OPTIMISED USE OF HIGH SPEED STEPPING MOTOR

BY IMPLEMENTATION OF ADEQUATE SPEED PROFILE

Prof. Pierre André, Engineering School "ENSMM"

Besançon,
( France)
NorbertVeignat, Product Manager,Portescap (La Chaux-de-Fonds, Switzerland)

I INTRODUCTION Then we present an easy way of calculating the

Steppe r motors are generally driven in open appropriate speed profile.
loop mode. The controller defines timing of the We will also point out the precautions to be
pulses sent to the driver. The motor must be taken if the motor is driven microstep
in mode.
strong enough to move the load without loss of
synchronism. II TRAPEZOIDAL SPEED PROFILE
The advantage of the stepper, related to its In many applications, trapezoidal speed
design, is to have enough stable positions per profiles are used. This is mainly because
revolution for positioning the load without the motors are working at low speed and because
need for an encoder. Driven in open loop, it is a such profiles are easy to implement using
positioning device by itself. existing dedicated integrated circuits. It is,
For fast motion its design presents some however, important to understand the limitations
disadvantages : at high speed, dynamic motor of such a speed profile.
torque is affected by the high number of The dynamic torque curve of a stepper
commutations per revolution, the iron structure, usually has a shape as shown in Fig. II-1.
and the electrical time constant.
Torque
All these reasons lead the designer to
Maximum torque
select stepper motors for use as a positioner
To
and move the load along a trapezoidal speed Usable torque
(~70% of the peak torque)
profile.
T1
The Disc Magnet stepper motor has the
ωo ω1 ω max Speed
advantage of lower iron losses than regular
hybrid stepper motors. It can therefore also be
Fig. II-1
used for fast incremental motion. To take
advantage of its dynamic possibilities one Say we want to reach the speed ω1 as

should first know the dynamic torque output of quickly as possible. When using a linear profile

the motor-and-driver and then adapt the velocity the motor torque is considered equal to T1 over

profile accordingly. the entire speed range, and we do not take

In this paper, a simple but reliable model is advantage of the motor's capability of delivering

used for calculating the dynamic torque output. much more torque at low speed. In

this example it would be better to use a

constant speed up to

(C) Portescap. All right reserved.

ω0 and then follow an exponential speed profile also show that up to the rated flux the

from ω0 to ω1. resistance does not change with induction.

III DYNAMIC TORQUE CALCULATION Figure III-1 illustrates the model we decided
As stated previously, to optimise the speed to use.
profile it is imperative to know the dynamic
R', L
torque of the motor-and-driver unit. There are two
ways of doing so. The first one is to measure it.
This requires the motor, the driver and some
R, L
special equipment. A second method consists
of calculating the torque.
Many papers have been publ ished on how
to calculate the dynamic torque of stepper
Fig. III-1 :
motors, but they often neglect the motor's iron Modelisation of iron losses
losses. At low speed this is correct, but at high
step rates iron losses are no longer negligible. III-1-2 Principle of Measurement
In this Chapter we present a model of iron Two kinds of measurements were made:
losses adapted to the Disc Magnet Motor, and - measure without a rotor
then apply it to the dynamic torque calculation. - measure with the rotor in place
Please note that this model has been It is important to understand the way these
simplified after many practical measurements. measurements were made, because after
These simplifications would not hold with other comparing both methods and their results we
motor technologies. were able to simplify our theoretical model.

III-1-2-1 Measurements without a rotor

III-1 MODELISATION OF IRON LOSSES The rotor of disc magnet motors has no

III-1-1 Principle iron and, with the permeability of the magnet

An alternating magnetic flux crossing an being close to that of air, it is possible to have

iron structure generates iron losses. There are the rotor locked or to take it out altogether. The

two kinds of iron losses: those due to procedure uses a voltage generator. By

hysteresis and losses due to eddy currents. measuring the voltage, the current and the

To simulate these losses we suppose that phase shift we can figure out the R' value at

the stator magnetic circuit is perfect and that we different frequencies (see fig. III-2).

have added a second winding with the same

number of turns as the motor winding. We
imagine this winding connected to a resistance
whose value will depend on the frequency of the
magnetic flux, i.e. the step rate. Measurements

(C) Portescap. All right reserved.

 simulation has shown that with formulas 1 and 2
R'
a variation of inductance will generate big
changes for R'.
A
ψ
U=U0sinωt V A.n.Ι
R, L
U Ι m(Z).Ι
dφ Φ
n
dt

=n.ω.Φ (L/R').(dΦ/dt)

ϕ ψ
Fig. III-2 : RΙ
Measurements without a rotor Re(Z). Ι

With : R winding resistance

Fig. III-3: Phasor diagram
L winding inductance III-1-2-2Measurements with the rotor
U voltage across the motor winding A DC motor is used to rotate the stepper
Ι current in the motor winding motor. The stepper windings are connected to a
φ magnetic flux high impedance oscilloscope (see fig.III-4).

ϕ phase shift between voltage and

R'

current
we have the following equations: magnet

U = RΙ+n(dφ / dt) E

0 = R'Ι'+n(dφ / dt)
Te
nΙ+nΙ' = φ / A

According to the phasor diagram fig. III-3 Fig. III-4 :

we can calculate R' using the following formulas: Measurements with the rotor

By using modulus :
By measuring the EMF versus speed we
2 find the losses due to R' generated by the
(LωΙ)
R' = L ⋅ ω / −1
2 2 magnetic flux from the magnet. With a perfect
(U cos ϕ − RΙ ) + (U sin ϕ)
magnetic circuit EMF would rise linearily with
By projecting on the real axis : speed. The value of R' is given by the equation:
2 2
2 2 R' = L ⋅ ω ⋅ E / (K ⋅ ω / N) − E
(L ω) ⋅ (1 ± 1 − 4 ⋅ ( ∆R / Lω) )
R' =
( 2 ⋅ ∆R) By measuring the torque required to drive
the stepper we determine the power loss in the
By using the power dissipated :
magnetic circuit due to flux crossing the winding

R' =
((Ucos ϕ −RΙ) 2 2
+ (Usin ϕ) ) and flux crossing only the edges of the teeth.

2 We can separate both losses (see fig. III-5).

UΙ cos ϕ− RΙ

The last formula is the one to use because

it does not take into account inductance. The

(C) Portescap. All right reserved.

 III-1-3-1Measurements without a rotor
R' Measure of the real part of the impedance (Fig.
III-6).
Measure of the imaginary part of the impedance
(Fig. III-7).
Calculation of R' from the results (fig. III-8).
Measures of Re(Z), motor PH632-508-002

25.00

20.00

Resistance in ohms
15.00
R (Z)
Polynomial (R (Z))

Fig. III-5 : 10.00

Magnetic flux due to the magnet 5.00

0.00

Flux crossing the motor winding 0 500 1000 1500 2000 2500 3000

Electrical frequency (Hz)

3500 4000 4500 5000

Flux crossing only the edges of the teeth

With : Te torque required to rotate the rotor Fig. III-6 :
α motor speed Measure of the real part of the impedance
Measures of Im(Z), motor PH632-508-002

PF2 power loss due to R' 0.01

0.009

PF1 power loss at the teeth edges 0.008

0.007

0.006

we can calculate R' with the method used for

Im(Z)

0.005 L (Z)
0.004 Linear (L (Z))

0.003

the EMF, then we can estimate and dissociate 0.002

0.001

the losses at the tooth edges.

0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

Electrical frequency (Hz)

The equations are :

Fig. III-7 :
2 2 2 Measure of the imaginary part of the impedance
E E. (Kω / N) − E
PF2 = =
R' L ⋅ω
and Measures of R', motor PH632-508-002

1600

PF1 = Te ⋅ α − PF 2 1400

1200

Note: 1000
R' in ohms

Meth. pui
800 Polynomial (Meth. pui )

This method allows to separate the losses

600

400

200

due to the flux created by the magnet and 0

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

crossing the motor winding, from losses due to

Electrical frequency (Hz)

the leakage flux of the magnet crossing only the Fig. III-8 :
From the measure of Re(Z) and
Jm(Z)
tooth edges.
we can calculate R'

III-1-3 Practical measurements and Note :

interpretation Up to 2500 Hz which means 10000
Measurements were made for all Portescap steps/sec, we canlinearize Re(Z) and
Jm(Z).
stepper motors. Here we only show the results
for the motor type escap  III-1-3-2Measurements with the rotor

PH632-508.002.

(C) Portescap. All right reserved.

 Evolution of the EFM, motor PH632-508-002 The evolution of the impedance of the motor
45

40 winding versus step rate shows the iron losses

35

30
due to flux created by the winding only.
EFM in volts

25 EMF (V)
Polynomial (EMF (V))
20

15 The losses in the common magnetic circuit

10

0
seen by the magnet and by the coils seem to
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

Electrical frequency
be pretty small. Therefore we can separate the

Fig. III-9 : losses created by the flux of the magnet from

Evolution of the EMF versus speed the losses created by the flux of the coil.
Then our model becomes very simple. The
Up to 2500 Hz the EMF is roughly linear,
losses due to the flux created by the winding
indicating that the losses created by the magnet
are simulated by modifying the real part of the
flux crossing the motor winding are negligible.
impedance. The losses due to the magnet flux
being located mainly at the tooth edges are
Torque necessary, motor PH632-508-002

40 taken care of by subtracting the torque required

35

30 to rotate the stepper from the torque we shall

Torque in mNm

25

20 Torque (nNm) calculate.

15

10 As mentioned above, we have linearized

5

0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
the motor resistance versus frequency and also
Electrical frequency (Hz)

the torque due to the losses created by the

Fig. III-10 : magnet flux, according to the following
Torque required to drive the stepper motor
equations.
Equation for the real part of the impedance :
Analysis of power distribution, motor PH632-508-002

 f 
R( Z) = R ⋅ ( λ − 1) ⋅ + 1
25

20  2500 
Power in Watts

with R : motor resistance at 20°C

15 Pmec
Circuit Losses
Teeth Losses
10

5
f : electrical frequency
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
λ : value defined for each motor size
(see Fig. III-12).
Electrical frequency (Hz)

Fig. III-11 :
Analysis of power distribution versus Equation of the torque to be subtracted from the
step rate calculated torque :
TF = a ⋅ ωmechanical + b
III-1-3-3Interpretating the Measurements
The coefficients a and b have been identified
The measure of EMF shows that the
for each Portescap motor (see fig. III-12).
magnetic flux crossing the motor winding does
not generate too much losses. The torque Motor a b λ
necessary to rotate the stepper is mainly due to Type

the losses located at the tooth edges. P110 1,5.10-6 1.10-3 8

-6 -3
P310 2,2.10 1.10 2

(C) Portescap. All right reserved.

 P430 8,5.10-6 2.10-3 4.5 Then, with the speed considered constant,

P520/PP520 13,4.10-6 3.10-3 4 the BEMF can be a time function just like the

P530 15,8.10-6 6.10-3 5 current reference. The actual current is

P532 9.10-6 4.10-3 5 calculated with the electrical equation. We

P630 76,3.10-6 10.10-3 6 distinguish two cases, i.e.the current regulation

P632 80,7.10-6 10.10-3 8 in regenerative mode or in non regenerative

PH632 47,7.10-6 10.10-3 8 mode. The current being a periodic function, we

P850 200.10-6 17.10-3 9 should have :

P852 436.10-6 33.10-3 9.5 i (t+T/2)=-i (t), especially i (t=0)=-i (T/2)

The calculation will be reiterated with i(t) shifted

Fig. III-12 :
versus the actual current until the case i(t=0) = -
Coefficients a, b,λ for each
i (T/2) is verified.
Portescap motor

III-2 Dynamic torque calculation

For calculating the dynamic torque, we
consider:
- the motor speed is constant during one step
- the time origin is the instant the current
reference crosses the zero value.

With : ω the rotor angular velocity

N the number of pole pairs

T π)/(ω⋅N)
the electrical period T=(2

ψ phase advance in electrical degrees

to phase advance in seconds

to=(ψ⋅T)/(360)

T is:
The average torque per phase
T /2
2
T=
T ⋅ω
⋅ ∫ i( t) ⋅ E( t)dt
0
with i(t) the actual instantaneous current and
E(t) the BEMF.

III-2-1 Calculation principle

Firstly, the torque is calculated for a given
speed and a given phase advance. At the time
t= 0, we suppose that the current equals the
current reference which is zero.

(C) Portescap. All right reserved.

 Figure III-13 illustrates the BEMF, the
current reference and the actual current as
calculated. Calculations are repeated until Ιinit =

-Ιfin.

E(t) in V.
Ic(t) and i(t) A

Iinit.
E(t)
T/2
Ic(t)
0 t0 Time
i(t)
Ifin.

Fig. III-13 :
Shape of the back-EMF (E(t)), the current
reference Ιc(t) and the actual current
i(t) for a
phase advance of tox360/T electrical degrees

Figure III-14 shows the algorithm used to

calculate the dynamic torque for a given speed
and a given phase advance. The phase advance
should of course be adjusted for maximum
torque.

Figure III-15 shows the dynamic torque

both measured and calculated. It should be
noted that the result will be less precise if we
neglect the iron losses.

(C) Portescap. All right reserved.

Input :
Motor Data :
K : Torque L : Inductance
R : Resistance N : Number of pole pairs
Driver Data :
U : Maximum voltage
Ιc : Current reference per phase Td : Chopper period

Variables :
ω : Motor Speed
ψ : Commutation angle (electrical degrees)

Ιc(t) = f(t)
Example formicrostep mode
Ιc(t) = Ιc.sin(N.ω.t)

E(t) = K.ω.sin(Nω(t-t0))

i(t=0)=0

t=0

yes no
Ι c(t) ≥ i(t)

yes no
Regenerative
 U − E( t )   − T d τ  −T /τ Mode ?
i( t + T d ) =    1 − e 
+ i( t ). e d
 R 

− E (t )
( ) + i(t ). e− t
 − U − E( t )   −t τ  −T /τ
i( t + T d ) =    1 − e d  + i( t ). e d i (t + t d ) = 1 − e− t d τ d /τ
 R  R

Mechanical power for t ∈ (0,t+T d) = i(t+t d ).E(t+t d ) . t d + Mecanichal power t ∈ (0, t)

no yes
t+T d = T/2

t=t+T d

no
i(t+Td)+i (0) ≤ Ι c MAX 1000
i(0)=-i(t+T d)+i (0)
2

Average torque per phase = (2/ T. ω).Mechanical power for t ∈ (0, T/2)

Fig. III-14 :
Dynamic torque algorithm for constant motor speed and commutation angle
(C) Portescap. All right reserved.
 Dynamic Torque

0.25
Equations :

0.2
During acceleration
Calculated Value

0.15
Measured Value
Tmotor = TM(1−ω / ωM )
dω
0.1
J = Tmotor
dt
Solving this differential equation results in :
0.05

0
dω
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
J = TM (1 − ω / ωM )
dt
Speed in step/s

TM
Fig. III-15 −t
Jω M
Dynamic torque of the PH632-508-002 with an ω( t) = ωM (1 − e )
L/R driver, U=36V, serial resistance 33 ohms
TM
−t
ωM Jω M
θ( t) = ωM [ t + J (e − 1)]
IV Adequate speed profile TM
In the previous paragraph we calculated the Another method is to proceed in discrete steps.
dynamic torque of the stepper associated with Let ωn be the average speed of step number n.
its driver. Knowing this torque and the The torque is considered constant from ω(n) to
kinematics of the mechanical transmission, like
ω(n+1).
friction and viscous torque, and reflected inertia,
ω(n) is:
Then the motor torque at the speed
we can calculate the velocity profile.
Tmotor (ωn) = TM . (1-ωn / ωM)

The acceleration α(n) to go fromωn to ωn+1 is:

IV-1 Exponential speed profile

α (n) = TM .(1 − ωn / ωM ) / J
Consider the useful motor torque to equal
seventy percent of the calculated torque, from ω(n +1) − ω(n) = α (n) . ∆t (n)
which we have subtracted the friction and th
with ∆t(n) being the time required for the n
viscous torque. step.
Let's assimilate this usable torque to a linear Now, if we expressω(n) and ω(n+1) in steps/s and
function of speed (see fig. IV-1).
α(n) in steps/s 2, we get:

Usable torque to accelerate 2

∆t ( n) =
TM
ω(n +1) − ω(n)
By substituting ∆t(n) we find:

ω(n +1) 2 − ω( n) 2 = 2α (n)

0 ωM Speed
or
Fig. IV-1
Torque available to calculate TM P
ω(n + 1) = ω(n) 2 + 2 ⋅ (1 − ω(n) / ωM ) ⋅
the speed profile J 2π

(C) Portescap. All right reserved.

where P is the number of steps/rev. This
recursion series is an easy way of calculating Equation for the acceleration :
the velocity profile step by step. d( ω)
J = TA(i) (ω)
IV-2 Adequate Speed profile dt
Let ωn be the speed of the nth step and ∆t(n) the
Depending on the driver it is not always
possible to consider the dynamic torque as a time required to make this step.
ω 
linear function of speed. In the following ( n+1) − ω (n)
J =T
A (i) (ω)
paragraph dynamic torque is assimilated to an  ∆ t (n ) 
association of straight lines (see fig. IV-2).
If the acceleration is in steps/s 2 and P the
We do not consider the same curves for
number of steps/rev., then:
the usable acceleration torque and deceleration TA( i) ( ω) P 2
torque. With friction being a negative torque, it ⋅ = accelerati on in steps / s
J 2π
has to be subtracted from the acceleration Let ω(n) be in steps/s.
torque and added to the deceleration or breaking th
The time required to move this step
n is :
torque.
2
∆t (n) =
ω(n + 1) + ω(n)
Acceleration torque

The expression ofω(n+1) can be written as:

T A0
T A1 TA (i) ( ω). P
ω (n + 1) 2 = ω(n) 2 + 2 ⋅
T A2 J⋅ 2 ⋅ π
T Ai
TAi+1  (ω − ω 
P n A(i)
ω(n+1)2 = ω(n)2 +  ⋅ ( TA(i) − TA(i+1) ) + TA()i 
ωA1 ω A2 ωAi ω Ai+1 Speed J.2. π (ωA(i+1) − ωA(i) 

From this formula and with the following

Deceleration torque
algorithm (fig. IV-3), we can calculate the
T D0 velocity profile.
T D1
T D2 n= 0
ω A0 = 0
T Di i=0
ωA i = ?
TDi+1 TAi = ?

ωD1 ω D2 ωDi ω Di+1 Speed n=n+1

i=i+1
ω n < ωAi + 1
Fig. IV-2 yes no
Torque versus speed of a stepper motor
assimilated to parts of straight lines ω (n+1)2 = ω n 2 + P ( ω n - ωA i ) . ( T A i - TAi+1) T A i
( ωAi+i - ω A i) +
J. π

Speed profile during acceleration

:
According to paragraph IV-1, the torque Speed of the step n+1
in step/s =ω(n+1)

at speed ω∈[ωA(i), ωA(i+1) ]

is:
ω − ωA(i) Fig. IV-3
TA(i) (ω) =
ω − ωA(i+1) − ωA(i)
[ ]
⋅ TA(i) − TA(i+1) + TA(i)

(C) Portescap. All right reserved.

We can then build the following table and the for acceleration and deceleration
speed profile (see fig. IV-4) :
Practical realisation
1st step ω 1 = ....... s/s So far we have calculated two tables, one
2nd step ω 2 = ....... s/s Speed in step/s
3rd step ω 3 = ....... s/s
each for acceleration and deceleration. Each
.. table has two columns, for the number of steps
.
. (or microsteps) to be made at an average speed
.
ωn, and for the average speed in steps/s (or
.
th
n step ω n = ....... s/s microsteps/s). Obviously, the total displacement
1 2 3 4 5 6 7 8.....
Fig. IV-4 is known and, with the algorithm IV-6, also the
Speed profile in steps/svs the step number
number of steps to accelerate to constant
speed. We now need to generate the pulses
Speed profile during deceleration
:
which are sent to the driver, and to count them
A curve similar to fig. IV-4 describes the
deceleration (see fig. IV-5) : in order to know the displacement at any time.
One solution is to use a microprocessor or a
Speed in step/s
DSP, and two timers. One of them is working as
a pulse generator and the other one as a
counter (see fig. IV-7).

ROM & RAM

Number of step 8 7 6 5 4 3 2 1
external Clock
clock

Fig. IV-5
Timer Counter
µP

Realisation of a displacement of X steps in the Motor Driver

minimum time : Fig. IV-7

Let XA: number of steps of acceleration Practical realisation

XD: number of steps of deceleration

XT: total displacement in steps
XA and XD, are calculated with the algorithm of
fig. IV-6.
XA = 0
XD = 0

XA = X A + 1

XD = XD + 1
no ω(XA) > ω (XD)
yes

yes no
X D + X A < XT

yes XA = Number of steps to accelerate

X D + X A = XT XD = Number of steps to decelerate

no X A-1 = Number of steps to accelerate

X D = Number of steps to decelerate

Fig. IV-6
Calculation of the number of steps
(C) Portescap. All right reserved.
 The table calculated previously has to be two frequencies corresponds to a speed jump of
converted into the number of periods of the 196 steps/s.
external clock. Firstly, the microprocessor If the motor makes 10 microsteps/step, the
resets both timers. Once a motion is required, a speed of 10000 steps/s is generated by loading
pulse is sent to the driver. The timer (pulse the value 10 into the timer. The closest smaller
generator) is loaded with the frequency required frequency we can generate by loading 11 is
next. The microprocessor checks the number of 90909 Hz. The jump between these two
steps already made and compares it to the frequencies equals 909 full steps/s, which is
displacement profile in order to know the next much higher than when working in half step
frequency to be put into the timer once the mode. Figure V-1 illustrates this example.
preceding pulse is sent. The next frequency to
Speed in pulse/s
load will be :
- the following period in the acceleration table
when accelerating
- the previous period in the deceleration table
when braking
- the same period when running at constant time

speed. Fig. V-1

Possible speed profile
V Precautions to be taken in micro- Motor working in half step mode
step mode Motor making10 microsteps/step
The microstep mode allows to increase the
motor resolution and to have a smoother motion. Theoretical demonstration
Nevertheless, if the number of microsteps/step Let: fh external clock frequency
and the peak speed are high, the electronics fn frequency generated by loading the
designer has to make sure the external clock value N into the timer
frequency of the timer is high enough to avoid fn+1 the closest smaller frequency
excessive jumps in the frequency between ∆f the frequency jump betweenn+1 f
microsteps, which at high speed could not be and nf
compatible with the dynamic torque of the fh fh
∆f = fn+1 - fn = − ∼−
motor. N ( N+ 1) N
2

Example:
If the frequency to be generated is X times the
A motor with 200 steps/rev. has to rotate at
value of fh, then ∆f is X2.fh/N 2 which as a speed
10'000 steps/s. The external clock frequency is
jump is X times larger.
1 MHz. If the motor is working in half step
Very often, the electronics designer is not aware
mode, the value 50 has to be loaded into the
of this phenomenon which can disturb the motor
timer to generate a frequency of 20 khz. The
behaviour at high speed.
closest smaller frequency we can generate is
CONCLUSION:
obtained by loading the value of 51.
This speed is 19607 Hz. The jump between the
(C) Portescap. All right reserved.
Today, stepper motor technologies like the Disc motor performance. Other methods exist, like
Magnet motor ofPortescap allow to use stepper the phase plane, which consists of calculating
motors not only as apositioner but also for fast the motor position and speed step by step.
incremental motion. Such motors can really Unfortunately, such a process requires to know
compete against traditionalBrushless DC perfectly well all the kinematics of the entire
motors without the need of an expensive drive system which is difficult in an industrial
electronics for closing the loop. To fully take environment.
advantage of such a motor, it is important to The paper allows to take full advantage of the
know with reasonable accuracy the dynamic Disc Magnet Motor. The calculation of the
torque of the motor associated to its driver. This dynamic torque as well as the optimum velocity
paper describes an easy way to introduce iron profile have been implemented in a software
losses into the model for determining the called "SOAP", developed by Portescap for use
dynamic torque. With a very simple algorithm, a by its application engineers in order to help
speed profile close to the optimum can be customers to optimise their incremental motion
calculated. It is clear that with such a method applications.
we always need a safety factor of roughly 70%
meaning that we are not using 100% of the

Bibliography:

• C. Oudet
"A New Family ofMultipolar P.M. Stepper Motors"
IMCSD, 1981
• M. Crivu, M. Jufer
"Auto-commutation de Moteurs Pas-à-pas sansCapteur"
6e colloque sur les moteurs pas-à-pas, EPFL, July 4-5, 1990
• M Jufer and R.Osseni
"Brushless DC Motor and Driver
Modeling with Iron Losses Effects"
IMCSD, 1988
• R. Osseni
"Modélisation et Auto-
asservissements deMoteurs Synchrones Autocommutés"
Thèse, EPFL, 1988
• L. Antognini
"Réglage etCommande deMoteurs Pas-à-pas"
Thèse # 547, EPFL, 1984
• P.P. Acarnley
"Stepping Motors"
IEE Control Engineering series 19
• M. Simon-Vermot and Dr.Karmous
"Optimisation of the Utilisation of a Stepper Motor"
Electric Drive Design and Applications, EPFL,oct. 19-20, 1994

(C) Portescap. All right reserved.