EngineeringReference andApplicationSolutions

A1

PLC Programmable Logic Controller

Controller Drive

BallscrewRotating Nut

Motor

Table

Drill Head

Drive/Controller

Motor

Joystick

Drive

Drive

Motor

Indexer

TransferMachine

Circuit Board

Rotary Indexer

A2

Motor TechnologiesIntroductionMotion control, in its widest sense, could relate toanything from a welding robot to the hydraulicsystem in a mobile crane. In the field of ElectronicMotion Control, we are primarily concerned withsystems falling within a limited power range,typically up to about 10HP (7KW), and requiringprecision in one or more aspects. This may involveaccurate control of distance or speed, very oftenboth, and sometimes other parameters such astorque or acceleration rate. In the case of the twoexamples given, the welding robot requires precisecontrol of both speed and distance; the cranehydraulic system uses the driver as the feedbacksystem so its accuracy varies with the skill of theoperator. This wouldn’t be considered a motioncontrol system in the strict sense of the term.

Our standard motion control system consists ofthree basic elements:

Fig. 1 Elements of motion control system

The control system. The actual task performed bythe motor is determined by the indexer/controller; itsets things like speed, distance, direction andacceleration rate. The control function may bedistributed between a host controller, such as adesktop computer, and a slave unit that acceptshigh-level commands. One controller may operatein conjunction with several drives and motors in amulti-axis system.

We’ll be looking at each of these system elementsas well as their relationships to each other.

The motor. This may be a stepper motor (eitherrotary or linear), a DC brush motor or a brushlessservo motor. The motor needs to be fitted withsome kind of feedback device unless it is a steppermotor.

Fig. 2 shows a system complete with feedback tocontrol motor speed. Such a system is known as aclosed-loop velocity servo system.

Fig. 2 Typical closed loop (velocity) servo system

Table of ContentsMotor Applications A3Step Motor Technology A4Linear Step Motor Technology A9Common Questions Regarding Step Motors A12DC Brush Motor Technology A13Brushless Motor Technology A17Hybrid Servo Technology A20Direct Drive Motor Technology A21Step Motor Drive Technology A23Microstepping Drive Technology A29Analog and Digital Servo Drives A31Brushless Servo Drive Technology A34Servo Tuning A36Feedback Devices A39Machine Control A45Control System Overview A46Understanding I/O Modules A48Serial & Parallel Communications A51Electrical Noise Symptoms & Solutions A52Emergency Stop A54System Selection Considerations A55Motor Sizing and Selection Software A57System Calculations – Move Profiles A58System Calculations – Leadscrew Drives A60System Calculations – Direct Drives A63System Calculations – Gear Drives A64System Calculations – Tangential Drives A65System Calculations – Linear Motors A66Glossary of Terms A68Technical Data A71Application Examples A72

The drive. This is an electronic power amplifier thatdelivers the power to operate the motor in responseto low-level control signals. In general, the drive willbe specifically designed to operate with a particularmotor type – you can’t use a stepper drive tooperate a DC brush motor, for instance.

Command Signals

High-Level Commands

Host Computer

or PLC

Indexer/ Controller

Drive Motor

Hybrid Stepper DC Servo Brushless

Servo Linear Stepper

Tachometer

Drive MotorController

Velocity Feedback

Overview

A3

AE

ngin

eeri

ng R

efe

rence

Motor TechnologiesApplication Areas of Motor TypesThe following section gives some idea of theapplications that are particularly appropriate foreach motor type, together with certainapplications that are best avoided. It should bestressed that there is a wide range ofapplications that can be equally well met bymore than one motor type, and the choice willtend to be dictated by customer preference,previous experience or compatibility withexisting equipment.

Cost-conscious applications will always beworth attempting with a stepper, as it willgenerally be hard to beat the stepper’s price.This is particularly true when the dynamicrequirements are not severe, such as “setting”type applications like positioning a guillotineback-stop or a print roller.

High-torque, low-speed, continuous-dutyapplications are also appropriate for stepmotors. At low speeds, it is very efficient interms of torque output relative to both size andinput power. Microstepping can improve low-speed applications such as a metering pumpdrive for very accurate flow control.

High-torque, high-speed, continuous-dutyapplications suit the servo motor, and in fact, astep motor should be avoided in suchapplications because the high-speed losses cancause excessive motor heating. A DC motorcan deliver greater continuous shaft power athigh speeds than a stepper of the same framesize.

Short, rapid, repetitive moves are the naturaldomain of steppers or hybrid servos due to theirhigh torque at low speeds, good torque-to-inertia ratio and lack of commutation problems.The brushes of the DC motor can limit itspotential for frequent starts, stops and directionchanges.

Low-friction, mainly inertial loads can beefficiently handled by the DC servo provided thestart/stop duty requirements are not excessive.This type of load requires a high ratio of peak tocontinuous torque and in this respect the servomotor excels.

Very arduous applications with a highdynamic duty cycle or requiring very highspeeds may require a brushless motor. Thissolution may also be dictated whenmaintenance-free operation is necessary.

Low-speed, high-smoothness applicationsare appropriate for microstepping or direct driveservos.

Applications in hazardous environments or ina vacuum may not be able to use a brushmotor. Either a stepper or a brushless motor iscalled for, depending on the demands of theload. Bear in mind that heat dissipation may bea problem in a vacuum when the loads areexcessive.

Will a stepper meet the torque/speed

requirements?

Do you need to run continuously at speeds above

2000 rpm?

Do you need to control torque?

Does the load change rapidly

during operation?

Do you need to detect position

loss OR measure actual load

position to correct for backlash?

Use a stepperUse a microstepping,

hybrid servo, or direct drive servo

Is quiet operation important?

Is low-speed smoothness important?

Is quiet operation important?

Use a microstepping with encoder feedback?

Is rapid settling important?

Use a brush servo.

Are there any other brush

servos in the system?

Must the motor EITHER 1) 2)

Will a brush servo meet

the torque/speed requirements?

Use a brushless servo.

Is there a hybrid servo which meets the torque/speed

requirements?

Try a hybrid servo with

encoder feedback if necessary.

Really quiet?

If there are other brushless motors, it may be better to be consistent

with this one. Otherwise use a brush servo.

Higher torque/speed technology.

Start Here

No

No

NoYes

Yes

Yes

No

Yes

YesYes

Yes

No No

No

No

Yes

No

No

No

Yes

No

Yes

No

Yes

No No

Yes

Will a brushless servo meet

the torque/speed requirements?

Be maintenance- free Operate in any environment

A4

Motor TechnologiesStepper MotorsStepper Motor BenefitsStepper motors have the following benefits:

• Low cost• Ruggedness• Simplicity in construction• High reliability• No maintenance• Wide acceptance• No tweaking to stabilize• No feedback components are needed• They work in just about any environment• Inherently more failsafe than servo motors.

There is virtually no conceivable failure within thestepper drive module that could cause the motor torun away. Stepper motors are simple to drive andcontrol in an open-loop configuration. They onlyrequire four leads. They provide excellent torque atlow speeds, up to 5 times the continuous torque ofa brush motor of the same frame size or double thetorque of the equivalent brushless motor. This ofteneliminates the need for a gearbox. A stepper-drivensystem is inherently stiff, with known limits to thedynamic position error.

A4

Stepper Motor DisadvantagesStepper motors have the following disadvantages:

• Resonance effects and relatively long settlingtimes

• Rough performance at low speed unless amicrostep drive is used

• Liability to undetected position loss as a result ofoperating open-loop

• They consume current regardless of loadconditions and therefore tend to run hot

• Losses at speed are relatively high and can causeexcessive heating, and they are frequently noisy(especially at high speeds).

• They can exhibit lag-lead oscillation, which isdifficult to damp. There is a limit to their availablesize, and positioning accuracy relies on themechanics (e.g., ballscrew accuracy). Many ofthese drawbacks can be overcome by the use ofa closed-loop control scheme.

Note: The Compumotor Zeta Series minimizes orreduces many of these different stepper motordisadvantages.

There are three main stepper motor types:• Permanent Magnet (P.M.) Motors• Variable Reluctance (V.R.) Motors• Hybrid Motors

Courtesy Airpax Corp., USA

N

N N N

NS

S SS

S

S

SS

SN

N NN

NSN

S

Fig. 1.1 “Canstack” or permanent magnet motor

Coil A

Coil B

Rotor

Stator cup A

Stator cup B

Output shaft

Variable Reluctance (V.R.) Motors. There is nopermanent magnet in a V.R. motor, so the rotorspins freely without “detent” torque. Torque outputfor a given frame size is restricted, although thetorque-to-inertia ratio is good, and this type of motoris frequently used in small sizes for applications suchas micro-positioning tables. V.R. motors are seldomused in industrial applications (having no permanentmagnet). They are not sensitive to current polarityand require a different driving arrangement than theother motor types.

Fig. 1.2 Variable reluctance motor

Permanent Magnet (P.M.) Motors. The tin-can or“canstack” motor shown in Fig. 1.1 is perhaps themost widely-used type in non-industrialapplications. It is essentially a low-cost, low-torque,low-speed device ideally suited to applications infields such as computer peripherals. The motorconstruction results in relatively large step angles,but their overall simplicity lends itself to economichigh-volume production at very low cost. The axial-air gap or disc motor is a variant of the permanentmagnet design which achieves higher performance,largely because of its very low rotor inertia.However this does restrict the applications of themotor to those involving little inertia. (e.g.,positioning the print wheel in a daisy-wheel printer).

A5

AE

ngin

eeri

ng R

efe

rence

Motor TechnologiesHybrid Motors. The hybrid motor shown in Fig. 1.3is by far the most widely-used stepper motor inindustrial applications. The name is derived from thefact that it combines the operating principles of theother two motor types (P.M. & V.R.). Most hybridmotors are 2-phase, although 5-phase versions areavailable. A recent development is the “enhancedhybrid” motor, which uses flux-focusing magnets togive a significant improvement in performance,albeit at extra cost.

Fig. 1.3 Hybrid stepper motor

Housing

Non-magneticStainlessSteel Shaft

Rotor

PrelubricatedBearing

Stator

Fig. 1.4 Simple 12 step/rev hybrid motor

The rotor of this machine consists of two polepieces with three teeth on each. In between thepole pieces is a permanent magnet that ismagnetized along the axis of the rotor, making oneend a north pole and the other a south pole. Theteeth are offset at the north and south ends asshown in the diagram.

The stator consists of a shell having four teeth thatrun the full length of the rotor. Coils are wound onthe stator teeth and are connected together inpairs.

With no current flowing in any of the motorwindings, the rotor will take one of the positionsshown in the diagrams. This is because thepermanent magnet in the rotor is trying to minimizethe reluctance (or “magnetic resistance”) of the fluxpath from one end to the other. This will occurwhen a pair of north and south pole rotor teeth arealigned with two of the stator poles. The torquetending to hold the rotor in one of these positions isusually small and is called the “detent torque”. Themotor shown will have 12 possible detent positions.

If current is now passed through one pair of statorwindings, as shown in Fig. 1.5(a), the resulting northand south stator poles will attract teeth of theopposite polarity on each end of the rotor. Thereare now only three stable positions for the rotor, thesame as the number of rotor teeth. The torquerequired to deflect the rotor from its stable positionis now much greater, and is referred to as the“holding torque”.

Fig. 1.5 Full stepping, one phase on

The operation of the hybrid motor is most easilyunderstood by looking at a very simple model thatwill produce 12 steps per rev. (Fig. 1.4).

N

S

N

N

S

SN

S

1A

2B2A

1B

N

S

N

N

S

S

N S

N

NN

S

SS

S N

S

S

S

N

N

N

S

N

S

SNN

NS

N

S

(a) (b)

(c) (d)

By changing the current flow from the first to thesecond set of stator windings (b), the stator fieldrotates through 90° and attracts a new pair of rotorpoles. This results in the rotor turning through 30°,corresponding to one full step. Reverting to the firstset of stator windings but energizing them in theopposite direction, we rotate the stator fieldthrough another 90° and the rotor takes another30° step (c). Finally, the second set of windings areenergized in the opposite direction (d) to give athird step position. We can now go back to thefirst condition (a), and after these four steps therotor will have moved through one tooth pitch. Thissimple motor therefore performs 12 steps per rev.Obviously, if the coils are energized in the reversesequence, the motor will go round the other way.

A6

Motor TechnologiesIf two coils are energized simultaneously (Fig. 1.6),the rotor takes up an intermediate position since itis equally attracted to two stator poles. Greatertorque is produced under these conditions becauseall the stator poles are influencing the rotor. Themotor can be made to take a full step simply byreversing the current in one set of windings; thiscauses a 90° rotation of the stator field as before. Infact, this would be the normal way of driving themotor in the full-step mode, always keeping twowindings energized and reversing the current ineach winding alternately.

Fig. 1.6 Full stepping, two phase on

N

N S

S

N

N

N

SS

S

S

N S

N

N

NN

S

SS

N

S N

S

S

SS

N

NN

S

S N

N

S

S NN

N

S

By alternately energizing one winding and then two(Fig. 1.7), the rotor moves through only 15° at eachstage and the number of steps per rev will bedoubled. This is called half stepping, and mostindustrial applications make use of this steppingmode. Although there is sometimes a slight loss oftorque, this mode results in much bettersmoothness at low speeds and less overshoot andringing at the end of each step.

Fig. 1.7 Half stepping

S N

S

S NN

N

S

S

S N

N

S

S NN

N

S

S

N

S

SNN

NS

motor and drive characteristics). In the half-stepmode, we are alternately energizing two phasesand then only one as shown in Fig. 1.9. Assumingthe drive delivers the same winding current in eachcase, this will cause greater torque to be producedwhen there are two windings energized. In otherwords, alternate steps will be strong and weak.This does not represent a major deterrent to motorperformance—the available torque is obviouslylimited by the weaker step, but there will be asignificant improvement in low-speed smoothnessover the full-step mode.

Clearly, we would like to produce approximatelyequal torque on every step, and this torque shouldbe at the level of the stronger step. We can achievethis by using a higher current level when there isonly one winding energized. This does not over-dissipate the motor because the manufacturer’scurrent rating assumes two phases to be energized(the current rating is based on the allowable casetemperature). With only one phase energized, thesame total power will be dissipated if the current isincreased by 40%. Using this higher current in theone-phase-on state produces approximately equaltorque on alternate steps (see Fig. 1.10).

Fig. 1.8 Full step current, 2-phase on

Current Patterns in the Motor WindingsWhen the motor is driven in its full-step mode,energizing two windings or “phases” at a time (seeFig. 1.8), the torque available on each step will bethe same (subject to very small variations in the

1 2 3 4

Phase 1

Phase 2

Fig. 1.9 Half step current

1 2 3 4 5 6 7 8

Phase 1

Phase 2

Fig. 1.10 Half step current, profiled

1 2 3 4 5 6 7 8

Phase 1

Phase 2

A7

AE

ngin

eeri

ng R

efe

rence

Motor TechnologiesWe have seen that energizing both phases withequal currents produces an intermediate stepposition half-way between the one-phase-onpositions. If the two phase currents are unequal, therotor position will be shifted towards the strongerpole. This effect is utilized in the microsteppingdrive, which subdivides the basic motor step byproportioning the current in the two windings. In thisway, the step size is reduced and the low-speedsmoothness is dramatically improved. High-resolution microstep drives divide the full motor stepinto as many as 500 microsteps, giving 100,000steps per revolution. In this situation, the currentpattern in the windings closely resembles two sinewaves with a 90° phase shift between them (seeFig. 1.11). The motor is now being driven very muchas though it is a conventional AC synchronousmotor. In fact, the stepper motor can be driven inthis way from a 60 Hz-US (50Hz-Europe) sine wavesource by including a capacitor in series with onephase. It will rotate at 72 rpm.

Fig. 1.11 Phase currents in microstep mode

Standard 200-Step Hybrid MotorThe standard stepper motor operates in the sameway as our simple model, but has a greater numberof teeth on the rotor and stator, giving a smallerbasic step size. The rotor is in two sections asbefore, but has 50 teeth on each section. The half-tooth displacement between the two sections isretained. The stator has 8 poles each with 5 teeth,making a total of 40 teeth (see Fig. 1.12).

Phase 1 Current: Zero

Phase 2 Current: Zero

+-

+-

Fig. 1.12 200-step hybrid motor

RotorStator

If we imagine that a tooth is placed in each of thegaps between the stator poles, there would be atotal of 48 teeth, two less than the number of rotorteeth. So if rotor and stator teeth are aligned at 12o’clock, they will also be aligned at 6 o’clock. At 3o’clock and 9 o’clock the teeth will be misaligned.However, due to the displacement between thesets of rotor teeth, alignment will occur at 3 o’clockand 9 o’clock at the other end of the rotor.

The windings are arranged in sets of four, andwound such that diametrically-opposite poles arethe same. So referring to Fig. 1.12, the north polesat 12 and 6 o’clock attract the south-pole teeth atthe front of the rotor; the south poles at 3 and 9o’clock attract the north-pole teeth at the back. Byswitching current to the second set of coils, thestator field pattern rotates through 45°. However, toalign with this new field, the rotor only has to turnthrough 1.8°. This is equivalent to one quarter of atooth pitch on the rotor, giving 200 full steps perrevolution.

Note that there are as many detent positions asthere are full steps per rev, normally 200. Thedetent positions correspond with rotor teeth beingfully aligned with stator teeth. When power isapplied to a stepper drive, it is usual for it toenergize in the “zero phase” state in which there iscurrent in both sets of windings. The resulting rotorposition does not correspond with a natural detentposition, so an unloaded motor will always move byat least one half step at power-on. Of course, if thesystem was turned off other than in the zero phasestate, or the motor is moved in the meantime, agreater movement may be seen at power-up.

Another point to remember is that for a givencurrent pattern in the windings, there are as manystable positions as there are rotor teeth (50 for a200-step motor). If a motor is de-synchronized, theresulting positional error will always be a wholenumber of rotor teeth or a multiple of 7.2°. A motorcannot “miss” individual steps – position errors ofone or two steps must be due to noise, spuriousstep pulses or a controller fault.

A8

Motor Technologies

Fig. 1.14 also shows that the rotor flux only has tocross a small air gap (typically 0.1mm or 0.004")when the rotor is in position. By magnetizing therotor after assembly, a high flux density is obtainedthat can be largely destroyed if the rotor isremoved. Stepper motors should therefore not bedismantled purely to satisfy curiosity, since theuseful life of the motor will be terminated.

Because the shaft of the motor passes through thecenter of the permanent magnet, a non-magneticmaterial must be used to avoid a magnetic short-circuit. Stepper shafts are therefore made ofstainless steel, and should be handled with care.Small-diameter motors are particularly vulnerable ifthey are dropped on the shaft end, as this willinvariably bend the shaft.

To produce a motor with a higher torque output,we need to increase the strength of both thepermanent magnet in the rotor and the fieldproduced by the stator. A stronger rotor magnetcan be obtained by increasing the diameter, givingus a larger cross-sectional area. However,increasing the diameter will degrade theacceleration performance of the motor becausethe torque-to-inertia ratio worsens (to a firstapproximation, torque increases with diametersquared but inertia goes up by the fourth power).Nevertheless, we can increase torque outputwithout degrading acceleration performance by

Occasionally a 5-lead motor may be encountered.These are not recommended since they cannot beused with conventional bipolar drives requiringelectrical isolation between the phases.

Looking at the motor longitudinal section (Fig. 1.14),we can see the permanent magnet in the rotor andthe path of the flux through the pole pieces and thestator. The alternating flux produced by the statorwindings flows in a plane at right angles to thepage. Therefore, the two flux paths are at right

Bifilar WindingsMost motors are described as being “bifilar wound”,which means there are two identical sets ofwindings on each pole. Two lengths of wire arewound together as though they were a single coil.This produces two windings that are electrically andmagnetically almost identical – if one coil were to bewound on top of the other, even with the samenumber of turns, the magnetic characteristicswould be different. In simple terms, whereas almostall the flux from the inner coil would flow throughthe iron core, some of the flux from the outer coilwould flow through the windings of the coilunderneath.

The origins of the bifilar winding go back to theunipolar drive (see Drive Technologies section,page A23). Rather than have to reverse the currentin one winding, the field may be reversed bytransferring current to a second coil wound in theopposite direction. (Although the two coils arewound the same way, interchanging the ends hasthe same effect.) So with a bifilar-wound motor, thedrive can be kept simple. However, thisrequirement has now largely disappeared with thewidespread availability of the more-efficient bipolardrive. Nevertheless, the two sets of windings dogive us additional flexibility, and we shall see thatdifferent connection methods can be used to givealternative torque-speed characteristics.

If all the coils in a bifilar-wound motor are broughtout separately, there will be a total of 8 leads (seeFig. 1.13). This is becoming the most commonconfiguration since it gives the greatest flexibility.However, there are still a number of motorsproduced with only 6 leads, one lead serving as acommon connection to each winding in a bifilarpair. This arrangement limits the motor’s range ofapplication since the windings cannot be connectedin parallel. Some motors are made with only 4leads, these are not bifilar-wound and cannot beused with a unipolar drive. There is obviously noalternative connection method with a 4-lead motor,but in many applications this is not a drawback andthe problem of insulating unused leads is avoided.

Fig. 1.13 Motor lead configurations

angles to each other and only interact in the rotorpole pieces. This is an important feature of thehybrid motor – it means that the permanent magnetin the rotor does not “see” the alternating field fromthe windings, hence it does not produce ademagnetizing effect. Unlike the DC servo motor, itis generally impossible to de-magnetize a steppermotor by applying excess current. However, toomuch current will damage the motor in other ways.Excessive heating may melt the insulation or thewinding formers, and may soften the bondingmaterial holding the rotor laminations. If thishappens and the laminations are displaced, theeffects can be the same as if the rotor had beende-magnetized

Fig. 1.14 Longitudinal section through singlestack motor

4-lead 5-lead 6-lead 8-lead

A9

AE

ngin

eeri

ng R

efe

rence

Motor Technologies

The forcer is equipped with 4 pole pieces eachhaving 3 teeth. The teeth are staggered in pitchwith respect to those on the platen, so thatswitching the current in the coils will bring the nextset of teeth into alignment. A complete switchingcycle (4 full steps) is equivalent to one tooth pitchon the platen. Like the rotary stepper, the linearmotor can be driven from a microstep drive. In thiscase, a typical linear resolution will be 12,500 stepsper inch.

The linear motor is best suited for applications thatrequire a low mass to be moved at high speed. In aleadscrew-driven system, the predominant inertia isusually the leadscrew rather than the load to bemoved. Hence, most of the motor torque goes toaccelerate the leadscrew, and this problembecomes more severe the longer the travelrequired. Using a linear motor, all the developedforce is applied directly to the load and theperformance achieved is independent of the lengthof the move. A screw-driven system can developgreater linear force and better stiffness; however,the maximum speed may be as much as ten timeshigher with the equivalent linear motor. Forexample, a typical maximum speed for a linearmotor is 100 in/sec. To achieve this with a 10-pitchballscrew would require a rotary speed of 6,000rpm. In addition, the linear motor can travel up to12 feet using a standard platen.

How the Linear Motor WorksThe forcer consists of two electromagnets (A and B)and a strong rare earth permanent magnet. Thetwo pole faces of each electromagnet are toothedto concentrate the magnetic flux. Four sets of teethon the forcer are spaced in quadrature so that onlyone set at a time can be aligned with the platenteeth.

The magnetic flux passing between the forcer andthe platen gives rise to a very strong force ofattraction between the two pieces. The attractiveforce can be up to 10 times the peak holding forceof the motor, requiring a bearing arrangement tomaintain precise clearance between the pole facesand platen teeth. Either mechanical roller bearingsor air bearings are used to maintain the requiredclearance.

When current is established in a field winding, theresulting magnetic field tends to reinforcepermanent magnetic flux at one pole face andcancel it at the other. By reversing the current, thereinforcement and cancellation are exchanged.Removing current divides the permanent magneticflux equally between the pole faces. By selectivelyapplying current to phase A and B, it is possible toconcentrate flux at any of the forcer’s four polefaces. The face receiving the highest fluxconcentration will attempt to align its teeth with theplaten. Fig. 1.17 shows the four primary states orfull steps of the forcer. The four steps result inmotion of one tooth interval to the right. Reversingthe sequence moves the forcer to the left.

adding further magnet sections or “stacks” to thesame shaft (Fig. 1.15). A second stack will enabletwice the torque to be produced and will double theinertia, so the torque-to-inertia ratio remains thesame. Hence, stepper motors are produced insingle-, two- and three-stack versions in eachframe size.

Fig. 1.15 Three-stack hybrid stepping motor

As a guideline, the torque-to-inertia ratio reduces bya factor of two with each increase in frame size(diameter). So an unloaded 34-size motor canaccelerate twice as rapidly as a 42-size, regardlessof the number of stacks.

Linear Stepping Motors

Fig. 1.16 Linear stepping motor

Platen Teeth

Field Windings

Platen

Phase A Electromagnet

Forcer

Permanent Magnet

BBAAPole Faces

SN

1 2 1 2

Air Gap

Phase B Electromagnet

The linear stepper is essentially a conventionalrotary stepper that has been “unwrapped” so that itoperates in a straight line. The moving componentis referred to as the forcer and it travels along afixed element or platen. For operational purposes,the platen is equivalent to the rotor in a normalstepper, although it is an entirely passive deviceand has no permanent magnet. The magnet isincorporated in the moving forcer together with thecoils (see Fig. 1.16).

A10

Motor TechnologiesStep Motor CharacteristicsThere are numerous step motor performancecharacteristics that warrant discussion. However,we’ll confine ourselves to those traits with thegreatest practical significance.

Fig. 1.18 illustrates the static torque curve of thehybrid step motor. This relates to a motor that isenergized but stationary. It shows us how therestoring torque varies with rotor position as it isdeflected from its stable point. We’re assuming thatthere are no frictional or other static loads on themotor. As the rotor moves away from the stableposition, the torque steadily increases until itreaches a maximum after one full step (1.8°). Thismaximum value is called the holding torque and itrepresents the largest static load that can beapplied to the shaft without causing continuousrotation. However, it doesn’t tell us the maximumrunning torque of the motor – this is always lessthan the holding torque (typically about 70%).

Fig. 1.18 Static torque-displacementcharacteristic

Repeating the sequence in the example will causethe forcer to continue its movement. When thesequence is stopped, the forcer stops with theappropriate tooth set aligned. At rest, the forcerdevelops a holding force that opposes any attemptto displace it. As the resting motor is displaced fromequilibrium, the restoring force increases until thedisplacement reaches one-quarter of a toothinterval. (See Fig. 1.18.) Beyond this point, therestoring force drops. If the motor is pushed overthe crest of its holding force, it slips or jumps rathersharply and comes to rest at an integral number oftooth intervals away from its original location. If thisoccurs while the forcer is travelling along the platen,it is referred to as a stall condition.

Fig. 1.17 The four cardinal states or full steps ofthe forcer

SN

SN

SN

B Aligned

A Aligned

B Aligned

SN

Flux LinesDirection of MMF due to electromagnet

A Aligned

Phase BPhase A

1

2

2

1To

rque

Cloc

kwis

eCo

unte

r Clo

ckw

ise 4 Motor Steps

MaxTorque

UnstableStable StableAngle

As the shaft is deflected beyond one full step, thetorque will fall until it is again at zero after two fullsteps. However, this zero point is unstable and thetorque reverses immediately beyond it. The nextstable point is found four full steps away from thefirst, equivalent to one tooth pitch on the rotor or1/50 of a revolution.

Although this static torque characteristic isn’t agreat deal of use on its own, it does help explainsome of the effects we observe. For example, itindicates the static stiffness of the system, (i.e.,how the shaft position changes when a torque loadis applied to a stationary motor). Clearly the shaftmust deflect until the generated torque matches theapplied load. If the load varies, so too will the staticposition. Non-cumulative position errors willtherefore result from effects such as friction or out-of-balance torque loads. It is important toremember that the static stiffness is not improvedby using a microstepping drive—a given load on theshaft will produce the same angular deflection. Sowhile microstepping increases resolution andsmoothness, it may not necessarily improvepositioning accuracy.

A11

AE

ngin

eeri

ng R

efe

rence

Motor TechnologiesUnder dynamic conditions with the motor running,the rotor must be lagging behind the stator field if itis producing torque. Similarly, there will be a leadsituation when the torque reverses duringdeceleration. Note that the lag and lead relate onlyto position and not to speed. From the statictorque curve (Fig. 1.18), clearly this lag or leadcannot exceed two full steps (3.6°) if the motor is toretain synchronism. This limit to the position errorcan make the stepper an attractive option insystems where dynamic position accuracy isimportant.

When the stepper performs a single step, thenature of the response is oscillatory as shown inFig. 1.19. The system can be likened to a mass thatis located by a “magnetic spring”, so the behaviorresembles the classic mass-spring characteristic.Looking at it in simple terms, the static torque curveindicates that during the step, the torque is positiveduring the full forward movement and so isaccelerating the rotor until the new stable point isreached. By this time, the momentum carries therotor past the stable position and the torque nowreverses, slowing the rotor down and bringing itback in the opposite direction. The amplitude,frequency and decay rate of this oscillation willdepend on the friction and inertia in the system aswell as the electrical characteristics of the motorand drive. The initial overshoot also depends onstep amplitude, so half-stepping produces lessovershoot than full stepping and microstepping willbe better still.

Fig. 1.19 Single step response

Attempting to step the motor at its naturaloscillation frequency can cause an exaggeratedresponse known as resonance. In severe cases,this can lead to the motor desynchronizing or“stalling.” It is seldom a problem with half-stepdrives and even less so with a microstepper. Thenatural resonant speed is typically 100-200 fullsteps/sec. (0.5-1 rev/sec).

Angl

e

Time

Under full dynamic conditions, the performance ofthe motor is described by the torque-speed curve asshown in Fig. 1.20. There are two operating ranges,the start/stop (or pull in) range and the slew (or pullout) range. Within the start/stop range, the motor canbe started or stopped by applying index pulses atconstant frequency to the drive. At speeds within thisrange, the motor has sufficient torque to accelerateits own inertia up to synchronous speed without theposition lag exceeding 3.6°. Clearly, if an inertial loadis added, this speed range is reduced. So the start/stop speed range depends on the load inertia. Theupper limit to the start/stop range is typically between200 and 500 full steps/sec (1-2.5 revs/sec).

Fig. 1.20 Start/stop and slew curves

To operate the motor at faster speeds, it isnecessary to start at a speed within the start/stoprange and then accelerate the motor into the slewregion. Similarly, when stopping the motor, it mustbe decelerated back into the start/stop rangebefore the clock pulses are terminated. Usingacceleration and deceleration “ramping” allowsmuch higher speeds to be achieved, and inindustrial applications the useful speed rangeextends to about 3000 rpm (10,000 full steps/sec).Note that continuous operation at high speeds isnot normally possible with a stepper due to rotorheating, but high speeds can be used successfullyin positioning applications.

The torque available in the slew range does notdepend on load inertia. The torque-speed curve isnormally measured by accelerating the motor up tospeed and then increasing the load until the motorstalls. With a higher load inertia, a loweracceleration rate must be used but the availabletorque at the final speed is unaffected.

Torq

ue

Steps per second

Start/ Stop

Range

Slew Range

Slew Curve

Start/Stop CurveHolding Torque

A12

Motor TechnologiesCommon Questions and AnswersAbout Step Motors

1. Why do step motors run hot?Two reasons: 1. Full current flows through themotor windings at standstill. 2. PWM drivedesigns tend to make the motor run hotter.Motor construction, such as laminationmaterial and riveted rotors, will also affectheating.

2. What are safe operating temperatures?The motors have class B insulation, which israted at 130°C. Motor case temperatures of90°C will not cause thermal breakdowns.Motors should be mounted where operatorscannot come into contact with the motor case.

3. What can be done to reduce motor heating?Many drives feature a “reduce current atstandstill” command or jumper. This reducescurrent when the motor is at rest withoutpositional loss.

4. What does the absolute accuracy specificationmean?This refers to inaccuracies, non-cumulative,encountered in machining the motor.

5. How can the repeatability specification bebetter than that of accuracy?Repeatability indicates how precisely aprevious position can be re-obtained. Thereare no inaccuracies in the system that affect agiven position, returning to that position, thesame inaccuracy is encountered.

6. Will motor accuracy increase proportionatelywith the resolution?No. The basic absolute accuracy andhysteresis of the motor remain unchanged.

7. Can I use a small motor on a large load if thetorque requirement is low?Yes, however, if the load inertia is more thanten times the rotor inertia, cogging andextended ringing at the end of the move will beexperienced.

8. How can end of move “ringing” be reduced?Friction in the system will help damp thisoscillation. Acceleration/deceleration ratescould be increased. If start/stop velocities areused, lowering or eliminating them will help.

9. Why does the motor stall during no loadtesting?The motor needs inertia roughly equal to itsown inertia to accelerate properly. Anyresonances developed in the motor are at theirworst in a no-load condition.

10. Why is motor sizing important, why not just gowith a larger motor?If the motor’s rotor inertia is the majority of theload, any resonances may become morepronounced. Also, productivity would suffer asexcessive time would be required to acceleratethe larger rotor inertia. Smaller may be better.

11. What are the options for eliminatingresonance?This would most likely happen with full stepsystems. Adding inertia would lower theresonant frequency. Friction would tend to

dampen the modulation. Start/stop velocitieshigher than the resonant point could be used.Changing to half step operation would greatlyhelp. Ministepping and microstepping alsogreatly minimize any resonant vibrations.Viscous inertial dampers may also help.

12. Why does the motor jump at times when it'sturned on?This is due to the rotor having 200 naturaldetent positions. Movement can then be ±3.6°in either direction.

13. Do the rotor and stator teeth actually mesh?No. While some designs used this type ofharmonic drive, in this case, an air gap is verycarefully maintained between the rotor and thestator.

14. Does the motor itself change if a microsteppingdrive is used?The motor is still the standard 1.8° stepper.Microstepping is accomplished byproportioning currents in the drive (higherresolutions result). Ensure the motor’sinductance is compatible.

15. A move is made in one direction, and then themotor is commanded to move the samedistance but in the opposite direction. Themove ends up short, why?Two factors could be influencing the results.First, the motor does have magnetic hysteresisthat is seen on direction changes. This is in thearea of 0.03°. Second, any mechanicalbacklash in the system to which the motor iscoupled could also cause loss of motion.

16. Why are some motors constructed as eight-lead motors?This allows greater flexibility. The motor can berun as a six-lead motor with unipolar drives.With bipolar drives, the windings can then beconnected in either series or parallel.

17. What advantage do series or parallelconnection windings give?With the windings connected in series, low-speed torques are maximized. But this alsogives the most inductance so performance athigher speeds is lower than if the windingswere connected in parallel.

18. Can a flat be machined on the motor shaft?Yes, but care must be taken to not damagethe bearings. The motor must not bedisassembled. Compumotor does not warrantythe user’s work.

19. How long can the motor leads be?For bipolar drives, 100 feet. For unipolardesigns, 50 feet. Shielded, twisted pair cablesare required.

20. Can specialty motors, explosion-proof,radiation-proof, high-temperature, low-temperature, vacuum-rated, or waterproof, beprovided?Compumotor is willing to quote on mostrequirements with the exception of explosionproof.

21. What are the options if an explosion-proofmotor is needed?Installing the motor in a purged box should beinvestigated.

A13

AE

ngin

eeri

ng R

efe

rence

Motor Technologies

This design was quickly improved, and by the endof the 19th century the design principles of DCmotors had become well established.

About that time; however, AC power supplysystems came into general use and the popularityof the DC motor declined in favor of the lessexpensive AC induction motor. More recently, theparticular characteristics of DC motors, notably highstarting torque and controllability, have led to theirapplication in a wide range of systems requiringaccurate control of speed and position. Thisprocess has been helped by the development ofsophisticated modern drive and computer controlsystems.

PrinciplesIt is well known that when a current-carryingconductor is placed in a magnetic field, itexperiences a force (Fig. 1.22).

Fig. 1.22 Force on a conductor in amagnetic field

DC Brush MotorsThe history of the DC motor can be traced back tothe 1830s, when Michael Faraday did extensivework with disc type machines (Fig. 1.21).

Fig. 1.21 Simple disc motor

The force acting on the conductor is given by:

F = I x B

where B = magnetic flux density and I = current

If this single conductor is replaced by a largenumber of conductors (i.e., a length of wire iswound into a coil), the force per unit length isincreased by the number of turns in the coil. This isthe basis of a DC motor.

Practical ConsiderationsThe problem now is that of using this force toproduce the continuous torque required in apractical motor.

To achieve maximum performance from the motor,the maximum number of conductors must beplaced in the magnetic field, to obtain the greatestpossible force. In practice, this produces a cylinderof wire, with the windings running parallel to the axisof the cylinder. A shaft is placed down this axis toact as a pivot, and this arrangement is called themotor armature (Fig. 1.23).

Fig. 1.23 DC motor armature

This is achieved by constructing the armature as aseries of small sections connected in sequence tothe segments of a commutator (Fig 1.24). Electricalconnection is made to the commutator by means oftwo brushes. It can be seen that if the armaturerotates through 1/6 of a revolution clockwise, thecurrent in coils 3 and 6 will have changed direction.As successive commutator segments pass thebrushes, the current in the coils connected to thosesegments changes direction. This commutation orswitching effect results in a current flow in the

As the armature rotates, so does the resultantmagnetic field. The armature will come to rest withits resultant field aligned with that of the stator field,unless some provision is made to constantlychange the direction of the current in the individualarmature coils.

CommutationThe force that rotates the motor armature is theresult of the interaction between two magneticfields (the stator field and the armature field). Toproduce a constant torque from the motor, thesetwo fields must remain constant in magnitude andin relative orientation.

Fig. 1.24 Electrical arrangement of the armature

N S

Conductive Disc

BrushMagnet

Force (F)

Magnetic Field (B)

Conductor Carrying Current (I) (Into Page)

Force on Conductor F = I x B

Resultant Field Due to Armature Current

Shaft

Armature

Direction of Current Into PageStator Field

2

1 3

4

5

6

Current In Out

A14

Motor Technologiesarmature that occupies a fixed position in space,independent of the armature rotation, and allowsthe armature to be regarded as a wound core withan axis of magnetization fixed in space. This givesrise to the production of a constant torque outputfrom the motor shaft.

The axis of magnetization is determined by theposition of the brushes. If the motor is to have similarcharacteristics in both directions of rotation, thebrush axis must be positioned to produce an axis ofmagnetization that is at 90° to the stator field.

DC Motor TypesSeveral different types of DC motor are currentlyin use.

Iron cored. (Fig. 1.25). This is the most commontype of motor used in DC servo systems. It is madeup of two main parts; a housing containing the fieldmagnets and a rotor made up of coils of wirewound in slots in an iron core and connected to acommutator. Brushes, in contact with thecommutator, carry current to the coils.

Fig. 1.25 Iron-cored motor

Brushless. The major limiting factor in theperformance of iron-cored motors is internalheating. This heat escapes through the shaft andbearings to the outer casing, or through the airgapbetween the armature and field magnets and fromthere to the casing. Both of these routes arethermally inefficient, so cooling of the motorarmature is very poor.

Fig. 1.28 Brushless motor

In the brushless motor, the construction of the ironcored motor is turned inside out, so that the rotorbecomes a permanent magnet and the statorbecomes a wound iron core.

The current-carrying coils are now located in thehousing, providing a short, efficient thermal path tothe outside air. Cooling can further be improved byfinning the outer casing and blowing air over it ifnecessary (to effectively cool an iron-cored motor, itis necessary to blow air through it.) The ease ofcooling the brushless motor allows it to produce amuch higher power in relation to its size.

The other major advantage of brushless motors istheir lack of a conventional commutator and brushgear. These items are a source of wear andpotential trouble and may require frequentmaintenance. By not having these components, thebrushless motor is inherently more reliable and canbe used in adverse environmental conditions.

To achieve high torque and low inertia, brushlessmotors do require rare earth magnets that aremuch more expensive than conventional ceramicmagnets. The electronics necessary to drive abrushless motor are also more complex than for abrush motor. A more thorough explanation ofbrushless motors is provided on page A17.

Losses in DC MotorsDC motors are designed to convert electrical powerinto mechanical power and as a consequence ofthis, during periods of deceleration or if externallydriven, will generate electrical power. However, allthe input power is not converted into mechanicalpower due to the electrical resistance of thearmature and other rotational losses. These lossesgive rise to heat generation within the motor.

S

S

NN

Backiron Return Path

Stator Lam Teeth

Magnets

Windings

CommutatorBrushes

Rotor Winding

Stator Magnets

Moving coil. There are two principle forms of thistype of motor. 1. The “printed” motor (Fig. 1.26),using a disc armature. 2. The “shell” type armature(Fig. 1.27).

Since these types of motors have no moving iron intheir magnetic field, they do not suffer from ironlosses. Consequently, higher rotational speeds canbe obtained with low power inputs.

Fig. 1.26 Disc-armature “printed” motor

Diagrams courtesy of Electro-Craft Ltd.

S

S

MotionAir gap

Magnet pole

Flux pathCoreArmature

(Hollow cup, shaped conductor array)

Fig. 1.27 Shell-armature motor

Permanent magnet (8 pole)

Motion

A15

AE

ngin

eeri

ng R

efe

rence

Motor TechnologiesShort circuit currents. As the brushes slide overthe commutator, the brush is in contact with twocommutator segments for a brief period. During thisperiod, the brush will short out the coil connectedto those segments (Fig. 1.30). This conditiongenerates a torque that opposes the main drivingtorque and increases with motor speed.

Fig. 1.30 Generation of short-circuit currents

All these losses will contribute heat to the motorand it is this heating that will ultimately limit themotor application.

Other Limiting ConsiderationsTorque ripple. The requirement for constant torqueoutput from a DC motor is that the magnetic fieldsdue to the stator and the armature are constant inmagnitude and relative orientation, but this ideal isnot achieved in practice. As the armature rotates,the relative orientation of the fields will changeslightly and this will result in small changes in torqueoutput called “torque ripple” (Fig. 1.31).

Fig. 1.31 Torque ripple components

This will not usually cause problems at high speedssince the inertia of the motor and the load will tendto smooth out the effects, but problems may ariseat low speeds.

Motors can be designed to minimize the effects oftorque ripple by increasing the number of windings,or the number of motor poles, or by skewing thearmature windings.

Motor losses can be divided into two areas: Thosethat depend on the load and those that depend onspeed (Fig. 1.29).

Fig. 1.29 Losses in a DC motor

Winding losses. These are caused by the electricalresistance of the motor windings and are equal toI2R (where I = armature current and R = armatureresistance).

As the torque output of the motor increases, Iincreases, which gives rise to additional losses.Consideration of winding losses is very importantsince heating of the armature winding causes anincrease in R, which results in further losses andheating. This process can destroy the motor if themaximum current is not limited. Furthermore, athigher temperatures, the field magnets begin tolose their strength. Hence, for a required torqueoutput the current requirement becomes greater.

Brush contact losses. These are fairly complex toanalyze since they depend upon several factors thatwill vary with motor operation. In general, brushcontact resistance may represent a high proportionof the terminal resistance of the motor. The result ofthis resistance will be increased heating due to I2Rlosses in the brushes and contact area.

Iron losses. Iron losses are the major factor indetermining the maximum speed that may beattained by an iron-cored motor. These fall into twocategories:

• Eddy current losses are common in allconductive cored components experiencing achanging magnetic field. Eddy currents areinduced into the motor armature as it undergoeschanges in magnetization. These currents arespeed-dependent and have a significant heatingeffect at high speeds. In practice, eddy currentsare reduced by producing the armature core as aseries of thin, insulated sections or laminations,stacked to produce the required core length.

• Hysteresis losses are caused by the resistanceof the core material to constant changes ofmagnetic orientation, giving rise to additional heatgeneration, which increases with speed.

Friction losses. These are associated with themechanical characteristics of the motor and arisefrom brush friction, bearing friction, and airresistance. These variables will generate heat andwill require additional armature current to offset thiscondition.

Motor losses

SpeedLoad

Winding losses

Iron losses

Friction losses

Brush losses

Short-cut circuit losses

Commutator Brush

Winding

Torque Ripple

Steady Torque O/P

Time

Torque

A16

Motor TechnologiesMotor EquationsUnlike a step motor, the DC brush motor exhibitssimple relationships between current, voltage,torque and speed. It is therefore worth examiningthese relationships as an aid to the application ofbrush motors.

The application of a constant voltage to theterminals of a motor will result in its accelerating toattain a steady final speed (n). Under theseconditions, the voltage (V) applied to the motor isopposed by the back emf (nKE) and the resultantvoltage drives the motor current (I) through themotor armature and brush resistance (Rs).

The equivalent circuit of a DC motor is shown inFig. 1.34.

Fig. 1.34 DC motor equivalent circuit

Rs = motor resistanceL = winding inductanceVg = back emf andRL represents magnetic losses.

The value of RL is usually large and so can beignored, as can the inductance L, which is generallysmall.If we apply a voltage (V) to the motor and a current(I) flows, then:

V = IRs + Vg

but Vg = nKE

so V = IRs + nKE (1)This is the electrical equation of the motor.If KT is the torque constant of the motor (typically inoz/in per Amp), then the torque generated by themotor is given by:

T = IKT (2)The opposing torque due to friction (TF) and viscousdamping (KD) is given by:

TM = TF + nKD

If the motor is coupled to a load TL ,then atconstant speed:

T = TL + TF + nKD (3)Equations (1), (2) and (3) allow us to calculate therequired current and drive voltage to meet giventorque and speed requirements. The values of KT,KE, etc. are given in the motor manufacturer’s data.

Demagnetization. The permanent magnets of aDC motor field will tend to become demagnetizedwhenever a current flows in the motor armature.This effect is known as “armature reaction” and willhave a negligible effect in normal use. Under highload conditions, however, when motor current maybe high, the effect will cause a reduction in thetorque constant of the motor and a consequentreduction in torque output.

Above a certain level of armature current, the fieldmagnets will become permanently demagnetized.Therefore, it is important not to exceed themaximum pulse current rating for the motor.

Mechanical resonances and backlash. It mightnormally be assumed that a motor and its load,including a tachometer or position encoder, are allrigidly connected together. This may, however, notbe the case.

It is important for a bi-directional drive or positioningsystem that the mechanics are free from backlash,otherwise, true positioning will present problems.

In high-performance systems, with highaccelerations, interconnecting shafts and couplingsmay deflect under the applied torque, such that thevarious parts of the system may have differentinstantaneous velocities that may be in oppositedirections. Under certain conditions, a shaft may gointo torsional resonance (Fig. 1.32).

Fig. 1.32 Torsional oscillation

Rs

RL Vg

L

v

I

Shaft

Load Motor Tach

Back emfAs described previously, a permanent magnet DCmotor will operate as a generator. As the shaft isrotated, a voltage will appear across the brushterminals. This voltage is called the backelectromotive force (emf) and is generated evenwhen the motor is driven by an applied voltage. Theoutput voltage is essentially linear with motor speedand has a slope that is defined as the motor voltageconstant, KE (Fig. 1.33). KE is typically quoted involts per 1000 rpm.

Fig. 1.33 Back-emf characteristic

Output volts

Shaft speed

A17

AE

ngin

eeri

ng R

efe

rence

Motor TechnologiesBrushless Motor OperationTo turn this motor into a brushless design, we muststart by eliminating the windings on the rotor. Thiscan be achieved by turning the motor inside out. Inother words, we make the permanent magnet therotating part and put the windings on the statorpoles. We still need some means of reversing thecurrent automatically – a cam-operated reversingswitch could be made to do this job (Fig. 1.36).Obviously such an arrangement with a mechanicalswitch is not very satisfactory, but the switchingcapability of non-contacting devices tends to bevery limited. However, in a servo application, wewill use an electronic amplifier or drive which canalso be used to do the commutation in response tolow-level signals from an optical or hall-effectsensor (see Fig. 1.37). This component is referredto as the commutation encoder. So unlike the DCbrush motor, the brushless version cannot bedriven by simply connecting it to a source of directcurrent. The current in the external circuit must bereversed at defined rotor positions. Hence, themotor is actually being driven by an alternatingcurrent.

Fig. 1.37 Brushless motor

A simple conventional DC brush motor (Fig. 1.35)consists of a wound rotor that can turn within amagnetic field provided by the stator. If the coilconnections were made through slip rings, thismotor would behave like a step motor (reversing thecurrent in the rotor would cause it to flip through180°). By including the commutator and brushes,the reversal of current is made automatically andthe rotor continues to turn in the same direction.

Fig. 1.36 “Inside out” DC motor

Going back to the conventional brush motor, arotor consisting of only one coil will exhibit a largetorque variation as it rotates. In fact, thecharacteristic will be sinusoidal, with maximumtorque produced when the rotor field is at rightangles to the stator field and zero torque at thecommutation point (see Fig. 1.38). A practical DCmotor has a large number of coils on the rotor,each one connected not only to its own pair ofcommutator segments but to the other coils aswell. In this way, the chief contribution to torque ismade by a coil operating close to its peak-torqueposition. There is also an averaging effect producedby current flowing in all the other coils, so theresulting torque ripple is very small.

Brushless MotorsBefore we talk about brushless motors in detail,let’s clear up a few points about terminology. Theterm “brushless” has become accepted as referringto a particular variety of servo motor. Clearly a stepmotor is a brushless device, as is an AC inductionmotor (in fact, the step motor can form the basis ofa brushless servo motor, often called a hybridservo, which is discussed later). However, the so-called “brushless” motor has been designed to havea similar performance to the DC brush servowithout the limitations imposed by a mechanicalcommutator.

Within the brushless category are two basic motortypes: trapezoidal and sine wave motors. Thetrapezoidal motor is really a brushless DC servo,whereas the sine wave motor bears a closeresemblance to the AC synchronous motor. To fullyexplain the difference between these motors, wemust review the evolution of the brushless motor.

Fig. 1.35 Conventional DC brush motor

Commutator-+

N S

-+

Reversing Switch

S

N

Drive

Commutation Encoder

NS

A18

Motor TechnologiesFig. 1.41 Position of rotor at commutation point

The torque characteristic in Fig. 1.39 indicates thatmaximum torque is produced when the rotor andstator fields are at 90° to each other. Therefore, togenerate constant torque we would need to keepthe stator field a constant 90° ahead of the rotor.Limiting the number of phases to three means thatwe can only advance the stator field in incrementsof 60° (Fig. 1.40). This means we must keep thestator field in the same place during 60° of shaftrotation. So we can’t maintain a constant 90°torque angle, but we can maintain an average of90° by working between 60° and 120°. Fig. 1.41shows the rotor position at a commutation point.When the torque angle has fallen to 60°, the statorfield is advanced from 1 to 2 so that the angle nowincreases to 120°, and it stays here during the next60° of rotation.

Fig. 1.38 3-phase brushless motor

We would like to reproduce a similar situation in thebrushless motor; however, this would require alarge number of coils distributed around the stator.This may be feasible, but each coil would require itsown individual drive circuit. This is clearlyprohibitive, so a compromise is made. A typicalbrushless motor has either two or three sets of coilsor “phases” (see Fig. 1.38). The motor shown in Fig.1.38 is a two-pole, three-phase design. The rotorusually has four or six rotor poles, with acorresponding increase in the number of statorpoles. This doesn’t increase the number ofphases—each phase has its turns distributedbetween several stator poles.

Fig. 1.39 Position-torque characteristic

Fig. 1.40 Stator field positions for differentphase currents

A2

C1

B2

B1

C2

A1

B2

C1

A2

B1

C2

A1

Direction of Rotor Field Relative to Stator Field

180°90°-

+Torque

0°

I

N

NS

S

NN

SS

C1

A2B2

B1

C2

A1

SS

NN

B2

C1

A2

B1

C2

A1

B2

C1

A2

B1

C2

A1

A1

C2

B1

A2

C1

B2

Stator Field

Stator Field

Stator Field

I

I

A1

C2

B1B2

A2

C1

C1

A2

B2B1

C2

A1

N

S

Average Lag = 90°

120°

Stator Field

Rotor Field

Rotation

Stator Field

60°

A19

AE

ngin

eeri

ng R

efe

rence

Motor TechnologiesThe Trapezoidal MotorWith a fixed current level in the windings, the use ofthis extended portion of the sinusoidal torquecharacteristic gives rise to a large degree of torqueripple. We can minimize the effect by manipulatingthe motor design to “flatten out” the characteristic –to make it trapezoidal, (Fig. 1.42). In practice, this isnot very easy to do, so some degree of non-linearitywill remain. The effect of this tends to be a slight“kick” at the commutation points, which can benoticeable when the motor is running very slowly.

Fig. 1.42 Trapezoidal motor characteristic

Torque ripple resulting from non-linearity in thetorque characteristic tends to produce a velocitymodulation in the load. However, in a system usingvelocity feedback the velocity loop will generallyhave a high gain. This means that a very smallincrease in velocity will generate a large error signal,reducing the torque demand to correct the velocitychange. So in practice, the output current from theamplifier tends to mirror the torque characteristic(Fig. 1.43) so that the resulting velocity modulationis extremely small.

Fig. 1.43 Current profile in velocity-controlledservo

The Sine Wave MotorIn the sine wave motor (sometimes called an ACbrushless servo), no attempt is made to modify thebasic sinusoidal torque characteristic. Such a motorcan be driven like an AC synchronous motor byapplying sinusoidal currents to the motor windings.These currents must have the appropriate phasedisplacement, 120° in the case of the three-phasemotor. We now need a much higher resolutiondevice to control the commutation if we wantsmooth rotation at low speeds. The drive needs togenerate 3 currents that are in the correctrelationship to each other at every rotor position. Sorather than the simple commutation encodergenerating a handful of switching points, we nowneed a resolver or high-resolution optical encoder.In this way, it’s possible to maintain a 90° torque

angle very accurately, resulting in very smooth low-speed rotation and negligible torque ripple. Asimplified explanation of why the sine wave motorproduces constant torque is given in the nextsection.

The drive for a sine wave motor is more complexthan for the trapezoidal version. We need areference table from which to generate thesinusoidal currents, and these must be multiplied bythe torque demand signal to determine theirabsolute amplitude. With a star-connected three-phase motor, it is sufficient to determine thecurrents in two of the windings—this willautomatically determine what happens in the third.As previously mentioned, the sine wave motorneeds a high-resolution feedback device. However,this device can also provide position and velocityinformation for the controller.

Why constant torque from a sine wavemotor?To understand this, it’s easier to think in terms of atwo-phase motor. This has just two sets ofwindings that are fed with sinusoidal currents at 90°to each other. If we represent shaft position by anangle θ, then the currents in the two windings are ofthe form Isinθ and Icosθ.

Going back to our original motor model, you’llremember that the fundamental torquecharacteristic of the motor is also sinusoidal. So fora given current I, the instantaneous torque valuelooks like:

T = I KT sinθWhere KT is the motor torque constant

By making the motor current sinusoidal as well, andin phase with the motor torque characteristic, thetorque generated by one phase becomes:

T1 = (I sinθ) KT sinθ= I KT sin2θ

Similarly, the torque produced by the other phaseis:

T2 = I KT cos2θThe total torque is:

T1 + T2 = I KT (sin2θ + cos2θ)but: sin2θ + cos2θ = 1 for any value of θtherefore: T1 + T2 = IKT

So for sinusoidal phase currents with a constantamplitude, the resultant torque is also constant andindependent of shaft position.

For this condition to remain true, the drive currentsmust accurately follow a sine-cosine relationship.This can only occur with a sufficiently highresolution in the encoder or resolver used forcommutation.

60°

60°

Current

(Torque)

A20

Motor Technologies

Stator

Rotor

Two MS StyleConnectors

Position FeedbackDevice Rotor

Bearing

Position FeedbackDevice Stator

Housing

The Hybrid ServoIn terms of their basic operation, the step motorand the brushless servo motor are identical. Theyeach have a rotating magnet system and a woundstator. The only difference is that one has morepoles than the other, typically two or three pole-pairs in the brushless servo and 50 in the stepper.You could use a brushless servo as a stepper – nota very good one, since the step angle would belarge. But by the same token, you can also use astepper as a brushless servo by fitting a feedbackdevice to perform the commutation. Hence the“hybrid servo”, so called because it is based on ahybrid step motor (Fig. 1.44). These have also beendubbed ‘stepping servos’ and ‘closed-loopsteppers’. We prefer not to use the term ‘stepper’at all since such a servo exhibits none of theoperating characteristics of a step motor.

The hybrid servo is driven in precisely the samefashion as the brushless motor. A two-phase driveprovides sine and cosine current waveforms inresponse to signals from the feedback device. Thisdevice may be an optical encoder or a resolver. Sincethe motor has 50 pole pairs, there will be 50 electricalcycles per revolution. This conveniently permits a 50-cycle resolver to be constructed from the same rotorand stator laminations as the motor itself.

A hybrid servo generates approximately the sametorque output as the equivalent step motor,assuming the same drive current and supplyvoltage. However, the full torque capability of themotor can be utilized since the system is operatingin a closed loop (with an open-loop step motor, it isalways necessary to allow an adequate torquemargin). The hybrid servo system will be moreexpensive than the equivalent step motor systems,but less costly than a brushless servo. As with thestep motor, continuous operation at high speed isnot recommended since the high pole count resultsin greater iron losses at high speeds. A hybrid servoalso tends to run quieter and cooler than its stepmotor counterpart; since it is a true servo, power isonly consumed when torque is required andnormally no current will flow at standstill. Low-speed smoothness is vastly improved over theopen-loop full step motor.

It is worth noting that the hybrid servo is entirelydifferent from the open-loop step motor operated in‘closed loop’ or ‘position tracking’ mode. In positiontracking mode, an encoder measures the loadmovement and final positioning is determined byencoder feedback. While this technique can providehigh positioning accuracy and eliminatesundetected position loss, it does not allow fulltorque utilization, improve smoothness or reducemotor heating.

Fig. 1.44 Hybrid servo motor with resolver feedback

A21

AE

ngin

eeri

ng R

efe

rence

Motor TechnologiesDirect Drive MotorsMotor Construction and OperationDirect drive systems couple the system’s loaddirectly to the motor without the use of belts orgears. In some situations, brushed or brushlesssservo motors may lack adequate torque orresolution to satisfy some applications’ needs.Therefore, mechanical means, such as gearreduction systems to increase torque andresolution, are used to meet system requirements.The Dynaserv Direct Drive can provide very hightorque in a modest package size and solves manyof the performance issues of the gear reducer. All ina system that is as easy to use as a steppingmotor.

Fig. 1.45 below shows the construction of theDynaserv DM Series direct drive motor comparedto a conventional motor with a gear reducer. Thegear reducer relies on large amounts of frictionalcontact to reduce the speed of the load. Thisgearing effectively increases torque and resolutionbut sacrifices speed and accuracy. The direct drivemotor is brushless and gearless so it eliminatesfriction from its power transmission Since thefeedback element is coupled directly to the load,system accuracy and repeatability are greatlyincreased and backlash is eliminated.

Fig. 1.45 Construction comparison

The motor contains precision bearings, magneticcomponents and integral feedback in a compactmotor package (see Fig. 1.46). The motor is anouter rotor type, providing direct motion of theoutside housing of the motor and thus the load.The cross roller bearings that support the rotorhave high stiffness, to allow the motor to beconnected directly to the load. In most cases, it isnot necessary to use additional bearings orconnecting shafts.

Fig. 1.46 Expanded motor view—Dynaserv Model DM

The torque is proportional to the square of the sumof the magnetic flux (Øm), of the permanent magnetrotor and the magnetic flux (Øc), of the statorwindings. See Fig. 1.47. High torque is generateddue to the following factors. First, the motordiameter is large. The tangential forces betweenrotor and stator act as a large radius, resulting inhigher torque. Secondly, a large number of smallrotor and stator teeth create many magnetic cyclesper motor revolution. More working cycles meansincreased torque.

Fig. 1.47 Dynaserv magnetic circuit

Hub

Rotor Core

Stator Core

Retaining Ring

Housing KitClamp Ring

LED KitEncoder Plate PDA Kit

Encoder

Housing

Core

T

Rotor

Excitation Coil

Permanent Magnet

Stator AStator B

Φ mΦm

Φc

Gear Reducer

Conventional Motor Direct Drive Motor

DC/AC Motor

Encoder Rotor Core

Encoder PCB

Bearing Rotating Element

Stator Core

Slit Plate

Stator Element

A22

Motor TechnologiesDirect Drive Motor AdvantagesHigh PrecisionDynaserv motors eliminate the backlash orhysteresis inevitable in using any speed reducer.Absolute positioning of 30 arc-sec is typical with arepeatability of ±2 arc-sec.

Faster Settling TimeThe Dynaserv system reduces machine cycle timesby decreasing settling times. This result is realizedbecause of the “gearless” design and sophisticated“I-PD” control algorithm.

High Torque at High SpeedThe torque/speed curve of the various Dynaservmodels is very flat. This results in high accelerationat high speeds (4.0 rps) with good controllability.

Smooth RotationThe very low velocity and torque ripple of theDynaserv contribute to its excellent speedcontrollability with a more than 1:1,000 speed ratio.

Optimum TuningDynaserv systems offer the user a tuning mode thatsimplifies the setting of optimum parameters for theactual load. Turning on the “test” switch on thefront panel of the drive produces a test signal.Using an oscilloscope, the gain settings are quicklyoptimized by adjusting the digital switches andpotentiometers on the front panel.

Clean OperationThe Dynaserv system is brushless and gearless,which results in a maintenance-free operation. Withpreparation, the Dynaserv can operate in class 10environments.

Fig. 1.48 Dynaserv velocity/torque ripple

Conditions• Load 30 x Rotor Inertia • Rotation: CW • Speed Mode15

10

5

3

0°0.2 0.4 0.6 0.8 1.0 1.2

Speed Ripple (DM1150A)

Revolution (rps)

Rip

ple

(%)

Torque Ripple (DM1015A)

Rotational Angle (degrees)

90° 180° 270° 360°

20

15.3

1514.7

0

5

5%

Torq

ue (N

• m

)

A23

AE

ngin

eeri

ng R

efe

rence

Drive TechnologiesStepping Motor DrivesThe stepper drive delivers electrical power to themotor in response to low-level signals from thecontrol system.

The motor is a torque-producing device, and thistorque is generated by the interaction of magneticfields. The driving force behind the stator field is themagneto-motive force (MMF), which is proportionalto current and to the number of turns in thewinding. This is often referred to as the amp-turnsproduct. Essentially, the drive must act as a sourceof current. The applied voltage is only significant asa means of controlling the current.

Input signals to the stepper drive consist of steppulses and a direction signal. One step pulse isrequired for every step the motor is to take. This istrue regardless of the stepping mode. So the drivemay require 200 to 101,600 pulses to produce onerevolution of the shaft. The most commonly-usedstepping mode in industrial applications is the half-step mode in which the motor performs 400 stepsper revolution. At a shaft speed of 1800 rpm, thiscorresponds to a step pulse frequency of 20kHz.The same shaft speed at 25,000 steps per revrequires a step frequency of 750 kHz, so motioncontrollers controlling microstep drives must beable to output a much higher step frequency.

Fig. 2.1 Stepper drive elements

The simplest type of switch set is the unipolararrangement shown in Fig. 2.2. It is referred to as aunipolar drive because current can only flow in onedirection through any particular motor terminal. Abifilar-wound motor must be used since reversal ofthe stator field is achieved by transferring current tothe second coil. In the case of this very simpledrive, the current is determined only by the motorwinding resistance and the applied voltage.

Fig. 2.2 Basic unipolar drive

Such a drive will function perfectly well at lowstepping rates, but as speed is increased, thetorque will fall off rapidly due to the inductance ofthe windings.

The logic section of the stepper drive is oftenreferred to as the translator. Its function is totranslate the step and direction signals into controlwaveforms for the switch set (see Fig. 2.1). Thebasic translator functions are common to mostdrive types, although the translator is necessarilymore complex in the case of a microstepping drive.However, the design of the switch set is the primefactor in determining drive performance, so we willlook at this in more detail.

Translator Switch Set

Step

Direction

Stepper Drive Elements

Motor

Phase 1

Phase 2

Step

Direction

2B1A 2A

V+

0V

TR1 TR2 TR3 TR4

1B

A24

Drive TechnologiesInductance/Water AnalogyFor those not familiar with the property ofinductance, the following water analogy may beuseful (Fig. 2.3). An inductor behaves in the sameway as a turbine connected to a flywheel. When thetap is turned on and pressure is applied to the inletpipe, the turbine will take time to accelerate due tothe inertia of the flywheel. The only way to increase

the acceleration rate is to increase the appliedpressure. If there is no friction or leakage loss in thesystem, acceleration will continue indefinitely for aslong as the pressure is applied. In a practical case,the final speed will be determined by the appliedpressure and by friction and the leakage past theturbine blades.

Applying a voltage to the terminals of an inductorproduces a similar effect. With a pure inductance(i.e., no resistance), the current will rise in a linearfashion for as long as the voltage is applied. Therate of rise of current depends on the inductanceand the applied voltage, so a higher voltage mustbe applied to get the current to rise more quickly. Ina practical inductor possessing resistance, the finalcurrent is determined by the resistance and theapplied voltage.

Once the turbine has been accelerated up tospeed, stopping it again is not a simple matter. Thekinetic energy of the flywheel has to be dissipated,and as soon as the tap is turned off, the flywheeldrives the turbine like a pump and tries to keep thewater flowing. This will set up a high pressureacross the inlet and outlet pipes in the reversedirection. The equivalent energy store in theinductor is the magnetic field. As this fieldcollapses, it tries to maintain the current flow bygenerating a high reverse voltage.

By including a one-way valve across the turbineconnections, the water is allowed to continuecirculating when the tap is turned off. The energystored in the flywheel is now put to good use inmaintaining the flow. We use the same idea in therecirculating chopper drive, in which a diode allowsthe current to recirculate after it has built up.

Going back to our simple unipolar drive, if we lookat the way the current builds up (Fig. 2.4) we cansee that it follows an exponential shape with its finalvalue set by the voltage and the winding resistance.To get it to build up more rapidly, we could increasethe applied voltage, but this would also increase thefinal current level. A simple way to alleviate thisproblem is to add a resistor in series with the motorto keep the current the same as before.

Fig. 2.3 Inductance water analogy

Current (Flow)

Voltage (Pressure)

Reverse Pressure When Flow Interrupted

Higher Pressure Causes Flywheel to Accelerate More Rapidly

I

Tap

1-Way Valve

Water Flow Equivalent to Current

Turbine

Kinetic Energy of Flywheel Equivalent to Energy Stored in Magnetic Field

Pressure Equivalent to Applied Voltage

A25

AE

ngin

eeri

ng R

efe

rence

Drive TechnologiesBipolar DriveAn obvious possibility is the simple circuit shown inFig. 2.6, in which two power supplies are usedtogether with a pair of switching transistors.Current can be made to flow in either directionthrough the motor coil by turning on one transistoror the other. However, there are distinct drawbacksto this scheme. First, we need two power supplies,both of which must be capable of delivering thetotal current for both motor phases. When all thecurrent is coming from one supply the other isdoing nothing at all, so the power supply utilizationis poor. Second, the transistors must be rated atdouble the voltage that can be applied across themotor, requiring the use of costly components.

Fig. 2.6 Simple bipolar drive

Fig. 2.8 Bipolar R-L drive

The standard arrangement used in bipolar motordrives is the bridge system shown in Fig. 2.7.Although this uses an extra pair of switchingtransistors, the problems associated with theprevious configuration are overcome. Only onepower supply is needed and this is fully utilized;transistor voltage ratings are the same as thatavailable for driving the motor. In low-powersystems, this arrangement can still be used withresistance limiting as shown in Fig. 2.8.

Fig. 2.7 Bipolar bridge

Unipolar Drive

Fig. 2.5 Basic unipolar drive

R-L DriveThe principle described in the Inductance/WaterAnalogy (p. A24) is applied in the resistance-limited(R-L) drive see Fig. 2.4. Using an applied voltage of10 times the rated motor voltage, the current willreach its final value in one tenth of the time. If youlike to think in terms of the electrical time constant,this has been reduced from L/R to L/10R, so we’llget a useful increase in speed. However we’repaying a price for this extra performance. Understeady-state conditions, there is 9 times as muchpower dissipated in the series resistor as in themotor itself, producing a significant amount of heat.Furthermore, the extra power must all come fromthe DC power supply, so this must be much larger.R-L drives are therefore only suited to low-powerapplications, but they do offer the benefits ofsimplicity, robustness and low radiated interference.

Fig. 2.4 Principle of the R-L drive

RL I

RL IR

V

2V I

2V

V

I V

R

2V

2R

2B1A 2A

V+

0V

TR1 TR2 TR3 TR4

1B

A drawback of the unipolar drive is its inability toutilize all the coils on the motor. At any one time,there will only be current flowing in one half of eachwinding. If we could utilize both sections at thesame time, we could get a 40% increase in amp-turns for the same power dissipation in the motor.

To achieve high performance and high efficiency,we need a bipolar drive (one that can drive currentin either direction through each motor coil) and abetter method of current control. Let’s look first athow we can make a bipolar drive.

TR1

V+

0V

V-

TR2

TR1

TR2

TR3

TR4

V+

0V

V+

0V

TR2

TR1

TR4

TR3

A26

Drive TechnologiesRecirculating Chopper DriveThe method of current control used in most stepperdrives is the recirculating chopper (Fig. 2.9). Thisapproach incorporates the four-transistor bridge,recirculation diodes, and a sense resistor. Theresistor is of low value (typically 0.1 ohm) andprovides a feedback voltage proportional to thecurrent in the motor.

Fig. 2.9 Recirculating chopper drive

Current is injected into the winding by turning onone top switch and one bottom switch, and thisapplies the full supply voltage across the motor.Current will rise in an almost linear fashion and wecan monitor this current by looking across thesense resistor. When the required current level hasbeen reached, the top switch is turned off and thestored energy in the coil keeps the currentcirculating via the bottom switch and the diode.Losses in the system cause this current to slowlydecay, and when a pre-set lower threshold isreached, the top switch is turned back on and thecycle repeats. The current is therefore maintained atthe correct average value by switching or“chopping” the supply to the motor.

This method of current control is very efficientbecause very little power is dissipated in theswitching transistors other than during the transientswitching state. Power drawn from the powersupply is closely related to the mechanical powerdelivered by the shaft (unlike the R-L drive, whichdraws maximum power from the supply atstandstill).

A variant of this circuit is the regenerative chopper.In this drive, the supply voltage is applied acrossthe motor winding in alternating directions, causingthe current to ramp up and down at approximatelyequal rates. This technique tends to require fewercomponents and is consequently lower in cost,however, the associated ripple current in the motoris usually greater and increases motor heating.

TR3TR1

D1 D2TR2 TR4

Vs

Rs

TR3TR1

D1 D2TR2 TR4

Vs

Rs

Injection Recirculation

Motor current

+ +

– –

A27

AE

ngin

eeri

ng R

efe

rence

Drive TechnologiesRegeneration and Power DumpingLike other rotating machines with permanentmagnets, the step motor will act as a generatorwhen the shaft is driven mechanically. This meansthat the energy imparted to the load inertia duringacceleration is returned to the drive duringdeceleration. This will increase the motor currentand can damage the power switches if the extracurrent is excessive. A threshold detector in thedrive senses this increase in current andmomentarily turns off all the bridge transistors(Fig. 2.10). There is now a path for the regeneratedcurrent back to the supply capacitor, where itincreases the supply voltage. During this phase, thecurrent is no longer flowing through the senseresistors, so the power switches must be turned onagain after a short period (typically 30µS) forconditions to be reassessed. If the current is still toohigh, the drive returns to the regenerative state.

Fig. 2.10 Current flow during regeneration

A small increase in supply voltage duringregeneration is acceptable, but if the rise is toogreat the switches may be damaged by over-voltage rather than excessive current. To resolvethis problem, we use a power dump circuit thatdissipates the regenerated power.

The circuit of a simple power dump is shown inFig. 2.11. A rectifier and capacitor fed with AC fromthe supply transformer provide a reference voltageequal to the peak value of the incoming AC. Undernormal conditions this will be the same as the drivesupply voltage. During excess regeneration, thedrive supply voltage will rise above this reference,and this will turn on the dump transistor connectingthe 33-ohm resistor across the power supply.When the supply voltage has decreased sufficiently,the transistor is turned back off. Although theinstantaneous current flowing through the dumpresistor may be relatively high, the average powerdissipated is usually small since the dump period isvery short. In applications where the regeneratedpower is high, perhaps caused by frequent andrapid deceleration of a high inertia, a supplementaryhigh-power dump resistor may be necessary.

Fig. 2.11 Power dump circuit

+V

Power supply capacitor

Powerdump circuit

AC in

C1 R1

1K

R2D1

D2

HV

R63310W

TR1

R5

100K

R3

R4

TR2

0V

Ω

A28

Drive TechnologiesStepper Drive Technology OverviewWithin the various drive technologies, there is aspectrum of performance. The uni-polar resistance-limited (R-L) drive is a relatively simple design, but itlacks shaft power performance and is very inefficient.A uni-polar system only uses half of the motorwinding at any instant. A bi-polar design allowstorque producing current to flow in all motorwindings, using the motor more efficiently, butincreasing the complexity of the drive. A bi-polar R-Ldrive improves shaft performance, but is still veryinefficient—generating a lot of wasted heat. Analternative to resistance-limiting is to control currentby means of chopper regulation. A chopperregulator is very efficient since it does not wastepower by dropping voltage through a resistor.However, good current control in the motor isessential to deliver optimum shaft power. Pulse widthmodulation (PWM) and threshold modulation are twotypes of chopper regulation techniques. PWMcontrols the average of the motor current and is verygood for precise current control, while thresholdmodulation controls current to a peak level.Threshold modulation can be applied to a widerrange of motors, but it does suffer greater loss ofperformance than PWM when the motor has a largeresistance or long motor cables are used. Bothchopper regulation techniques can use recirculatingcurrent control, which improves the powerdissipation in the motor and drive and overall systemefficiency. As system performance increases, thecomplexity and cost of the drive increases.

Stepper drive technology has evolved—being drivenby machine builders that require more shaft power insmaller packages, higher speed capability, betterefficiency, and improved accuracy. One trend of thetechnology is towards microstepping, a techniquethat divides each full step of the motor into smallersteps. This is achieved electronically in the drive byproportioning the current between the motorwindings. The higher the resolution, the moreprecision is required in the current control circuits. Inits simplest form, a half-step system increases theresolution of a standard 1.8° full-step motor to 400steps/rev. Ministepping drives have more precisecurrent control and can increase the resolution to4,000 steps/rev. Microstep drives typically haveresolutions of 50,000 steps/rev, and in addition toimproved current control, they often haveadjustments to balance offsets between each phaseof the motor and to optimize the current profile forthe particular motor being used.

Full-Step and Half-Step SystemsFull-step and half-step systems do not have theresolution capability of the ministepping ormicrostepping systems. However, the drivetechnology is not as complex and the drives arerelatively inexpensive. Full-step and half-step systemswill not have the same low-speed smoothness ashigher resolution systems.

An inherent property of a stepper motor is its low-speed resonance, which may de-synchronize amotor and cause position loss. Full-step and half-step drives are more prone to resonance effects andthis may limit their application in low-speed systems.Full-step and half-step systems can be operated atspeeds above the motor’s resonant speed withoutloss of synchronization. For this reason, full-step andhalf-step systems are normally applied in high-speed,point-to-point positioning applications. In these typesof applications, the machine designer is primarilyconcerned with selecting a motor/drive systemcapable of producing the necessary power output.

Since power is the product of torque and speed, ahigh-torque system with low-speed capability maynot produce as much power as a low-torque, high-speed system. Sizing the system for torque only maynot provide the most cost-effective solution, selectinga system based on power output will make the mostefficient use of the motor and drive.

Step motor systems typically require the motor toaccelerate to reach high speed. If a motor wasrequested to run instantaneously at 3000 rpm, themotor would stall immediately. At slow speeds, it ispossible to start the motor without position loss byapplying unramped step pulses. The maximum speedat which synchronization will occur without ramping iscalled the start/stop velocity. The start/stop velocity isinversely proportional to the square-root of the totalinertia. The start/stop capability provides a benefit forapplications that require high-speed point-to-pointpositioning—since the acceleration to the start/stopvelocity is almost instantaneous, the move-time willbe reduced. No additional time is required toaccelerate the motor from zero to the start/stopvelocity. While the move-time can be reduced, it isgenerally more complicated for the controller orindexer to calculate the motion profile and implementa start/stop velocity. In most applications, using start/stop velocities will eliminate the need to run the motorat its resonant frequency and prevent de-synchronization.

Velo

city

Time

Velo

city

Time

Ministep SystemsApplications that require better low-speedsmoothness than a half-step system shouldconsider using a microstepping or ministeppingsolution. Microstepping systems, with resolutionsof 50,000 steps/rev, can offer exceptionalsmoothness, without requiring a gear-reducer.Ministepping systems typically do not have wave-trimming capability or offset adjustment to achievethe optimum smoothness, but offer a greatimprovement over full-step and half-step systems.Ministepping systems have resolutions between1,000 and 4,000 steps/rev.

The motor is an important element in providinggood smoothness. Some motor designs areoptimized for high-torque output rather thansmooth rotation. Others are optimized forsmoothness rather than high torque. Ministeppingsystems are typically offered with a motor as a“packaged” total solution, using a motor that hasbeen selected for its premium smoothnessproperties.

Ministep systems are sometimes selected toimprove positional accuracy. However, with anopen-loop system, friction may prevent thetheoretical unloaded accuracy from being achievedin practice.

A29

AE

ngin

eeri

ng R

efe

rence

Drive TechnologiesMicrostepping DrivesAs we mentioned earlier, subdivision of the basicmotor step is possible by proportioning the currentin the two motor windings. This produces a seriesof intermediate step positions between the one-phase-on points. It is clearly desirable that theseintermediate positions are equally spaced andproduce approximately equal torque when themotor is running.

Accurate microstepping places increased demandson the accuracy of current control in the drive,particularly at low current levels. A small phaseimbalance that may be barely detectable in a half-step drive can produce unacceptable positioningerrors in a microstep system. Pulse-widthmodulation is frequently used to achieve higheraccuracy than can be achieved using a simplethreshold system.

The phase currents necessary to produce theintermediate steps follow an approximatelysinusoidal profile as shown in Fig. 2.12. Howeverthe same profile will not give the optimum responsewith all motors. Some will work well with asinusoidal shape, whereas others need a morefilled-out or trimmed-down shape (Fig. 2.12). So amicrostep drive intended to operate with a varietyof motors needs to have provision for adjusting thecurrent profile. The intermediate current levels areusually stored as data in an EPROM, with somemeans of selecting alternative data sets to givedifferent profiles. The change in profile may bethought of in terms of adding or subtracting athird-harmonic component to or from the basic sinewave.

Fig. 2.12 Microstep current profile

Due to this dependence on motor type forperformance, it is usual for high-resolutionmicrostep systems to be supplied as a matchedmotor-drive package.

The Stepper Torque/Speed CurveWe have seen that motor inductance is the factorthat opposes rapid changes of current andtherefore makes it more difficult to drive a stepperat high speeds. Looking at the torque-speed curvein Fig. 2.13, we can see what is going on. At lowspeeds, the current has plenty of time to reach therequired level and so the average current in themotor is very close to the regulated value from thedrive. Changing the regulated current setting orchanging to a drive with a different current ratingwill affect the available torque accordingly.

Fig 2.13 Regulated and voltage-limited regionsof the torque-speed curve

In the case of high-resolution microstep drivesproducing 10,000 steps per rev or more, the bestperformance will only be obtained with a particulartype of motor. This is one in which the stator teethare on a 7.5° pitch, giving 48 equal pitches in 360°.In most hybrid steppers, the stator teeth have thesame pitch as the rotor teeth, giving equalincrements of 7.2°. This latter arrangement tends togive superior torque output, but is less satisfactoryas a microstepper since the magnetic poles are“harder” – there is no progressive transfer of toothalignment from one pole to the next. In fact, withthis type of motor, it can be quite difficult to find acurrent profile that gives good static positioningcombined with smooth low-speed rotation. Analternative to producing a 7.5°-pitch stator is toincorporate a slight skew in the rotor teeth. Thisproduces a similar effect and has the benefit ofusing standard 7.2° laminations throughout.Skewing is also used in DC brush motors as ameans of improving smoothness.

As speed increases, the time taken for the currentto rise becomes a significant proportion of theinterval between step pulses. This reduces theaverage current level, so the torque starts to fall off.As speed increases further, the interval betweenstep pulses does not allow the current time to reacha level where the chopping action can begin. Underthese conditions, the final value of current dependsonly on the supply voltage. If the voltage isincreased, the current will increase more rapidly andhence will achieve a higher value in the availabletime. So this region of the curve is described as“voltage limited”, as a change in the drive currentsetting would have no effect. We can conclude thatat low speeds the torque depends on the drivecurrent setting, whereas at high speeds it dependson the drive supply voltage. It is clear that high-speed performance is not affected by the drivecurrent setting. Reducing the current simply“flattens out” the torque curve without restrictingthe ability to run at high speeds. When performanceis limited by the available high-speed torque, thereis much to be said for running at the lowest currentthat gives an adequate torque margin. In general,dissipation in motor and drive is reduced and low-speed performance in particular will be smootherwith less audible noise.

Sinewave Filled out Trimmed

Average Current During Pulse

Speed

Drive with Higher Supply

Voltage

Voltage-Limited RegionRegulated Region

Torq

ue

A30

Drive TechnologiesHowever, having doubled the effective number ofturns in the winding means that we have alsoincreased the inductance by a factor of 4. Thiscauses the torque to drop off much more rapidly asspeed is increased, and as a result, the seriesmode is most useful at low speeds. The maximumshaft power obtainable in series is typically half thatavailable in parallel (using the same current settingon the drive).

Connecting the two half-windings of an 8-leadmotor in parallel allows the current to divide itselfbetween the two coils. It does not change theeffective number of turns and the inductancetherefore remains the same. So at a given drivecurrent, the torque characteristic will be the samefor two half-windings in parallel as for one of thewindings on its own. For this reason, “parallel” inthe context of a 6-lead motor refers to the use ofone half-winding only.

As has already been mentioned, the current ratingof a step motor is determined by the allowabletemperature rise. Unless the motor manufacturer’sdata states otherwise, the rating is a “unipolar”value and assumes both phases of the motor areenergized simultaneously. So a current rating of 5Ameans that the motor will accept 5A flowing in eachhalf-winding.

When the windings of an 8-lead motor areconnected in parallel, half of the total resistance isproduced. For the same power dissipation in themotor, the current may now be increased by 40%.Therefore, the 5A motor will accept 7A with thewindings in parallel, giving a significant increase inavailable torque. Conversely, connecting thewindings in series will double the total resistanceand the current rating is reduced by a factor of 1.4,giving a safe current of 3.5A for our 5A-motor inseries.

As a general rule, parallel is the preferredconnection method as it produces a flatter torquecurve and greater shaft power (Fig. 2.15). Series isuseful when high torque is required at low speeds,and it allows the motor to produce full torque froma lower-current drive. Care should be taken to avoidoverheating the motor in series since its currentrating is lower in this mode. Series configurationsalso carry a greater likelihood of resonance due tothe high torque produced in the low-speed region.

With a bipolar drive, alternative possibilities exist forthe motor connections as shown in Fig. 2.14. An8-lead motor can be connected with the two halvesof each winding either in series or in parallel. Witha 6-lead motor, either one half-winding or bothhalf-windings may be connected in series. Thealternative connection schemes produce differenttorque-speed characteristics and also affect themotor’s current rating.

Fig. 2.14 Series & parallel connections

Fig. 2.15 Series & parallel torque/speed curves

ParallelSeries

1A 1B 2A 2B 1A 1B 2A 2B

SeriesParallel

Speed

Torq

ue

Compared with using one half-winding only,connecting both halves in series requires the drivecurrent to flow through twice as many turns. For thesame current, this doubles the “amp-turns” andproduces a corresponding increase in torque. Inpractice, the torque increase is seldom as high as100% due to the non-linearity of the magneticmaterial. Equally, the same torque will be produced athalf the drive current when the windings are in series.

A31

AE

ngin

eeri

ng R

efe

rence

Drive TechnologiesDC Brush Motor DrivesLinear and Switching AmplifiersLinear amplifiers – this type of amplifier operates insuch a way that, depending on the direction ofmotor rotation, either TR1 or TR2 will be in serieswith the motor and will always have a voltage (V)developed across it (Fig. 2.16).

This characteristic is the primary limitation on theuse of linear amplifiers (since there will always bepower dissipated in the output stages of theamplifier). To dissipate this power, large transistorsand heat sinks will be required, making this type ofamplifier unsuitable for use in high power systems.However, the linear amplifier does offer the benefitof low radiated electrical noise.

Fig. 2.16 Linear servo amplifier

Overview – The Analog DriveIn the traditional analog drive, the desired motorvelocity is represented by an analog input voltageusually in the range ±10 volts. Full forward velocityis represented by +10v, and full reverse by -10v.Zero(D) volts represents the stationary condition andintermediate voltages represent speeds inproportion to the voltage.

The various adjustments needed to tune an analogdrive are usually made with potentiometers. With alittle experience, this can usually be performedquite quickly, but in some difficult applications itmay take longer. Repeating the adjustments onsubsequent units may take the same time unlessthere is an easy way of duplicating the poten-tiometer settings. For this reason, some proprietarydrives use a plug-in “personality card” that may befitted with preset components. However, this notonly increases the cost but may preclude thepossibility of fine tuning later.

Overview – The Digital DriveAn alternative to the analog system is the digitally-controlled drive in which tuning is performed bysending data from a terminal or computer. This leadsto easy repetition between units and, since suchdrives are invariably processor-based, facilitatesfully-automatic self tuning. The input signal to such adrive may also be an analog voltage but can equallytake the form of step and direction signals, like astepper drive.

Digital drives are used more in conjunction withbrushless servo motors than with DC brush motors.Such drives are almost wholly digital with theexception of the power stage that actually deliverscurrent to the motor. Velocity feedback is derivedeither from an encoder or resolver and again isprocessed as digital information. It becomes logicalto incorporate a position controller within such adrive, so that incoming step and direction signalscan be derived from a conventional stepper-typeindexer. Equally, the positioner may be controlledby simple commands using a high-level motioncontrol language – see the X-code products in thiscatalog.

A Comparison of Analog and Digital DrivesThe analog drive offers the benefit of lower costand, in the case of a drive using tach feedback,very high performance. The wide bandwidth of thebrush tach allows high gains to be used withoutinducing jitter at standstill, resulting in a very “stiff”system.

The digital drive, while more costly, is comparativelyeasy to set up and adjustments can be quicklyrepeated across several units. Automatic self-tuningcan be a distinct advantage where the loadparameters are unknown or difficult to measure.The digital drive also offers the possibility ofdynamic tuning – sometimes vital where the loadinertia changes dramatically during machineoperation. Such an option is clearly out of thequestion with a potentiometer-adjusted drive.

Switching amplifiers – this amplifier is the mostcommonly used type for all but very low-powerrequirements and the most commonly usedmethod of output control is by pulse widthmodulation (PWM).

Power dissipation is greatly reduced since theoutput transistors are either in an “on” or an “off”state. In the “off” state, no current is conducted andso no power is dissipated. In the “on” state thevoltage across the transistors is very low (1-2 volts),so that the amount of power dissipated is small.

Such amplifiers are suitable for a wide range ofapplications (including high power applications).

The operation of a switching or chopper amplifier isvery similar to that of the chopping stepper drivealready described. Only one switch set is requiredto drive a DC brush motor, making the drivesimpler. However, the function of a DC drive is toprovide a variable current into the motor to controlthe torque. This may be achieved using eitheranalog or digital techniques.

Analog and Digital Servo DrivesUnlike stepper drives, amplifiers for both brush andbrushless servo motors are either analog or digital.The analog drive has been around for many years,whereas the digital drive is a relatively recentinnovation. Both types have their merits.

+ Ve

TR1

- Ve

TR2

V

M

A32

Drive Technologies

Motor velocity is measured by a tach generatorattached to the motor shaft. This produces avoltage proportional to speed that is compared withthe incoming velocity demand signal, and the resultof this comparison is a torque demand. If the speedis too low, the drive delivers more current, which inturn creates torque to accelerate the load. Similarly,if the speed is too high or the velocity demand isreduced, current flow in the motor will be reversedto produce a braking torque.

This type of amplifier is often referred to as a four-quadrant drive. This means that it can produceboth acceleration and braking torque in eitherdirection of rotation. If we draw a diagramrepresenting direction of rotation in one axis anddirection of torque in the other (see Fig. 2.18), youwill see that the motor can operate in all four“quadrants”. By contrast, a simple variable-speeddrive capable of running only in one direction andwith uncontrolled deceleration would be describedas single-quadrant.

Fig. 2.18 Four-quadrant operation

The velocity amplifier in Fig. 2.17 has a high gain sothat a small velocity difference will produce a largeerror signal. In this way, the accuracy of speedcontrol can be made very high even when there arelarge load changes.

A torque demand from the velocity amplifieramounts to a request for more current in the motor.The control of current is again achieved by afeedback loop that compares the torque demandwith the current in the motor. This current ismeasured by a sense resistor R, which produces avoltage proportional to motor current. This innerfeedback loop is frequently referred to as a torqueamplifier since its purpose is to create torque inresponse to a demand from the velocity amplifier.

The torque amplifier alone may be used as thebasis of a servo drive. Some types of positioncontroller generate a torque output signal ratherthan a velocity demand, and there are alsoapplications in which torque rather than speed is ofprimary interest (winding material onto a drum, forinstance). Most analog drives can be easilyconfigured either as velocity or torque amplifiers bymeans of a switch or jumper links. In practice, theinput signal is still taken to the same point, but thevelocity amplifier is bypassed.

Analog DC Drive OperationThe elements of an analog velocity amplifier areshown in Fig. 2.17. The function of the system is tocontrol motor velocity in response to an analoginput voltage.

Fig. 2.17 Elements of an analog servo system

M T

BA

R

Torque Feedback Loop

Velocity Feedback Loop

Velocity Control Signal

Torque Control Signal

Drive Amplifier

Torque

Torque

CW

CCW

CCW CW

Braking CCW

Accelerating CW

Braking CW

Accelerating CCW

A33

AE

ngin

eeri

ng R

efe

rence

Drive Technologies

Digital Servo Drive OperationFig. 2.19 shows the components of a digital drivefor a servo motor. All the main control functions arecarried out by the microprocessor, which drives aD-to-A convertor to produce an analog torquedemand signal. From this point on, the drive is verymuch like an analog servo amplifier.

Feedback information is derived from an encoderattached to the motor shaft. The encoder generatesa pulse stream from which the processor candetermine the distance travelled, and by calculatingthe pulse frequency it is possible to measurevelocity.

The digital drive performs the same operations asits analog counterpart, but does so by solving aseries of equations. The microprocessor isprogrammed with a mathematical model (or“algorithm”) of the equivalent analog system. Thismodel predicts the behavior of the system. Inresponse to a given input demand and outputposition. It also takes into account additionalinformation like the output velocity, the rate ofchange of the input and the various tuning settings.

The tuning of a digital servo is performed either bypushbuttons or by sending numerical data from acomputer or terminal. No potentiometeradjustments are involved. The tuning data is usedto set various coefficients in the servo algorithm andhence determines the behavior of the system. Evenif the tuning is carried out using pushbuttons, thefinal values can be uploaded to a terminal to alloweasy repetition.

In some applications, the load inertia variesbetween wide limits – think of an arm robot thatstarts off unloaded and later carries a heavy load atfull extension. The change in inertia may well be afactor of 20 or more, and such a change requiresthat the drive is re-tuned to maintain stableperformance. This is simply achieved by sendingthe new tuning values at the appropriate point in theoperating cycle.

Fig. 2.19 Digital servo drive

To solve all the equations takes a finite amountof time, even with a fast processor – this time istypically between 100µs and 2ms. During thistime, the torque demand must remain constantat its previously-calculated value and there willbe no response to a change at the input oroutput. This “update time” therefore becomes acritical factor in the performance of a digitalservo and in a high-performance system it mustbe kept to a minimum.

M E

AmplifierPWM Control

D to A ConverterMicroprocessor

Step

Direction

Tuning

RS-232C

Encoder

A34

Drive TechnologiesBrushless Motor DrivesThe trapezoidal driveFig. 2.20 shows a simplified layout of the drive for athree-phase trapezoidal motor. The switch set isbased on the familiar H-bridge, but uses three bridgelegs instead of two.The motor windings areconnected between the three bridge legs as shown,with no connection to the star point at the junction ofthe windings. By turning on the appropriate twotransistors in the bridge, current can be made to flowin either direction through any two motor windings.At any particular time, the required current pathdepends on rotor position and direction of rotation,so the bridge transistors are selected by logic drivenfrom the commutation encoder.

A PWM recirculating chopper system controls thecurrent in the same way as in the DC brush drivedescribed previously. The required currentfeedback information is provided by senseresistors connected in series with two of the motorleads. The voltage signals derived from theseresistors must be decoded and combined toprovide a useful current reference, and the circuitthat does this also uses the commutation encoderto determine how to interpret the information. Infact, this is not a simple process because the

relatively small feedback voltage (about 1V) mustbe separated from the large voltage excursionsgenerated by the chopping system (240V in thecase of a typical high-power drive).

The input stages of the brushless drive follow thesame pattern as a conventional analog brush drive(using a high-gain velocity amplifier that generatesthe torque demand signal). Velocity feedback canbe derived in a number of ways, but it is clearlydesirable to use a brushless method in conjunctionwith a brushless motor. Some motors incorporatea brushless tach generator that producesmulti-phase AC outputs. These signals have tobe processed in a similar way to the currentfeedback information using additional data from atach encoder. Again, this is not a particularlystraightforward process and it is difficult to obtaina smooth, glitch-free feedback signal. A moresatisfactory alternative is to use a high-resolutionoptical encoder and convert the encoder pulsefrequency to an analog voltage. The encodercan also be used as the feedback device for aposition controller.

Fig. 2.20 Simplified trapezoidal brushless servo drive

-

+Velocity Input

Velocity Feedback

Logic PWM & Circuit

Communication Encoder

Motor

Torque AmpVelocity Amp

-

+

Current Sense

Selector

A35

AE

ngin

eeri

ng R

efe

rence

Drive TechnologiesThe Sine Wave DriveSine wave brushless motors can be two- or three-phase, and the drive we'll look at is for the two-phase version (see Fig. 2.21). This uses two H-bridges to control current in the two motorwindings, and the power section of this driveclosely resembles a pair of DC brush drives. Bycontrast with the previous example, this drive usesa digital processor-based control section that takesits input in the form of step and direction signals.We need to generate currents in the two motorwindings that follow a sine and cosine pattern asthe shaft rotates. The drive shown in Fig. 2.21 usesa brushless resolver and a resolver-to-digitalconverter (RDC) to detect the shaft position. Fromthis, we will get a number that can be fed to areference table to determine the instantaneouscurrent values for that particular shaft position. Bearin mind that the reference table will only indicaterelative currents in the two windings—the absolute

values will depend on the torque demand at thetime. So the processor must multiply the sine andcosine values by the torque demand to get the finalvalue of current in each phase. The resultingnumbers are fed to D-to-A converters that producean analog voltage proportional to demandedcurrent. This is fed to the two PWM chopperamplifiers.

Commutation information for a sine wave drive mayalso be derived from an absolute or incrementaloptical encoder. An incremental encoder will be lessexpensive for the same resolution, but requiressome form of initialization at power-up to establishthe required 90° torque angle.

A “pseudo-sine wave” drive using feedback from alow-resolution absolute encoder can offer a cost-effective alternative. The pseudo sine wave drivegives superior performance to the trapezoidal driveat lower cost than the standard high-resolution sinewave system.

Fig. 2.21 Two-phase sine wave brushless drive

Micro- processor

D-to-A Converter

D-to-A Converter

PWM Control

PWM Control

H - Bridge

H - Bridge

Resolver to Digital Converter

Step

DirectionMotor

Resolver

A36

Servo TuningTuning a Servo SystemAny closed-loop servo system, whether analog ordigital, will require some tuning. This is the processof adjusting the characteristics of the servo so thatit follows the input signal as closely as possible.

Why is tuning necessary?A servo system is error-driven, in other words, theremust be a difference between the input and theoutput before the servo will begin moving to reducethe error. The “gain” of the system determines howhard the servo tries to reduce the error. A high-gainsystem can produce large correcting torques whenthe error is very small. A high gain is required if theoutput is to follow the input faithfully with minimalerror.

Now a servo motor and its load both have inertia,which the servo amplifier must accelerate anddecelerate while attempting to follow a change atthe input. The presence of the inertia will tend toresult in over-correction, with the system oscillatingor “ringing” beyond either side of its target (Fig. 3.1).This ringing must be damped, but too muchdamping will cause the response to be sluggish.When we tune a servo, we are trying to achieve thefastest response with little or no overshoot.

Fig. 3.1 System response characteristics

In practice, tuning a servo means adjustingpotentiometers in an analog drive or changing gainvalues numerically in a digital system. To carry outthis process effectively, it helps to understandwhat’s going on in the drive. Unfortunately, thetheory behind servo system behavior, though wellunderstood, does not reveal itself to most of uswithout a struggle. So we’ll use a simplifiedapproach to explain the tuning process in a typicalanalog velocity servo. Bear in mind that thissimplified approach does not necessarily accountfor all aspects of servo behavior.

A Brief Look at Servo TheoryA servo is a closed-loop system with negativefeedback. If you make the feedback positive, youwill have an oscillator. So for the servo to workproperly, the feedback must always remainnegative, otherwise the servo becomes unstable. Inpractice, it’s not as clear-cut as this. The servo canalmost become an oscillator, in which case itovershoots and rings following a rapid change atthe input.

So why doesn’t the feedback stay negative all thetime?To answer this, we need to clarify what we mean by“negative”. In this context, it means that the inputand feedback signals are in antiphase. If the input isdriven with a low frequency sinewave, the feedbacksignal (which will also be a sinewave) is displaced inphase by 180°. The 180°-phase displacement isachieved by an inversion at the input of theamplifier. (In practice it’s achieved simply byconnecting the tach the right way round – connectit the wrong way and the motor runs away.)

The very nature of a servo system is such that itscharacteristics vary with frequency, and thisincludes phase characteristics. So feedback thatstarts out negative at low frequencies can turnpositive at high frequencies. The result can beovershoot, ringing or ultimately continuousoscillation.

We’ve said that the purpose of servo tuning is toget the best possible performance from the systemwithout running into instability. We need tocompensate for the characteristics of the servocomponents and maintain an adequate stabilitymargin.

What determines whether the system will be stableor not?Closed-loop systems can be difficult to analyzebecause everything is interactive. The output getsfed back to the input in antiphase and virtuallycancels it out, so there seems to be nothing left tomeasure. The best way to determine what’s goingon is to open the loop and then see what happens.

Fig. 3.2 Closed-loop velocity servo

Fig. 3.3A Measuring open-loop characteristics

Critical Damping

Underdamped ResponseOutput

Time

Overdamped Response

Motor

Tach

+-

Oscillator

Scope

Motor

Tach

+-

Servo Amplifier

Feedback Signal

Velocity Input

A37

AE

ngin

eeri

ng R

efe

rence

Servo TuningMeasuring the open-loop characteristic allows us tofind out what the output (and therefore thefeedback) signal will be in response to a particularinput. We need to measure the gain and phaseshift at different frequencies, and we can plot theresults graphically. For a typical servo system, theresults might look like this:

Fig. 3.3B Open-loop gain and phase characteristics

The gain scale is in decibels (dB), which is alogarithmic scale (a 6dB decrease corresponds to areduction in amplitude of about 50%). The 0dB linerepresents an open-loop gain of one (unity), so atthis frequency the input and output signals will havethe same amplitude. The falling response in the gaincharacteristics is mainly due to the inertia of themotor itself.

The phase scale is in degrees and shows the phaselag between input and output. Remember that thefeedback loop is arranged to give negativefeedback at low frequencies, (i.e., 180° phasedifference). If the additional phase lag introduced bythe system components reaches 180°, thefeedback signal is now shifted by 360° andtherefore back in phase with the input. We need tomake sure that at no point do we get a feedbacksignal larger than the original input and in phasewith it. This would amount to positive feedback,producing an ever-increasing output leading tooscillation.

Fortunately, it is possible to predict quite accuratelythe gain and phase characteristics of most servosystems, provided that you have the necessarymathematical expertise and sufficient data aboutthe system. So in practice, it is seldom necessary tomeasure these characteristics unless you have aparticular stability problem that persists.

Frequency

Phase Shift

Gain

+30

db 0

+10

+20

-10

0

-90

-180

We’ve said that a problem can occur when there isa phase shift of 180° round the loop. When thishappens, the open-loop gain must be less than one(1) so that the signal fed back is smaller than theinput. So here is a basic requirement for a stablesystem:

The open-loop gain must be less than unitywhen the phase shift is 180°.When this condition is only just met (i.e., the phaseshift is near to180° at unity gain) the system will ringafter a fast change on the input.

Fig. 3.4 Underdamped response

Characteristics of a PracticalServo SystemTypical open-loop gain and phase characteristics ofan unloaded drive-motor-tach system will looksomething like Fig. 3.5.

Fig. 3.5 Characteristics of a practical system

Input Output

Ringing at unity-gain frequency

Gain

Phase

+dB

0

-dB

-180°

0

-360°

B

Shaft resonance 2 kHz typical

Crossover frequency 40–300 Hz typical

The first thing we notice is the pronounced spike inthe gain plot at a frequency of around 2kHz. This iscaused by shaft resonance, torsional oscillation inthe shaft between the motor and the tach. Observethat the phase plot drive dramatically through thecritical 180° line at this point. This means that theloop gain at this frequency must be less than unity(0dB), otherwise the system will oscillate.

The TIME CONSTANT control determines thefrequency at which the gain of the amplifier starts toroll off. You can think of it rather like the treblecontrol on an audio amplifier. When we adjust thetime constant control, we are changing the high-frequency gain to keep the gain spike at 2kHz justbelow 0dB. With too high a gain (time constant toolow) the motor will whistle at about 2kHz.

A38

Servo TuningThe second point of interest is the CROSSOVERFREQUENCY, which is the frequency at which thegain curve passes through 0dB (unity gain). Thisfrequency is typically between 40 and 300Hz. Onthe phase plot, ß (beta) is the phase margin at thecrossover frequency. If ß is very small, the systemwill overshoot and ring at the crossover frequency.So ß represents the degree of damping – thesystem will be heavily damped if ß is large.

The DAMPING control increases the phase marginat the crossover frequency. It operates by applyinglead compensation, sometimes called accelerationfeedback. The compensation network creates aphase lead in the region of the crossover frequency,which increases the phase margin and thereforeimproves the stability.

Increasing the damping also tends to reduce thegain at the 2kHz peak, allowing a higher gain to beused before instability occurs. Therefore, the timeconstant should be re-adjusted after the dampinghas been set up.

What’s the effect of adding load inertia?An external load will alter both the gain and phasecharacteristics. Not only will the overall gain bereduced owing to the larger inertia, but anadditional gain spike will be introduced due totorsional oscillation between the motor and theload. This gain spike may well be larger than theoriginal 2kHz spike, in which case the motor willstart to buzz at a lower frequency when the timeconstant is adjusted.

The amplitude of this second spike will depend onthe compliance or stiffness of the couplingbetween motor and load. A springy coupling willproduce a large gain spike; this means having toreduce the gain to prevent oscillation, resulting inpoorer system stiffness and slower response. So,if you’re after a snappy performance, it’s importantto use a torsionally-stiff coupling between themotor and the load.

Gain

Phase

+dB

0

-dB

-180°

0

-360°

Motor plus load

Motor alone (no load)

Fig. 3.6 Characteristics of a system with inertial load

A39

AE

ngin

eeri

ng R

efe

rence

Servo TuningTachometersA permanent magnet DC motor may be used as atachometer. When driven mechanically, this motorgenerates an output voltage that is proportional toshaft speed (see Fig. 4.1). The other mainrequirements for a tachometer are that the outputvoltage should be smooth over the operating rangeand that the output should be stabilized againsttemperature variations.

Small permanent magnet DC “motors” arefrequently used in servo systems as speed sensingdevices. These systems usually incorporatethermistor temperature compensation and makeuse of a silver commutator and silver loadedbrushes to improve commutation reliability at lowspeeds and at the low currents, which are typical ofthis application.

To combine high performance and low cost, DC-servo motor designs often incorporate atachometer mounted on the motor shaft andenclosed within the motor housing (Fig. 4.1).

Fig. 4.1 Tachometer output characteristics

Fig. 4.3 Principle of optical encoder

An incremental encoder generates a pulse for agiven increment of shaft rotation (rotary encoder), ora pulse for a given linear distance travelled (linearencoder). Total distance travelled or shaft angularrotation is determined by counting the encoderoutput pulses.

An absolute encoder has a number of outputchannels, such that every shaft position may bedescribed by its own unique code. The higher theresolution the more output channels are required.

The Basics of Incremental EncodersSince cost is an important factor in most industrialapplications, and resetting to a known zero pointfollowing power failure is seldom a problem, therotary incremental encoder is the type most favoredby system designers. Its main element is a shaftmounted disc carrying a grating, which rotates withthe grating between a light source and a maskeddetector. The light source may be a light emittingdiode or an incandescent lamp, and the detector isusually a phototransistor or more commonly aphoto-voltaic diode. Such a simple system,providing a single low-level output, is unlikely to befrequently encountered, since quite apart from itslow output signal, it has a DC offset that istemperature dependent, making the signal difficultto use (Fig. 4.4).

Fig. 4.4 Encoder output voltage

Fig. 4.2 Motor with integral tachometer

Optical EncodersIn servo control systems, where mechanicalposition is required to be controlled, some form ofposition sensing device is needed. There are anumber of types in use: magnetic, contact,resistive, and optical. However, for accurateposition control, the most commonly used device isthe optical encoder. There are two forms of thisencoder – absolute and incremental.

Optical encoders operate by means of a grating,that moves between a light source and a detector.When light passes through the transparent areas ofthe grating, an output is seen from the detector.For increased resolution, the light source iscollimated and a mask is placed between thegrating and the detector. The grating and the maskproduce a shuttering effect, so that only when theirtransparent sections are in alignment is lightallowed to pass to the detector (Fig. 4.3).

In practice, two photodiodes are used with twomasks, arranged to produce signals with 180°phase difference for each channel, the two diodeoutputs being subtracted so as to cancel the DCoffset (Fig. 4.5). This quasi-sinusoidal output maybe used unprocessed, but more often it is eitheramplified or used to produce a square wave output.Hence, incremental rotary encoders may have sinewave or square wave outputs, and usually have upto three output channels.

GratingCollimated Light Source

Mask Detector

Shaft Speed

Output Volts

Motor

Tachometer

DC Offset

VOutput Voltage

Shaft Rotation

Feedback Devices

A40

Servo TuningFeedback DevicesCMOS (Complimentary Metal-OxideSemiconductor) – Available for compatibility withthe higher logic levels, it normally used with CMOSdevices.

Line driver – These are low-output impedancedevices designed for driving signals over a longdistance, and are usually used with a matchedreceiver.

Complementary outputs – Outputs derived fromeach channel give a pair of signals, 180° out ofphase. These are useful where maximum immunityto interference is required.

Noise problemsThe control system for a machine is normallyscreened and protected within a metal cabinet. Anencoder may be similarly housed. However, unlesssuitable precautions are taken, the cableconnecting the two can be a source of trouble dueto its picking up electrical noise. This noise mayresult in the loss or gain of signal counts, giving riseto incorrect data input and loss of position.

Fig. 4.7 Corruption of encoder signal by noise

Fig. 4.5 Output from dual photodiode system

A two-channel encoder, as well as giving positionof the encoder shaft, can also provide informationon the direction of rotation by examination of thesignals to identify the leading channel. This ispossible since the channels are normally arrangedto be in quadrature (i.e., 90° phase shifted:Fig. 4.6).

For most machine tool or positioning applications, athird channel known as the index channel or Zchannel is also included. This gives a single outputpulse per revolution and is used when establishingthe zero position.

Fig. 4.6 Quadrature output signals

Fig. 4.6 shows that for each complete square wavefrom channel A, if channel B output is alsoconsidered during the same period, four pulseedges will occur. This allows the resolution of theencoder to be quadrupled by processing the A andB outputs to produce a separate pulse for eachsquare wave edge. For this process to be effective,however, it is important that quadrature ismaintained within the necessary tolerance so thatthe pulses do not run into one another.

Square wave output encoders are generallyavailable in a wide range of resolutions (up to about5000 lines/rev), and with a variety of different outputconfigurations, some of which are listed below.

TTL (Transistor-Transistor Logic) – This is acommonly available output, compatible with TTLlogic levels, and normally requiring a 5 volt supply.TTL outputs are also available in an open-collectorconfiguration which allows the system designer tochoose from a variety of pull-up resistor value.

Fig. 4.7, shows how the introduction of two noisepulses has converted a four-pulse train into one ofsix pulses.

A number of techniques are available to overcomeproblems due to noise. The most obvious resolutionis to use shielded interconnecting cables.

However, since the signals may be at a low level(5 volts) and may be generated by a high-impedance source, it may be necessary to go tofurther lengths to eliminate the problem.

The most effective way to resolve the problem isto use an encoder with complementary outputs(Fig. 4.8) and connect this to the control system bymeans of shielded, twisted-pair cable.

The two outputs are processed by the controlcircuitry so that the required signal can bereconstituted without the noise.

Fig. 4.8 Complementary output signals

Output 1

Output 2

DC Offset

Shaft Rotation

Combined Output (1-2)

0

V1

V1

Channel B

Channel A

90 Phase Shift (± Tolerance)

Channel A

- Ve Noise Pulse

+ Ve Noise Pulse

Channel A

Noise Spike

A

A41

AE

ngin

eeri

ng R

efe

rence

Servo TuningFeedback DevicesIf the A signal is inverted and is fed with the A signalinto an OR gate (whose output depends on onesignal or the other being present), the resultantoutput will be a square wave (Fig. 4.9).

Fig. 4.9 Reduction of noise in a complementarysystem

The simple interconnection of encoder andcontroller with channel outputs at low level may besatisfactory in electrically “clean” environments orwhere interconnections are very short. In caseswhere long interconnections are necessary orwhere the environment is “noisy”, complementaryline driver outputs will be needed, andinterconnections should be made with shielded,twisted-pair cable.

Factors Affecting AccuracySlew rate (speed) – An incremental rotary encoderwill have a maximum frequency at which it willoperate (typically 100KHz), and the maximumrotational speed, or slew rate, will be determined bythis frequency. Beyond this, the output will becomeunreliable and accuracy will be affected.

Consider a 600-line encoder rotated at 1rpm (givesan output of 10Hz). If the maximum operatingfrequency of the encoder is 50KHz, its speed will belimited to 5000 times this (i.e., 50KHz • 10Hz =5000 rpm).

If an encoder is rotated at speeds higher than itsdesign maximum, there may be conditions set upthat will be detrimental to the mechanicalcomponents of the assembly. This could damagethe system and affect encoder accuracy.

Quantization error – All digital systems havedifficulty, interpolating between output pulses.Therefore, knowledge of position will be accurateonly to the grating width (Fig. 4.10).

Fig. 4.10 Encoder quantization error

Quantization error = F(1,2N) (N = # of lines/diskrotation)

EccentricityThis may be caused by bearing play, shaft run out,incorrect assembly of the disc on its hub or the hubon the shaft. Eccentricity may cause a number ofdifferent error conditions.

a) Amplitude Modulation – In a sine wave encoder,eccentricity will be apparent as amplitudemodulation (Fig. 4.11).

Fig. 4.11 Amplitude modulation caused byeccentricity

b) Frequency modulation – As the encoder isrotated at constant speed, the frequency of theoutput will change at a regular rate (Fig. 4.12). Ifviewed on an oscilloscope, this effect will appear as“jitter” on the trace.

Fig. 4.12 Encoder frequency modulation

c ) Inter-channel jitter – If the optical detectors forthe two encoder output channels are separated byan angular distance on the same radius, then any“jitter” will appear at different times on the twochannels, resulting in “inter-channel jitter”.

Environmental ConsiderationsLike electrical noise, other environmental factorsshould be considered before installing an opticalencoder.

In particular, temperature and humidity should beconsidered (consult manufacturers’ specifications).

In environments contaminated with dust, oil vaporor other potentially damaging substances, it may benecessary to ensure that the encoder is enclosedwithin a sealed casing.

Channel A

Inverted A

OR

Quantization Error

Quantization Error = F (1, 2N) (N = no. of lines/disk rotation)

Nominal Signal Level

Signal Amplitude

Nominal Frequency (f )1

Increased Frequency (f )2

A42

Servo TuningFeedback DevicesMechanical ConstructionShaft encoder (Fig. 4.13). In this type of encoder,which may be either incremental or absolute, theelectronics are normally supported on a substantialmounting plate that houses the bearings and shaft.The shaft protrudes from the bearings on the“outside” of the encoder, for connection to therotating system, and on the “inside”, so that theencoder disc may be mounted in the appropriateposition relative to the light source and detector.The internal parts are covered by an outer casing,through which the interconnecting leads pass.

Fig. 4.13 Shaft encoder

The kit encoder will usually be less robust than theshaft encoder, but this need not be a problem if themotor is mounted so that the encoder is protected.If this cannot be done, it will normally be possible tofit a suitable cover over the encoder.

A typical kit encoder will include a body, on whichwill be mounted the electronic components and ahub and disc assembly for fitting to the shaft.Some form of cover will also be provided, mainly toexclude external light and provide some mechanicalprotection.

Linear encoder. These encoders are used where itis required to make direct measurement of linearmovement. They comprise a linear scale (whichmay be from a few millimeters to a meter or more inlength), and a read head. They may be incrementalor absolute and their resolution is expressed in linesper unit length (normally lines/inch or lines/cm).

For use in extreme environmental or industrialconditions, the whole enclosure may be specified toa more substantial standard (heavy duty) withsealed bearings and sealing between the mountingplate and cover. Also the external electricalconnections may be brought out through a highquality connector.

Modular or kit encoder (Fig. 4.14). These areavailable in a number of forms, their principaladvantage being that of reduced cost.

Fig. 4.14 Modular encoder

Since many servo motors have a double-endedshaft, it is a simple matter to fit a kit encoder onto amotor.

Shaft

Mounting Plate

Cover

Interconnecting Leads

Cover

Hub and Disc

Electronics

Body

Interconnecting Leads

A43

AE

ngin

eeri

ng R

efe

rence

Servo TuningFeedback Devices

Fig. 4.16 Incremental disk

Basics of Absolute EncodersAn absolute encoder is a position verification devicethat provides unique position information for eachshaft location. The location is independent of allother locations, unlike the incremental encoder,where a count from a reference is required todetermine position.

Fig. 4.15 Absolute disk

Fig. 4.17 Absolute encoder output

The disk pattern of an absolute encoder is inmachine readable code, usually binary, grey codeor a variety of grey. The figure above represents asimple binary output with four bits of information.The current location is equivalent to the decimalnumber 11. Moving to the right from the currentposition, the next decimal number is 10 (1-0-1-0binary). Moving to the left from the current position,the next position would be 12 (1-1-0-0).Fig. 4.18 Multi-turn absolute encoders

Gearing an additional absolute disk to the primaryhigh-resolution disk provides for turns counting, sothat unique position information is available overmultiple revolutions.Here is an example of how an encoder with 1,024counts per revolution becomes an absolute devicefor 524,288 discrete positions.The primary high-resolution disk has 1,024 discretepositions per revolution. A second disk with 3tracks of information will be attached to the high-resolution disk geared 8:1. The absolute encodernow has 8 complete turns of the shaft or 8,192discrete positions. Adding a third disk geared 8:1will provide for 64 turns of absolute positions. Intheory, additional disks could continue to beincorporated. But in practice, most encoders stopat or below 512 turns. Encoders using thistechnique are called multi-turn absolute encoders.This same technique can be incorporated in a rackand pinion style linear encoder resulting in longlengths of discrete absolute locations.

Advantages of Absolute EncodersRotary and linear absolute encoders offer a numberof significant advantages in industrial motion controland process control applications.

No Position Loss During Power Down orLoss of PowerAn absolute encoder is not a counting device like anincremental encoder, because an absolute systemreads actual shaft position. The lack of power doesnot cause the encoder lose position information.

Bit

1

1

0

1

Current Position

1011 = Decimal 11

Bearing

SealsAdditional Turns

Stages

High Resolution Main Disk

In an absolute encoder, there are several concentrictracks, unlike the incremental encoder, with itssingle track. Each track has an independent lightsource. As the light passes through a slot, a highstate (true “1”) is created. If light does not passthrough the disk, a low state (false “0”) is created.The position of the shaft can be identified throughthe pattern of 1’s and 0’s.

The tracks of an absolute encoder vary in slot size,moving from smaller at the outside edge to largertoward the center. The pattern of slots is alsostaggered with respect to preceding andsucceeding tracks. The number of tracksdetermines the amount of position information thatcan be derived from the encoder disk – resolution.For example, if the disk has ten tracks, theresolution of the encoder would usually be 1,024positions per revolution or 210.

For reliability, it is desirable to have the disksconstructed of metal rather than glass. A metal diskis not as fragile, and has lower inertia.

A44

Servo TuningFeedback DevicesWhenever power is supplied to an absolute system,it can read the current position immediately. In afacility where frequent power failures are common,an absolute encoder is a necessity.Operation in Electrically Noisy EnvironmentsEquipment such as welders and motor startersoften generate electrical noise that can often looklike encoder pulses to an incremental counter.Electrical noise does not alter the discrete positionthat an absolute system reads.High-speed Long-distance Data TransferUse of a serial interface such as RS-422 gives theuser the option of transmitting absolute positioninformation over as much as 4,000 feet.Eliminate Go Home or Referenced StartingPointThe need to find a home position or a referencepoint is not required with an absolute encodingsystem since an absolute system always knows itslocation. In many motion control applications, it isdifficult or impossible to find a home referencepoint. This situation occurs in multi-axis machinesand on machines that can't reverse direction. Thisfeature will be particularly important in a “lights-out”manufacturing facility. Significant cost savings isrealized in reduced scrap and set-up time resultingfrom a power loss.Provide Reliable Position Information inHigh-speed ApplicationsThe counting device is often the factor limiting theuse of incremental encoders in high-speedapplications. The counter is often limited to amaximum pulse input of 100 KHz. An absoluteencoder does not require a counting device orcontinuous observation of the shaft or load location.This attribute allows the absolute encoder to beapplied in high-speed and high-resolutionapplications.

ResolversA resolver is, in principle, a rotating transformer.If we consider two windings, A and B (Fig. 4.19),and if we feed winding B with a sinusoidal voltage,then a voltage will be induced into winding A. If werotate winding B, the induced voltage will be atmaximum when the planes of A and B are paralleland will be at minimum when they are at rightangles. Also, the voltage induced into A will varysinusoidally at the frequency of rotation of B so thatEOA = Ei Sinø. If we introduce a third winding (C),positioned at right angles to winding A, then as werotate B, a voltage will be induced into this windingand this voltage will vary as the cosine of the angleø, so that EOC = Ei Cosø

Fig. 4.19 Resolver principle

Referring to Fig 4.20, we can see that if we are ableto measure the relative amplitudes of the twowinding (A & C) outputs at a particular point in thecycle, these two outputs will be unique to thatposition.

Fig. 4.20 Resolver output

The information output from the two phases willusually be converted from analog to digital form, foruse in a digital positioning system, by means of aresolver-to-digital converter (Fig. 4.21). Resolutionsup to 65,536 counts per revolution are typical ofthis type of system.

Fig. 4.21 Resolver-to-digital converter

In addition to position information, speed anddirection information may also be derived. Theresolver is an absolute position feedback device.Within each electrical cycle, Phase A and Phase Bmaintain a constant (fixed) relationship.

The excitation voltage Ei may be coupled to therotating winding by slip rings and brushes, thoughthis arrangement is a disadvantage when used with abrushless motor. In such applications, a brushlessresolver may be used so that the excitation voltage isinductively coupled to the rotor winding (Fig. 4.22).

Fig. 4.22 Brushless resolver

1 Electrical Cycle

E Sin

360°

i

E Cos i

Phase Comparitor

Sine Multiplier

Cosine Multiplier

Voltage Control

Oscillator

Up/Down Counter

DC Signal (Velocity)

Integrator

Integrator

Digital Output (Shaft Angle)

Winding A

Winding B

EOA

Ei

q

Stator Phase 1

Stator Phase 2

Rotor

A45

AE

ngin

eeri

ng R

efe

rence

Control SystemsMachine ControlMany industrial designers are concerned withcontrolling an entire process. Motion control is oneimportant and influential aspect of completemachine control. The primary elements of machinecontrol include:

Fig. 5.1 Primary machine control elements

have not historically concentrated on motioncontrol, plug-in indexers or those that communicateover BCD are preferable. Because these indexerboards often include their own microprocessor,they prevent slow polling rates, but incorporate aseparate programming language. In general, thiscompromise is acceptable for all but the mostcomplicated motion/machine applications.

Bus-based SystemsBus-based machine control systems are commonin today's industrial environment. STD, VME andPC-AT are only a few of the numerous options.Most of these options can operate through astandard operating system (DOS, OS/2, OS/9) thatcan be used to program add-on cards for I/O,motion and communication interfaces. Flexiblegraphical operator interfaces remain one of thecomputer's major advantages.

Some successful examples of bus-based machinecontrol applications include gear grinding anddressing. PCB placement machines, hard diskmanufacturing, and automotive glass bending.Wherever intensive communications or dataprocessing are required, the benefits of the busstructure can be realized.

There are some disadvantages to the bus-basedmachine control system that relate to the amount ofintegration between the motion and I/O structure.Separate cards are required for each, resulting in aneed for software integration of differentprogramming languages. Motion control operations,such as servo loops, should be polled and updatedon a more immediate basis than auxiliary I/O or theoperator interface. The programmer must developthis polling hierarchy to thread the system together.

Integrated controllersA more integrated approach to machine controluses a stand-alone architecture that builds in thesame essential elements of I/O, motion, operatorinterface, and communication. This approach usesa single software and hardware platform to controlan entire machine application. The polling of servoloops, I/O points, and the operator interface arehandled internally, invisible to the user. A commonsoftware language is provided to integrate themotion and I/O actuation. This pre-tested approachallows a typical machine control application to bedeveloped with a minimum of effort and cost.The total application cost is the major considerationwhen selecting an integrated machine controller.While the initial hardware cost is typically higherthan other solutions, the software investment andmaintenance of a single language is an overridingand positive factor. Software development andmaintenance costs for any machine controlapplication can dwarf the initial hardware expense.The integrated approach offers a more economicalsolution.

Machine Control

Displays Keyboards

Touchscreens

Mainframes MIS SPC

Sensors Gauges Meters

Data Acquisition Proportional Valves

Switches Indicators Readout Actuators

Servos Steppers

HydraulicsM

otio

n Co

ntro

l

Digital I/O

Analog I/O

Networks

Operator Interface

Motion Control: For precise programmable loadmovement using a servo motor, stepper motor,or hydraulic actuators. Feedback elements areoften employed.

Analog and Digital I/O: For actuation of anexternal process, devices such as solenoids,cutters, heaters, valves, etc.

Operator Interface: For flexible interaction withthe machine process for both setup and on-linevariations. Touchscreens, data pads, andthumbwheels are examples.

Communications Support: For processmonitoring, diagnostics and data transfer withperipheral systems.

There are many different machine controlarchitectures that integrate these elements. Eachresults in varying levels of complexity andintegration of both motion and non-motionelements. PLC-based, bus-based and integratedsolutions are all commercially available. Yourselection of a machine control strategy will often bebased on performance, total application cost, andtechnology experience.

PLC-based ControlThe PLC-based architecture is utilized for I/Ointensive control applications. Based upon banks ofrelays that are scanned, or polled, by a centralprocessor, the PLC provides a low-cost option forthose familiar with its ladder logic programminglanguage. Integration of the motion, I/O, operatorinterface, and communication are usually supportedthrough additional cards that are plugged in itsbackplane.

The addition of scanning points decreases thepolling rate of any individual point, and can thuslead to lower machine response. Because PLCs

A46

Control SystemsX Code Programming X Code has been designed to allow motion controlequipment to be programmed by users with little orno computer experience. Although the languageincludes more than 150 commands, depending onthe product, it is only necessary to learn a smallpercentage of these to write simple programs.

Most command codes use the initial letter of thefunction name, which makes them easy toremember. Here are some examples of frequentlyused commands.

V – velocity in revs/secD – distance in stepsA – acceleration rate in revs/sec2

G – go; start the moveT – time delay in seconds

A typical command string might look like this:

V10 A50 D4000 G T2 G

This would set the velocity to 10 revs/sec,acceleration to 50 revs/sec2 and distance to 4000steps. The 4000-step move would be performedtwice with a 2-second wait between moves.

Please refer to specifications of X Code productsfor a list of all the available X Code commands.

Single-axis and Multi-axis ControllersA single-axis controller can, as the name implies,only control one motor. The controller in anintegrated indexer/drive comes into this category.However, such units are frequently used in systemsusing more than one motor where the operationsdo not involve precise synchronization betweenaxes.

A multi-axis controller is designed to control morethan one motor and can very often performcomplex operations such as linear or circularinterpolation. These operations require accuratesynchronization between axes, which is generallyeasier to achieve with a central controller.

A variant of the multi-axis controller is themultiplexed unit, which can control several motorson a time-shared basis. A printing machine havingthe machine settings controlled by stepper motorscould conveniently use this type of controller whenthe motors do not need to be movedsimultaneously.

Hardware-based ControllersControl systems designed without the use of amicroprocessor have been around for many yearsand can be very cost-effective in simplerapplications. They tend to lack flexibility and aretherefore inappropriate where the move parametersare continually changing. For this reason, thehardware-based controller has now given wayalmost exclusively to systems based on amicroprocessor.

Control System OverviewThe controller is an essential part of any motioncontrol system. It determines speed, direction,distance and acceleration rate – in fact all theparameters associated with the operation that themotor performs. The output from the controller isconnected to the drive’s input, either in the form ofan analog voltage or as step and direction signals.In addition to controlling one or more motors, manycontrollers have additional inputs and outputs thatallow them to monitor other functions on a machine(see Machine Control, p. A45).

Controllers can take a wide variety of forms. Someexamples are listed below.

Standalone – This type of controller operateswithout data or other control signals from externalsources. A standalone unit usually incorporates akeypad for data entry as well as a display, andfrequently includes a main power supply. It will alsoinclude some form of nonvolatile memory to allow itto store a sequence of operations. Many controllersthat need to be programmed from a terminal orcomputer can, once programmed, also operate instandalone mode.

Bus-based – A bus-based controller is designed toaccept data from a host computer using a standardcommunications bus. Typical bus systems includeSTD, VME and IBM-PC bus. The controller willusually be a plug-in card that conforms to thestandards for the corresponding bus system. Forexample, a controller operating on the IBM-PC busresides within the PC, plugging into an expansionslot and functioning as an intelligent peripheral.

PLC-based – A PLC-based indexer is designed toaccept data from a PLC in the form of I/Ocommunication. Typically, the I/O information is inBCD format. The BCD information may select aprogram to execute, a distance to move, a timedelay, or any other parameter requiring a number.The PLC is well suited to I/O actuation, but poorlysuited to perform complex operations such as mathand complicated decision making. The motioncontrol functions are separated from the PLC'sprocessor and thus do not burden its scan time.

X Code-based – X Code is a command languagespecifically developed for motion control andintended for transmission along an RS-232C link.Controllers using this language either accept real-time commands from a host computer or executestored sequences that have been previouslyprogrammed. The simplicity of RS-232Ccommunication allows the controller to beincorporated into the drive itself, resulting in anintegrated indexer/drive package.

A47

AE

ngin

eeri

ng R

efe

rence

Control SystemsFig. 5.2 Processor-based controller

I/O Interface

Microprocessor

RS-232C Communications

Interface

Nonvolatile RAM

Program Memory ROM

Programmable Pulse

Generator

XCode

Commands

Inputs

Outputs

Step

Direction

Output to Drive

Processor-based ControllersThe flexibility offered by a microprocessor systemmakes it a natural choice for motion control.Fig. 5.2 shows the elements of a typical step anddirection controller that can operate either inconjunction with a host computer or as a stand-alone unit.

All the control functions are handled by themicroprocessor whose operating program isstored in ROM. This program will include aninterpreter for the command language, which maybe X Code for example.

X Code commands are received from the hostcomputer or terminal via the RS-232Ccommunications interface. These commands aresimple statements that contain the required speed,distance and acceleration rate, etc. The processorinterprets these commands and uses theinformation to control the programmable pulsegenerator. This in turn produces the step anddirection signals that will control a stepper or servodrive.

The processor can also switch outputs andinterrogate inputs via the I/O interface. Outputscan initiate other machine functions such aspunching or cutting, or simply activate drive panelindicators to show the program status. Inputs maycome from sources such as operator pushbuttonsor directional limit switches.

When the controller is used in a standalone mode,the required motion sequences are programmedfrom the host and stored in nonvolatile memory(normally battery-backed RAM). These sequencesmay then be selected and executed from switchesvia the I/O interface or from a separate machinecontroller such as a PLC.

A48

Control SystemsUnderstanding Input and Output Modules

Most motion controllers/indexers offerprogrammable inputs and outputs to control andinteract with other external devices and machineelements.

Programmable Output ExampleAfter indexing a table to a preset position, energizea programmable output to activate a knife that willcut material on the table.

Programmable Input ExampleAfter indexing a table in a pick and placeapplication, the indexer waits for an input signalfrom a robot arm, signaling the indexer that a parthas been located on the table.

The primary reason for using I/O modules is tointerface 5VDC logic signals from an indexer toswitches and relays on the factory floor, whichtypically run on voltage levels ranging from 24VDCto 220VAC. Solid-state I/O modules are essentiallya relay, utilizing light emitting diode (LED) and atransistor along with a signal conditioning circuit toactivate a switch. These I/O modules isolate (nodirect connection) the internal microprocessorcircuitry of an indexer from oversized DC and ACvoltages. The lack of a physical connectionbetween the indexer and external devices, protectsthe indexer from hazardous voltage spikes andcurrent surges.

DC Input and Output ModulesAs with all DC devices, this is a polarized, + and –input module. Since current will flow in only onedirection, care must be taken to observe thesepolarities during installation.

DC input modules typically feature an input signalconditioning circuit. This circuit requires the input toremain on/off for a minimum of 5 milliseconds

Fig. 5.3 Typical DC input connection diagram

DC Input Operational SequenceAs switch #1 closes, current flows through thelimiting resistor (1K ohm), and then into the LED.The light issued by the LED due to this forwardcurrent flow in turn simulates the photo transistor.Hence the term “opto” or optically isolated. Thephototransistor then drives the base of the secondtransistor to a high level, bringing its output, orcollector, to a low level.

The operation of the DC output model is similar tothe DC input module. A 5VDC signal from anindexer is used to activate an LED. The output ofthe module is defined as open collector.

Fig. 5.4 represents a typical DC output schematic.Note the diode across the relay coil. These shouldalways be installed to eliminate the leading inductivekick caused by the relay. A typical part number forsuch a diode is 1N4004. Failure to provide thisprotection can cause noise problems or thedestruction of the output device.

before recognizing the switch. This eliminates ashort voltage spike or “de-bounce” contact closureless than 5 milliseconds in duration. However, a 0.1microfarad, ceramic disc capacitor across theactual switching contacts is still recommended toprevent switch bounce that can be as long as 10-80 milliseconds.

10 to 24VDC Floating Source

+

-

10 to 24 mAScrew

Terminals

1K Coupling

LED

Switch #1

4KV Isolation Barrier

Photo Transistor

Signal Conditioning

Ground

Logic Signal

Optional Circuit Board

Indicating LED

+5VDC+

-

Fig. 5.4 Typical DC output connection diagram

Indicator+5VDC

(equivalent circuit)

Logic LED Photo Transistor

Amplifier

Output Transistor

Screw Terminals

2.4 Amps

24VDC

+

-

Load 10 (solenoid)

Freewheeling Diode

(+)

(-)

A49

AE

ngin

eeri

ng R

efe

rence

Control Systems

Fig. 5.5 Typical AC input connection diagram

The module will only turn on or off at points A, B, orC; where the voltage is zero.

AC output modules do have leakage current, whichmay “turn-on” small current loads. To solve

AC output modules feature a Triac power device asits output. A Triac output offers three distinctadvantages.

1. Zero voltage turn on eliminates in-rush currentsto the load.

2. Zero voltage turn off eliminates inductive kickproblems.

3. A snubber across the output.

Fig. 5.6 Zero voltage turn on and off

Fig. 5.7 Typical AC output connection diagram

potential problems, add a parallel resistor acrossthe load, 5K, 5W for 120VAC and a 10K, 10W for240VAC.

AC Input and Output ModulesAC modules are not polarized devices. This makesit virtually impossible to install a unit backwards.

AC input modules operate like DC input moduleswith the addition of a bridge rectifier to change AC

voltage to DC levels. AC input modules also includetransient protection to filter out spikes from the ACline (caused by lightning strikes, arc welders, etc.).

120 VAC 60Hz

Power Line

Screw Terminals

8mA

Pushbutton Rectifying Bridge

LED

14K Photo Transistor

Signal Conditioning

Ground

Logic Signal

Indicator

+5VDC

A B C

Indicator+5VDC

(equivalent circuit)

Logic LED Photo Transistor

4KV Isolation Barrier

Zero Voltage Circuit

Triac Snubber

C

R

Screw Terminals

Load (motor)

120VAC 60Hz

Power Line

A50

Control SystemsSerial and Parallel CommunicationsSerial and parallel communications are methods oftransferring data from a host computer to aperipheral device such as a Compumotor indexer.In the case of a Compumotor indexer, the dataconsist of parameters such as acceleration,

velocity, move distance, and move directionconfigured in ASCII characters. Bothcommunication techniques are generally bi-directional allowing the host to both transmit andreceive information from a peripheral device.

SerialSerial communication transmits data one bit at atime on a single data line. Single data bits aregrouped together into a byte and transmitted at apredetermined interval (baud rate). Serialcommunication links can be as simple as a 3-lineconnection; transmit (Tx), receive (Rx) and ground(G). This is an advantage from a cost standpoint,but usually results in slower communications thanparallel communications. Common serial interfacesinclude RS-232C, RS-422, RS-485, RS-423.

TroubleshootingProcedure for troubleshooting 3-wire RS-232Ccommunication.

1. Verify that the transmit of the host is wired to thereceive of the peripheral, and receive of the hostis wired to the transmit of the peripheral. Note:Try switching the receive and transmit wires oneither the host or peripheral if you fail to get anycommunication.

2. Some serial ports require handshaking. You canestablish 3-wire communication by jumperingRTS to CTS (usually pins 4 and 5) and DSR toDTR (usually pins 6 and 20).

3. Configure the host and peripheral to the samebaud rate, number of data bits, number of stopbits, and parity.

4. If you receive double characters (e.g., typing “A”and receiving “AA”), your computer is set to halfduplex mode. Change to full duplex mode.

5. Use DC common or signal ground as yourreference, NOT earth ground.

6. Cable lengths should not exceed 50 ft. unlessyou are using some form of line driver, opticalcoupler, or shield. As with any control signal, besure to shield the cable to earth ground at oneend only.

7. To test terminal or terminal emulation softwarefor proper 3-wire communication, unhook theperipheral device and transmit a character. Anechoed character should not be received. If acharacter is received, you are in half duplexmode. Jumper the host’s transmit and receivelines and send another character. You shouldreceive the echoed character. If not, consult themanufacturer of the host’s serial interface forproper pin outs.

Fig. 5.8 Serial Communications

ParallelParallel communication requires handshaking andtransmits data one byte (8 bits) at a time. Whendata are transferred from the host processor to aperipheral device, the following steps take place.

1. The host sets a bit on the bus signalling to theperipheral that a byte of data has been sent.

2. The peripheral receives data and sets a bit onthe bus, signalling to the host that data havebeen received.

The advantage of communicating in parallel vs.serial is faster communications. However, sinceparallel communications require morecommunication lines, the cost can be higher thanserial communications.

Parallel bus structures include:

IEEE-488, IBM PC, VME, MULTIBUS, Q and STD.

TroubleshootingProcedure for troubleshooting parallelcommunication.

1. Make certain the address setting of theperipheral device is configured properly.

2. Confirm that multiple boards are not set to thesame address (and each board is sealedproperly into a slot).

3. Verify that peripheral subroutines to reset theboard, write data, and read data work properly.Follow the handshaking procedure outlined inthe device’s user manual.Note: Compumotor bus-based indexers comecomplete with a diskette that includes pretestedprograms to verify system functions androutines for simple user program development.

Fig. 5.9 Parallel Communications

Time (baud rate)

Data bits

Star

t bit

Parit

y bi

t

Stop bits

0

1

0

0

0

0

0

1

Data bus

Signals A = 0100 1 = 0011

0001 0001

IEEE-488 IBM PC VME Bus STD Bus Multi Bus

A51

AE

ngin

eeri

ng R

efe

rence

Control Systems

117 75 u118 76 v119 77 w120 78 x121 79 y122 7A z123 7B 124 7C I125 7D 126 7E127 7F DEL

Serial and Parallel CommunicationsADDRESS: Multiple devices are controlled on thesame bus, each with a separate address or unitnumber. This address allows the host tocommunicate individually to each device.ASCII: American Standard Code for InformationInterchange. This code assigns a number to eachnumeral and letter of the alphabet. In this manner,information can be transmitted between machinesas a series of binary numbers.BAUD RATE: Number of bits transmitted persecond. Typical rates include 300; 600; 1,200;2,400; 4,800; 9,600, 19,200. This means at 9,600baud, 1 character can be sent nearly everymillisecond.DATA BITS: Since the ASCII set consists of 128characters, computers may transmit only 7 bits ofdata. Most computers do, however, support an 8-bit extended ASCII character set.DCE: Data Communications Equipment transmitson pin 3 and receives on pin 2.DTE: Data Terminal Equipment. Transmits on pin 2and receives on pin 3.FULL DUPLEX: The terminal will display onlyreceived or echoed characters.HALF DUPLEX: In half duplex mode, a terminal willdisplay every character transmitted. It may alsodisplay the received character.HANDSHAKING SIGNALS:RTS: Request To Send DTR: Data Terminal ReadyCTS: Clear To Send IDB: Input Data BufferDSR: Data Set Ready ODB: Output Data Buffer

NULL MODEM: A simple device or set ofconnectors that switches the receive and transmitlines a 3-wire RS-232C connector.PARITY: An RS-232C error detection scheme thatcan detect an odd number of transmission errors.SERIAL POLLING: Method of checking the statusof the IEEE-488 device. By reading the status byte,the host can determine if the device is ready toreceive or send characters.START BITS: When using RS-232C, one or twobits are added to every character to signal the endof a character.TEXT/ECHO (ON/OFF): This setup allows receivedcharacters to be re-transmitted back to the originalsending device. Echoing characters can be used toverify or “close the loop” on a transmission.XON/XOFF: Two ASCII characters supported insome serial communication programs. If supported,the receiving device transmits an XOFF character tothe host when its character buffer is full. The XOFFcharacter directs the host to stop transmittingcharacters to the device. Once the buffer empties,the device will transmit an XON character to signalthe host to resume transmission.

DEC HEX GRAPHIC DEC HEX GRAPHIC DEC HEX GRAPHIC DEC HEX GRAPHIC

ASCII TableDEC HEX GRAPHIC000 00 NUL001 01 SOH002 02 STX003 03 ETX004 04 EOT005 05 ENQ006 06 ACK007 07 BEL008 08 BS009 09 HT010 0A LF011 0B VT012 0C FF013 0D CR014 0E SO015 0F S1016 10 DLE017 11 DC1018 12 DC2019 13 DC3020 14 DC4021 15 NAK022 16 SYN023 17 ETB024 18 CAN025 19 EM026 1A SUB027 1B ESC028 1C FS029 1D GS

030 IE RS031 1F US032 20 SPACE033 21 !034 22 "035 23 #036 24 $037 25 %038 26 &039 27 '040 28 (041 29 )042 2A *043 2B +044 2C ,045 2D -046 2E .047 2F /048 30 0049 31 1050 32 2051 33 3052 34 4053 35 5054 36 6055 37 7056 38 8057 39 9058 3A :

088 58 X089 59 Y090 5A Z091 5B [092 5C /093 5D ]094 5E V095 5F -096 60 ’097 61 a098 62 b099 63 c100 64 d101 65 e102 66 f103 67 g104 68 h105 69 i106 6A j107 6B k108 6C l109 6D m110 6E n111 6F o112 70 p113 71 q114 72 r115 73 s116 74 t

~

059 3B ;060 3C <061 3D =062 3E >063 3F ?064 40 @065 41 A066 42 B067 43 C068 44 D069 45 E070 46 F071 47 G072 48 H073 49 I074 4A J074 4B K075 4C L076 4D M077 4E N078 4F O080 50 P081 51 Q082 52 R083 53 S084 54 T085 55 U086 56 V087 57 W

A52

Control Systems

Multiple devices on the same circuit should begrounded together at a single point.

Furthermore, power supplies and programmablecontrollers often have DC common tied to Earth(AC power ground). As a rule, it is preferable tohave indexer signal ground or DC common floatingwith respect to Earth. This prevents noisyequipment that is grounded to Earth from sendingnoise into the indexer. The Earth groundconnection should be made at one point only asdiscussed in “Ground Loops” on p. A53.

In many cases, optical isolation may be required tocompletely eliminate electrical contact between theindexer and a noisy environment. Solid-state relaysprovide this isolation.

Externally Conducted NoiseExternally conducted noise is similar to power linenoise, but the disturbances are created on signaland ground wires connected to the indexer. Thiskind of noise can get onto logic circuit ground orinto the processor power supply and scramble theprogram. The problem here is that controlequipment often shares a common DC ground thatmay run to several devices, such as a DC powersupply, programmable controller, remote switchesand the like. When some noisy device, particularly arelay or solenoid, is on the DC ground, it may causedisturbances within the indexer.

The solution for DC mechanical relays andsolenoids involves connecting a diode backwardsacross the coil to clamp the induced voltage “kick”that the coil will produce. The diode should be ratedat 4 times the coil voltage and 10 times the coilcurrent. Using solid-state relays eliminates thiseffect altogether.

Fig. 5.11 Coil Suppression Methods

Electrical Noise . . .Sources, Symptoms and SolutionsNoise related difficulties can range in severity fromminor positioning errors to damaged equipmentfrom runaway motors crashing blindly through limitswitches. In microprocessor controlled equipment,the processor is constantly retrieving instructionsfrom memory in a controlled sequence. If anelectrical disturbance occurs, it could cause theprocessor to misinterpret an instruction, or accessthe wrong data. This is likely to be catastrophic tothe program, requiring a processor reset. MostCompumotor indexers are designed with awatchdog timer that shuts down the system if theprogram is interrupted. This prevents the morecatastrophic failures.

Sources of NoiseBeing invisible, electrical noise can be verymysterious, but it invariably comes from thefollowing sources:

• Power line disturbances• Externally conducted noise• Transmitted noise• Ground loops

Some common electrical devices generateelectrical noise.

• Coil driven devices: conducted and power linenoise

• SCR-fired heaters: transmitted and power linenoise

• Motors and motor drives: transmitted and powerline noise

• Welders (electric): transmitted and power linenoise

Power line disturbances are usually easy to solvedue to the wide availability of line filtering equipmentfor the industry. Only the most severe situations callfor an isolation transformer. Line filtering equipmentis required when other devices connected to thelocal power line are switching large amounts ofcurrent, especially if the switching takes place athigh frequency. Corcom and Teal are twomanufacturers of suitable power line filters.

Also, any device having coils is likely to upset theline when it is switched off. Surge suppressors suchas MOVs (General Electric) can limit this kind ofnoise. A series RC network across the coil is alsoeffective, (resistance; 500 to 1,000 Ω, capacitance;0.1 to 0.2µF). Coil-driven devices (inductive loads)include relays, solenoids, contactors, clutches,brakes and motor starters.

Fig. 5.10 Typical RC Network

AC or DC

R

Inductive Load

C

DC

Diode

AC or DC

Varistor (MOV)

A53

AE

ngin

eeri

ng R

efe

rence

Control SystemsTransmitted NoiseTransmitted noise is picked up by externalconnections to the indexer, and in severe cases,can attack an indexer with no external connections.The indexer enclosure will typically shield theelectronics from this, but openings in the enclosurefor connection and front panel controls may “leak”.As with all electrical equipment, the indexer chassisshould be scrupulously connected to Earth tominimize this effect.When high current contacts open, they draw anarc, producing a burst of broad spectrum radiofrequency noise that can be picked up on anindexer limit switch or other wiring. High current andhigh voltage wires have an electrical field aroundthem, and may induce noise on signal wiring(especially when they are tied in the same wiringbundle or conduit).When this kind of problem occurs, considershielding signal cables or isolating the signals. Aproper shield surrounds the signal wires to interceptelectrical fields, but this shield must be tied to Earthto drain the induced voltages. At the very least,wires should be run in twisted pairs to limit straightline antenna effects.Most Compumotor cables have shields tied toEarth, but in some cases the shields must begrounded at installation time. Installing the indexerin a NEMA electrical enclosure ensures protectionfrom this kind of noise, unless noise-producingequipment is also mounted inside the enclosure.Connections external to the enclosure must beshielded.Even the worst noise problems, in environmentsnear 600 amp welders and 25kW transmitters, havebeen solved using enclosures, conduit, opticalisolation, and single-point ground techniques.

Ground LoopsGround Loops create the most mysterious noiseproblems. They seem to occur most often insystems where a control computer is usingRS-232C communication. Garbled transmissionand intermittent operation symptoms are typical.The problem occurs in systems where multipleEarth ground connections exist, particularly whenthese connections are far apart.

ExampleSuppose a Model 500 is controlling an axis, and thelimit switches use an external power supply. TheModel 500 is controlled by a computer in anotherroom. If the power supply Common is connected toEarth, ground loop problems may occur (mostcomputers have their RS-232C signal common tiedto Earth). The loop starts at the Model 500’s limitswitch ground, goes to Earth through the powersupply to Earth at the computer. From there, theloop returns to the Model 500 through RS-232Csignal ground. If a voltage potential exists betweenpower supply Earth and remote computer Earth,ground, current will flow through the RS-232Cground creating unpredictable results.The way to test for and ultimately eliminate a groundloop is to lift or “cheat” Earth ground connections inthe system until the symptoms disappear.

Defeating NoiseThe best time to handle electrical noise problems isbefore they occur. When a motion system is in thedesign process, the designer should consider thefollowing system wiring guidelines (listed by order ofimportance).

1. Put surge suppression components on allelectrical coils: resistor/capacitor filters, MOVs,Zener and clamping diodes.

2. Shield all remote connections and use twistedpairs. Shields should be tied to Earth at oneend.

3. Put all microelectronic components in anenclosure. Keep noisy devices outside. Monitorinternal temperature.

4. Ground signal common wiring at one point.Float this ground from Earth if possible.

5. Tie all mechanical grounds to Earth at one point.Run chassis and motor grounds to the frame,frame to Earth.

6. Isolate remote signals. Solid-state relays or optoisolators are recommended.

7. Filter the power line. Use common RF filters(isolation transformer for worst-case situations).

A noise problem must be identified before it can besolved. The obvious way to approach a problemsituation is to eliminate potential noise sources untilthe symptoms disappear, as in the case of groundloops. When this is not practical, use the aboveguidelines to troubleshoot the installation.

ReferencesInformation about the equipment referred to may beobtained by calling the numbers listed below.

• Corcom line filters (312) 680-7400

• Opto-22 optically isolated relays (714) 891-5861

• Crydom optically isolated relays (213) 322-4987

• Potter Brumfield optically isolated relays(812) 386-1000

• General Electric MOVs (315) 456-3266

• Teal power line isolation filters (800) 888-8325

A54

Control SystemsStopping in an EmergencyFor safety reasons, it is often necessary toincorporate some form of emergency stop systeminto machinery fitted with stepper or servo motors.There are several reasons for needing to stopquickly.

• To prevent injury to the operator if he makes amistake or operates the machinery improperly.

• To prevent damage to the machine or to theproduct as a result of a jam.

• To guard against machine faults. You shouldconsider all the possible reasons for stopping tomake sure that they are adequately covered.

How should you stop the system?There are several ways to bring a motor to a rapidstop. The choice depends partly on whether it ismore important to stop in the shortest possible timeor to guarantee a stop under all circumstances. Forinstance, to stop as quickly as possible meansusing the decelerating power of the servo system.However, if the servo has failed or control has beenlost, this may not be an option open to you. In thiscase, removing the power will guarantee that themotor stops; but if the load has a high inertia, thismay take some time. If the load is moving verticallyand can back-drive the motor, this introducesadditional complications. In extreme cases wherepersonal safety is at risk, it may be necessary tomechanically lock the system even at the expenseof possible damage to the machine.

Emergency Stop Methods1. Full-torque controlled stop.Applying zero velocity command to a servo amplifierwill cause it to decelerate hard to zero speed incurrent limit, in other words, using the maximumavailable torque. This will create the fastest possibledeceleration to rest. In the case of a digital servowith step and direction inputs, cutting off the steppulses will produce the same effect.

The situation is different for a stepper drive. Thestep pulse train should be decelerated to zerospeed to utilize the available torque. Simply cuttingoff the step pulses at speeds above the start-stoprate will de-synchronize the motor and the fulldecelerating torque will no longer be available. Thecontroller needs to be able to generate a rapiddeceleration rate independent of the normalprogrammed rate, to be used only for overtravellimit and emergency stop functions.

2. Disconnect the motor.Although this method is undoubtedly safe, it is nothighly recommended as a quick-stop measure. Thetime taken to stop is indeterminate, since itdepends on load inertia and friction, and in high-performance systems the friction is usually kept to aminimum. Certain types of drives may be damagedby disconnecting the motor under power. Thismethod is particularly unsatisfactory in the case of avertical axis, since the load may fall under gravity.

3. Remove the AC input power from the drive.On drives that incorporate a power dump circuit, adegree of dynamic braking is usually provided whenthe power is removed. This is a better solution thandisconnecting the motor, although the powersupply capacitors may take some time to decayand this will extend the stopping distance.

4. Use dynamic braking.A motor with permanent magnets will act as agenerator when driven mechanically. By applying aresistive load to the motor, a braking effect isproduced that is speed-dependent. Deceleration istherefore rapid at high speeds, but falls off as themotor slows down.

A changeover contactor can be arranged to switchthe motor connections from the drive to theresistive load. This can be made failsafe by ensuringthat braking occurs if the power supply fails. Theoptimum resistor value depends on the motor, butwill typically lie in the 1-3 ohms range. It must bechosen to avoid the risk of demagnetization atmaximum speed as well as possible mechanicaldamage through excessive torque.

5. Use a mechanical brake.It is very often possible to fit a mechanical brakeeither directly on the motor or on some other part ofthe mechanism. However, such brakes are usuallyintended to prevent movement at power-down andare seldom adequate to bring the system to a rapidhalt, particularly if the drive is delivering full currentat the time. Brakes can introduce friction even whenreleased, and also add inertia to the system – botheffects will increase the drive power requirements.

What is the best stopping method?It is clear that each of the methods outlined abovehas certain advantages and drawbacks. This leadsto the conclusion that the best solution is to use acombination of techniques, ideally incorporating ashort time delay.

We can make use of the fact that a contactor usedfor dynamic braking will take a finite time to dropout, so it is possible to de-energize the contactorcoil while commanding zero speed to the drive. Thisallows for a controlled stop to occur under fulltorque, with the backup of dynamic braking in theevent that the amplifier or controller has failed.

WARNING! – there is a risk that decelerating aservo to rest in full current limit could result inmechanical damage, especially if a high-ratiogearbox is used. This does not necessarily ensure asafe stop, be sure that the mechanism canwithstand this treatment.

A mechanical brake should also be applied to avertical axis to prevent subsequent movement. Analternative to the electrically-operated brake is thedifferential drag brake, which will prevent the loadfrom falling but creates negligible torque in theopposite direction.

A55

AE

ngin

eeri

ng R

efe

rence

System CalculationsSystem Selection ConsiderationsApplication ConsiderationsAccuracyAn accuracy specification defines the maximumerror in achieving a desired position. Some types ofaccuracy are affected by the application. Forexample, repeatability will change with the frictionand inertia of the system the motor is driving.

Accuracy in a rotary motor is usually defined interms of arcminutes or arcseconds (the terms

arcsecond and arcminute are equivalent to secondand minute, respectively). There are 1,296,000seconds of arc in a circle. For example, anarcsecond represents 0.00291 inches ofmovement on a circle with a 50-foot radius. This isequivalent to about the width of a human hair.

Stepper AccuracyThere are several types of performance listed underCompumotor’s motor specifications: repeatability,accuracy, relative accuracy, and hysteresis.

RepeatabilityThe motor’s ability to return to the same positionfrom the same direction. Usually tested by movingthe motor one revolution, it also applies to linearstep motors moving to the same place from thesame direction. This measurement is made with themotor unloaded, so that bearing friction is theprominent load factor. It is also necessary to ensurethe motor is moving to the repeat position from adistance of at least one motor pole. Thiscompensates for the motor’s hysteresis. A motorpole in a Compumotor is 1/50 of a revolution.

AccuracyAlso referred to as absolute accuracy, thisspecification defines the quality of the motor’smechanical construction. The error cancels itselfover 360° of rotation, and is typically distributed in asinusoidal fashion. This means the error willgradually increase, decrease to zero, increase in theopposite direction and finally decrease again uponreaching 360° of rotation. Absolute accuracycauses the size of microsteps to vary somewhatbecause the full motor steps that must be traversedby a fixed number of microsteps varies. The stepscan be over or undersized by about 4.5% as aresult of absolute accuracy errors.

Relative Accuracy

Also referred to as step-to-step accuracy, thisspecification tells how microsteps can change insize. In a perfect system, microsteps would all beexactly the same size, but drive characteristics andthe absolute accuracy of the motor cause the stepsto expand and contract by an amount up to therelative accuracy figure. The error is not cumulative.

HysteresisThe motor’s tendency to resist a change indirection. This is a magnetic characteristic of themotor, it is not due to friction or other externalfactors. The motor must develop torque toovercome hysteresis when it reverses direction. Inreversing direction, a one revolution move will showhysteresis by moving the full distance less thehysteresis figure.

Servo & Closed-Loop Stepper AccuracyRepeatability, accuracy and relative accuracy inservos and closed-loop stepper systems relate asmuch to their feedback mechanisms as they do tothe inherent characteristics of the motor and drive.

ServosCompumotor servos use resolver feedback todetermine their resolution and position. It isessentially the resolution of the device reading theresolver position that determines the highestpossible accuracy in the system. Digiplan servosuse encoder feedback to determine their resolutionand position. In this case, it is the encoder’sresolution that determines the system’s accuracy.The positional accuracy is determined by the drive’sability to move the motor to the position indicatedby the resolver or encoder. Changes in friction orinertial loading will adversely affect the accuracyuntil the system is properly tuned.

Closed-Loop SteppersCompumotor closed-loop stepper systems use anencoder to provide feedback for the control loop.The encoder resolution determines the system’saccuracy. When enabled, the controlling indexerattempts to position the motor within the specifieddeadband from the encoder. Typically, this meansthe motor will be positioned to within one encoderstep. To do this satisfactorily, the motor must havemore resolution than the encoder. If the step size ofthe motor is equal to or larger than the step size ofthe encoder, the motor will be unable to maintain aposition and may become unstable. In a systemwith adequate motor-to-encoder resolution, themotor is able to maintain one encoder step ofaccuracy with great dependability. This is acontinuous process that will respond to outsideevents that disturb the motor’s position.

A56

System CalculationsSelection ConsiderationsApplication ConsiderationsLoad characteristics, performance requirements,and coupling techniques need to be understoodbefore the designer can select the best motor/drivefor the job. While not a difficult process, severalfactors need to be considered for an optimumsolution. A good designer will adjust thecharacteristics of the elements under his control –including the motor/drive and the mechanicaltransmission type (gears, lead screws, etc.) – tomeet the performance requirements. Someimportant parameters are listed below.

TorqueRotational force (ounce-inches) defined as a linearforce (ounces) multiplied by a radius (inches). Whenselecting a motor/drive, the torque capacity of themotor must exceed the load. The torque any motorcan provide varies with its speed. Individual speed/torque curves should be consulted by the designerfor each application.

InertiaAn object’s inertia is a measure of its resistance tochange in velocity. The larger the inertial load, thelonger it takes a motor to accelerate or deceleratethat load. However, the speed at which a motorrotates is independent of inertia. For rotary motion,inertia is proportional to the mass of the objectbeing moved times the square of its distance fromthe axis of rotation.

FrictionAll mechanical systems exhibit some frictional force,and this should be taken into account when sizingthe motor, as the motor must provide torque toovercome any system friction. A small amount offriction is desirable since it can reduce settling timeand improve performance.

Torque-to-Inertia RatioThis number is defined as a motor’s rated torquedivided by its rotor inertia. This ratio is a measure ofhow quickly a motor can accelerate and decelerateits own mass. Motors with similar ratings can havedifferent torque-to-inertia ratios as a result ofvarying construction.

Load Inertia-to-Rotor Inertia RatioFor a high performance, relatively fast system, loadinertia reflected to the motor should generally notexceed the motor inertia by more than 10 times.Load inertias in excess of 10 times the rotor inertiacan cause unstable system behavior.

Torque MarginWhenever possible, a motor/drive that can providemore motor torque than the application requiresshould be specified. This torque marginaccommodates mechanical wear, lubricanthardening, and other unexpected friction.Resonance effects, while dramatically reduced withthe Compumotor microstepping system, can causethe motor’s torque to be slightly reduced at somespeeds. Selecting a motor/drive that provides atleast 50% margin above the minimum neededtorque is good practice.

VelocityBecause available torque varies with velocity,motor/drives must be selected with the requiredtorque at the velocities needed by the application.In some cases, a change in the type of mechanicaltransmission used is needed to achieve therequired performance.

ResolutionThe positioning resolution required by theapplication will have an effect on the type oftransmission used and the motor resolution. Forinstance, a leadscrew with 4 revolutions per inchand a 25,000-step-per-revolution motor/drivewould give 100,000 steps per inch. Each stepwould then be 0.00001 inches.

Duty CycleSome motor/drives can produce peak torque forshort time intervals as long as the RMS or averagetorque is within the motor’s continuous duty rating.To take advantage of this feature, the applicationtorque requirements over various time intervalsneed to be examined so RMS torque can becalculated.

Solving Duty Cycle Limitation ProblemsOperating a motor beyond its recommended dutycycle results in excessive heat in the motor anddrive. The drive cycle may be increased byproviding active cooling to the drive and the motor.A fan directed across the motor and anotherdirected across the drive’s heatsink will result inincreased duty cycle capability.

In most cases, it is possible to tell if the duty cycle isbeing exceeded by measuring the temperature ofthe motor and drive. Refer to the specifications forindividual components for their maximum allowabletemperatures.

Note: Motors will run at case temperatures up to100°C (212°F)—temperatures hot enough to burnindividuals who touch the motors.

To Improve Duty Cycle:• Mount the drive with heatsink fins running

vertically• Fan cool the motor• Fan cool the drive• Put the drive into REMOTE POWER SHUTDOWN

when it isn’t moving, or reduce current• Reduce the peak current to the motor

(if possible)• Use a motor large enough for the application

A57

AE

ngin

eeri

ng R

efe

rence

System CalculationsA wide range of applications can be solvedeffectively by more than one motor type. However,some applications are particularly appropriate foreach motor type. Compumotor’s Motor Sizing andSelection Software package is designed to helpyou easily identify the proper motor size and typefor your specific motion control application.

This software helps calculate load inertias andrequired torques—information that is reflectedthrough a variety of machine transmissions andreductions, including leadscrews, gears, belts, andpulleys. This software then produces graphs of theresults, allowing you to select the proper motorfrom a comprehensive, detailed database of morethan 200 motor models.

IBM® PC-compatible, Motor Sizing & Selectionsoftware also generates a number of application-specific reports, including profiles and speed/

AE

ngin

eeri

ng R

efe

rence

Motor Sizing and Selection Softwaretorque curves that are based on user-providedinformation. This advanced graphics package isVGA/EGA compatible and allows data entry witheither a mouse or keypad.

Contact your local Automotive Technology Centerto obtain a copy.

A58

System Calculations

ExampleYou need to move 6" in 2 seconds

a = -d = = 6.75

v = = 4.51.5 (6 inches)(2 seconds)

inchessecond

4.5 (6 inches)(2 seconds)2

inchessecond2

Move ProfileBefore calculating torque requirements of anapplication, you need to know the velocities andaccelerations needed. For those positioningapplications where only a distance (X) and a time(S) to move that distance are known, thetrapezoidal motion profile and formulas given beloware a good starting point for determining yourrequirements. If velocity and accelerationparameters are already known, you can proceed toone of the specific application examples on thefollowing pages.

Move distance X in time S.

Assume that:1. Distance X/4 is moved in time S/3 (Acceleration)2. Distance X/2 is moved in time S/3 (Run)3. Distance X/4 is moved in time S/3 (Deceleration)

The graph would appear as follows:

The acceleration (a), velocity (v) and deceleration (d)may be calculated in terms of the knowns, X and S.

v = at = x =

a = -d = = = =2Xt2

X( 4 )2 X2 x 9S2S( 3 )

24.5XS2

S3

1.5XS

4.5XS2

Common Move Profile Considerations

TT

T

S/3 2S/3 S

S/3 S/3 S/3

0

V

a d

Velo

city

time

Distance: ______________ Inches of Travel ________________________ revolutions of motor

Move Time: ____________________________________________________ seconds

Accuracy: ______________________________________________________ arcminutes, degrees or inches

Repeatability: ___________________________________________________ arcseconds, degrees or inches

Duty Cycle

on tme: ____________________________________________________ seconds

off time: ___________________________________________________ seconds

Cycle Rate: ____________________________________________________ sec. min. hour

Motor/Drive SelectionBased on Continuous Torque RequirementsHaving calculated the torque requirements for anapplication, you can select the motor/drive suitedto your needs. Microstepping motor systems(S Series, Zeta Series OEM650 Series, LN Series)have speed/torque curves based on continuousduty operation. To choose a motor, simply plot totaltorque vs. velocity on the speed/torque curve. Thispoint should fall under the curve and allowapproximately a 50% margin for safety. An S106-178 and an S83-135 curve are shown here.

Note: When selecting a ZETA Series product, a50% torque margin is not required.

ExampleAssume the following results from load calculations:F = 25 oz-in Friction torqueA = 175 oz-in Acceleration torqueT = 200 oz-in Total torqueV = 15 rev/sec Maximum velocity

You can see that the total torque at the requiredvelocity falls within the motor/drive operating rangefor both motors by plotting TT .

The S83-135 has approximately 250 oz-in availableat V max (25% more than required). The S106-178has 375 oz-in available, an 88% margin.

In this case, we would select the S106-178motor/drive to assure a sufficient torque marginto allow for changing load conditions.

0 10 20 30 40 50

TT = 200

oz-in

RPS (Vmax)

106-178

83-135

0

300

600

900

1200

1500

A59

AE

ngin

eeri

ng R

efe

rence

System Calculations

How to Use a Step MotorHorsepower CurveHorsepower (HP) gives an indication of the motor’stop usable speed. The peak or “hump” in ahorsepower curve indicates a speed that givesmaximum power. Choosing a speed beyond the peakof the HP curve results in no more power: the powerattained at higher speeds is also attainable at a lowerspeed. Unless the speed is required for theapplication, there is little benefit to going beyond thepeak as motor wear is faster at higher speeds.

Applications requiring the most power the motor cangenerate, not the most torque, should use a motorspeed that is just below the peak of the HP curve.

Motor/Drive SelectionBased on peak torque requirementsServo-based motor/drives have two speed/torquecurves: one for continuous duty operation andanother for intermittent duty. A servo system canbe selected according to the total torque andmaximum velocity indicated by the continuous dutycurve. However, by calculating the root meansquare (RMS) torque based on your duty cycle, youmay be able to take advantage of the higher peaktorque available in the intermittent duty range.

TRMS =

Where:

• Ti is the torque required over the time interval ti

• means “the sum of”

ExampleAssume the following results from your loadcalculations.

TF = 25 oz-in Friction Torque

TA = 775 oz-in Acceleration Torque

TT = 800 oz-in Total Torque

Vmax = 20 rps Maximum Velocity

Motion Profile

M

tiM

M

Duty CycleIndex 4 revs in 0.3 seconds, dwell 0.3 secondsthen repeat.

If you look at the S106-178 speed/torque curve,you’ll see that the requirements fall outside thecurve.

T1 = Torque reqired to accelerate the load fromzero speed to maximum speed (TF + TA)

T2 = Torque required to keep the motor movingonce it reaches max speed (TF)

T3 = Torque required to decelerate from maxspeed to a stop (TA - TF)

T4 = Torque required while motor is sitting still atzero speed (Ø)

t1 = Time spent accelerating the loadt2 = Time spent while motor is turning at

constant speedt3 = Time spent decelerating the loadt4 = Time spent while motor is at rest

Ti2 ti

T12t1 + T2

2 t2 + T32 t3 + T4

2 t4

t1 + t2 + t3 + t4

TRMS =

(800)2(.1) + (25)2(.1) + (750)2(.1) + (0)2(.3)

(.1) + (.1) + (.1) + (.3)=

TRMS = 447 oz. in.

Now plot TRMS and TT vs. Tmax on the speed/torquecurve.

The drawing below resembles thespeed/torque curve for the Z606 motor.

The Z606 motor will meet the requirements.RMS torque falls within the continuous dutycycle and total torque vs. velocity falls withinthe intermittent range.

0

35

oz-in

(.24)

(N-m) (HP)

Power

0 6 12 18 24 30

70 (.49)

105 (.73)

140 (1.98)

175 (1.22)

.035

.070

.105

.140

.175

Speed-rps

Torq

ue

0

Torque

Horsepower

t1

0

20 rps

t2 t3 t4

0.1 0.2 0.3 0.4 0.5 0.6 t

600

(800)

1800

10 20 30 40 50 60

(506)

T

T

T

RMS

A60

System CalculationsLeadscrew DrivesLeadscrews convert rotary motion to linear motionand come in a wide variety of configurations.Screws are available with different lengths,diameters, and thread pitches. Nuts range from thesimple plastic variety to precision ground versionswith recirculating ball bearings that can achieve veryhigh accuracy.

The combination of microstepping and a qualityleadscrew provides exceptional positioningresolution for many applications. A typical 10-pitch(10 threads per inch) screw attached to a 25,000step/rev. motor provides a linear resolution of0.000004" (4 millionths, or approximately 0.1micron) per step.

A flexible coupling should be used between theleadscrew and the motor to provide some damping.The coupling will also prevent excessive motorbearing loading due to any misalignment.

Microscope PositioningApplication Type: X/Y Point to Point

Motion: Linear

Description: A medical research lab needs toautomate their visual inspection process. Eachspecimen has an origin imprinted on the slide withall other positions referenced from that point. Thesystem uses a PC-AT Bus computer to reduce datainput from the operator, and determines the nextdata point based on previous readings. Each datapoint must be accurate to within 0.1 microns.

Machine Objectives• Sub-micron positioning• Specimen to remain still during inspection• Low-speed smoothness (delicate equipment)• Use PC-AT Bus computerMotion Control Requirements• High resolution, linear encoders• Stepper (zero speed stability)• Microstepping• PC-AT Bus controllerCompumotor Solution: Microstepping motorsand drives, in conjunction with a precision ground40 pitch leadscrew table, provide a means of sub-micron positioning with zero speed stability.Conventional mechanics cannot provide 0.1 micronaccuracies without high grade linear encoders. It isnecessary for the Compumotor Model AT6400indexer, which resides directly on the computerbus, to provide full X, Y, Z microscope control andaccept incremental encoder feedback.

Microstepping motors

Encoders

A61

AE

ngin

eeri

ng R

efe

rence

System Calculations

Diameter

In. Steel Brass Alum.

2.75 25.1543 26.9510 8.6468 oz-in2

3.00 35.6259 38.1707 12.2464 oz-in2

3.25 49.0699 52.5749 16.8678 oz-in2

3.50 66.0015 70.7159 22.6880 oz-in2

3.75 86.9774 93.1901 29.8985 oz-in2

4.00 112.5956 120.6381 38.7047 oz-in2

4.25 143.4951 153.7448 49.3264 oz-in2

4.50 180.3564 193.2390 61.9975 oz-in2

4.75 223.9009 239.8939 76.9659 oz-in2

5.00 274.8916 294.5267 94.4940 oz-in2

Leadscrew Efficiencies

Efficiency (%)Type High Median Low

Ball-nut 95 90 85Acme with metal nut* 55 40 35Acme with plastic nut 85 65 50

* Since metallic nuts usually require a viscouslubricant, the coefficient of friction is both speedand temperature dependent.

Coefficients of Static Friction Materials

(Dry Contact Unless Noted) µS

Steel on Steel 0.58Steel on Steel (lubricated) 0.15Aluminum on Steel 0.45Copper on Steel 0.22Brass on Steel 0.19Teflon on Steel 0.04

Leadscrew Application DataInertia of Leadscrews per InchDiameter

In. Steel Brass Alum.

0.25 0.0017 0.0018 0.0006 oz-in2

0.50 0.0275 0.0295 0.0094 oz-in2

0.75 0.1392 0.1491 0.0478 oz-in2

1.00 0.4398 0.4712 0.1512 oz-in2

1.25 1.0738 1.1505 0.3691 oz-in2

1.50 2.2266 2.3857 0.7654 oz-in2

1.75 4.1251 4.4197 1.4180 oz-in2

2.00 7.0372 7.5399 2.4190 oz-in2

2.25 11.2723 12.0774 3.8748 oz-in2

2.50 17.1807 18.4079 5.9059 oz-in2

Other Leadscrew Drive Applications• XY Plotters• Facsimile transmission• Tool bit positioning• Cut-to-length machinery• Back gauging• Microscope drives• Coil winders• Slides• Pick-and-Place machines• Articulated arms

Precision GrinderA bearing manufacturer is replacing someequipment that finishes bearing races. The oldequipment had a two-stage grinding arrangementwhere one motor and gearbox provided a rough cutand a second motor with a higher ratio gearboxperformed the finishing cut. The designer would liketo simplify the mechanics and eliminate one motor.He wants to use a single leadscrew and exploit thewide speed range available with microstepping toperform both cuts. This will be accomplished bymoving a cutting tool mounted on the end of theleadscrew into the workpiece at two velocities; aninitial velocity for the rough cut and a much reducedfinal velocity for the finish cut.

The torque required to accelerate the load andovercome the inertia of the load and the rotationalinertia of the leadscrew is determined to be 120 oz-in. The torque necessary to overcome friction ismeasured with a torque wrench and found to be 40oz-in. A microstepping motor with 290 oz-in oftorque is selected and provides adequate torquemargin.

This grinder is controlled by a programmablecontroller (PC) and the environment requires thatthe electronics withstand a 60°C environment. Anindexer will provide the necessary velocities andaccelerations. The speed change in the middle ofthe grinding operation will be signaled to the PCwith a limit switch, and the PC will in turn programthe new velocity into the indexer. Additionally, theindexer Stall Detect feature will be used inconjunction with an optical encoder mounted onthe back of the motor to alert the PC if themechanics become “stuck.”

A62

System Calculations

diameter. Assume that the leadscrew has an Acmethread and uses a plastic nut. Motor inertia is givenas 6.56 oz-in2. In this example, we assume ahorizontally oriented leadscrew where the force ofgravity is perpendicular to the direction of motion.In non-horizontal orientations, leadscrews willtransmit varying degrees of influence from gravity tothe motor, depending on the angle of inclination.Compumotor Sizing Software automaticallycalculates these torques using vector analysis.

1. Calculate the torque required to overcomefriction. The coefficient of static friction for steel-to-steel lubricant contact is 0.15. The median value ofefficiency for an Acme thread and plastic nut is0.65. Therefore:

F = µsW = 0.15 (200 lb) = 480 oz

TFriction = = = 23.51 oz-in

2. Compute the rotational inertia of the load and therotational inertia of the leadscrew:

JLoad = = x = 3.24 oz-in2

JLeadscrew = =

= 80.16 oz-in2

3. The torque required to accelerate the load maynow be computed since the motor inertia wasgiven:TAccel = (JLoad + JLeadscrew + JMotor)

= (4.99 + 80.16 + 6.56(oz-in2))

= 149 oz-in

TTotal = TFriction + TAccel

TTotal = 23.51 oz-in + 149 oz-in = 172.51 oz-in

W(2πp)2

16 ozlb

200 lb(2π5)2

in2

Leadscrew FormulasThe torque required to drive load W using aleadscrew with pitch (p) and efficiency (e) has thefollowing components:

TTotal = TFriction + TAcceleration

TFriction =

Where:F = frictional force in ouncesp = pitch in revs/ine = leadscrew efficiency

F = µs W for horizontal surfaces where µs =coefficient of static friction and W is the weight ofthe load. This friction component is often called“breakaway”.Dynamic Friction: F = µDW is the coefficient to usefor friction during a move profile. However, torquecalculations for acceleration should use the worstcase friction coefficient, µs.

TAccel = (JLoad + JLeadscrew + JMotor)

ω = 2πpvJLoad = ; JLeadscrew =

Where:T = torque, oz-inω = angular velocity, radians/sect = time, secondsv = linear velocity, in/secL = length, inchesR = radius, inchesρ = density, ounces/in3

g = gravity constant, 386 in/sec2

The formula for load inertia converts linear inertiainto the rotational equivalent as reflected to themotor shaft by the leadscrew.

ProblemFind the torque required to accelerate a 200-lb steelload sliding on a steel table to 2 inches per secondin 100 milliseconds using a 5 thread/inch steelleadscrew 36 inches long and 1.5 inches in

480 oz2π 5 rev

rev in

F2πpe

1g

ωt

w(2πp)2

πLρR4

2

(16 oz)lb

F2πpe x x 0.65

1g

ωt

(ω = 2π 5in )(2 in

sec ) = 20πsec

πLρR4

2π2

(36 in) (4.48 oz) (0.75 in)4 in3

1386 in/sec2

20π.

sec0.1sec

e

Leadscrew DrivesVertical or Horizontal Application:ST – Screw type, ball or acme ST =e – Efficiency of screw e = %µS – Friction coefficient µS =L – Length ofscrew L = inchesD – Diameter of screw D = inchesp—Pitch p = threads/inchW – Weight of load W = lbs.F—Breakaway force F = ouncesDirectly coupled to the motor? yes/noIf yes, CT – Coupling typeIf no, belt & pulley or gearsRadius of pulley or gear inchesGear: Number of teeth – Gear 1 Number of teeth – Gear 2Weight of pulley or gear ouncesWeight of belt ounces

A63

AE

ngin

eeri

ng R

efe

rence

System Calculations

Hollow Cylinder

w2

πLρ2

1g

ωt

JLoad = (R21 + R2

2)

Where W, the weight, is known

or

JLoad = (R42 – R4

1)

Where ρ, the density, is known

W = πLρ (R22 – R2

1)

T = (JLoad + JMotor)

ProblemCalculate the motor torque required to accelerate asolid cylinder of aluminum 5" in radius and 0.25"thick from rest to 2.1 radians/sec (0.33 revs/sec) in0.25 seconds. First, calculate JLoad using thedensity for aluminum of 1.54 oz/in3.

JLoad = = = 378 oz-in2πLρR4

2π x 0.25 x 1.54 x 54

2Assume the rotor inertia of the motor you will use is37.8 oz-in2.

TTotal = (JLoad + JMotor) xωt

= x (378 + 37.8) x1386

2.10.25

= 9.05 oz-in

1g

5.96R

LR 2

1

Solid Cylinder (oz-in2)Inertia: JLoad =

Where weight and radius are known

Inertia (oz-in2) JLoad =

Where ρ, the material density is known

Weight W = πLρR2

Inertia may be calculated knowing either the weightand radius of the solid cylinder (W and R) or itsdensity, radius and length (ρ, R and L.)

The torque required to accelerate any load is:

T (oz-in) = Ja

a = = 2π (accel.) for Accel. in rps2

Where:a = angular acceleration, radians/sec2

ω2 = final velocity, radians/secω1 = initial velocity, radians/sec

t = time for velocity change, secondsJ = inertia in units of oz-in2

The angular acceleration equals the time rate ofchange of the angular velocity. For loadsaccelerated from zero, ω1 = 0 and a =TTotal = (JLoad + JMotor)

TTotal represents the torque the motor must deliver.The gravity constant (g)in the denominatorrepresents accelerationdue to gravity (386 in/sec2) and convertsinertia from units of oz-in2 to oz-in-sec2.

WR2

2

πLρR4

2

ω2 - ω1

t

ωt

ωt1

g

R

L

Directly Driven LoadsThere are many applications where the motionbeing controlled is rotary and the low-speedsmoothness and high resolution of a Compumotorsystem can be used to eliminate gear trains or othermechanical linkages. In direct drive applications, amotor is typically connected to the load through a

flexible or compliant coupling. This couplingprovides a small amount of damping and helpscorrect for any mechanical misalignment.

Direct drive is attractive when mechanical simplicityis desirable and the load being driven is ofmoderate inertia.

Direct Drive Formulas

5.96R

LR 2

1R

L

R – Radius R = inchesR(1) – Inner radius R(1) = inchesR(2) – Outer radius R(2) = inchesL – Length L = inchesW – Weight of disc W = ouncesρ – Density/Material ρ = ounces/inch3

g – Gravity constant g = 386 in/sec2

A64

System Calculations

R

W

G1

R1

R2

G2

N1

N2

Gears

Gear Driven LoadsR – Radius R = inchesR(1) – Radius gear #1 R(1) = inchesR(2) – Radius gear #2 R(2) = inchesN(1) – Number of teeth G#1 N(1) =N(2) – Number of teeth G#2 N(2) =G – Gear ratio N(1) G =

N(2)W – Weight of load W = ouncesW(1) – Weight G#1 W(1) = ouncesW(2) – Weight G#2 W(2) = ouncesL – Length L = inchesF – Friction F =BT – Breakaway torque BT = ounce/inches

JLoad = R4Load

WLoad

2NGear 2( NGear 1 )

2

JLoad = R2Load

orπLLoadρLoad

2NGear 2( NGear 1 )

2

NGear 2( NGear 1 )2

JGear1 = R2Gear1

WGear1

2

JGear2 = R2Gear2

WGear2

2

TTotal = (JLoad + JGear1 + JGear2 + JMotor)1g

ωt

Gear Drive Formulas

Gear DrivesTraditional gear drives are more commonly usedwith step motors. The fine resolution of amicrostepping motor can make gearingunnecessary in many applications. Gears generallyhave undesirable efficiency, wear characteristics,backlash, and can be noisy.

Gears are useful, however, when very large inertiasmust be moved because the inertia of the load

reflected back to the motor through the gearing isdivided by the square of the gear ratio.

In this manner, large inertial loads can be movedwhile maintaining a good load-inertia to rotor-inertiaratio (less than 10:1).

Where:

J = inertia, oz-in (gm-cm2) “as seen by the motor”T = torque, oz-in (gm-cm)

W = weight, oz (gm)R = radius, in. (cm)N = number of gear teeth (constant)L = length, in (cm)ρ = density, oz/in3 (gm/cm3)ω = angular velocity, radians/sec @ motor shaftt = time, secondsg = gravity constant, 386 in/sec2

R

W

G1

R1

R2

G2

N1

N2

Gears

A65

AE

ngin

eeri

ng R

efe

rence

System Calculations

R– Radius R = inches

W – Weight (include weight of belt or chain) W = ounces

W(P) – Weight of pulley or material W(P) ounces

F – Breakaway force F = ounces

V – Linear velocity V = inches/sec

CT – Coupling type CT =

SL – Side load SL =

Tangential Drives

Tangential Drive Formulas

TTotal = TLoad + TPulley + TBelt + TMotor + TFriction

TTotal =

(JLoad + JPulley + JBelt + JMotor) +TFriction

JLoad = WLR2

JPulley = (Remember to multiply by 2 if there are 2 pulleys.)

1g

ωt

WpR2

2

JBelt = WBR2

TFriction = FR

ω =VR

Where:T = torque, oz-in (gm-cm)ω = angular velocity, radians/sect = time, seconds

WL = weight of the load, ozWP = pulley weight, ozWB = belt or rack weight, oz

F = frictional force, oz (gm)R = radius, in (cm)V = linear velocityg = gravity constant, 386 in/sec2

ρ = density, oz/in3

ProblemWhat torque is required to accelerate a 5-lb load toa velocity of 20 inches per second in 10milliseconds using a flat timing belt? The motordrives a 2-inch diameter steel pulley 1/2-inch wide.The timing belt weighs 12 oz. Load static friction is30 ozs. Motor rotor inertia is 10.24 oz-in.2

JBelt = WBR2 = 12 oz (1 in)2 = 12 oz-in2

TFriction = F x R = 30 oz x 1 in = 30 oz-in

ω = = 20 x = 20

TTotal = (80 + 7.04 + 12 + 10.24) + 30

TTotal = 596.2 oz-in

JLoad = WLR2 = 5 lb x 16 x (1 in)2 = 80 oz-in2oz

lb

JPulley = = π x 0.5 in x2(πLρR4)

2

= 7.04 oz-in2

VR

insec

1 rad1 in

radsec

1386

20.01

(4.48 oz/in3) (1 in)4

R

W

A66

System Calculations

1. Motor Sizing AXIS 1A. Weight of payload (lbs) 10.0B. Fixed forces, if any (lbs) 0C. Known move distance (in) 40

time (sec) 1.0D. Angle from horizontal (degrees) 0

2. Total length of travel (inches) 40

3. Desired repeatability (in) 001

4. Desired resolution (in) .0005

5. Necessary settling time after move100 ms to within .001 inches

6. Life expectancy: Percent duty cycle 20%

Estimated number of moves/year 200,000

7. Is the center of gravity significantly changed? no

8. What is the environment? clean [√] dirty [ ]

Specifics9. Operating temperature range 65° to 85°F

10. Can air be available? yes

C. Minimum acceleration rateA = Vmax (53.2 in/sec) = 212.8 in/sec2

Accel. time (.250 sec)A = Minimum acceleration (212.8 in/sec2)

386 in/sec2 per 1 G = 0.551 g's

Step 3: Calculate maximum acceleration rate ofL20 (using constant acceleration indexer).

Based on the speed/force curve below, the L20has 14.0 lbs of force at 53.2 in/sec (Vmax).

Linear Step MotorsThere are many characteristics to consider whendesigning, selecting and installing a completemotion control system. The applications dataworksheet and the application considerationsdetailed below will help determine if a linear motorsystem is recommended for a given application. Alinear motor, when properly specified, will providethe optimum performance and the greatestreliability.

Step 1: Total mass to be acceleratedMtotal = Mload (10.0) + Mforcer (2.0) = 12.0 lbs.

Step 2: Acceleration rateA. Average velocity = move distance

move time= (40 inches) (1.0 sec)= 40.0 in/sec

B. Maximum velocity(Based on trapezoidal move profile)Vmax = 1.33 x Vavg (40.0 in/sec) = 53.2 in/sec

Application Data Worksheet #1Application: Single Axis √ Multi-Axis X-Y Gantry

Description of system operation: A part is moved in and out of a

machine very quickly. The part comes to rest at the same point in the

machine each time. An operator sets this distance with a thumbwheel

switch.

The SolutionActual and assumed factors that contribute to thesolution are:

1. Force (F) = mass (M) x acceleration (A)Note: mass units are in pounds

2. Acceleration due to gravity(1g=386 inches/sec2)

3. L20 forcer weighs 2.0 lb.4. Attractive force between L20 forcer and platen

= 200 lbs.5. Trapezoidal velocity profile:

Accel time = 1.0 sec/4= 0.250 sec.

Vmax = 1.33 x Vavg

20 (9.08)

16 (7.26)

12 (6.45)

8 (3.83)

4 (1.82)

0 20

(50.8)40

(101.6)60

(152.4)80

(203.2)100

(254.0)Speed in (cm)

Forc

e lb

s (k

g) 14.0 lbs

53.2 ips

Amax = = 1.16 g's

Step 4: Non-damped safety marginIf all available force could be used, the maximumcalculated acceleration rate:

Force (14.0 lb)Mtotal (12.0 lbs.)

The calculated acceleration rate should be reducedby 50% (100% non-damped safety margin) nettingan acceleration rate for the L20 of 0.58 g's. Theapplication requires a 0.55 g's acceleration rate.The L20 meets the requirements of this application.

Force vs. Speed L-L20

Sketch the proposed mechanical configuration:

20"

4 lbs.

40

Time (sec)

1.0 4

53.2

0 1.0

Velocity (in/sec)

1/4 1/4 1/4 1/4

Vmax

Vave

A67

AE

ngin

eeri

ng R

efe

rence

System CalculationsAccuracyIn linear positioning systems, some applicationsrequire high absolute accuracy, while manyapplications require a high degree of repeatability.These two variables should be reviewed toaccurately evaluate proper system performance.

In the “teach mode”, a linear motor can bepositioned and subsequently learn the coordinatesof any given point. After learning a number of pointsin a sequence of moves, the user will be concernedwith the ability of the forcer to return to the sameposition from the same direction. This scenariodescribes repeatability.

In a different application, a linear motor is used toposition a measuring device. The size of an objectcan be measured by positioning the forcer to apoint on the object. Determining the measuredvalue is based on the number of steps required toreach the point on the object. System accuracymust be smaller than the tolerance on the desiredmeasurement.

Open-loop absolute accuracy of a linear step motoris typically less than a precision grade leadscrewsystem. If a linear encoder is used in conjunctionwith a linear motor, the accuracy will be equivalentto any other transmission system.

The worst-case accuracy of the system is thesum of these errors:

Accuracy = A + B + C + D + E + F

A = Cyclic Error – The error due to motormagnetics that recurs once every pole pitchas measured on the body of the motor.

B = Unidirectional Repeatability – The errormeasured by repeated moves to the samepoint from different distances in the samedirection.

C = Hysteresis – The backlash of the motorwhen changing direction due to magneticnon-linearity and mechanical friction.

D = Cumulative Platen Error – Linear error ofthe platen as measured on the body of themotor.

E = Random Platen Error – The non-linearerrors remaining in the platen after the linearis disregarded.

F = Thermal Expansion Error – The errorcaused by a change in temperatureexpanding or contracting the platen.

Velocity RippleVelocity ripple is most noticeable when operatingnear the motor’s resonant frequency. Rotarystepping motor’s have this tendency as well, but itis usually less noticeable due to mechanical lossesin the rotary-to-linear transmission system, whichdampens the effects. Velocity ripple due toresonance can be reduced with the electronicaccelerometer damping option (-AC).

Platen MountingThe air gap between the forcer and the platensurface can be as small as 0.0005 inches. Properlymounting the platen is extremely important. Whenheld down on a magnetic chuck, the platen is flatand parallel within its specifications, however, in itsfree state, slight bows and twists may cause theforcer (L20) to touch the platen at several places.Compumotor recommends mounting the platenusing all its mounting holes on a ground flat piece ofsteel, such as an I-beam, U-channel or tube.

EnvironmentDue to the small air gap between the forcer andplaten, care should be taken to keep the platenclean. A small amount of dirt or adhesive material(such as paint) can cause a reduction in motorperformance. When appropriate, mounting themotor upside down or on its side will help keepforeign particles off the platen. Protective boots thatfold like an accordion as the motor travels can alsobe used to assist in keeping the platen clean.

Linear Step MotorsLife ExpectancyThe life of a mechanical bearing motor is limited bywearing of the platen surface over which thebearings roll. Factors that affect wear and life of amechanical bearing system include:

A. High velocities – Life is inversely proportional tovelocity cubed. Increasing velocity raises thetemperature of the platen due to eddy currentlosses in the solid platen material. (In normalhigh-speed, high duty cycle operation over asmall piece of platen, the platen can becomealmost too hot to touch.)

B. Load on the forcer – Load has some effect onthe life expectancy of the linear motor. Users areurged to adhere to the load specifications foreach motor.

Yaw, Pitch and RollIn applications such as end effector devices orwhere the load is located far from the motor’scenter of gravity, the stiffness characteristics of theforcer must be considered. Moment producingforces tend to deflect the forcer, and if strongenough, will cause the motor to stall or be removedfrom the platen. Yaw, pitch and roll specificationsare used to determine the maximum torque you canapply to the forcer.

A68

System CalculationsGlossary of TermsAbsolute PositioningRefers to a motion control systememploying position feedback devices(absolute encoders) to maintain a givenmechanical location.

Absolute ProgrammingA positioning coordinate referencedwherein all positions are specified relativeto some reference, or “zero” position. Thisis different from incremental programming,where distances are specified relative tothe current position.

AC ServoA general term referring to a motor drivethat generates sinusoidal shaped motorcurrents in a brushless motor wound as togenerate sinusoidal back EMF.

AccelerationThe change in velocity as a function oftime. Acceleration usually refers toincreasing velocity and decelerationdescribes decreasing velocity.

AccuracyA measure of the difference betweenexpected position and actual position of amotor or mechanical system. Motoraccuracy is usually specified as an anglerepresenting the maximum deviation fromexpected position.

Ambient TemperatureThe temperature of the cooling medium,usually air, immediately surrounding themotor or another device.

ASCIIAmerican Standard Code for InformationInterchange. This code assigns a numberseries of electrical signals to each numeraland letter of the alphabet. In this manner,information can be transmitted betweenmachines as a series of binary numbers.

BandwidthA measure of system response. It is thefrequency range that a control system canfollow.

BCDBinary Coded Decimal is an encodingtechnique used to describe the numbers 0through 9 with four digital (on or off) signallines. Popular in machine tool equipment,BCD interfaces are now giving way tointerfaces requiring fewer wires – such asRS-232C.

BitAbbreviation of Binary Digit, the smallestunit of memory equal to 1 or 0.

Back EMFThe voltage produced across a winding ofa motor due to the winding turns being cutby a magnetic field while the motor isoperating. This voltage is directlyproportional to rotor velocity and isopposite in polarity to the applied voltage.Sometimes referred to as counter EMF.

Block DiagramA simplified schematic representingcomponents and signal flow through asystem.

Bode PlotA graph of system gain and phase versusinput frequency which graphicallyillustrates the steady state characteristicsof the system.

Break FrequencyFrequency(ies) at which the gain changesslope on a Bode plot (break frequenciescorrespond to the poles and zeroes of thesystem).

Brushless DC ServoA general term referring to a motor drivethat generates trapezoidal shaped motorcurrents in a motor wound as to generatetrapezoidal Back EMF.

ByteA group of 8 bits treated as a whole, with256 possible combinations of one’s andzero’s, each combination representing aunique piece of information.

CommutationThe switching sequence of drive voltageinto motor phase windings necessary toassure continuous motor rotation. Abrushed motor relies upon brush/barcontact to mechanically switch thewindings. A brushless motor requires adevice that senses rotor rotational position,feeds that information to a drive thatdetermines the next switching sequence.

Closed LoopA broadly applied term relating to anysystem where the output is measured andcompared to the input. The output is thenadjusted to reach the desired condition. Inmotion control, the term describes asystem wherein a velocity or position (orboth) transducer is used to generatecorrection signals by comparison todesired parameters.

Critical DampingA system is critically damped when theresponse to a step change in desiredvelocity or position is achieved in theminimum possible time with little or noovershoot.

Crossover FrequencyThe frequency at which the gain interceptsthe 0 dB point on a Bode plot (used inreference to the open-loop gain plot).

Daisy-ChainA term used to describe the linking ofseveral RS-232C devices in sequencesuch that a single data stream flowsthrough one device and on to the next.Daisy-chained devices usually aredistinguished by device addresses, whichserve to indicate the desired destination fordata in the stream.

DampingAn indication of the rate of decay of asignal to its steady state value. Related tosettling time.

Damping RatioRatio of actual damping to criticaldamping. Less than one is anunderdamped system and greater thanone is an overdamped system.

Dead BandA range of input signals for which there isno system response.

DecibelA logarithmic measurement of gain. If G isa system’s gain (ratio of output to input),then 20 log G = gain in decibels (dB).

Detent TorqueThe minimal torque present in anunenergized motor. The detent torque of astep motor is typically about 1% of itsstatic energized torque.

Direct Drive ServoA high-torque, low-speed servo motor witha high resolution encoder or resolverintended for direct connection to the loadwithout going through a gearbox.

Duty CycleFor a repetitive cycle, the ratio of on timeto total cycle time.

Duty cycle =

EfficiencyThe ratio of power output to power input.

Electrical Time ConstantThe ratio of armature inductance toarmature resistance.

EncoderA device that translates mechanical motioninto electronic signals used for monitoringposition or velocity.

Form FactorThe ratio of the RMS value of a harmonicsignal to its average value in one half-wave.

FrictionA resistance to motion. Friction can beconstant with varying speed (Coulombfriction) or proportional to speed (viscousfriction).

GainThe ratio of system output signal to systeminput signal.

Holding TorqueSometimes called static torque, it specifiesthe maximum external force or torque thatcan be applied to a stopped, energizedmotor without causing the rotor to rotatecontinuously.

On Time

(On Time + Off Time)

A69

AE

ngin

eeri

ng R

efe

rence

System CalculationsGlossary of TermsHomeA reference position in a motion controlsystem derived from a mechanical datumor switch. Often designated as the “zero”position.

Hybrid ServoA brushless servo motor based on aconventional hybrid stepper. It may useeither a resolver or encoder forcommutation feedback.

HysteresisThe difference in response of a system toan increasing or a decreasing input signal.

IEEE-488A digital data communications standardpopular in instrumentation electronics. Thisparallel interface is also known as GPIB, orGeneral Purpose Interface Bus.

Incremental MotionA motion control term that describes adevice that produces one step of motionfor each step command (usually a pulse)received.

Incremental ProgrammingA coordinate system where positions ordistances are specified relative to thecurrent position.

InertiaA measure of an object’s resistance to achange in velocity. The larger an object’sinertia, the larger the torque that isrequired to accelerate or decelerate it.Inertia is a function of an object’s massand its shape.

Inertial MatchFor most efficient operation, the systemcoupling ratio should be selected so thatthe reflected inertia of the load is equal tothe rotor inertia of the motor.

IndexerSee PMC.

I/OAbbreviation of input/output. Refers toinput signals from switches or sensors andoutput signals to relays, solenoids etc.

Lead Compensation AlgorithmA mathematical equation implemented bya computer to decrease the delaybetween the input and output of a system.

LimitsProperly designed motion control systemshave sensors called limits that alert thecontrol electronics that the physical end oftravel is being approached and thatmotion should stop.

Logic GroundAn electrical potential to which all controlsignals in a particular system arereferenced.

Mechanical Time ConstantThe time for an energized DC motor toreach 2/3rds of its set velocity. Based on afixed voltage applied to the windings.

Mid-range InstabilityDesignates the condition resulting fromenergizing a motor at a multiple of itsnatural frequency (usually the third orderscondition). Torque loss and oscillation canoccur in underdamped open-loopsystems.

MicrosteppingAn electronic control technique thatproportions the current in a step motor’swindings to provide additional intermediatepositions between poles. Producessmooth rotation over a wide speed rangeand high positional resolution.

Open CollectorA term used to describe a signal outputthat is performed with a transistor. Anopen collector output acts like a switchclosure with one end of the switch atground potential and the other end of theswitch accessible.

Open LoopRefers to a motion control system whereno external sensors are used to provideposition or velocity correction signals.

Opto-isolatedA method of sending a signal from onepiece of equipment to another without theusual requirement of common groundpotentials. The signal is transmittedoptically with a light source (usually a LightEmitting Diode) and a light sensor (usuallya photosensitive transistor). These opticalcomponents provide electrical isolation.

PMCProgrammable motion controller,primarily designed for single- or multi-axis motion control with I/O as anauxiliary function.

PoleA frequency at which the transferfunction of a system goes to infinity.

Pulse RateThe frequency of the step pulses appliedto a motor driver. The pulse ratemultiplied by the resolution of the motor/drive combination (in steps perrevolution) yields the rotational speed inrevolutions per second.

PWMPulse Width Modulation. A method ofcontrolling the average current in amotors phase windings by varying theon-time (duty cycle) of transistorswitches.

RampingThe acceleration and deceleration of amotor. May also refer to the change infrequency of the applied step pulse train.

Rated TorqueThe torque producing capacity of amotor at a given sped. This is themaximum torque the motor can deliverto a load and is usually specified with atorque/speed curve.

RegenerationUsually refers to a circuit in a driveamplifier that accepts and drains energyproduced by a rotating motor eitherduring deceleration or free-wheelshutdown.

Registration MoveChanging the predefined move profilethat is being executed, to a differentpredefined move profile following receiptof an input or interrupt.

RepeatabilityThe degree to which the positioningaccuracy for a given move performancedrepetitively can be duplicated.

ResolutionThe smallest positioning increment thatcan be achieved. Frequently defined asthe number of steps required for amotor’s shaft to rotate one completerevolution.

ResolverA feedback device with a constructionsimilar to a motor’s construction (statorand rotor). Provides velocity and positioninformation to a drive’s microprocessoror DSP to electronically commutate themotor.

ParallelRefers to a data communication formatwherein many signal lines are used tocommunicate more than one piece of dataat the same time.

Phase AngleThe angle at which the steady state inputsignal to a system leads the output signal.

Phase MarginThe difference between 180° and thephase angle of a system at its crossoverfrequency.

PLCProgrammable logic controller; a machinecontroller that activates relays and other I/O units from a stored program. Additionalmodules support motion control and otherfunctions.

OutputInput

Phase Angle

ResonanceDesignates the condition resulting from energizing amotor at a frequency at or close to the motor’snatural frequency. Lower resolution, open-loopsystems will exhibit large oscillations from minimalinput.

RingingOscillation of a system following a sudden changein state.

RMS Torque

For an intermittent duty cycle application, the RMSTorque is equal to the steady- state torque thatwould produce the same amount of motor heatingover long periods of time.

TRMS = (Ti2 ti)

ti

Where: Ti = Torque during interval iti = Time of interval i

RS-232CA data communications standard that encodes astring of information on a single line in a timesequential format. The standard specifies theproper voltage and time requirements so thatdifferent manufacturers devices are compatible.

ServoA system consisting of several devices whichcontinuously monitor actual information (position,velocity), compares those values to desiredoutcome and makes necessary corrections tominimize that difference.

SlewIn motion control, the portion of a move made at aconstant non-zero velocity.

Static TorqueThe maximum torque available at zero speed.

Step AngleThe angle the shaft rotates upon receipt of a singlestep command.

StiffnessThe ability to resist movement induced by anapplied torque. Is often specified as a torquedisplacement curve, indicating the amount a motorshaft will rotate upon application of a knownexternal force when stopped.

SynchronismA motor rotating at a speed correctly correspondingto the applied step pulse frequency is said to be insynchronism. Load torques in excess of themotor’s capacity (rated torque) will cause a loss ofsynchronism. The condition is not damaging to astep motor.

TorqueForce tending to produce rotation.

Torque ConstantKT = The torque generated in a DC motor per unitAmpere applied to its windings.

KT = T oz-in A amp

Simplified for a brushless motor at 90° commutationangle.

Torque RippleThe cyclical variation of generated torque at afrequency given by the product of motor angularvelocity and number of commutator segments ormagnetic poles.

Torque-to-Inertia RatioDefined as a motor’s holding torque divided by theinertia of its rotor. The higher the ratio, the higher amotor’s maximum acceleration capability will be.

Transfer FunctionA mathematical means of expressing the output toinput relationship of a system. Expressed as afunction of frequency.

TriggersInputs on a controller that initiate or “trigger” thenext step in a program.

TTLTransistor-Transistor Logic. Describes a commondigital logic device family that is used in mostmodern digital electronics. TTL signals have twodistinct states that are described with a voltage – alogical “zero” or “low” is represented by a voltage ofless than 0.8 volts and a logical “one” or “high” isrepresented by a voltage from 2.5 to 5 volts.

Voltage ConstantKE = The back EMF generated by a DC motor at adefined speed. Usually quoted in volts per 1000rpm.

ZeroA frequency at which the transfer function of asystem goes to zero.

MM

A70

Glossary of Terms

A71

Rotary Inertia Conversion TableDon’t confuse mass-inertia with weight-inertia: mass inertia = wt. inertia

gTo convert from A to B, multiply by entry in Table.

A N-m N-cm dyn-cm kg-m kg-cm g-cm oz-in ft-lbs in-lbs

N-m 1 102 107 0.1019716 10.19716 1.019716-104 141.6119 0.737562 8.85074

N-cm 10-2 1 105 1.019716-10-3 0.1019716-3 1.019712-102 1.41612 7.37562-10-3 8.85074-10-2

dyn-cm 10-7 10-5 1 1.019716-10-8 1.01972-10-6 1.01972-10-3 1.41612-10-5 7.37562-10-8 8.85074-10-7

kg-m 9.80665 9.80665-102 9.80665-107 1 102 105 1.38874-103 7.23301 86.79624

kg-cm 9.80665-10-2 9.80665 9.80665-105 10-2 1 103 13.8874 7.23301-10-2 0.86792

g-cm 9.80665-10-5 9.80665-10-3 9.80665-102 10-5 10-3 1 1.38874-10-2 7.23301-10-5 8.679624-10-4

oz-in 7.06155-10-3 0.706155 7.06155-104 7.20077-10-4 7.20077-10-2 72,0077 1 5.20833-10-3 6.250-10-2

ft-lbs 1.35582 1.35582-102 1.35582-107 0.1382548 13.82548 1.382548-104 192 1 12

in-lbs 0.112085 11.2985 1.12985-106 1.15212-10-2 1.15212 1.15212-103 16 8.33333-10-2 1

B

Densities of Common MaterialsMaterial oz/in3 gm/cm3

Aluminum (cast or hard-drawn) 1.54 2.66

Brass (cast or rolled 60% CU; 40% Zn) 4.80 8.30

Bronze (cast, 90% CU; 10% Sn) 4.72 8.17

Copper (cast or hand-drawn) 5.15 8.91

Plastic 0.64 1.11

Steel (hot or cold rolled, 0.2 or 0.8% carbon) 4.48 7.75

Hard Wood 0.46 0.80

Soft Wood 0.28 0.48

Torque Conversion TableTo convert from A to B, multiply by entry in Table.

Technical Data

B lb-ft-s2

A kg-m2 kg-cm2 g-cm2 kg-m-sec2 kg-cm-sec2 g-cm-sec2 oz-in2 oz-in-s2 lb-in-2 lb-in-s2 lb-ft2 (slug-ft-2)

kg-m2 1 104 107 0.10192 10.1972 1.01972-104 5.46745-104 1.41612-102 3.41716-103 8.850732 23.73025 0.73756

kg-cm2 10-4 1 103 1.01972-10-5 1.01972-10-3 1.01972 5.46745 1.41612-10-2 0.341716 8.85073-10-4 2.37303-10-3 7.37561-10-5

g-cm2 10-7 10-3 1 1.01972-10-8 1.01972-10-6 1.01972-10-3 5.46745-10-3 1.41612-10-5 3.41716-10-4 8.85073-10-7 2.37303-10-6 7.37561-10-8

kg-m-s2 9.80665 9.80665-104 9.80665-107 1 102 105 5.36174-105 1.388674-103 3.35109-104 86.79606 2.32714-102 7.23300

kg-cm-s2 9.80665-10-2 9.80665-102 9.80665-105 10-2 1 103 5.36174-103 13.88741 3.35109-102 0.86796 2.327143 7.23300-10-2

g-cm-s2 9.80665-10-5 0.980665 9.80665-102 10-5 10-3 1 5.36174 1.38874-10-2 0.335109 8.67961-10-4 2.32714-10-3 7.23300-10-5

oz-in2 1.82901-10-5 0.182901 1.82901-102 1.86506-10-6 1.86506-10-4 0.186506 1 2.59008-10-3 6.250-10-2 1.61880-10-4 4.34028-10-4 1.34900-10-5

oz-in-s2 7.06154-10-3 70.6154 7.06154-104 7.20077-10-4 7.20077-10-2 72.00766 3.86089-102 1 24.13045 6.250-10-2 0.167573 5.20833-10-3

lb-in2 2.92641-10-4 2.92641 2.92641-103 2.98411-10-5 2.98411-10-3 2.98411 16 4.14414-10-2 1 2.59008-10-3 6.94444-10-3 2.15840-10-4

lb-in-s2 0.112985 1.12985-103 1.12985-106 1.15213-10-2 1.152126 1.15213-103 6.17740-103 16 3.86088-102 1 2.681175 8.3333-10-2

lb-ft2 4.21403-10-2 4.21403-102 4.21403-105 4.29711-10-3 0.429711 4.297114-102 2.304-103 5.96755 144 0.372971 1 3.10809-10-2

lb-ft-s2

(slug ft2) 1.35583 1.35582-104 1.35582-107 0.138255 13.82551 1.38255-104 7.41289-104 192 4.63306-103 12 32.1740 1

Calculate Horsepower

Horsepower =

Torque = oz-in

Speed = revolutions per second

* The horsepower calculation uses the torqueavailable at the specified speed

1 Horsepower = 746 watts

Most tables give densities in lb/ft3. To convert to oz/in3

divide this value by 108. To convert lb/ft3 to gm/cm3

divide by 62.5. The conversion from oz/in3 to gm/cm3

is performed by multiplying oz/in3 by 1.73.

Reference: Elements of Strength of Materials,S. Timoshinko and D.H. Young, pp. 342-343.

Torque x Speed16,800

AE

ngin

eeri

ng R

efe

rence

A72

Application Examples

Feed-to-lengthApplications in which a continuous web, strip, orstrand of material is being indexed to length, mostoften with pinch rolls or some sort of grippingarrangement. The index stops and some processoccurs (cutting, stamping, punching, labeling,etc.).

Application No. Page1: BBQ Grill-Making Machine .................... A73

2: Film Advance .......................................... A74

3: On-the-Fly Welder .................................. A75

X/Y Point-to-pointApplications that deal with parts handlingmechanisms that sort, route, or divert the flow ofparts.

Application No. Page4: Optical Scanner ...................................... A76

5: Circuit Board Scanning........................... A77

Metering/DispensingApplications where controlling displacement and/or velocity are required to meter or dispense aprecise amount of material.

Application No. Page6: Telescope Drive ...................................... A78

7: Engine Test Stand .................................. A79

8: Capsule Filling Machine .......................... A80

Indexing/ConveyorApplications where a conveyor is being driven in arepetitive fashion to index parts into or out of anauxiliary process.

Application No. Page9: Indexing Table ........................................ A81

10: Rotary Indexer ........................................ A82

11: Conveyor ................................................ A83

ContouringApplications where multiple axes of motion areused to create a controlled path, (e.g., linear orcircular interpolation).

Application No. Page12: Engraving Machine ................................. A84

13: Fluted-Bit Cutting Machine ..................... A85

Tool FeedApplications where motion control is used to feed acutting or grinding tool to the proper depth.

Application No. Page14: Surface Grinding Machine ...................... A86

15: Transfer Machine .................................... A87

16: Flute Grinder ........................................... A88

17: Disc Burnisher ........................................ A89

WindingControlling the process of winding material arounda spindle or some other object.

Application No. Page18: Monofilament Winder .............................. A90

19: Capacitor Winder ................................... A91

FollowingApplications that require the coordination of motionto be in conjunction with an external speed orposition sensor.

Application No. Page20: Labelling Machine ................................... A92

21: Window Blind Gluing .............................. A93

22: Moving Positioning Systems ................... A94

Injection MoldingApplications where raw material is fed by gravityfrom a hopper into a pressure chamber (die ormold). The mold is filled rapidly and considerablepressure is applied to produce a molded product.

Application No. Page23: Plastic Injection Molding ......................... A95

Flying CutoffApplications where a web of material is cut whilethe material is moving. Typically, the cutting devicetravels at an angle to the web and with a speedproportional to the web.

Application No. Page24: Rotating Tube Cutting ............................ A96

Summary of Application Examples

A73

AE

ngin

eeri

ng R

efe

rence

Application Examples1. BBQ Grill-Making Machine

Application Type: Feed-to-LengthMotion: Linear

Application Description: A manufactuer wasusing a servo motor to feed material into amachine to create barbeque grills, shopping carts,etc. The process involves cutting steel rods andwelding the rods in various configurations.However, feed-length was inconsistent becauseslippage between the drive roller and the materialwas too frequent. Knurled nip-rolls could not beused because they would damage the material.The machine builder needed a more accuratemethod of cutting the material at uniform lengths.The customer used a load-mounted encoder toprovide feedback of the actual amount of materialfed into the cutting head.

Machine Objectives:• Compnesate for material slippage• Interface with customer’s operator panel• Smooth repeatable operation• Variable length indexes• High reliability

Motion Control Requirements:• Accurate position control• Load-mounted encoder feedback• High-speed indexing• XCode languageApplication Solution: By using the globalposition feedback capability of the BLHX drive, themachine builder was able to close the positionloop with the load-mounted encoder, while thevelocity feedback was provided by the motor-mounted encoder and signal processing. The two-encoder system provides improved stability andhigher performance than a single load-mountedencoder providing both position and velocityfeedback. The load-mounted encoder wascoupled to friction drive nip-rollers close to the cuthead.

Product Solutions:

Controller/Drive Motor

BLHX75BN ML3450B-10

BLHX150BN Servo Drive

Cutting Head

Nip-Roll and Load Mounted

Encoder

Motor and Drive Roll

Spool

A74

Application Examples

Drive/Indexer

Motor

2. Film Advance

Application Type: Feed-to-LengthMotion: Linear

Tangential drives consist of a pulley or pinionwhich, when rotated, exerts a force on a belt orracks to move a linear load. Common tangentialdrives include pulleys and cables, gears andtoothed belts, and racks and pinions.

Tangential drives permit a lot of flexibility in thedesign of drive mechanics, and can be veryaccurate with little backlash. Metal chains shouldbe avoided since they provide little or no motordamping.

Application Description: A movie camera isbeing modified to expose each frame undercomputer control for the purpose of generatingspecial effects. A motor will be installed in thecamera connected to a 1/2-inch diameter, 2-inchlong steel film drive sprocket and must index oneframe in 200 milliseconds. The frame spacing is 38mm (1.5").

Machine Requirements:• Index one frame within 200 milliseconds• Indexer must be compatible with BCD interface• Fast rewind and frame indexingMotion Control Requirements:• Little to no vibration at rest—∴ Stepper• Minimum settling time• Preset and slew moves

Application Solution:In this application, the move distance and time areknown, but the required acceleration is not known.The acceleration may be derived by observingthat, for a trapezoidal move profile with equalacceleration, slew and deceleration times, 1/3 ofthe move time is spent accelerating and 1/3 of thetotal distance is travelled in that time (a trapezoidalmove).

It is determined that the acceleration required is107.4 rps2 at a velocity of 7.166 rps. Assume thatthe film weighs 1 oz. and total film friction is 10 oz-in. The rotor, sprocket, and film inertia is calculatedto be 0.545 oz-in/sec2. Solving the torque formulaindicates that the motor for this application mustprovide 11.9 oz-in to drive the film and pulley (referto Direct Drive Formulas on p. A63).

An indexer is selected to be connected to a BCDinterface in the camera electronics. Preset andSlew modes on the indexer are then controlled bythe camera electronics to provide fast rewind andframe indexing.

Product Solutions:

Drive/Indexer Motor

SX S57-51-MO

A75

AE

ngin

eeri

ng R

efe

rence

Application Examples3. On-the-Fly Welder

Application Type: Feed-to-LengthMotion: Linear

Description: In a sheet metal fabrication process,an unfastened part rides on a conveyor beltmoving continuously at an unpredictable velocity.Two spot-welds are to be performed on each part,4 inches apart, with the first weld 2 inches from theleading edge of the part. A weld takes one second.

Machine Objectives• Standalone operation• Position welder according to position and

velocity of each individual part• Welding and positioning performed without

stopping the conveyor• Welding process must take 1 second to

completeMotion Control Requirements• Programmable I/O; sequence storage• Following• Motion profiling; complex following• High linear acceleration and speedApplication Solution:This application requires a controller that canperform following or motion profiling based on aprimary encoder position. In this application, thecontroller will receive velocity and position datafrom an incremental encoder mounted to a rolleron the conveyor belt carrying the unfastened parts.The conveyor is considered the primary drivesystem. The secondary motor/drive systemreceives instructions from the controller, based ona ratio of the velocity and position informationsupplied by the primary system encoder. The linearmotor forcer carries the weld head and is mountedon an overhead platform in line with the conveyor.

Linear motor technology was chosen to carry theweld head because of the length of travel. Thelinear step motor is not subject to the same linearvelocity and acceleration limitations inherent insystems converting rotary to linear motion. For

example, in a leadscrew system, the inertia of theleadscrew frequently exceeds the inertia of theload and as the length of the screw increases, sodoes the inertia. With linear motors, all the forcegenerated by the motor is efficiently applieddirectly to the load; thus, length has no effect onsystem inertia. This application requires a 54-inchplaten to enable following of conveyor speeds over20 in/sec.

Application Process1. A sensor mounted on the weld head detects

the leading edge of a moving part and sends atrigger pulse to the controller.

2. The controller receives the trigger signal andcommands the linear motor/drive to ramp up totwice the speed of the conveyor. This providesan acceleration such that 2 inches of the partpasses by the weld head by the time the weldhead reaches 100% of the conveyor velocity.

3. The controller changes the speed ratio to 1:1,so the weld head maintains the speed of theconveyor for the first weld. The weld takes 1second.

4. The following ratio is set to zero, and thewelder decelerates to zero velocity over 2inches.

5. The controller commands the linear forcer torepeat the same acceleration ramp as in step1above. This causes the weld head to positionitself, at an equal velocity with the conveyor, 4inches behind the first weld.

6. Step 3 is repeated to make the second weld.7. Once the second weld is finished, the controller

commands the linear forcer to return the weldhead to the starting position to wait for the nextpart to arrive.

Product Solutions:

Indexer Drive Motor Encoder

Model 500 L Drive PO-L20-P54 -E

Weld Head

Linear Motor

Indexer

Microstepping Drive

Encoder (Mounted

to Conveyor)

Spot Welds

A76

Application Examples4. Optical Scanner

Application Type: X-Y Point-to-PointMotion: Rotary

Application Description: A dye laser designerneeds to precisely rotate a diffraction grating undercomputer control to tune the frequency of thelaser. The grating must be positioned to an angularaccuracy of 0.05°. The high resolution of themicrostepping motor and its freedom from“hunting” or other unwanted motion when stoppedmake it ideal.

Machine Requirements:• System must precisely rotate a diffraction

grating to tune the frequency of the laser• PC-compatible system control• Angular accuracy of 0.05º• IEEE-488 interface is requiredMotion Control Requirements:• High resolution—∴ Microstepper• Little to no vibration at rest—∴ Stepper• No “hunting” at the end of move—∴ Stepper• Limited space is available for motor—∴ small

motor is required

Application Solution:The inertia of the grating is equal to 2% of theproposed motor’s rotor inertia and is thereforeignored. Space is at a premium in the cavity and asmall motor is a must. A microstepping motor,which provides ample torque for this application, isselected.

The laser’s instrumentation is controlled by acomputer with an IEEE-488 interface. An indexerwith an IEEE-488 interface is selected. It ismounted in the rack with the computer and iscontrolled with a simple program written in BASICthat instructs the indexer to interrupt the computerat the completion of each index.

Product Solutions:

Indexer Drive Motor

Model 4000 LN Drive LN57-51

Drive

Motor

Model 4000

A77

AE

ngin

eeri

ng R

efe

rence

Application Examples5. Circuit Board Scanning

Application Type: X-Y Point-to-PointMotion: Linear

Application Description: An Original EquipmentManufacturer (OEM) manufactures X-Ray Scanningequipment used in the quality control of printedcircuit boards and wafer chips.

The OEM wants to replace the DC motors,mechanics and analog controls with an automatedPC-based system to increase throughput andeliminate operator error. The host computer willinteract with the motion control card using a “C”language program. The operator will have theoption to manually override the system using ajoystick.

This machine operates in an environment wherePWM (pulse width modulation) related EMIemission is an issue.

Machine Requirements:• 2-Axis analog joystick• Joystick button• Travel limits• Encoder feedback on both axesDisplay Requirements:• X and Y position coordinatesOperator Adjustable Parameters:• Dimensions of sample under test• (0,0) position—starting pointMotion Control Requirements:• AT-based motion controller card• Replace velocity control system (DC motors) and

mechanics with more accurate and automatedpositioning scheme

• Manual Joystick control• Continuous display of X & Y axis position• User-friendly teach mode operations• Low EMI amplifiers (drives)Application Solution:The solution of this application uses the existingPC by providing a PC-based motion controller and

the AT6400 to control both axes. A microsteppingdrive is used because its linear amplifiertechnology produces little EMI. The PC monitor isthe operator interface.

A “C” language program controls the machine.

Machine operation begins with a display to the operatorof a main menu. This main menu lets the operatorselect between three modes: Automated Test, JoystickPosition and Teach New Automated Test.

In Automated Test mode, the PC displays a menuof preprogrammed test routines. Each of theseprograms has stored positions for the different testlocations. This data is downloaded to the controllerwhen a test program is selected. The controllercontrols the axes to a home position, moves toeach scan position, and waits for scan completionbefore moving to the next position.

In Joystick Position mode, the controller enablesthe joystick allowing the operator to move in bothX and Y directions using the joystick. The AT6400waits for a signal from the PC to indicate that thejoystick session is over.

When Teach mode is selected, the PC downloads ateach program to the controller (written by the user).After the axes are homed, the controller enables thejoystick and a “position select” joystick button. Theoperator then jogs axes to a position and pressesthe “position select” button. Each time the operatorpresses this “position select” button, the motioncontroller reads this position into a variable andsends this data to the PC for memory storage.These new position coordinates can now be storedand recalled in Automated Test mode.

Product Solutions:

Controller Drive Motor Accessories

AT6400-AUX1 LN Drive LN57-83-MO -EDaedal X-Y Table

Joystick

Motor

Joystick

Drive

Drive

Motor

Indexer

A78

Application Examples6. Telescope Drive

Application Type: Metering/DispensingMotion: Rotary

Traditional gear drives are more commonly usedwith step motors. The fine resolution of amicrostepping motor can make gearingunnecessary in many applications. Gears generallyhave undesirable efficiency, wear characteristics,backlash, and can be noisy.

Gears are useful, however, when very large inertiasmust be moved because the inertia of the loadreflected back to the motor through the gearing isdivided by the square of the gear ratio.

In this manner large inertial loads can be movedwhile maintaining a good load inertia-to-rotorinertia ratio (less than 10:1).

Application Description: An astronomerbuilding a telescope needs to track celestial eventsat a slow speed (15°/hour) and also slew quickly(15° in 1 second).

Machine Requirements:• Smooth, slow speed is required–∴ microstepper• High data-intensive application–∴ bus-based

indexer• Future capabilities to control at least 2 axes of

motion• Visual C++ interfaceMotion Control Requirements:• High resolution• Very slow speed (1.25 revolutions per hour)—

microstepping• AT bus-based motion controller card• Dynamic Link Library (DDL) device driver must

be provided with indexer. This helps Windows™programmers create Windows-basedapplications (i.e., Visual C++) to interface withthe indexer

Application Solution:A 30:1 gearbox is selected so that 30 revolutionsof the motor result in 1 revolution (360°) of thetelescope. A tracking velocity of 15°/hourcorresponds to a motor speed of 1.25 revs/hour orabout 9 steps/sec. on a 25,000 steps/rev. Moving15° (1.25 revolutions) in 1 second requires avelocity of 1.25 rps.

The inverse square law causes the motor to see 1/900 of the telescope’s rotary inertia. The equationsare solved and the torque required to acceleratethe telescope is 455 oz-in. The step pulsesrequired to drive the motor are obtained from alaboratory oscillator under the operator’s control.

Product Solutions:

Indexer Drive Motor

AT6200-AUX1* S Drive S106-178

* To control up to four axes, refer to the AT6400.

Drive

Computer (Indexer installed

in a PC)

Motor

A79

AE

ngin

eeri

ng R

efe

rence

Application Examples7. Engine Test Stand

Application Type: Metering/DispensingMotion: Rotary

Application Description: A jet enginemanufacturer is building a test facility for makingoperational measurements on a jet engine. Thethrottle and three other fuel flow controls need tobe set remotely. While the application only calls fora rotary resolution of 1 degree (1/360 rev.), thesmoothness and stiffness of a microsteppingsystem is required.

Motor speeds are to be low and the inertias of thevalves connected to the motors are insignificant.The main torque requirement is to overcome valvefriction.

Machine Requirements:• Low wear• Remote operation• High reliabilityMotion Control Requirements:• Motor velocity is low• High stiffness at standstill• Slow-speed smoothness• Four axes of control• Homing function

Application Solution:Each valve is measured with a torque wrench.Two valves measure at 60 oz-in and the other twomeasure at 200 oz-in. Two high-power and twolow-power microstepping motor/drives systemsare selected. These choices provideapproximately 100% torque margin and result in aconservative design.

The operator would like to specify each valveposition as an angle between 0° and 350°.Home position switches are mounted on the testrig and connected to each indexer to allow forpower-on home reference using the indexer’shoming feature.

Product Solutions:

Indexer Drive Motor

AT6400* S Drive S57-102

* A standalone indexer could also be used(instead of a bus-based indexer), refer to theModel 4000.

Drive

Motor

MotorDrive

Drive

Motor

MotorDrive

Computer (Indexer installed in a PC)

A80

Application Examples8. Capsule Filling Machine

Application Type: Metering/DispensingMotion: Linear

Application Description: The design requires amachine to dispense radioactive fluid into capsules.After the fluid is dispensed, it is inspected and thedata is stored on a PC. There is a requirement toincrease throughput without introducing spillage.

Machine Requirements:• Increase throughput• No spilling of radioactive fluid• Automate two axes• PC compatible system control• Low-cost solution• Smooth, repeatable motionMotion Control Requirements:• Quick, accurate moves• Multi-axis controller• PC bus-based motion control card• Open-loop stepper if possible• High-resolution motor/drive (microstepping)

Application Solution:The multi-axis indexer is selected to control andsynchronize both axes of motion on one cardresiding in the IBM PC computer. An additionalfeature is the integral I/O capability that’snecessary to activate the filling process. Thehorizontal axis carrying the tray of capsules isdriven by a linear motor. The simple mechanicalconstruction of the motor makes it easy to apply,and guarantees a long maintenance-free life. Thevertical axis raises and lowers the filling head andis driven by a microstepping motor and aleadscrew assembly. A linear motor was alsoconsidered for this axis, but the fill head wouldhave dropped onto the tray with a loss of powerto the motor. Leadscrew friction and the residualtorque of the step motor prevents this occurrence.

Product Solutions:

Indexer Drive Motor

AT6200 Axis 1: ZETA Drive S57-51Axis 2: ZETA Drive PO-L20-P18

Filling Heads

Hose

Tray of Empty Capsules

Full Capsules

Platen

Linear Motor

Motor with Leadscrew

Top View

Side View

Drive

Computer (Indexer installed in a PC)

Drive

A81

AE

ngin

eeri

ng R

efe

rence

Application Examples9. Indexing Table

Application Type: Indexing/ConveyorMotion: Linear

Application Description: A system is requiredto plot the response of a sensitive detector thatmust receive equally from all directions. It ismounted on a rotary table that needs to beindexed in 3.6° steps, completing each indexwithin one second. For set-up purposes, the tablecan be positioned manually at 5 rpm. The tableincorporates a 90:1 worm drive.

Machine Requirements:• Low-EMI system• Repeatable indexing• Remote operation• Table speed of 5 rpmMotion Control Requirements:• Jogging capability• Sequence select functionality• Capable of remote drive shutdown

Application Solution:The maximum required shaft speed (450 rpm) iswell within the capability of a stepper, which is anideal choice in simple indexing applications.Operating at a motor resolution of 400 steps/rev,the resolution at the table is a convenient 36,000step/rev. In this application, it is important thatelectrical noise is minimized to avoid interferencewith the detector. Two possible solutions are touse a low-EMI linear drive or to shut down thedrive after each index (with a stepper driving a90:1 worm gear there is no risk of position lossduring shutdown periods).

Product Solutions:

Indexer Drive Motor

Model 500 LN Drive LN57-102

* The SX drive/indexer and PK2 drive are otherproducts that have been used in these types ofapplications.

Drive Indexer

Radiation Source

Rotary Stage

Motor

Detector

A82

Application Examples10. Rotary Indexer

Application Type: Indexing ConveyorMotion: Rotary

Application Description: An engineer for apharmaceutical company is designing a machineto fill vials and wants to replace an old styleGeneva mechanism. A microstepping motor willprovide smooth motion and will prevent spillage.

The indexing wheel is aluminum and is 0.250-inchthick and 7.5" in diameter. Solving the equation forthe inertia of a solid cylinder indicates that thewheel has 119.3 oz-in2. The holes in the indexingwheel reduce the inertia to 94 oz-in2. The vialshave negligible mass and may be ignored for thepurposes of motor sizing. The table holds 12 vials(30° apart) that must index in 0.5 seconds anddwell for one second. Acceleration torque iscalculated to be 8.2 oz-in at 1.33 rps2. A triangularmove profile will result in a maximum velocity of0.33 rps. The actual torque requirement is lessthan 100 oz-in. However, a low load-to-rotorinertia ratio was necessary to gently move the vialsand fill them.

Machine Requirements:• Smooth motion• PLC control• Variable index lengthsMotion Control Requirements:• Smooth motion• Sequence select capability• I/O for sequence select• Programmable acceleration and decelerationApplication Solution:The index distance may be changed by theengineer who is controlling the machine with aprogrammable controller. Move parameters will bechanging and can therefore be set via BCD inputs.The indexer can be “buried” in the machine andactivated with a remote START input.

Product Solutions:

Drive Indexer Motor

SX Drive Indexer* S83-135

* The 6200, AT6200, and Model 500 are otherindexer products that have been used in thesetypes of applications.

PLC Programmable Logic Controller

Controller Drive

A83

AE

ngin

eeri

ng R

efe

rence

Application Examples11. Conveyor

Application Type: Indexing/ConveyorMotion: Linear

Tangential drives consist of a pulley or pinionwhich, when rotated, exerts a force on a belt orracks to move a linear load. Common tangentialdrives include pulleys and cables, gears andtoothed belts, and racks and pinions.

Tangential drives permit a lot of flexibility in thedesign of drive mechanics, and can be veryaccurate with little backlash. Metal chains shouldbe avoided since they provide little or no motordamping.

Application Description: A machine visionsystem is being developed to automatically inspectsmall parts for defects. The parts are located on asmall conveyor and pass through the camera’sfield of view. The conveyor is started and stoppedunder computer control and the engineer wants touse a system to drive the conveyor because it isnecessary for the part to pass by the camera at aconstant velocity.

It is desired to accelerate the conveyor to a speedof 20 inches/sec. in 100 milliseconds. A flat timingbelt weighing 20 ozs. is driven by a 2-inch diameteraluminum pulley 4 inches wide (this requires amotor velocity of 3.2 rps). The maximum weight ofthe parts on the pulley at any given time is 1 lb.and the load is estimated to have an inertia of 2 oz-in2. Static friction of all mechanical components is30 oz-in. The required motor toque wasdetermined to be 50.9 oz-ins (refer to Direct DriveFormulas on p. A63).

Machine Requirements:• Computer-controlled system• High accuracy• Low backlashMotion Control Requirements:• Accurate velocity control• Linear motion• High resolution• AT bus-based motion control cardApplication Solution:A computer controls the entire inspectionmachine. A bus-based compatible indexer cardwas selected. A microstepping motor/drive systemthat supplied 100 oz-in of static torque was alsochosen to complete the application.

Product Solutions:

Indexer Drive Motor

PC21* S Drive S57-83

* The AT6200 and AT6400 are other PC-basedindexer products that are often used in thesetypes of applications.

Drive

Motor

Computer (Indexer installed

in a PC)

A84

Application Examples12. Engraving Machine

Application Type: ContouringMotion: Linear

Application Description: An existing engravingmachine requires an upgrade for accuracy beyond0.008 inches, capability and operatingenvironment. Using a personal computer as thehost processor is desirable.

Machine Requirements:• Positional accuracy to 0.001 inches• Easy-to-use, open-loop control• CNC machining capability• Interface-to-digitizer pad• Compatibility with CAD systemsMotion Control Requirements:• High resolution• Microstepping• G-Code compatibility• IBM PC compatible controller

Application Solution:A four-axis motion controller resides on the bus ofan IBM compatible computer, allowing fullintegrated control of four axes of motion. Axes 3and 4 are synchronized to prevent table skew.CompuCAM’s G-Code package allows the user toprogram in industry-standard machine toollanguage (RS274 G-Code) or to import CAD fileswith CompuCAM-DXF. Open-loop microsteppingdrives with precision leadscrews give positionalaccuracies better than the desired ±0.001 inch.This simple retrofit to the existing hardware greatlyimproved system performance.

Product Solutions:

Indexer Drives Motor

AT6400* S Drives S83-135

* The Model 4000 (standalone) and AT6450 areservo controller products that have also beenused in these types of applications.

IBM PC with Indexer

Axis 4

Axis 1

Axis 2

Digitizer Pad

Drives

Motor

Motor

Motor

Motor

Axis 3

A85

AE

ngin

eeri

ng R

efe

rence

Application Examples13. Fluted-Bit Cutting Machine

Application Type: ContouringMotion: Linear

Application Description: The customermanufactures a machine that cuts a metal cylinderinto fluted cutting bits for milling machines. Themachine operation employed a mechanical camfollower to tie the bit’s rotation speed to thetraverse motion of the bit relative to the cuttingtool. The cut depth was manually adjusted using ahand crank.

This arrangement was acceptable when thecompany had a bit for the cam they wanted togrind. Unfortunately, custom prototype bits madeof titanium or other high-tech metals required thatthey make a cam before they could machine thebit, or do those parts on a $10,000 CNC screwmachine. Both of these alternatives were tooexpensive for this customer.

Machine Requirements:• Machine must be capable of making low-

volume custom bits as well as high-volumestandard bits—an be economical for bothprocesses.

• Quick set-up routine• Operator interface for part entryMotion Control Requirements:• Smooth motion• Four axes of coordinated motion• 2 axes of linear interpolation• Math capabilitiesApplication Solution:Controlled by a multi-axis step and directioncontroller, microstepping motors and drives areattached to four axes for smooth, programmablemotion at all speeds.

• Axis 1: Alignment• Axis 2: Chamfer (cutting depth)• Axis 3: Traverse• Axis 4: RotationTo allow for the flexibility required to cut a bit at adesired pitch, the traverse and rotation axes (axes3 and 4) are synchronized along a straight line.The controller’s linear interpolation allows thisfunctionality. Both the alignment and chamfer axes(axes 1 and 2) remain stationary during the cuttingprocess.

The controller’s operator input panel and mathcapabilities allow the operator to enter the bitdiameter, desired pitch, depth, and angle. Usingthese part specifications, the controller generatesall motion profiles and stores them in nonvolatilebattery-backed RAM. Programming isaccomplished with the controller’s menu-drivenlanguage. The typical process is as follows:

1. Axis 1 aligns the center line of the bit to thecutting tool.

2. Axis 2 lowers the cutting tool to the desiredcutting depth (chamfer).

3. Axis 3 traverses the bit along the cutting tool.

4. While axis 3 traverses, axis 4 rotates the bit tocreate the desired pitch.

Product Solutions:

Indexer Drives Motor

Model 4000* S Drives S83-135

* The Model AT6400 and AT6450 are othercontrollers that have been used in these typesof applications.

Drive

Drive

Drive

Drive

Indexer

Motor (Axis 2 - Chamfer)

Cutting Tool

Bit

Motor (Axis 1 - Alignment)

Motor (Axis 4 - Rotation)

Motor (Axis 3 - Traverse)

A86

Application Examples14. Surface Grinding Machine

Application Type: Tool FeedMotion: Linear

Application Description: A specialty machineshop is improving the efficiency of its surfacegrinding process. The existing machine is soundmechanically, but manually operated. Automatingthe machine will free the operator for other tasks,which will increase overall throughput of themachine shop.

Machine Requirements:• Allow flexibility to machine various parts• Easy set up for new parts• Automate all three axes• Keep operator informed as to progress• Low-cost solution• High-resolution grindingMotion Control Requirements:• Nonvolatile memory for program storage• Teach mode• Multi-axis controller• Interactive user configurable display• Open-loop stepper if possible• High resolution motor/drive (microstepping)

Application Solution:A four-axis motion controller with a user-configurable front panel is required for thisapplication. An indexer with a sealed, backlitdisplay would be ideal for the application’sindustrial environment (machine shop). Thecontroller’s Teach mode and sizable nonvolatilememory allows for easy entry and storage of newpart programs. Microstepping drives, which plentyof power, resolution, and accuracy are selectedinstead of more expensive closed-loop servosystems. The operator utilizes the controller’s jogfunction to position the grinding head at the proper“spark off” height. From this point, the controllertakes over and finishes the part while the operatorworks on other critical tasks. Increasing the partsrepeatability and throughput of the processjustified the cost of automating the machine.

Product Solutions:

Indexer Drive Motor

Model 4000* S Drive S83-93

* The AT6400 PC-based indexer has also beenused to solve similar applications.

Motor

Grinding Wheel

Motors

Indexer

Control Panel

Safety Guard

A87

AE

ngin

eeri

ng R

efe

rence

Application Examples15. Transfer Machine

Application Type: Tool FeedMotion: Linear

Application Description: A stage of a transfermachine is required to drill several holes in acasting using a multi-head drill. The motor has todrive the drill head at high speed to within 0.1" ofthe workpiece and then proceed at cutting speedto the required depth. The drill is then withdrawn atan intermediate speed until clear of the work, thenfast-retracted and set for the next cycle. Thecomplete drilling cycle takes 2.2 seconds with a0.6-second delay before the next cycle.

Due to the proximity of other equipment, the lengthin the direction of travel is very restricted. Anadditional requirement is to monitor the machinefor drill wear and breakage.

Machine Requirements:• Limited length of travel• Limited maintenance• Monitor and minimize drill damage• High-speed drillingMotion Control Requirements:• Packaged drive controller• Complex motion profile• High speed• High duty cycle

Application Solution:The combined requirements of high speed, highduty cycle and monitoring the drill wear all point tothe use of a servo motor. By checking the torqueload on the motor (achieved by monitoring drivecurrent), the drilling phase can be monitored (anincreased load during this phase indicates that thedrill is broken).

This type of application will require a ballscrewdrive to achieve high stiffness together with highspeed. One way of minimizing the length of themechanism is to attach the ballscrew to themoving stage and then rotate the nut, allowing themotor to be buried underneath the table. Sinceaccess for maintenance will then be difficult, abrushless motor should be selected.

Product Solutions:

Drive/Controller Motor

APEX6152 606 Motor

BallscrewRotating Nut

Motor

Table

Drill Head

Drive/Controller

A88

Application Examples

Application Solution:This is a natural application for stepper motors,since the speeds are moderate and the solutionmust be minimum-cost. The grinding processrequires that the two axes move at accuratelyrelated speed, so the controller must be capableof performing linear interpolation. The smalldynamic position error of the stepper systemensures that the two axes will track accurately atall speeds.

Product Solutions:

OperatorController Drive Motor Interface

6200* S Drive S83-135 RP240

* The Model 4000-FP has also been used tosolve similar applications.

16. Flute GrinderApplication Type: Tool FeedMotion: Linear

Application Description: A low-cost machinefor grinding the flutes in twist drills requires twoaxes of movement—one moves the drill forwardsunderneath the grinding wheel, the other rotatesthe drill to produce the helical flute. At the end ofthe cut, the rotary axis has to index the drill roundby 180° to be ready to grind the second flute. Thelinear speed of the workpiece does not exceed 0.5inches/sec.

Machine Requirements:• Two-axis control• Low cost• Easy set-up and change over of part programs• Smooth, accurate cutting motionMotion Control Requirements:• Two-axis indexer• Linear interpolation between axes• Nonvolatile program storage• Flexible data pad input• Moderate speeds• Programmable I/O

Twist Drill

Grinding Wheel

Rotary Head

X-Axis Motor

Controller Drive

Drive

Axis Motor

Operator Interface

Axis Drive

X-Axis Drive

A89

AE

ngin

eeri

ng R

efe

rence

Application Examples17. Disc Burnisher

Application Type: Tool FeedMotion: Rotary

Application Description: Rigid computer discsneed to be burnished so that they are flat to withintight tolerances. A sensor and a burnishing headmove together radially across the disc. When ahigh spot is sensed, both heads stop while theburnishing head removes the raised material. Thesurface speed of the disc relative to the headsmust remain constant, and at the smallestdiameter, the required disc speed is 2400 rpm.The machine operates in a clean environment, andtakes approximately one minute to scan anunblemished disk.

Machine Requirements:• High-speed burnishing• Surface speed of disc relative to the heads must

remain constant• Clean environment—∴ no brushed servo

motorsMotion Control Requirements:• Variable storage, conditional branching and

math capabilities• Linear interpolation between the head axes

(axes #1 and #2)• Change velocity on-the-fly• Programmable inputs

Application Solution:The drive for the disc requires continuousoperation at high speed, and a brushless solutionis desirable to help maintain clean conditions. Thenatural choice is a brushless servo system. Thespeed of this axis depends on head position andwill need to increase as the heads scan from theoutside to the center. To successfully solve thisapplication, the multi-axis indexer requires variablestorage, the ability to perform math functions, andthe flexibility to change velocity on-the-fly.

The sense and burnishing heads traverse at lowspeed and can be driven by stepper motors.Stepper motors—since the sense and burnishingheads need to start and step at the same time,linear interpolation is required.

Product Solutions:

Controller Drive #1 Drive #2 Drive #3

Model 4000* S Drive S Drive Z Drive

Motor #1 Motor #2 Motor #3

S83-93 S83-93 Z60

* The AT6400 PC-based indexer has also beenused in these types of applications.

Motor

Disc

Axis 3 Disc Drive Motor

Axis 2 Burnishing Head

Axis 1 Sensing Head

Motor

Multi Axis Controller (4000)

DriveDrive Drive

A90

Application Examples18. Monofilament Winder

Application Type: WindingMotion: Rotary

Application Description: Monofilament nylon isproduced by an extrusion process that results inan output of filament at a constant rate. Theproduct is wound onto a bobbin that rotates at amaximum speed of 2000 rpm. The tension in thefilament must be held between 0.2 lbs. and 0.6 lbsto ensure that it is not stretched. The windingdiameter varies between 2" and 4".

The filament is laid onto the bobbin by a ballscrew-driven arm, which oscillates back and forth atconstant speed. The arm must reverse rapidly atthe end of the move. The required ballscrew speedis 60 rpm.

Machine Requirements:• Controlled tension on monofilament• Simple operator interface• High throughputMotion Control Requirements:• 2 axes of coordinated motion• Linear interpolation• Constant torque from motor

Application Solution:The prime requirement of the bobbin drive is toprovide a controlled tension, which meansoperating in Torque mode rather than Velocitymode. If the motor produces a constant torque,the tension in the filament will be inverselyproportional to the winding diameter. Since thewinding diameter varies by 2:1, the tension will fallby 50% from start to finish. A 3:1 variation intension is adequate, so constant-torque operationis acceptable. (To maintain constant tension,torque must be increased in proportion to windingdiameter.)

This requirement leads to the use of a servooperating in torque mode (the need for constant-speed operation at 2000 rpm also makes astepper unsuitable). In practice, a servo in Velocitymode might be recommended, but with anoverriding torque limit, the programmed velocitywould be a little more than 2000 rpm. In this way,the servo will normally operate as a constant-torque drive. However, if the filament breaks, thevelocity would be limited to the programmedvalue.

The traversing arm can be adequately driven by asmaller servo.

Product Solutions:

Indexer Drive Motor

6250* BL30 ML2340

* The AT6450 PC-based servo controller and theAPEX20/APEX40 servo controllers have alsobeen used in this type of application.

Bobbin

Servo

ControllerDrive

Drive

Torque Motor

A91

AE

ngin

eeri

ng R

efe

rence

Application Examples19. Capacitor Winder

Application Type: WindingMotion: Linear

Application Description: The customer windsaluminum electrolytic capacitors. Six reels, twowith foil (anode and cathode) and four with paper,are all wound together to form the capacitor. Afterwinding the material a designated number of turns,the process is stopped and anode and cathodetabs are placed on the paper and foil. The tabsmust be placed so that when the capacitor iswound, the tabs end up 90° (±0.1°) from eachother. This process is repeated until the requirednumber of tabs are placed and the capacitorreaches its appropriate diameter.

The previous system used a PLC, conventional DCdrives, and counters to initiate all machinefunctions. DIP switches were used to change andselect capacitor lengths. Lengthy set-up andcalibration procedures were required for properoperation. In addition, material breakage wascommon, resulting in extensive downtime. Anoperator had to monitor the machine at all times toconstantly adjust the distances for accurate tabplacement.

Machine Requirements:• Constantly monitor the linear feed length of the

paper and foil and calculate the constantlychanging capacitor circumference as a functionof that length

• A complete motion control package is requiredto eliminate the need for a PLC and separatemotion cards

• Reduce time and complexity of set-up (too muchwiring in previous system)

• Reduce machine downtime caused by materialbreakage

Motion Control Requirements:• Following• Two axes of coordinated motion• Math capability• AT-based control card

Application Solution:Precise motion control of the material feed axesdemands closed-loop servo commands. Actuationof external cylinders and solenoids requires bothanalog and digital I/O. A flexible operator interfaceis needed for diagnostics and other alterations ofmachine function. Motion, I/O, and an operatorinterface should be provided with a machinecontroller.

The first motorized axis (mandril) pulls all sixmaterials together and feeds an appropriatedistance. An encoder is placed on this motor aswell as on the materials as they are fed into themandril. The controller constantly compares thetwo encoders to get an exact measurement oflinear distance, and compensates for materialstretching.

When the linear distance is achieved, the firstmotor comes to an abrupt stop while a secondaxis places a tab. The controller then initiates acold weld (pressure weld) of the tab onto thepaper and foil.

To avoid material breakage, constant tension isapplied to each of the six reels via air cylinders.Sensors are installed on all axes so that if a breakoccurs, the controller can stop the process.

A computer makes this process easy to use andset up. PC/AT-based support software allows theuser to build his controller command program.

The operator sets the diameter of the appropriatecapacitor, the operating speed and the number ofcapacitors (all via the keyboard). After thisprocess, the machine runs until a malfunctionoccurs or it has completed the job.

Product Solutions:

Controller Drive Motor Accessories

AT6250* BL30 ML2340 -E Encoder

* The 6250 standalone 2-axis servo controller andAPEX20/APEX40 servo drives have also beenused in these types of applications.

Opto I/O Rack

I/O to Limits, Cylinders and Solenoids

Encoder

Capacitor Wound Onto

Spindle

Spindel Axis Motor

and Encoder

Tab Feeder Axis Motor

Anode Tab

Reel

Cathode Tab Reel

Paper Reel

Paper Reel

Paper Reel

Paper Reel

Cathode Foil Reel

Anode Foil Reel

Computer (Indexer installed in a PC)

Drive Drive

Motor Output

Encoder Input

A92

Application Examples20. Labelling Machine

Application Type: FollowingMotion: Linear

Application Description: Bottles on a conveyorrun through a labelling mechanism that applies alabel to the bottle. The spacing of the bottles onthe conveyor is not regulated and the conveyorcan slow down, speed up, or stop at any time.

Machine Requirements:• Accurately apply labels to bottles in motion• Allow for variable conveyor speed• Allow for inconsistent distance between bottles• Pull label web through dispenser• Smooth, consistent labelling at all speedsMotion Control Requirements:• Synchronization to conveyor axis• Electronic gearbox function• Registration control• High torque to overcome high friction• High resolution• Open-loop stepper if possible

Application Solution:A motion controller that can accept input from anencoder mounted to the conveyor and referenceall of the speeds and distances of the label roll tothe encoder is required for this application. Aservo system is also required to provide thetorque and speed to overcome the friction of thedispensing head and the inertia of the large roll oflabels. A photosensor connected to aprogrammable input on the controller monitors thebottles’ positions on the conveyor. The controllercommands the label motor to accelerate to linespeed by the time the first edge of the labelcontacts the bottle. The label motor moves at linespeed until the complete label is applied, and thendecelerates to a stop and waits for the next bottle.

Product Solutions:

Controller Motor

APEX6152* APEX604

* The ZXF single-axis servo controller has alsobeen used in these types of applications.

Velo

city

Registration Input

Primary Axis

Secondary Axis

Time

Start Photocell

Encoder

Servo

Drive/Controller

A93

AE

ngin

eeri

ng R

efe

rence

Application Examples21. Window Blind Gluing

Application Type: FollowingMotion: Linear

Application Description: A window blindmanufacturer uses an adhesive to form a seamalong the edge of the material. It is critical that theglue be applied evenly to avoid flaws; however, thespeed that the material passes beneath thedispensing head is not constant. The glue needsto be dispensed at a rate proportional to thevarying speed of the material.

Machine Requirements:• Allow for varying material speed• Dispense glue evenly• Allow for multiple blind lengthsMotion Control Requirements:• Synchronization to material speed• Velocity following capabilities• Sequence storageApplication Solution:A step and direction indexer/follower and amicrostepping motor/drive are used to power adisplacement pump. The indexer/follower isprogrammed to run the motor/drive at a velocityproportional to the primary velocity of the material,based on input from a rotary incremental encoder.This assures a constant amount of glue along thelength of the material.

When the start button is depressed, the glue willbegin dispensing and can be discontinued with thestop button. If a new speed ratio is desired, FORcan be changed with either the front panelpushbutton, thumbwheels, or with the RS-232Cserial link.

Program

Two following commands are used.

FOR Sets the ratio between the secondarymotor resolution and the primaryencoder resolution

FOL Sets the ratio of the speed betweenthe primary and secondary motor

One input will be configured to start motion, asecond input will be used to stop motion. Themotor has 10000 steps/revolution. The encoderthat is placed on the motor pulling the materialhas 4000 pulses/revolution. It is desired to havethe motor dispensing the glue turning twice asfast as the encoder sensing the material.

FOR2.5 Set the motor to encoder ratio

FOL2ØØ The following speed ratio is 200% ortwice as fast

A1Ø Set acceleration to 10 rps2

AD1Ø Set deceleration to 10 rps2

MC The controller is placed in Continuousmode

Product Solutions:

Drive/Controller Motor

SXF Drive/Controller* S57-102

* The Model 500 single-axis controller and theS Drive have also been used in these types ofapplications.

Encoder

Motor

Drive/Controller

A94

Application Examples22. Moving Positioning System

Application Type: FollowingMotion: Linear

Application Description: In a packagingapplication, a single conveyor of boxes ridesbetween 2 conveyors of product. The productmust be accurately placed in the boxes fromalternate product conveyors without stopping thecenter conveyor of boxes. The line speed of theboxes may vary. When the product is ready, thecontroller must decide which box the product canbe placed into and then move the product intoalignment with the moving box. The product mustbe moving along side of the box in time for theproduct to be pushed into the box.

Machine Requirements:• Reliable product packaging on the fly• Standalone operation• Multiple product infeeds• Continuous operation without stopping the box

conveyorMotion Control Requirements:• Programmable I/O• Sequence storage• Complex following capabilities• Moving positioning system functionality• Multitasking

Application Solution:A standalone multiple-axis controller provides thecontrol for this application. The controller canperform motion profiling based on an externalencoder that is mounted on the center conveyor ofboxes. The two product conveyors are driven byservo motors for high speeds and accelerations.The controller looks for a product ready signalfrom a sensor mounted on the product infeedconveyor and then makes a move based on thestatus of the boxes on the box conveyor and thestatus of the product on the other productconveyor. The controller is multitasking the controlof the two product conveyors and the externalencoder input, as well as a sensor input to monitorthe status of the boxes. Thus the controller caninstantaneously decide into which box the productshould be placed and where that box is located.The controller then accelerates the product intoalignment with the appropriate box in time for theproduct to be completely placed in the box, andcontinues to monitor the other rest of the productand box positions.

Product Solutions:

Controller Drive Motor Encoder

Model 500 L Drive L20 -E Encoder

DriveController Drive Drive Drive

Product Synchronization

Product Infeed

ProductBox Conveyor

Product Synchronization

Product Infeed

Product

A95

AE

ngin

eeri

ng R

efe

rence

Application Examples23. Plastic Injection Molding

Application Type: Injection MoldingMotion: Linear

Application Description: A manufacturer ofinjection molding machines wants a system thatwill close a molding chamber, apply pressure tothe molding chamber for 5 seconds and then openthe mold. This action needs to be synchronizedwith other machine events. When the moldingchamber is open the motor must be ‘parked’ at adesignated position to allow clearance to removethe molded part. The manufacturer would like anelectronic solution (this is the only hydraulic axis onthe current machine).

Machine Requirements:• Electronic solution• Computer-controlled solution• 4000N (900lbs.) forceMotion Control Requirements:• Position and torque control• Serial link to computer and other drives• Ability to change pressure and dwell

Application Solution: A BLHX75BP brushlessservo drive with an ML345OB-25 motor and anETS8O-BO4LA Electro-Thrust Electric Cylinderwere used. The motor drives the rod inside thecylinder and extends/retracts the top moldingchamber. During this portion of the machine cycle,the servo drive must control the position of themotor. When the top molding chamber closes onthe bottom molding chamber, a pressure must beapplied. While pressure is being applied to themold the position of the motor is not important.However, the motor must control the pressure onthe molding chamber by applying a torque fromthe motor. A regular positioning servo can onlyapply torque by generating a position error—tryingto control torque through position is not veryaccurate and can create instabilities. The BLHXservo was chosen because it can switch betweenposition control and torque control on-the-flywithout instability or saturation and then, while intorque control mode, directly controls motortorque.

Product Solutions:

Controller/Drive Motor Actuator

BLHX75BN ML3450B-10 -ET580-BO4LA

Top Mold Chamber

Bottom Mold Chamber

Drive

"HOME" Position

"PARK" Position

"CLOSE" Position

ML3450B-25 Motor

Electric Cylinder

A96

Application Examples24. Rotating Tube Cutter

Application Type: Flying CutoffMotion: Linear

Application Description: Metal tubing feeds offof a spool and needs to be cut into predeterminedlengths. A rotating blade mechanism is used to cutthe tube, and the blade mechanism must spinaround the tube many times in order to completethe cut. The throughput of this machine must bemaximized, so the tubing cannot be stopped whilethis cut is being made. Therefore, to make a cleancut on the tube, the blade must move along withthe tube while the cut is being performed.

Machine Requirements:• Standalone operation• Move cutting mechanism with the tubing to

make the cut without stopping• Simple user interface to set different tube

lengths• High accuracy on cutMotion Control Requirements:• Programmable I/O• Program storage• Position following• High acceleration and speed

Application Solution:A single-axis servo controller/drive was chosen tosolve this application. An external encodermonitors the tube output and sends thisinformation back to the servo system. The servosystem tracks the length of the tube that is beingfed past the cutting blade. Once the appropriateamount of material has been fed past the blade,the servo accelerates the cutting device up to thespeed of the tube, sends an output to start thecutter, and then follows the tube speed exactly.

Product Solutions:

Drive/Controller Motor

APEX6152 APEX610

Controller/Drive

Coil of Tube

Cutting Mechanism

Leadscrew

Table

RP240

Encoder

Motor