ELEC4613 ELECTRIC DRIVE SYSTEMS

Application areas:

Variable torque:
Vehicle and transformation systems,
conveyors, lifts,
excavators,
materials handling equipment,

Variable-speed:

Pumps, mill drives, compressors

Wind generators,
Electric vehicle, railway, other transportation systems (e.g., in
mining)
Electric Ship
Electrically actuated aircraft, new developments

Position servo:

Wire bonding machines (3-axis drive), Machine tools drives,

Rolling mill drives
Material handing equipment
Automation systems,
Robots

Section 1 – Introduction and 1 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS

Section 1 – Introduction and 2 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS

Section 1 – Introduction and 3 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS

Section 1 – Introduction and 4 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS

Section 1 – Introduction and 5 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS

Introduction to Electrical Drives

Why (electronically) variable speed?

1. Operation with efficiency and energy saving

Examples in hybrid electric vehicle and other automotive electric actuators, variable speed
pumps, air conditioning, wind generators, and so on.

2. Speed and position control may be essential to the process.

Examples in tracking systems, process automation and control such as in wire bonding of
ICs, manufacturing industries and guidance systems.

3. Some processes are enhanced if speed control is available.

Examples in transportation, electric traction, material handling.

4. Reduction of transients and stresses due to abrupt motion and ON/OFF operations.

Examples in metal, textile and paper rolling mills.

5. Disturbance rejection from supply and load

Rejection of disturbance means higher productivity.

6. Matching motor and load characteristics

Servo, traction, pump and many other types of loads need to be matched with machine
characteristics.

Section 1 – Introduction and 6 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS

Components of a typical electric drive system

Utility Electric Power

Interaction Machine Electronic
Mitigation Theory
Converters

Sensors/ Electric Control

Observers Theory
Drive

Mechanical Real-time
Mechanics System Control
Modeling

Figure 1.1

Design of an electric drive requires specialist knowledge and integration

of:

1. Machine theory
2. Mechanics
3. Power Electronics
4. Electromechanical systems modeling
5. Control theory
6. Microprocessors, ASIC and DSP based real-time control
implementation
7. Intelligent sensors, state and parameter observers
8. Utility interaction alleviation and EMI mitigation techniques

Section 1 – Introduction and 7 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS

AC/DC
supply
Load
coupling

Power Load
Motion Motor
Motion Converter
Controllers
references

Motion sensors

Figure 1.2 Typical structure of a variable-speed drive system.

The power converter may include AC-DC, DC-DC, DC-AC and AC-AC power
electronic converter circuits.

For a frictionless translational system

d 2x dv
f ma
= = m =
dt 2
m
dt
;=
v ∫ adt; =x ∫ vdt
For a frictionless rotational system

d 2θ dω
α J =
T J=
=
dt 2
J
dt
ω
; = ∫ α dt; =
θ ∫ ωdt

Section 1 – Introduction and 8 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS

Position Controller Speed Controller Torque Controller

∗
θ + ω∗ + iT* + Machine
− − −
iT
θ ω

φ* + Load
−
if va, vb, vc
d/dt
Torque and Flux E
iT Producing
ia, ib, ic
Currents

Figure 1.3

Mech
Output Power
in Watts

10, 000,000 NSPSM

SPSM

1000,000

DCM
100,000 SCIM

1,000

100
PMSM

10

1
100 1,000 10,000 100k 300k
Rev/min

Figure 1.4
Quadrant Operation

Section 1 – Introduction and 9 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS

Figure 1.5

Section 1 – Introduction and 10 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
Elements of mechanics and representation of rotational mechanical load

Torque is the product of the tangential force F on the surface of the cylinder and radius r. This
tangential force creates a shear stress σ N/m2 on the surface of the cylinder. The shear stress is
the tangential force per unit surface area (infinitesimally small) of the cylinder. For a given shear
stress, the total force acting on the surface of the cylinder is

2π
=F ∫ 0
θ ⋅ σ 2π RLσ
LRd= N

Torque, T = 2π R 2 Lσ Nm

Figure 1.6

This implies that for a maximum shear stress of the material of the cylinder, the maximum
torque that the rotor can withstand is proportional to its volume. In other words, if a machine is
to develop a higher torque, its volume must be appropriately increased. Note that the maximum
shear stress is determined by the material of the cylinder. For an electrical machine, the material
is (laminated) silicon steel. Copper conductors which develop the torque are embedded in slots
formed at the periphery of the cylindrical surface (outer or inner) by stacking together many
appropriately shaped laminations. The maximum shear stress is experienced by the steel when
the machine is accelerated (or decelerated) at the maximum rate which is determined by the
maximum current capability of the conductors. Note also that the force developed on the
conductors is generally given by F = BLi, where B is magnetic field (T) and i is the current
(Amp) in the conductor. While maximum i, (or electric loading) is determined by the cross
section of the copper conductors, the maximum air-gap B field, (or magnetic loading) around the
conductors (about 0.9 T) is limited by the saturation field of the steel.

T
The developed torque/rotor volume, TPRV = = 2σ Nm/m3.
π R2 L

This equation implies that for given maximum σ and T, the machine dimensions R and L can be
freely chosen to satisfy this equation. The radius R, which also largely determines the moment of
inertia of and the energy stored in the moving system, may affect its selection .

The developed torque T (in Nm) on a cylindrical structure accelerates the structure with an
angular acceleration. When the load torque, which may increase with speed, balances the
developed torque, the cylindrical structure stops accelerating and rotates at a steady angular
speed ω rad/sec. The power associated with the rotating structure is given by P = Tω in Watts.

Section 1 – Introduction and 11 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
Moment of inertia, J

Rotational systems are described in terms of torque T in Nm, angular displacement θ in radians,
angular velocity ω in rad/sec, angular acceleration α in rad/sec2 and angular moment of inertia J
in kg-m2. Consider the rigid body of figure below, which rotates around a vertical axis with an
angular acceleration, α. Any mass element dm has a linear acceleration a given by

a = rα m/sec2 (1.4)

r
dm
F

Figure 1.7

where r is the radial distance of dm from the axis and

dω d 2θ
α= = . (1.5)
dt dt 2

The tangential force on dm is F = adm = rαdm (1.6)

Torque on dm is T = r×F = αr2dm (1.7)

The torque on the whole rigid body is given by

T = αr 2 dm = α r 2 dm
∫ ∫ (1.8)

because α is the same everywhere in the rigid body.

The quantity ∫ r 2 dm is called the moment of inertia J of the rigid body. Thus

J= ∫ r 2 dm kg-m2 (1.9)
In some drive systems the moment of inertia is variable, just as the travelling mass may be
variable in a lift, hoist or a conveyor drive. Examples are centrifuges, reel drives, paper and
textile mills, and robot arms. For such systems, the applied torque creates a rate of change of
angular momentum, in the absence of other opposing torques. Thus

Section 1 – Introduction and 12 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
d dJ dω
T= ( Jω ) = ω +J (1.10)
dt dt dt

For this course, the only constant moment of inertia J will be considered. Thus,

dω
T =J Nm (1.11)
dt

In rotational system other than direct drives, because of the gear box and other forms of load
couplings such as pulleys, timing belt, lead-screw and so on, the load angular position may be
different from the motor angular position. From 1.11, torque T applied to a purely inertia
(rotating mass) load JL produces acceleration according to

d 2θ L Nm (1.12)
=T J=
L J Lα L
dt 2

where JL = moment of inertia of a rotating mass (load) in kg-m2.

θL = angular position of the rotating mass (load)

in radians

αL = angular acceleration in rad/sec2.

This torque is also the counter torque which applies against the torque applied to the inertia load.

Moment of inertia of a rigid hollow cylinder

r2
dr
r1 r

Figure 1.8

The inner and outer radii of the cylinder are r1 and r2 respectively. Its length is L and the density
of the material is δ kg/m3.

For an annular ring of thickness dr at a radius r, dm = δ 2πrLdr

M r2
∴J
= ∫0 r 2 dm
= ∫r
1
r 2δ 2π rLdr

π
=
2
{
δ L r24 − r14 } (1.13)

Thus, the moment of inertia of cylindrical structures increases as the fourth power of the outer
radius!
Section 1 – Introduction and 13 F. Rahman (EET, UNSW)
basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
Viscous friction, D

Viscous friction load develops torque which is proportional to angular velocity. Some of the
applied torque applied to a mechanical system is also absorbed in overcoming viscous friction,
as found in bearings, dash-pots for damping mechanisms and so on.

Figure 1.9 A dash-pot

Simple viscous friction found in gear-boxes, dash-pots for damping of mechanical systems and
so on develops counter torque which is proportional to angular velocity. Thus,

dθ L (1.14)
TDL = D
dt
where D = viscous damping friction in Nm/rad/sec.

+TF, Nm D

ω, rad/sec

-TF, Nm

Figure 1.10

Section 1 – Introduction and 14 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
Other friction

+Ts, Nm

ω, Rad/sec

-Ts, Nm

Figure 1.11 Static Friction: T = Ts ⋅ sign( ω ) ω = 0

+Tc

-Tc

Figure 1.12 Coulomb Friction: T

= TC ⋅ sign( ω )

Fixed load torque

TL, Nm

ω, rad/sec

Figure 1.13 Hoist and lift type Load

For such a load

TL = K (1.15)

Note that this type of load opposes the motor torque in quadrant 1 (+ve T, +ve ω) in which the
load speed is positive (say, clockwise). In quadrant 2, where ω is negative, this load torque

Section 1 – Introduction and 15 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
actually acts in the same direction as the motor torque to accelerate the load in the –ve
(counterclockwise) direction.

Note that pure inertia loads, once accelerated to a speed, requires no torque to maintain speed. In
the absence of any friction or load torque, the counter toque applied to the motor thus becomes
zero.

Note also that there are friction loads which are more complex functions of speed. For example,
automotive vehicles develop counter torque which is proportional to the square of the speed of
the vehicle. We will not consider such loads in this course. We will also not consider loads with
variable moment of inertia. Practical loads may behave as indicated in Figure 1.14.

+T, Nm +T, Nm +T, Nm

ω, rad/sec ω, rad/sec ω, rad/sec

(a) (b) (c)

Figure 1.14 T-ω characteristics of typical loads

Restraining torque

Restraining load torque is expressed as spring constant Km. This is a counter torque which is
proportional to the angular displacement θL of the load shaft from a reference angle. Thus,

Figure 1.15

TL Dω + K mθ L
= (1.16)

The sum of all load torques referred to the motor shaft must be equal and opposite (according to
Newton’s law) to the applied motor torque. Note the referred load torques may not be the same
as the actual load torques, because of the gear box, pulley, lead-screw or rack and pinion
coupling that may exist between the motor and the load shafts. Thus

Section 1 – Introduction and 16 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
d 2θ m dθm (1.17)
T= J m 2
+ Dm + K mθ m + TLm
dt dt

d ωm (1.18)
= Jm + Dmωm + K mθ m + TLm
dt

T = shaft torque developed by the motor, Nm

Jm = moment of inertia referred to the motor shaft, kg-m2
Dm = damping coefficient referred to the motor, Nm/rad/sec
TLm = fixed load torque referred to motor shaft, Nm
ωm = rotational speed of motor, rad/sec
θm = motor shaft angle, rad.
Km = coefficient of restraining torque, referred to motor shaft, Nm/rad

Note that all items in the RHS of 1.18 are load or counter torque, TL; TLm represents the fixed
load torque which is independent of speed or angular position of motor shaft. The spring
constant term Kmθm is present only when there is a detent/spring or restraining coupling to load.
For continuously rotating systems, such a coupling can not be present. However, when the
coupling shafts between the motor, load and speed/position sensor are flexible, a torsional torque
term must be included in 1.18. This torque arises because of includes relative or torsional
displacements, between the motor and other driven shafts.

Load Coupling: Gear drive

Gearing is the most common coupling between motor and load shafts. It can be represented as
shown in Fig. 1.16.
f2

2 Td2
Td1
1
ω1 J1 J2
ω2
r1
r2

f2

Figure 1.16

In the absence of any friction, backlash and slippage,

dω1 (1.19)
Td 1 − r1 f1 = J 1
dt

and r f = J d ω 2 (1.20)
2 2 2
dt
Section 1 – Introduction and 17 F. Rahman (EET, UNSW)
basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
where f1 and f2 are the tangential forces at the point of contact, exerted by wheels 2 and 1
respectively. J1 and J2 are the moment of inertia associated with wheels 1 and 2 respectively. In
steady-state, f1 = f2 and r1ω1 = r2ω2.
dω dω (1.21)
∴ Td 1 = J 1 1 + r1 f1 = J 1 1 + r1 f 2
dt dt

d ω1 r d ω2 (1.22)
= J1 + 1 J2
dt r2 dt

  r1 
2
 d ω
∴ Td 1 =  J 1 +   J 2  1 (1.23)
  r2   dt

Thus a reflected inertia J '2 can be thought of rotating at the speed of the driving wheel 1. The
reflected inertia J '2 is given by,
2 2
 r1 '  ω2  (1.24)
=J  r  J2
= 2  ω  J2
 2  1
The total equivalent moment of inertia of the driving wheel is thus
2
ω  2 (1.25)
J 1e = J 1 +  2  J 2 = J 1 + a J 2
ω1 

Note the similarity with transformer turns ratio and the transferred values of primary and
secondary parameters. Normally, a ≤ 0.1 in one stage of gearing. In the gear-coupled drive
arrangement of fig 1.17, where the motor and load rotate at speeds ωm and ωL respectively,

ωL
ωm a=
Motor ωm
Jm
ωL
Load
JL

Figure 1.17
the load inertia reflected to the motor is: J Lm = a 2 J L (1.26)

When friction and other load torques are also present, the torque balance equation is
d ωm (1.27)
( Jm + a2 J L )
T=
dt
+ DLmωm + TLm

where DLm and TLm are the load friction coefficient and load torques referred to the motor. It can
be shown that with gearing,

Section 1 – Introduction and 18 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
 ELEC4613 ELECTRIC DRIVE SYSTEMS

DLm = a 2 DL and T = aT (1.28)

Lm L

Condition for maximum load acceleration

Neglecting friction and load torques

Tm
= (J m + a2 J L ) ddtω
m
(1.29)

Tm  J  dω L  J  (1.30)
=  J L + m  =  J L + m  α L
a  a 2  dt  a2 
d ωL
where α L load
= = accleration
dt

From the above,

aTm (1.31)
αL=
2
a JL + Jm

For maximum load acceleration,

dα L {a 2 J L + J m } Tm − aTm ( 2aJ L )
=0 =
da Denomonator

Thus, for maximum acceleration of a pure inertia load,

Jm = a2J L . (1.32)

which means that referred load inertia must be equal to the motor inertia if maximum load
acceleration of the load is to be achieved. From 1.31 and 1.32, the maximum load acceleration is
then given by
Tm (1.33)
α L max =
2 J LJm

Section 1 – Introduction and 19 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
Other couplings
ML
r r
Jm Js
Jm M1
ωm

J = Mr 2 Μ
Lead screw
Rack and pinion
Pulley

Figure 1.18

Distance travelled by the mass M per rev of leadscrew or pinion = L = 2πr.

2
 L 
J =M 
 2π 

Power and energy of rotational systems

The power of a rotational system is given by

P = T ω Watts (1.34)

For a rotational system, ignoring any restraining torque, the torque balance equation is,

dω (1.35)
T = J + Dω + TL
dt

Multiplying both sides by ω and integrating from zero to a time t,

t t t t
∫0
ωTdt =∫ J ω d ω +
0 ∫
0
Dω 2 dt + ∫ 0
ωTL dt

1 t Joules (1.36)
Wm =J ω 2 + ∫ Dω 2 dt + WL
2 0

The left hand side represents the total mechanical energy delivered to the system. The first term
1 2
in the right hand system is analogous to 1 mv 2 or Li or 1 CV 2 . It thus represents the kinetic
2 2 2
energy stored in the mechanical system which rotates at angular speed ω rad/sec. The second
t 2
term in the right hand side is analogous to ∫0 Ri dt . It represents the energy lost to the
surroundings due to viscous friction with air or other fluidic medium.
t
The third term on the right is analogous to
∫ vidt and represents the energy delivered to the
0
mechanical load in changing its speed during time t. The torque-speed characteristic of the load
determines this quantity.

Section 1 – Introduction and 20 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
Note that equations 1.26 and 1.28 can also be obtained from considerations of power invariance,
on the motor and load sides with PmL = PL in the steady-state, when the coupling (say, the gear
box) is lossless. In other words, for a purely inertia load JL, across a lossless coupling,

1 2 1 ωL2
J LmωLm = J LωL2 , so that=
J Lm = J L a2 J L (1.37)
2 2 ωm2

For a purely viscous load DL across a lossless coupling,

2 ωL2
DLmωLm = DLωL2 , so that=
DLm =2
DL a 2 DL (1.38)
ωm

For a fixed load torque across a lossless coupling

ωL
ωLmTLm = ωLTL , so that=
TLm = T aTL (1.39)
ωm L

Steady-state torque-speed characteristics of typical loads

Mechanical loads can be of many different types. Fluid couplings generally have viscous
friction, with torque proportional to speed. Friction torque of bearing and gear friction of servo
motors are also of this type. Small servo motors also exhibit static friction. Machine tool, robotic
and mill drives often encounter combinations of such loads.

Pumps, fans, blowers and compressors often have rising torque characteristics with speed in
which the load torque may be proportional to speed squared.

Traction and coiler loads generally have falling torque characteristics. These are hyperbolic in
nature which imply constant-power operation (P =Tω = constant). Excavators operate with
constant torque at different speeds.

Lifts, hoists and winches have constant torque characteristics, regardless of the direction of
speed.

While most load torques oppose the motor developed torques, hoist and lift drives can assist the
motor when overhauling.

Figure 1.19 summarizes characteristics of typical loads.

Section 1 – Introduction and 21 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
w
rad/sec

ωo

T, Prated = ωoTrated
Nm
Trated

(a) Torque - speed and power requirements of servo drives in machine tool and automation

ωmax

w Constant Power,
rad/sec P = Tω = constant

Prated = ωbaseTrated
ωb

T, Nm Trated

(b) Torque - speed and power requirements of typical traction drives (e.g., electric vehicles, trains and
coilers in paper, textile and other mill drives.
Figure 1.19

Section 1 – Introduction and 22 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS

w
rad/sec T = kω2
P = kω3

T, Nm

(c) Torque - speed and power requirements of fan, compressor and pump drives.

ω
Rad/ T=C
sec P = +/− Cω

0
T, Nm

(d) Torque - speed and power requirements of hoist, lift and winch drives

Figure 1.19 Typical load characteristics

Real loads often have some combinations of above types. For example, a vehicle exhibits static,
traction and viscous friction torque all at once.

The torque characteristic of a load as a function of speed can be easily determined by running
the load at different speeds and measuring the power to drive the load at each speed.

The load of a servo motor within its speed range is often a pure moment of inertia, J, where the
viscous and other friction torques are negligible. A machine tool feed drive is a good example.
The moment of inertia of a drive can be obtained by recording the coasting speed profile, when
the viscous friction torque is known. The coasting speed profile is the trace of the speed of the
drive from a top speed after the power to the drive is switched off.

Section 1 – Introduction and 23 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
Motor torque-speed characteristics
The number of types of motors in the market place is vast. It is often said that each type of motor
has its niche area of application. Very often, every application calls for its unique motor to be
developed.

The steady-state torque-speed characteristics of available motors are often of three categories
shown in figure 1.20.

Synchronous
ω, rad/sec Motor

Asynchronous
/Shunt Motor
Series Motor
−T, Nm Trated T, Nm Tmax

− ω, rad/sec

Figure 1. 20

All synchronous motors (rotor excited salient and non-salient, such as reluctance, permanent
magnet, stepping etc) have a flat torque-speed characteristic. They run at a fixed speed
determined by the frequency of the supply and can be loaded up to maximum torque, Tmax,
largely determined by the supply voltage.

Asynchronous motors (the induction motor, dc shunt motor etc) have a slightly drooping torque-
speed characteristic. The rated torque, Trated, for both type of the motor is usually about one-third
of the maximum torque, Tmax, which the motor can develop. For asynchronous motors, the rotor
flux is supplied by the same source as the stator or armature. If the flux is separately supplied or
controlled, the rising part of the torque-speed characteristic does not exist. Instead, a stall torque
is defined which is found by extrapolating the slightly drooping torque-speed characteristic to
intersect the zero speed axis.

The series motor is characterized by a constant power type (hyperbolic) T-ω curve. This is
obtained by reducing the rotor flux of the motor, by reducing field. For DC machines, this is
achieved by reducing the field current, while for induction motors, the field can be reduced by
limiting the stator voltage at the rated value while increasing the frequency of the supply above
the rated value.

Wherever possible, the motor T-ω characteristics should be matched closely with those of the
load. This often results in the lowest rating of the power supply circuit for a given drive. It
should be noted however that with modern power electronic power converters and recently
developed control strategies, T-ω characteristics of virtually all motors can be made to fit with
any type of load characteristic. Nevertheless, proper matching of T-ω characteristics of motor
and load reduces the size of the motor and the power converter circuit.

Section 1 – Introduction and 24 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
Stability of steady-state operation

When the motor and the load operate in the steady-state at a constant speed, the sum of the
motor and opposing load torques must be zero. Thus

Tm − TL = 0 (1.40)

The motor and the load T-ω characteristics in the steady-state were indicated earlier. The above
condition is satisfied at the operating point at which the motor and the load torque characteristics
intersect, ie, the motor exactly develops the load torque TL and runs at a constant speed ω while
doing so. If the motor or load T-ω change for some reason, for example, due to a change of load
conditions or a change of the voltage applied to the motor, an oscillatory response around the
operating point (TL-ω) may ensue. These oscillations may or may not die down depending on the
location of the operating point and the T-ω characteristics of the motor and the load at this
operating point (or point of intersection).

If the load has appreciable viscous friction, these oscillations may die out quickly. The absence
of such friction represents a poorly damped situation where oscillations will die down only if the
slopes of the motor and load T-ω characteristics at the point of intersection are favourable. If Tm
– TL is non-zero and the system friction is negligible, the torque balance equation becomes

dω
Tm − TL = J (1.41)
dt

In the above equation, both Tm and TL are functions of speed as indicated earlier. For small
perturbations around an operating point, Tm = Tm + ∆Tm , TL = TL + ∆TL , ω= ω + ∆ω .

For small perturbations, the torque balance equation 1.41 becomes

d∆ω dTm dT
J = ∆ω − L ∆ω (1.42)
dt dω dω

J d∆ω
+ ∆ω = 0 (1.43)
K dt
d
where K = (TL − Tm ) (1.44)
dω

The solution for ∆ω from 1.43 will be stable (i.e., bounded) if K > 0, (ie, the root of the
characteristic equation of the differential equation 1.40 is negative. If K > 0, the disturbances in
J
ω will decay with a time-constant τ = . Otherwise they will grow, leading to instability.
K

Consider the T-ω characteristic of the induction motor as indicated in figure 1.21. It is slightly
droopy and has a part with a positive (rising T-ω) slope. The load characteristic is assumed to be
positive, as indicated. Three load characteristics are considered. Load characteristics 1 and 2
dT dT
intersect the motor T-ω characteristic such that K = L − m is positive. Load characteristic
dω dω

Section 1 – Introduction and 25 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
3 however intersects the motor T-ω characteristic such that at the point of intersection, K < 0.
For load 3, the motor cannot drive the load at a stable speed and will eventually coast down to
zero speed.

ω 1
2
rad/sec
Motor
3

T, Nm

Figure 1.21 Intersection of motor and three different load characteristics: 1, 2 and 3.

Three possible cases for K are summarized below.

ω Tm Tm TL
ω ω TL
TL
Tm

T T T
K > 0: stable operation K = 0: unstable equilibrium K = < 0: unstable
Figure 1.22

Quadrant operation of a drive

Drives often are operated in both directions, clockwise and counterclockwise. Torque-speed
characteristics are thus defined in both the first and the third quadrants of the T-ω plane, as
indicated in figures 1.23 and 1.24. When a drive runs at a speed, it also stores energy which can
be treated as a source of energy when the drive is to be braked. This source can use the motor as
a generator to extract and return energy of the load to the electrical supply which can be AC or
DC. Dynamic braking usually dissipates this energy stored in the load, while regenerative
braking returns the energy of the mechanical system to the electrical source.

While the motor is driven in quadrants 1 (or 3) as the load requires, it can also be driven in
quadrants 2 (or 4) by proper choice of operation of the power converter circuit supplying the
motor. For instance, when the motor is operated in the forward direction in quadrant 1 and it is
required to be braked in order to slow down or come to stop quickly, the motor-converter system
can be operated in quadrant 2 so that the motor torque (or current) is reversed while the speed is
still positive (i.e., the motor runs in the forward direction). The motor now acts as a generator
with a voltage which is higher than the voltage supplied by the converter. The power converter

Section 1 – Introduction and 26 F. Rahman (EET, UNSW)

basic concepts in drives July 2011
ELEC4613 ELECTRIC DRIVE SYSTEMS
thus allows return of energy from the motor to the supply, thereby slowing the drive speed down
more quickly than would be possible if the motor was simply switched off during braking.

Q2 Q1
Ia + ω, Ia
2 + Va 1

Va Ea Va Ea

FWD Braking + ω, − T, + Ea, − Ia FWD Motoring, + ω, + T, + Ea, +Ia

Q3 Q4 + T, +Ia
− T, - Ia Ia Ia
2 1

Ea Va
Va Ea
− ω,
− Va
REV Motoring, − ω, − T, − Ea, − Ia REV Braking, − ω, + T, − Ea, + Ia

Figure 1.23

The amplitude of the reverse current through the motor during braking can be much higher than
the continuous current rating of the motor, only for a short time. It can be maintained at the
desired level at all times by continuous adjustment of the power converter so that the drive slows
down predictably under constant maximum reverse torque. The figure 1.23 indicates such
quadrant operation of a drive.

Reversing the converter connections across to motor is often carried out electronically, by
having two back-to-back converters in parallel across the motor. As shown in figure 1.24,
Converter 1 is for forward motoring with, +ve torque and +ve speed. Converter 2 is for reverse
motoring with –ve torque and –ve speed. Converter 2 is used for forward braking. Likewise,
converter is used for reverse braking. Such four quadrant operation will be described further
later.

1 2

~ Va Ea

ω, T

Figure 1.24

Section 1 – Introduction and 27 F. Rahman (EET, UNSW)

basic concepts in drives July 2011