Unit 4

www.Vidyarthiplus.
com
UNIT-IV
INVERTERS
www.Vidyarthiplus.com
Single-Phase Inverters
Half-Bridge Inverter
One of the simplest types of inverter. Produces a square wave output.
(cont’d)
Full Bridge (H-bridge) Inverter
Two half-bridge inverters combined.

Allows for four quadrant operation.
(cont’d)
Quadrant 1: Positive step-down converter
(forward motoring)
Q1-On; Q2 - Chopping; D3,Q1 freewheeling
(cont’d)
Quadrant 2: Positive step-up converter
(forward regeneration)
Q4 - Chopping; D2,D1 freewheeling
(cont’d)
Quadrant 3: Negative step-down converter
(reverse motoring)
Q3-On; Q4 - Chopping; D1,Q3 freewheeling
(cont’d)
Quadrant 4: Negative step-up converter
(reverse regeneration)
Q2 - Chopping; D3,D4 freewheeling
(cont’d)
Phase-Shift Voltage Control - the output of
the H-bridge inverter can be controlled by
phase shifting the control of the
component half-bridges. See waveforms
on next slide.
(cont’d)
(cont’d)
The waveform of the output voltage vab is a quasi-
square wave of pulse width φ. The Fourier series of vab
is given by:
4Vd   nφ  
vab = ∑ sin  2   cos ( nω t )
n =1,3,5... nπ   
The value of the fundamental, a1= 4Vd
sin (φ / 2 )
π
The harmonic components as a function of phase
angle are shown in the next slide.
(cont’d)
Three-Phase Bridge
Inverters
Three-phase bridge inverters are widely
used for ac motor drives. Two modes of
operation - square wave and six-step. The
topology is basically three half-bridge
inverters, each phase-shifted by 2π/3,
driving each of the phase windings.
Three-Phase Bridge Inverters

(cont’d)

(cont’d)

(cont’d)
The three square-wave phase voltages can
be expressed in terms of the dc supply
voltage, Vd, by Fourier series as:
2Vd
va 0 =
π
∑
n =1,3,5...
(−1) n +1 cos(nωt )
2Vd 2π
vb 0 =
π
∑
n =1,3,5...
(−1) n +1
cos(nωt − )
3
2Vd 2π
vc 0 =
π
∑
n =1,3,5...
(−1) n +1
cos(nωt + )
3

(cont’d)
The line voltages can then be expressed as:
2 3Vd
vab = va 0 − vb 0 =
π
∑
n =1,3,5...
cos(ω t + π / 6) − cos(nωt + π 6)
2 3Vd
vbc = vb 0 − vc 0 =
π
∑
n =1,3,5...
cos(ω t − π / 2) − cos(nωt − π 2)
2 3Vd
vca = vc 0 − va 0 =
π
∑
n =1,3,5...
cos(ω t + 5π / 6) − cos(nωt − 5π 6)

(cont’d)
The line voltages are six-step waveforms and
have characteristic harmonics of 6n± ±1,
where n is an integer. This type of inverter is
referred to as a six-step inverter.
The three-phase fundamental and harmonics

are balanced with a mutual phase shift of
2π/3.

(cont’d)
If the three-phase load neutral n is isolated from the the
center tap of the dc voltage supply (as is normally the
case in an ac machine) the equivalent circuit is shown
below.

(cont’d)
In this case the isolated neutral-phase

voltages are also six-step waveforms with
the fundamental component phase-shifted
by π/6 from that of the respective line
voltage. Also, in this case, the triplen
harmonics are suppressed.

(cont’d)
For a linear and balanced 3Φ load, the line currents
are also balanced. The individual line current
components can be obtained from the Fourier series
of the line voltage. The total current can be obtained
by addition of the individual currents. A typical line
current wave with inductive load is shown below.

(cont’d)
The inverter can operate in the usual inverting or
motoring mode. If the phase current wave, ia, is
assumed to be perfectly filtered and lags the phase
voltage by π/3 the voltage and current waveforms are
as shown below:

The inverter can also operate in rectification or regeneration
mode in which power is pushed back to the dc side from the ac
side. The waveforms corresponding to this mode of operation
with phase angle = 2π/3 are shown below:

(cont’d)
The phase-shift voltage control principle
described earlier for the single-phase
inverter can be extended to control the
output voltage of a three-phase inverter.

(cont’d)

(cont’d)
The three waveforms va0,vb0, and vc0 are of
amplitude 0.5Vd and are mutually phase-
shifted by 2π/3.
The three waveforms ve0,vf0, and vg0 are of

similar but phase shifted by φ.

(cont’d)
The transformer’s secondary phase voltages,
vA0, vB0, and vc0 may be expressed as follows:
v A0 = mvad = m(va 0 − vd 0 )
vB 0 = mvbe = m(vb 0 − ve 0 )
vC 0 = mvcf = m(vc 0 − v f 0 )
where m is the transformer turns ratio
(= Ns/Np). Notewww.Vidyarthiplus.com
that each of these waves is a
function of φ angle.

(cont’d)
The output line voltages are given by:
v AB = v A0 − vB 0
vBC = vB 0 − vC 0
vCA = vC 0 − v A0
While the component voltage waves va0, vd0, vA0 … etc. all
contain triplen harmonics, they are eliminated from the
line voltages because they are co-phasal. Thus the line
voltages are six-step waveforms with order of harmonics
= 6n±1 at a phase angle φ.

(cont’d)
The Fourier series for vA0 and vB0 are given
by:
4mVd   nφ  
v A0 = ∑ sin  2   cos ( nω t )
n =1,3,5... nπ   
4mVd   nφ  
vB 0 = ∑ sin  2   cos ( nω t − 2π / 3)
n =1,3,5... nπ   

(cont’d)
The Fourier series for vAB is given by:
vAB = vA0 − vB 0
4mVd   nφ     2π 
= ∑ sin  2   cos ( nω t ) − cos n  ω t − 3 
n =1,5,7,11... nπ      
Note that the triplen harmonics are removed

in vAB although they are present in vA0 and
vB0. www.Vidyarthiplus.com
PWM Technique
While the 3Φ 6-step inverter offers simple

control and low switching loss, lower order
harmonics are relatively high leading to high
distortion of the current wave (unless
significant filtering is performed).
PWM inverter offers better harmonic control

of the output than 6-step inverter.
PWM Principle
The dc input to the inverter is “chopped” by
switching devices in the inverter. The
amplitude and harmonic content of the ac
waveform is controlled by the duty cycle of
the switches. The fundamental voltage v1
has max. amplitude = 4Vd/π for a square
wave output but by creating notches, the
amplitude of v1 is reduced (see next slide).
PWM Principle (cont’d)
PWM Techniques
Various PWM techniques, include:
• Sinusoidal PWM (most common)

• Selected Harmonic Elimination (SHE)
PWM
• Space-Vector PWM
• Instantaneous current control PWM
• Hysteresis band current control PWM
• Sigma-delta modulation
Sinusoidal PWM
The most common PWM approach is

sinusoidal PWM. In this method a
triangular wave is compared to a
sinusoidal wave of the desired
frequency and the relative levels of the
two waves is used to control the
switching of devices in each phase leg
of the inverter.
Sinusoidal PWM
(cont’d)
Single-Phase (Half-Bridge) Inverter
Implementation
Sinusoidal PWM (cont’d)
when va0> vT T+ on; T- off; va0 = ½Vd

va0 < vT T- on; T+ off; va0 = -½Vd
Sinusoidal PWM
(cont’d)
Sinusoidal PWM (cont’d)

Definition of terms:
Triangle waveform switching freq. = fc (also called
carrier freq.)
Control signal freq. = f (also called modulation
Peak amplitude
freq.) of control signal
Amplitude modulation ratio, m = Vp

VT
Peak amplitude
Frequency modulation ratio, of triangle wave
mf (P)= www.Vidyarthiplus.com
fc / f
Multiple Pulse-Width Modulation
• In multiple-pulse modulation, all pulses are

the same width
• Vary the pulse width according to the
amplitude of a sine wave evaluated at the
center of the same pulse
Generate the gating signal
2 Reference Signals, vr, -vr

Comparing the carrier and reference signals

• Generate g1 signal by comparison with vr
• Generate g4 signal by comparison with -vr
Comparing the carrier and reference signals
Potential problem if Q1 and Q4 try to turn ON

at the same time!
If we prevent the problem
Output voltage is low when g1 and g4 are

both high
This composite signal is difficult to generate
Generate the same gate pulses with one

sine wave
Alternate scheme
rms output voltage
• Depends on the modulation index, M
pδ 2pδ
V =V →V ∑ m
o S
π S
π
m =1
Where δm is the width of the mth pulse
Fourier coefficients of the output voltage
B =∑
n
2p4V
nπ
m =1
S
sin
nδ 
m
4  (
sin n α +
m
3δ
4
m
) (
− sin n π + α +
m
δ 
m

4  )
n = 1, 3, 5,..
Harmonic Profile
Compare with multiple-pulse case for p=5
Distortion Factor is considerably less

Series-Resonant Inverter
Operation
T1 fired, resonant pulse of current

flows through the load. The current
falls to zero at t = t1m and T1 is “self –
commutated”.
T2 fired, reverse resonant current
flows through the load and T2 is also
“self-commutated”.
The series resonant circuit must be
underdamped,
R2 < (4L/C)
Operation in Mode 1 – Fire T1
di1 1
L + Ri1 + ∫ i1dt + vC (0) = VS
dt C
i1 (0) = 0
vC (0) = −VC
R
− t
i1 (t ) = A1e 2L
sin ωr t
1
 1 R2  2
ωr =  − 2
 LC 4 L 
di1 Vs + Vc
= = A1
dt t =0 ωr L
Vs + Vc −α t
i1 (t ) = e sin ωr t
ωr L
R
α=
2L
To find the time when the current is

maximum, set the first derivative = 0
di1
=0
dt
 Vs + Vc 
( cos ωr t ) = 0
−α t −α t
 −α e sin ω r t + ω r e
 ωr L 
.....
ωr
= tan ωr tm
α
−1 ωr t m
tan = ωr t m
α
1 −1 ωr
tm = tan
ωr 2
To find the capacitor voltage, integrate the

current
t
1
vC1 (t ) = ∫ i1 (t )dt − Vc
C0
1  Vs + Vc  −α t
t
vC1 (t ) = ∫   ( e sin ωr t ) dt − VC
C 0  ωr L 
...
vC1 (t ) = −(Vs + VC )e −α t (α sin ωr t + ωr cos ωr t ) / ωr + Vs
π
0 ≤ t ≤ t1m ( ) The current i1 becomes = 0 @ t=t1m
ωr
απ
−
ωr
vC1 (t1m ) = VC1 = (Vs + VC ) e + Vs
Operation in Mode 2 – T1, T2 Both OFF
t2m
Operation in Mode 3 – Fire T2
di3 1
L + Ri3 + ∫ i3 dt + vC3 (0) = 0
dt C
i3 (0) = 0
vC3 (0) = −VC2 = −VC1
VC1 −α t
i3 (t ) = e sin ωr t
ωr L
t
1
vC3 (t ) = ∫ i3 dt − VC1
C0
−VC1 e−α t (α sin ωr t + ωr cos ωr t )
vC3 (t ) =
ωr
π
0 ≤ t ≤ t3 ( )
ωr
m
π
−α
ωr
vC3 (t3m ) = VC3 = VC = VC1 e
π
−α
ωr
vC1 (t1m ) = VC1 = (VS + VC )e + VS
.
.
1
VC = VS z
e −1
ez
VC1 = VS z
e −1
VC + VS = VC1
Space Vector Modulation

• Space Vector Diagram
r jβ r
V3 V2
OPO PPO
SECTOR
II
r r SECTOR III r SECTOR I
ω
• Active vectors: V1 to V6 V ref
(stationary, not rotating) r r
V4 θ V1
α
r OPP PPP OOO POO
• Zero vector: V0 r
V0
SECTOR IV SECTOR VI
• Six sectors: I to VI SECTOR V
OOP r r POP
V5 V6

• Space Vectors
• Three-phase voltages
v AO (t ) + v BO (t ) + vCO (t ) = 0 (1)
• Two-phase voltages
 2π 4π  v AO (t )
vα (t )  2 cos 0 cos cos
= 3 3  v (t ) 
v (t ) 
3 sin 0 sin 2π 4π   BO  (2)
 β   sin  vCO (t ) 
 3 3 
• Space vector representation
r
V (t ) = vα (t ) + j vβ (t ) (3)
(2) → (3)
r 2
V (t ) = [ v AO (t ) e j 0 + v BO (t ) e j 2π / 3 + vCO (t ) e j 4π / 3 ] (4)
3
where e jx = cos x + j sin x

• Space Vectors (Example)
Switching state [POO] → S1, S6 and S2 ON

2 1 1
v AO (t ) = Vd , v BO (t ) = − Vd v
and CO (t ) = − Vd (5)
3 3 3
r jβ r
(5) → (4) V3 V2
OPO PPO
r 2 SECTOR
V1 = Vd e j 0 (6) SECTOR III
II
r ω SECTOR I
3 V ref
r r
Similarly, V4 θ V1
α
OPP PPP OOO POO
r 2 j ( k −1)
π r
Vk = Vd e 3
(7) V0
3 SECTOR IV SECTOR VI
SECTOR V
k = 1, 2, ..., 6. OOP r r POP
V5 V6

• Active and Zero Vectors
P
Switching State Vector
Space Vector On-state Switch
(Three Phases) Definition
S1 S3 S5
Zero r [PPP] S1 , S3 , S5 r
Vector
V0 V0 = 0
[OOO] S 4 , S6 , S 2
A
r r 2
Vd B V1 [POO] S1 , S6 , S2 V1 = Vd e j 0
C 3
r π
r 2 j
V2 [PPO] S1 , S3 , S2 V2 = Vd e 3
S4 S6 S2 3
r 2π
r 2 j
V3 [OPO] S 4 , S3 , S 2 V3 = Vd e 3
N 3
Active
Vector r 3π
r 2 j
V4 [OPP] S4 , S3 , S5 V4 = Vd e 3
• Active Vector: 6 3
• Zero Vector: 1 r r 2 j
4π
V5 [OOP] S 4 , S 6 , S5 V5 = Vd e 3
3
• Redundant switching r r 2 j
5π
states: [PPP] and [OOO] V6 [POP] S1 , S6 , S5 V6 = Vd e 3

3

• Reference Vector Vref
r jβ r
V3 V2
• Definition OPO PPO
r
Vref = Vref e jθ SECTOR
II
SECTOR III r SECTOR I
ω
V ref
• Rotating in space at ω r r
V4 θ V1
α
ω = 2π f (8)
OPP PPP OOO POO
r
V0
• Angular displacement
SECTOR IV SECTOR VI
t
θ (t ) = ∫0 ω dt (9)
SECTOR V
OOP r r POP
V5 V6

• Relationship Between Vref and VAB
• Vref is approximated by two active r

V2
and a zero vectors
• Vref rotates one revolution, Tb r r

Vref
V2
VAB completes one cycle Ts
θ r
• Length of Vref corresponds to V1
Ta r
magnitude of VAB Ts
V1

• Dwell Time Calculation r
V2
• Volt-Second Balancing
r r r r
Vref Ts = V1 Ta + V2 Tb + V0 T0 Tb r r
 (10) Ts
V2 Vref
Ts = Ta + Tb + T0
r r r
• Ta, Tb and T0 – dwell times for V1 , V2 and V0 θ r
V1
Ta r
V1
Ts
• Ts – sampling period
• Space vectors
r jθ
r 2 r 2 j
π
r
,
Vref = Vref e , V1 = Vd V2 = Vd e 3 and V0 = 0 (11)
3 3
(11) → (10)
 2 1
 Re : V ref
(cos θ ) Ts
= V d
Ta
+ Vd Tb
3 3

Im : Vref (sin θ ) Ts = 1 Vd Tb (12)
 3

• Dwell Times
Solve (12)
 3 Ts Vref π
Ta = sin ( −θ )
 Vd 3
 3 Ts Vref
Tb = sin θ 0 ≤ θ < π /3 (13)
 Vd


T0 = Ts − Ta − Tb

• Vref Location versus Dwell Times
r
V2
SECTOR I
r
Tb r V ref
V2
Ts Q
θ r
V1
Ta r
V1
Ts
r π π π π π
V ref Location θ =0 0 <θ < θ= <θ < θ=
6 6 6 3 3
Ta > 0 Ta = 0
Dwell Times Ta > Tb Ta = Tb Ta < Tb
Tb = 0 Tb > 0

• Modulation Index
 π
T
 a = Ts m a sin ( −θ )
3
Tb = Ts ma sin θ (15)

T0 = Ts − Tb − Tc
3 Vref (16)
ma =
Vd

• Modulation Range
r jβ r
V3 V2
OPO PPO
• Vref,max SECTOR
II
SECTOR III r SECTOR I
ω
2 3 Vd V ref
Vref , max = Vd × = (17) r r
3 2 3 V4 θ V1
α
OPP PPP OOO POO
r
V0
SECTOR IV SECTOR VI
(17) → (16) SECTOR V
OOP r r POP
• ma,max = 1 → V5 V6
• Modulation range: 0 ≤ ma ≤ 1 (18)

• Switching Sequence Design
• Basic Requirement:
Minimize the number of switchings per
sampling period Ts
• Implementation:
Transition from one switching state to
the next involves only two switches in
the same inverter leg.

• Seven-segment Switching Sequence
r r r r r r r
V0 V1 V2 V0 V2 V1 V0
OOO POO PPO PPP PPO POO OOO
• Selected vectors: v AN
V0, V1 and V2 Vd
0
• Dwell times: vBN

Vd
Ts = T0 + Ta + Tb 0
vCN Vd
0
T0 Ta Tb T0 Tb Ta T0
4 2 2 2 2 2 4
Ts
• Total number of switchings: 6


• Undesirable Switching Sequence
• Vectors V1 and V2 swapped
r r r r r r r
V0 V2 V1 V0 V1 V2 V0
OOO PPO POO PPP POO PPO OOO
vAN
Vd
0
vBN
Vd
0
vCN Vd
0
T0 Tb Ta T0 Ta Tb T0
4 2 2 2 2 2 4
Ts
• Total number of switchings: 10


• Switching Sequence Summary (7–segments)
Sector Switching Sequence
r r r r r r r
V0 V1 V2 V0 V2 V1 V0
I
OOO POO PPO PPP PPO POO OOO
r r r r r r r
V0 V3 V2 V0 V2 V3 V0
II
OOO OPO PPO PPP PPO OPO OOO
r r r r r r r
V0 V3 V4 V0 V4 V3 V0
III
OOO OPO OPP PPP OPP OPO OOO
r r r r r r r
V0 V5 V4 V0 V4 V5 V0
IV
OOO OOP OPP PPP OPP OOP OOO
r r r r r r r
V0 V5 V6 V0 V6 V5 V0
V
OOO OOP POP PPP POP OOP OOO
r r r r r r r
V0 V1 V6 V0 V6 V1 V0
VI
OOO POO POP PPP POP POO OOO
Note: The switching sequences for the odd and ever sectors are different.

• Simulated Waveforms
Sector
VI VI
V V
IV IV
III III
II II
I I
v AB
Vd
0
π 2π 3π
v AO
2V d / 3
0
iA
0
π 2π 3π
f1 = 60Hz, fsw = 900Hz, ma = 0.696, Ts = 1.1ms


• Waveforms and FFT
v AB
Vd
2π
v AO
2Vd / 3
iA
π 2π 3π
V ABn / Vd
V AB1 = 0.566Vd
www.Vidyarthiplus.com n

• Waveforms and FFT (Measured)

• Waveforms and FFT (Measured)
VAB n / Vd VAB n / Vd
n =1 n =1
n=2
ma ma
( f1 = 60Hz and Ts = 1 / 720 sec )

• Even-Order Harmonic Elimination
r r r r r r r r r r r r r r
V0 V5 V4 V0 V4 V5 V0 V0 V4 V5 V0 V5 V4 V0
v AN v AN
Vd Vd
vBN vBN
Vd Vd
vCN Vd vCN Vd
v AB v AB
− Vd − Vd
Type-A sequence Type-B sequence

(starts and ends with [OOO]) (starts and ends with [PPP])

r r
V3 V2
r 30° r
V4 V1
30 °
r r
V5 V6
Space vector Diagram

VABn
v AB Vd
v AO
• Measured waveforms and FFT

V AB n / Vd
n =1
ma
( f1 = 60Hz and Ts = 1 / 720 sec )

• Five-segment SVM
r r r r r r r r r r
V0 V1 V2 V1 V0 V0 V2 V1 V2 V0
OOO POO PPO POO OOO PPP PPO POO PPO PPP
v AN Vd Vd
0
vBN Vd Vd
0
vCN Vd
0
T0 Ta Ta T0 T0 Tb Tb T0
Tb Ta
2 2 2 2 2 2 2 2
Ts Ts
(a) Sequence A (b) Sequence B

• Switching Sequence ( 5-segment)
Sector Switching Sequence (A)
r r r r r
V0 V1 V2 V1 V0
I vCN = 0
OOO POO PPO POO OOO
r r r r r
V0 V3 V2 V3 V0
II vCN = 0
OOO OPO PPO OPO OOO
r r r r r
V0 V3 V4 V3 V0
III v AN = 0
OOO OPO OPP OPO OOO
r r r r r
V0 V5 V4 V5 V0
IV v AN = 0
OOO OOP OPP OOP OOO
r r r r r
V0 V5 V6 V5 V0
V v BN = 0
OOO OOP POP OOP OOO
r r r r r
V0 V1 V6 V1 V0
VI v BN = 0
OOO POO POP POO OOO

• Simulated Waveforms ( 5-segment)
v g1 2π / 3
vg3
vg5 2π 4π
v AB
Vd
0
2π 4π
iA
0
2π 4π
• f1 = 60Hz, fsw = 600Hz, ma = 0.696, Ts = 1.1ms
• No switching for a 120° period per cycle.

• Low switching frequency but high harmonic distortion

Unit 4

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Unit 4

Uploaded by

Copyright:

Available Formats

www.Vidyarthiplus.

Two half-bridge inverters combined.

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

The three-phase fundamental and harmonics

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

In this case the isolated neutral-phase

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

The three waveforms ve0,vf0, and vg0 are of

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

Three-Phase Bridge Inverters

Note that the triplen harmonics are removed

While the 3Φ 6-step inverter offers simple

PWM inverter offers better harmonic control

PWM Principle (cont’d)

Various PWM techniques, include:

• Sinusoidal PWM (most common)

The most common PWM approach is

Sinusoidal PWM (cont’d)

when va0> vT T+ on; T- off; va0 = ½Vd

Sinusoidal PWM (cont’d)

Amplitude modulation ratio, m = Vp

Multiple Pulse-Width Modulation

• In multiple-pulse modulation, all pulses are

Generate the gating signal

2 Reference Signals, vr, -vr

Comparing the carrier and reference signals

Comparing the carrier and reference signals

Potential problem if Q1 and Q4 try to turn ON

If we prevent the problem

Output voltage is low when g1 and g4 are

This composite signal is difficult to generate

Generate the same gate pulses with one

rms output voltage

• Depends on the modulation index, M

Where δm is the width of the mth pulse

Fourier coefficients of the output voltage

Compare with multiple-pulse case for p=5

Distortion Factor is considerably less

T1 fired, resonant pulse of current

Operation in Mode 1 – Fire T1

To find the time when the current is

To find the capacitor voltage, integrate the

Operation in Mode 2 – T1, T2 Both OFF

Operation in Mode 3 – Fire T2

Space Vector Modulation

Space Vector Modulation

Space Vector Modulation

Switching state [POO] → S1, S6 and S2 ON

Space Vector Modulation

states: [PPP] and [OOO] V6 [POP] S1 , S6 , S5 V6 = Vd e 3

Space Vector Modulation