The Smith-Kettlewell Technical File

A Quarterly Publication of
The Smith-Kettlewell Eye Research Institute’s
Rehabilitation Engineering Research Center

William Gerrey, Editor

Issue: sktf-Summer-1993

Original support provided by:
The Smith-Kettlewell Eye Research Institute
and the National Institute on Disability and Rehabilitation Research

Note: This archive is provided as a historical resource. Details regarding products, suppliers, and other contact information are original and may be outdated.

Questions about this archive can be sent to








Stepper motors are brushless devices whose shafts can

be rotated in discrete "steps." Common applications for

these motors are: driving the head carriage in many disk

drives, as well as positioners in printers and

numerically controlled milling machines.

"Variable-reluctance" and "hybrid" types are discussed


Surplus stores have bins full of stepper motors,

some affixed with gears and others with lead screws.

You'll know a permanent-magnet stepper motor by the way

the shaft feels as you turn it; the shaft wants to move

in jumps as the permanent magnets seek pole pieces of the

stators. (There is room for some confusion here, since a

simple DC motor with a permanent-magnet field can also

feel like this, but steppers will usually have more leads

than a simple DC motor.)

The main design parameters worth considering are:

the "resolution" (the number of degrees per step) and the

static torque (will it be powerful enough to stay where

you put it?). Many steppers come with gear boxes, making

the attainable resolution quite fine. When toying with

surplus ones, you can guess the resolution by noting how

many times the rotor magnets "home in" on stator poles.

In general, stepper motors have more than one

stator. (There is no electrical connection to the

rotor.) Each stator is a ring of iron surrounding a coil

of wire--the "stator winding." "Pole pairs" are created

by attaching L-shaped pieces either side of the ring.

This creates interleaving fingers on the inside of the

circle; half of these fingers (of one magnetic polarity)

point in one direction, and the other set, positioned in

between the first set of fingers, points the other way

(these being of the opposite magnetic polarity). Thus,

when energized, the inner circumference of each stator

has alternating magnetic poles--a north followed by a

south, then north, and so on. Having multiple stators,

we will designate them by letters A, B, C, and so on.

Unlike motors with "brushes," it is the business

of driver electronics to be the "commutator." The

simplest driving system is to energize one stator (stator

A), then energize stator B, then C (if there is a C),

etc. Depending on the type, this is not always done,

because energizing two stators simultaneously can

increase the available torque.

The Variable-Reluctance Type

Typically, these motors have three stators. The rotor

has "teeth" of soft iron with the same "pitch" (called

the "pole pitch") as pole pairs of the stator. When a

stator is energized, the rotor moves to the nearest

position that minimizes the air gap--the position that

establishes the best magnetic circuit.

Possibly to isolate the magnetic circuits, the

rotors may have multiple stacks of rotor teeth, one for

each stator. Such a motor was described in a textbook as

a 3-phase variable-reluctance stepper made by Warner

Electric. (Published by the Institute of Electrical

Engineering, London and New York, the book is "Stepping

Motors" by P.P. Acarnley, c1982.)

We dismantled an old 8-inch disk drive and

retrieved the motor used to drive the head; this looks

identical to the one in the book by Acarnley. With the

brand name being Applied Motion Products, this is a

3-stator 24-volt device. One lead of all three stators

goes to a common (black) lead; the free ends of the

stators come out as brown, red and orange leads (colors

representing 1, 2, and 3, respectively). The winding

resistance is listed on the motor as being 19 ohms (per


Two driving schemes are discussed in Acarnley:

The simplest is the "full-step" system; first A is

energized, then A is turned off and B is energized, then

C (by itself) is energized, then A again. With our

cannibalized motor, this worked very well, and it was

observed that 24 steps complete a revolution. However,

as described in Acarnley's book, there is a "half-step"

driving scheme which goes as follows:

First A is energized by itself; this causes poles

of the rotor to "home in" on poles of A. Then B is

added--causing the rotor to make a "half step" forward

(in the direction of B). Then A is turned off; this

causes the rotor to home in on B. Then C is added,

causing the rotor to advance a half step toward C. When

B is turned off, the rotor advances a half step to home

in on C. When A is next energized, the rotor advances

(in the direction it has been going) a half step away

from C--teeth in the rotor are about to align with poles

of A, but shifted by one period of the pole pitch. When

C is turned off, poles in the rotor home in on A, except

that the shaft has made one-eighth of a revolution; 48

steps will be required for each complete revolution. Of

course, any such sequence can be run in reverse to change

the direction of rotation.

For some reason, our liberated disk-head motor

does not make the half steps properly. This may be due

to alignment of stacks in the rotor, or this motor may

just not be made for it.

The driving circuit for a "whole-step" process can

be very simple: A "one-of-three" up-down counter could

drive power transistors in sequence--the counter

advancing in accordance with a clock signal.

The Hybrid Type

So-called "hybrid stepper motors" have a "single

stack" of permanent magnets on their rotors. These often

have two sets of stators with "N" number of pole pairs.

Their rotors have a corresponding number of radially

mounted permanent magnets with alternating polarity--a

south pole followed by a space, then a north pole

followed by a space, then another south pole, and so on.

The amount of rotation for each step is one-fourth the

"pole pitch"; the rotor can be moved to any of four

positions for any stator pole pair.

The two sets of stators are displaced from one

another; the poles of one set are aligned with spaces

between poles of the other. For analysis, let us call

these stators "A" and "B."

The traditional driving scheme is a four-step

switching process. A graph of the drive signals

indicates that two flip-flops generate square waves which

are out of phase by 90 degrees; these flip-flops are

driving the two stators. We can go through the process

"step by step" as follows:

When a stator is energized, its pole pieces take

on alternating magnetic polarity--north followed by a

space, south followed by a space, another north pole,

etc. Suppose we energize both stators with the same

polarity; there will be two north poles adjacent to one

another, followed by two adjacent south poles, and so on.

(By adjacent, I mean "across the aisle," from one stator

to the other.) Let us consider the current north pole of

stator A, the starting place of our analysis.

Magnets of the rotor will align so that each

permanent-magnet south pole is attracted to adjacent

stator north poles. Thus, the rotor "starts" by being

displaced from the pole pieces of either stator.

If we change the polarity of the drive to stator A

(and proceed around the stator from where we started), we

have a south pole, followed by two adjacent north poles,

followed by adjacent south poles; i.e., coincident poles

of the same polarity have been shifted in the direction

of stator B. Now, by changing the polarity of stator B

and noting polarity from our starting position we have:

two adjacent south poles, followed by adjacent north

poles, followed by adjacent south poles, etc.

Changing A again, we polarize stator A as we

started with. We now have north, followed by adjacent

south poles, then two norths, etc. Finally, we change B

back to the polarity with which we began; establishing

two norths, two souths, which is the order we started


Each time we change one polarity, we shift the

location of "like poles" of the two stators, and magnets

of the rotor are attracted to these. Of course, this

sequence can be reversed, giving us the ability to run

the motor in either direction. Note that skipping steps

can get you into trouble; for example, if you change

polarity of both stators, the rotor will be in a quandary

as to which way to go and the torque is theoretically


If each stator has twelve pole pairs, the rotor

will move four steps per pole pair, making 48 steps per

revolution (rotating 7.5 degrees per step).

Two winding configurations are available: The

stator windings in "bipolar" stepper motors have two

leads per stator. Bipolar stepper motors with two

stators have four leads; the winding of one stator is fed

by one pair of leads, while the winding of the other is

fed by the other pair. The "unipolar" version has

center-tapped windings; the connections come out as three

leads per stator--a motor with two stators has six leads

in all.

Bipolar motors require more complicated switching

than unipolar ones, since polarity of the voltages

applied to the windings must be reversed; this requires

eight driver transistors. In driving unipolar motors,

the polarity of the electromagnets is changed by using

one-half of the winding or the other; twice as much

magnet wire is required to get the same number of

ampere-turns in the motor, but only four driver

transistors are needed.

The literature discusses timing issues of driver

circuits; these discussions have been put to rest by the

emergence of driver chips which do the work for us. The

simplest driver circuit, used for the unipolar style, is

as follows: The center taps of both stator windings go

to the positive supply; the ends of the windings go to

open-collector outputs of flip-flops, a flip-flop for

each stator. The bipolar circuit requires polarity

reversals: Two flip-flops, one for each stator, drive

complementary transistor pairs whose outputs can either

pull to the positive supply or to ground. The stators,

which only have two leads each, can thus be powered in

either direction. (One of the timing issues is that

without safeguards, both transistors in a complementary

pair might turn on at the same time and burn themselves


Comparing performance of the two winding styles,

choosing motors of identical physical size, unipolar

motors have about 30% less torque at low-to-moderate

speeds. This is due to the unavoidable requirement of

reducing wire size in order to fit as many turns each

side of the center tap as can fit on the whole winding of

a bipolar unit. Torque of the two types is equivalent at

high speeds.

Inductance of the windings causes a decrease in

torque as speed increases. You can compensate for this

by increasing the power supply voltage by a factor of

four and placing a resistor in series with each stator

winding. They recommend a resistor value of three times

the DC resistance of the winding. (This is called the

"L/4R circuit," counting the winding resistance.)

Naturally, this will increase power drain, and the power

rating of the resistors will have to be 2 times I squared

times R, using carbon power resistors.

In the unipolar circuit, a resistor is placed in

series with the center tap; since only half the winding

is powered at any given time, resistors on each end are

not required. The winding resistances listed in

specifications for unipolar motors are for half the

winding--measured either side of center tap.

We cannibalized a Seagate hard drive and found two

stepper motors. Oddly enough, the rotor that spins the

two hard disks is a very crude stepper capable of four

steps per turn. It has two 3-ohm stators; three leads

connect to these, making them look like a center-tapped


Of much more interest is the stepper that drives

the head carriage; this is capable of a resolution too

high to count by hand. It has two bipolar stators of

27.5 ohms each.

Lying about the place, we found a head motor of

exactly the same physical dimensions--also a

Seagate--which has five stators. This appears to have a

similar resolution, although I don't relish the task of

driving it. The winding resistance, all identical, is 19

ohms for each stator.

Now that IC's to drive stepper motors can be

gotten, they become an attractive electromagnetic device.

The signals required for the driver are pulses, which

can come from a variable-frequency clock, and a

direction-determining signal. If you couple this with a

counter, you can tell the system how far to go and when

to stop the motor.

Addendum Based on New


I just came back from our major electronics show of

the West Coast, WESCON. I saw some stepper motors of the

types described here, all right, but there are new kinds.

It seems that they are able to get about 20 times

the position resolution using lasers to cut pole pieces

of the stators. A company called Semix had a unit

capable of 800 steps per turn in the whole-step

mode--1600 steps per revolution in the half-step mode.

Another exhibitor, Motion Science, Inc., has an

eighteen-inch long table full of permanent-magnet poles,

with a shuttle containing stators riding atop an air

bearing (air escaping from holes underneath the shuttle

to make it nearly frictionless). He was claiming a

movement of only 2 microns per step.

Well, it looks like we can look forward to some

very fine control with these new kinds. They were all

bipolar types using permanent magnets.


id="motorsurvey" name="motorsurvey">

Once known as Airpax and now a division of Philips,

the products listed here are but a few representative

samples--all of the "hybrid" type (using permanent-magnet

rotors). The Philips motors all come in 5-volt and

12-volt models, as well as bipolar and unipolar. Often,

there are related models with different resolutions, 24

versus 48 steps per turn.

Series K82600 and K82800--

These models have gear boxes with 1/4th-inch output

shafts. Their motor bodies are 2.187 inches in diameter.

The over-all length--up to the mounting surface from

which the shaft emerges--is 2.48 inches. The shaft

protrudes one inch, but for some of this distance (a bit

less than half an inch), there is a bushing around the

shaft that serves as a bearing. (There is a "flat" on

the shaft for a setscrew to rest against.) On opposite

corners of the gear boxes are 8-32 mounting studs;

mounting holes on the remaining corners presumably accept

8-32 bolts.

All in this series are unipolar; they have six

leads and the stator windings are center-tapped. The

K82600 units have 24 steps per revolution (one step moves

the rotor 15 degrees); the K82800 ones have 48 (one step

moves the rotor 7.5 degrees). Speaking for the basic

motors (before their gear boxes), the holding torque of

the K82800 units is 13 inch-ounces, while the K82600 ones

have a holding torque of 10 inch-ounces.

They carry a hyphenated suffix; -P1 means 5 volts,

while -P2 means 12 volts. The winding resistance of 5V

units is approximately 7.5 ohms, and that of 12V units is

about 45 ohms. The inductances (for 5V and 12V units,

respectively) are 7mH and 32mH for 82600 ones; for the

82800 series, inductances are 8.5mH and 44mH,


Gear boxes are specified by the last two digits;

for example, K82616 and K82816 each have gear reductions

of 5 to 1. Available gear ratios are as follows:

The K82601 and K82801 units have no gear boxes;

the ratio is listed as 1 to 1. Remembering that a 6 or 8

specifies 24 or 48 steps: 11 means 2 to 1, 16 is 5 to 1,

21 is 10 to 1, 24 is 15 to 1, 27 is 20 to 1, 30 is 25 to

1, 31 is 30 to 1, 36 is 50 to 1, 37 is 60 to 1, 39 is 75

to 1, and 45 means 150 to 1.

Now, remember that the gear reduction affects both

the torque and the angle of rotation per step. For

example, a 10 to 1 gear box--such as that of the

K82821--creates a holding torque of 130 inch-ounces, and

the rotation per step is 0.75 degrees. The K82845-P2 is

a 12V unit whose steps rotate the shaft 0.05dg.

The 35M0--

B Series--These also have gear boxes. They are 1.86

inches deep--including the gear box. The gear box is egg

shaped (2.44 inches in its longest direction). A

1/8-inch diameter shaft emerges nearest the narrow end.

Ears either side of the gear box are provided for


They are available in unipolar and bipolar

configurations--either available as 5- or 12-volt units.

Three step angles are available--18dg (20 steps per

revolution), 15dg (24 steps), and 7.5dg (48 steps).

(Remember, torque and resolution specs are for the motor

before the gear box.)

For the unipolar ones, the 18dg, 15dg, and 7.5dg

units have holding torques of 1.9, 2.4, and 2.6

ounce-inches, respectively. As you would expect, the

holding torques for bipolar ones are higher--2.3, 2.8,

and 3.5 ounce-inches, respectively.

The 35M020B has 20 steps per revolution of the

rotor, the 35M024B has 24 steps per revolution, and the

35M048B has 48 steps. A 1 or a 2 following the B means

5V or 12V, respectively. Following that designation, a U

or B specifies unipolar or bipolar. Thus, a 35020B1B

unit is a 5-volt bipolar 20-step motor.

Gear ratios are another suffix--a hyphen X,

followed by a number: 24 is 15 to 1, 27 is 20 to 1, 31

is 30 to 1, 37 is 60 to 1, 39 is 75 to 1, 45 is 150 to 1,

52 is 300 to 1, and 64 is 1350 to 1. Thus, a

35M020B1B-x31 is a 5V bipolar 20-step motor carrying a 30

to 1 gear box. (Note that this would give you 600 steps

per revolution. Do you sense a clock in the works?)

For both winding styles, 5V units have 12.5-ohm

resistances, while the 12V units are 72 ohms.

Inductances vary through the list of combinations.

Rounding off, unipolar 5V motors in this series present 6

or 7 millihenries, while 12V unipolar ones range from 30

to 36 millihenries. Bipolar ones have about twice the

inductance--14mH for 5V motors and about 75mH for 12V


The L82400 Basic Stepper Motor--

This has a 1.6-inch diameter mounting flange at the

output end. It is about 0.92 inches deep and the motor

body is 1.654 inches in diameter. The shaft diameter is

1/8 inch. The holding torque of the unipolar version is

10.4 ounce-inches; for the bipolar version, this torque

is 12.4. It has 48 steps per revolution.

The L82401 is unipolar; the L82402 is bipolar.

The suffix -P1 means 5 volts; -P2 means 12 volts.

The winding resistances for both bipolar and

unipolar ones are 9.1 ohms for the 5V one and 52.4 ohms

for the 12V one. Inductances for unipolar ones are 7.5mH

and 46.6mH for 5- and 12-volt units, respectively. For

bipolar, inductances are 14.3mH and 77.9mH.

The Basic L82701 Stepper--

Slightly more than an inch deep, this model has two

mounting holes on a roughly diamond-shaped flange. The

diameter of the body is 2.28 inches. Its shaft diameter

is 1/4 inch.

This motor has 48 steps per revolution. It only

comes in a uni-polar version. The holding torque is 15


For 5V and 12V versions (specified by -P1 and -P2

suffixes), the winding resistances are 6.3 ohms and 36

ohms, respectively. The corresponding inductances are

7mH and 38mH.

Linear Actuators

These are the cutest little darlings! They look like

a small can with mounting ears at one end. Running

through the center is a leadscrew--a threaded shaft with

a rather coarse pitch to the thread. The rotor has no

shaft; it is tapped to match the thread of the leadscrew.

The leadscrew is secured to something which is to

be moved to and fro; i.e., the idea is to keep it from

turning. When the rotor inside the device turns, it

operates the leadscrew so as to make it move in or


The 9200 Series of Linear Actuators--

The diameter of these devices is 1.06 inches, and the

body is 1.05 inches deep. The body has a bushing which

protrudes beyond the mounting face by about 0.6 inches.

The output end of the leadscrew has a 4-40

thread--terminated by a shoulder--so that it can be

screwed to the piece being moved.

The part number is preceded either by K or L; the

K has a maximum travel of 1/2 inch, while the L model

(provided with a longer leadscrew) has a maximum travel

of 1-7/8 inch.

Available units are unipolar; this is shown by the

last digit in the number, which is always 1. Then, the

third digit in the number determines the distance

traveled per step: 4 means 0.004 inches per step, 2

means 0.002 inches, and 1 means 0.001 inches per step.

Thus, an L9211 is a unipolar linear actuator whose range

is 1-7/8 inches (specified by the L), and whose

resolution is one-thousandth of an inch per step. A

K9241 is a short-range (1/2 inch) unit whose resolution

is four-thousandths per step.

Finally, 5V and 12V actuators are available,

specified by the suffix -P1 and -P2, respectively.

The coil resistances are 15 ohms for the 5V

version and 84 ohms for the 12V one. Winding inductances

are 5mH and 29mH, respectively.


The catalogue lists two addresses as follows:

  • Philips Technologies, Airpax Mechatronics

    Group (International Division), 604 West Johnson Ave., PO

    Box 590, Cheshire, CT 06410; Phone: (203) 271-6000.

  • Airpax, 2001 Gateway Place, San

    Jose, CA 95110; Phone: (408) 453-7373.


id="chipsurvey" name="chipsurvey">


Three Motorola chips, as well as a couple of

experimenters' kits, are presented here. It seems that

currently available devices are designed to drive 2-phase

bipolar hybrid motors, examples of which are listed in

the previous article.

The Motorola MC3479P

For about $4, this is available from at least two

sources. The Jameco Cat. No. is 25216. It can also be

gotten from JDR Micro Devices under the catalog number of

its own part designation--MC3479P.

This is said to be appropriate for driving 6- or

12-volt motors. Actually, by associating a resistor with

a bias pin, the outputs act like current sources, so they

can drive 5-volt stators of units listed in the previous

article. With this particular chip, the maximum drive

current available is 350mA.

This is a 16-pin chip. Pins 4, 5, 12 and 13 are

connected to the substrate; a heat sink can be soldered

to these, or soldering these pins to a ground plane on

the board can aid in dissipating the heat. These pins

are the negative supply terminal on the chip and can be

considered as ground.

Provision is made so that the outputs can be diode

protected; the four outputs go to diodes internal to the

chip (the anodes of these diodes being at the output

pins). The cathodes of these four clamping diodes go to

a common pin (pin 1), and pin 1 is connected to the VCC

(pin 16) through a zener diode, a regular diode, or even

a jumper. If a 3-volt zener is used (such as a 1N5221),

its cathode goes to pin 1 and its anode goes to pin 16.

Outputs 2 and 3 go to one stator--the leads called

L2 and L1, respectively. Outputs 14 and 15 go to the

other stator--these leads are called L4 and L3,


This chip is capable of supplying two kinds of

drive signals. The conventional scheme is to always

supply both stators simultaneously. The Motorola

literature calls this the "whole-step" drive. The

advantage is that the "holding torque" is consistent for

each step. This is true because magnets of the rotor are

always attracted to like poles of both stators. Four

steps per pole pair are made (see the previous article).

The so-called "half-step" drive alternates

energizing both stators with powering one or the other

singly. This causes the rotor to move eight steps per

pole pair of a stator. The disadvantage is that, when

only one stator is energized, the holding torque is not

as high as when both are powered.

There is an open-collector output, pin 11, that

when the output drivers are in the "phase A" state, the

collector brings this pin low. Phase A is defined as


Connections L1 and L3, pins 3 and 15, are high.

L2 and L4, pins 2 and 14, are low. In other words, the

coil between pins 2 and 3 will see negative on pin 2 and

positive on pin 3. The coil between pins 14 and 15 will

see negative on pin 14 and positive on pin 15.

Motorola defines "clockwise" as first changing the

polarity of the L1-L2 stator, then reversing the current

through the L3-L4 stator. Thus, in the whole-step mode,

the first clock pulse causes pin 2 (L2) to go high and

pin 3 to go low. In the half-step mode, this clock pulse

would de-energize the L1-L2 stator, and a subsequent

pulse would cause its reversal.

There is a pin (pin 6) which has three functions:

First, it is a "bias pin." This pin determines

the base currents of the NPN transistors which sink

current at the negative ends of the coils; these

transistors are current sources. Normally, this pin is

taken to ground through a resistor, the value of which

determines the drive current.

Pin 6 is also called the "set pin"; if less than 5

microamps is drawn from it--if it is open-circuited--the

internal logic presets to the "Phase A" condition. When

pin 6 is open, pin 11 (the "phase-A output detection pin)

goes low, and all other inputs are ignored until current

is drawn from pin 6 again.

Finally, opening pin 6 causes the outputs to

assume a high-impedance condition (simultaneously setting

the chip for the "phase A" state).

Unlike the other inputs, pin 6 is not to be

operated at the 5V logic level. Unless it is permanently

connected to ground through a resistor, pin 6 should go

through that resistor to an open-collector output. This

could be an open-collector gate or just an NPN transistor

as follows:

Pin 6 goes through a resistor to the collector of

a 2N2222 whose emitter is grounded. The base of the 2222

is then operated through a 10K resistor. (A small

capacitor in parallel with this 10K resistor will make

the transistor turn off faster.)

The output sink current is determined by: IBS

(the current in microamps drawn from pin 6, the "bias/set

pin") equals 0.86 times the output current in milliamps.

Pin 6 wants to see a current sink of greater than 5uA.

In normal operation, the voltage on pin 6 will be VM (the

main supply) minus 0.7V. Thus, the appropriate RB (the

resistor from pin 6 to ground) will be the quantity VM

minus 0.7V, divided by IBS.

Suppose I wish to drive a 2-phase bipolar stepper

motor with drive currents of 200mA, and suppose I supply

the circuit with a healthy 9V battery. This 200mA times

0.86 equals 172uA. Next, I want the value of a

resistor--from pin 6 to ground--which will draw 172

microamps: The quantity 9 minus 0.7 volts, or 8.3 volts,

divided by 1.72 times 10 to the minus fourth, equals 43K.

One suggested use of pin 6 is to reduce current

drain when the motor is not being operated. If the

stepper is positioning something which does not require

the full holding torque to keep the mechanical assembly

in place, the current drawn from pin 6 can be reduced

while the system is dormant.

Besides the "bias/set" pin, this chip has four

inputs. They are intended to be operated with 5-volt

logic; the maximum allowable voltage is 5.5V. They are

CMOS compatible, however, with the maximum hysteresis

being 0.4V.

The "clock input" (pin 7) advances the logic that

controls the output drivers, and it does so on positive

transitions. When the "set pin" is open-circuited,

pulses of the clock--as well as all other input

signals--have no effect.

The direction of the logic sequence is determined

by pin 10. When high, driver outputs operate in a

sequence defined as "clockwise"; when pin 10 is low, the

sequence is defined as "counterclockwise." (The actual

direction of rotation also depends on how the leads to

the stators are wired to the chip.)

Pin 9 selects the whole-step vs. half-step mode.

When low, the motor will move four steps per pole pair.

When pin 9 is high, the motor can make twice as many

steps because there are times when only one stator is

energized. (In the half-step mode, the holding torque

will not be uniform, since the rotor, during these

"half-steps," will be attracted to one stator only.)

Pin 8 is called the "output impedance control."

Although it always should be committed to 5V or ground,

the setting of pin 8 is only relevant when driving motors

in the half-step mode. Its operation gives you a choice:

In the half-step mode, one or the other of the

stators will not be energized one-fourth of the time.

When pin 8 is low, the two outputs of a non-energized

coil will be in a high-impedance state. When pin 8 is at

logic 1, the PNP transistors will be turned on; the

outputs will have a low impedance with respect to VM.

(The importance of this selection has eluded your editor

thus far.)

The SAA1042 and SAA1042A

These are different from the above 3479 in two

respects. First, they have mounting ears by which they

can be attached to heat sinks. Second, they are not

quite as versatile; they have fewer features. However,

since they are capable of supplying 500mA, they are worth

listing here.

The SAA1042 is for 6V and 12V motors; the SAA1042A

can run 24V motors. Running 5V motors is no problem,

since they, too, have current-sourcing outputs. Their

pin assignments are the same; the 1042A just has higher


Their pin arrangements are slightly different from

the 3479: they have a separate supply pin for the logic.

Coil-output drivers are on pins 1 and 3, 14 and 16.

This means that the external jumper or diode for clamping

the output voltages goes between pins 2 and 15, a small


Pin 6, the "bias/set" pin is the same, except that

the relationship between bias current and drive current

is not such a nice linear one as it is with the 3479.

Some sample values for RB off pin 6 were gotten from a

graph in the literature, and the corresponding drive

currents are listed as follows:

  • For 6V:
    • 60K--50mA
    • 50K--70mA
    • 40K--100mA
    • 24K--200mA
    • 11K--500mA
  • For 12V:
    • 130K--50mA
    • 100K--80mA
    • 90K--100mA
    • 56K--200mA
    • 38K--300mA
    • 29K--400MA
    • 22K--500mA
  • For 24V (using the SAA1042A):
    • 160K--130mA
    • 110K--200mA
    • 75K--300mA
    • 60K--400mA
    • 46K--500mA

For over-voltage clamping of the outputs,

they recommend a 3.9V zener between pins 2 and 15

(cathode at pin 2). Appropriate part numbers would be:

1N5520, 1N4622, 1N4730, or 1N748.

As with the MC3479, pin 8 determines half-step

versus whole-step mode; bringing it high puts the chip in

the "half-step" mode--eight steps per pole pair. Pin 10

is the clockwise versus counterclockwise pin; bringing it

high means that the outputs work in the "clockwise"

direction in response to the clock. ("Clockwise" means

that the L1-L2, pins 1 and 3, change first; they reverse

polarity in the "whole-step" mode, or de-energize the

coil in the "half-step" mode.)

If pin 6 is open-circuited, the chips (both 1042

and 1042A) are set to "state A"; pin 3 will be high and

pin 1 will be low, while pin 16 will be high and pin 14

will be low. (Note that pin 3 is L1, pin 1 is L2; pin 16

is L3 and pin 14 is L4.)

An advantage in these chips is that the logic need

not be restricted to 5V. The logic VCC pin, pin 11, can

be up to 20 volts. (This restriction of 20V applies to

the "logic supply" of both chips.) The SAA1042A allows a

"VM" (on pin 15) of 30V, while the SAA1042 allows an

absolute maximum VM of 20V.


As with any driver system, power dissipation is an

issue. The SAA chips list a maximum power dissipation of

2 watts; it is the editor's impression that this is

without a heat sink.

The literature for the MC3479P does not give a

maximum package dissipation directly. They say that

dissipation should be such as not to exceed junction

temperatures of 150 degrees C. They then say that

"thermal conductivity," junction to ambient, is 45

degrees per watt without a heat sink (hmmmmmmm).

Calculating the power dissipated by the chip is a

bit of a trick, since output transistors needn't be

saturated. However, if you know the motor coil

resistances and the currents you intend to run them with,

you know the voltages across the windings; subtracting a

winding's voltage from the full supply and multiplying

the difference by the winding current, you will know how

much each driver output pair will dissipate while that

winding is energized. In the "whole-step" mode, both

coil drivers are always on simultaneously, so the power

dissipated by the whole package will be twice the winding

current times voltage dropped across the driver

transistors. We are only partly finished; power

dissipation must now be linked with junction temperature.

Suppose our power calculations lead us to a figure

of 2 watts of dissipation. Multiplying 2 watts by 45dg

per watt, we get 90dg C. They just add an arbitrary 25dg

C. to this--an assumed figure for the ambient

temperature--thus obtaining a figure of 115dg C. With

the 3479 in mind, this is less than the 150dg C. which

they allow without a heat sink.

  • MC3479P Specs:
    • Recommended Operating Voltage--7.2 to


    • Maximum Voltage--18V.
    • Clamp Voltage (at diode

      cathodes)--At VM, or up to VM plus 4.5V.

    • Drive Current--Up to 350mA.
    • Maximum Logic Input Voltage--5.5V.
    • Input Hysteresis--0.4V max.
    • Clock--Positive-edge triggered.



  • MC3479P (without heat-sink mounting):
    • 4, 5, 12, 13--Ground
    • 16--VM/VCC (for both logic and

      motor drive)

    • 1--Clamp
    • 2--Output L2
    • 3--Output L1
    • 14--Output L4
    • 15--Output L3
    • 11--Phase A Indicator (open

      collector brings pin 11 low when outputs are in "phase


    • 6--Bias and Set
    • 7--Clock (positive-edge triggered)
    • 9--NOT Whole/Half Step (when low,

      chip operates in "full-step" mode, moving the motor four

      steps per pole pair)

    • 10--Clockwise vs. Counterclockwise

      (also called "NOT Counterclockwise," logic high is said

      to make operation clockwise)

    • 8--Output Impedance Control (in

      half-step mode, when pin 8 is low, non-energized outputs

      will be in tristate; when pin 8 is high, non-energized

      outputs will be low-impedance with respect to pin 16)

  • SAA1042 and SAA1042A:
  • 4, 5, 12, 13--Nonexistent

    (coinciding with ears for heat sink)

  • 9--Ground (along with heat-sink


  • 11--Logic VCC
  • 15--VM (V with a subscript M) for


  • 1--L2 Output
  • 3--L1 Output
  • 16--L3 Output
  • 14--L4 Output
  • 2--Clamp (to be clamped at pin 15)
  • 6--Bias/Set
  • 7--Clock (positive-edge triggered)
  • 10--Clockwise/Counterclockwise

    (high for clockwise)

  • 8--Whole/Half step

Exemplary Circuits Using the MC3479Pand a Bipolar

Stepper Motor

These should get you started. We cannibalized a

Seagate hard drive, and we were able to run its head to

and fro across the disk just for fun.

The basic driver here makes it just a variable

speed motor with selectable direction. An embellishment

is then added to make the shaft oscillate (rotate to and

fro). This is just the sort of hookup you could use to

make a mechanical monkey beat a drum and tap his foot.

For those who remember the fabulous window

displays set up by F.A.O. Schwartz at Christmas time,

compare how much easier it is to build this monkey

without gears and a reciprocating mechanism. With that

in mind, why not design a mobile window display for your

business or something.

Just a clock and the driver chip are required to

run the motor, as long as the direction terminal, pin 10

of the 3479, is committed to VCC or ground.

The oscillating shaft requires a primitive scheme

for starting the mechanism at an initial position. That

way, the monkey won't poke himself in the eye with his

drumstick or put his foot through your work bench.

Basic Variable-Speed Motor--

A supply of 9V or 12V is used. As you might expect,

VCC must be bypassed heavily because of hash from the

motor--parallel combinations of 220uF and 0.1uF at either

ends of the rails will do (negatives of the electrolytics

at ground). This supply does not have to be well

regulated, however.

Pin 1 of a 555 is grounded; pins 4 and 8 go to

VCC. Located close to the chip is 0.1uF between pins 1

and 8. Also, pin 5, the voltage-control pin, is bypassed

to ground by 0.1uF.

Pins 2 and 6 are tied together and go through

0.01uF to ground. Pins 2 and 6 also go through 100K to

pin 7. Pin 7 goes through 47K, in series with a 500K

rheostat to VCC.

The 555's output, pin 3, goes to the clock input,

pin 7, of the MC3479P motor driver chip. This clock

input is positive-edge triggered.

On the MC3479P, pin 4 is grounded. Pin 16 goes to

VCC--heavily bypassed nearby. Pin 6 goes through 47K to

ground. Pin 8 goes to VCC (although in "full-step mode,"

pin 8 can be committed anywhere). Pin 9 is grounded;

this selects the "whole-step mode." Finally, committing

pin 10 to ground, or to VCC, selects direction of

rotation (which is also affected by hookup of the stator


Pin 1 goes to the anode of a 1N4001 diode; the

cathode goes to pin 16. One stator goes between pins 2

and 3, while the other stator goes between pins 14 and

15. Experimentation will determine which way the motor

runs when the chip runs "clockwise" (with pin 10

committed to VCC).

Circuit for Making the Shaft Oscillate--

The motor is made to run in one direction until a

switch is tripped; thereafter, the excursions are

determined by selecting outputs from a counter. A

flip-flop sets the initial direction by bringing pin 10

of the driver high; closure of the switch when the

mechanism arrives changes the direction and enables a

4040 12-bit counter. (The switch can be a simple spring

leaf, or an opto-isolator.)

A 4011 quad NAND gate is used to create a

NOT-R/NOT-S flip-flop. Pin 7 of the 4011 is grounded,

while pin 14 goes to VCC. Pin 3, the output of one gate,

goes to pin 5, an inpinput of the second. Pin 4, the

output of the second gate, goes to pin 2, an input of the

first. Pin 1 is NOT-Reset and pin 6 is NOT-Set.

One side of the switch is grounded. The other

side goes to pin 1 of the 4011. Pin 1 also goes through

100K to VCC. Pin 6 of the 4011 goes through 910K to VCC,

as well as going through 0.1uF to ground.

The clear terminal of the 4040 counter, pin 11,

goes to pin 4 of the 4011. Pin 3 of the 4011 goes to its

own pin 12, another gate input. Pin 11, the output of

this gate, goes to pin 10 of the 3479 driver chip.

The pin 13 input of this gate receives inverted

squarewave pulses from the desired 4040 output. Thus,

pin 13 of the 4011 goes to its own pin 10. Its pins 8

and 9 are tied together and are taken to an output of the

4040--pin 13 or pin 12 being likely choices.

Pin 3 of the 555 goes to pin 10, the clock input,

of the 4040 counter (positive-edge triggering for this

pin). Pin 8 of the 4040 is grounded; pin 16 goes to VCC.

Located close to the chip, pin 16 is bypassed to pin 8

by 0.1uF.

The amount of rotation back and forth is

determined by the choice of output on the 4040 12-bit

counter. Pin 12 causes the shaft of our disk-drive motor

to go approximately one full turn. The following are the

divider outputs of the 4040 in order: Pin 9 divides the

clock by two--the shaft would go two steps in either

direction. Next come: 7, 6, 5, 3, 2, 4, 13, 12, 14, 15

and 1.

The All Electronics SMKIT-2 Stepper Motor


A company called All Electronics, an outfit that sells

various oddments, has a stepper-motor kit. Called the

SMKIT-2, it comes with a motor, a driver (which must be

assembled on a PC board), and a power supply--all of this

selling for about $25. The driver allows you to run the

motor step-by-step or at variable speed.

We have not purchased this kit, so we have no

verbal description for its assembly. But, it sure sounds

like a screamin' deal.

Address List

  • All Electronics, PO Box 567, Van Nuys, CA

    91408; Phone: (818) 904-0524; (800) 826-5432.

  • Jameco Electronics, 1355 Shoreway

    Road, Belmont, CA 94002; Tel: (415) 592-8097

  • JDR Micro Devices: 2233 Samaritan

    Drive, San Jose, CA 95124; Phone: (800) 538-5000, or

    (408) 559-1200.