corner will stick uphigher than the can and block the beam. This isan extremely fast event. The beam is broken foronly 15 milliseconds. The sensor’s ON delay is30ms. The programmable logic controller (PLC)

processing this information needs about 40ms toaccount for a 20ms ON delay for the input filterand a 20ms scan time. The sensor internalmonostable circuit can react in less than 10ms. If the “one shot” time is preset to 60ms, the PLCcan properly record the event. From the PLCperspective it is like the target has been in frontof the sensor for the entire 70ms time period.

Documents

Solis State Sensors