Issuu

Electronic Power Control VOLUME 2 : ELECTRONIC MOTOR CONTROL

JEAN POLLEFLIET

ACADEMIA PRESS

The Author: Jean.Pollefliet@telenet.be

Academia Press

P. Van Duyseplein 8

B - 9000 Gent (Belgium)

Info@academiapress.be

www.academiapress.be

Editor Academia Press is part of Lannoo editors, books and multimedia division of Editions Lannoo nv

Jean Pollefliet Electronic Power Control. Volume 2: Electronic Motor Control Gent, Academia Press, 2015, 410 pages. First edition 1986 Eighth edition 2015

Illustrations and lay-out: Jean Pollefliet Translation: Paul Fogarty, Rotterdam University Illustrations cover:

© Photo IBA: scanner Proteus ® PLUS with HEIDENHAIN encoders (p. 18.26) © LEM: isolated current and voltage measurement in the industry (p. 17.24) © Maxon Motor: Mars Rover with 39 Maxonmotors (p. 20.34) © Siemens: paper rolling machine (p. 19.45)

ISBN 978-90-382-252-65 D/2015/4804/111 U 2352

All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher or the author.

To my wife Gilberte

PREFACE This book first appeared in 1986 and after 29 years has reached the eighth edition. From the seventh edition the book was also available in English. Every edition saw continuous updating rearranging as well as addition of material and chapters. At the same time attention was also paid to the didactic aspects. This is not just important for students but also for the large group of people who use the book for self study. New in the eighth edition is a brief study of standing waves in transmission lines, of importance for a longue line between frequency converter and three phase motor. Also new is an introduction to the principles of 3-level inverters. In this edition we continue to use the tradition of white and green pages. The green pages contain the mathematical derivations which in the first case are not necessary for studying the electronics. Once a sufficiently high level and the desire for specialist knowledge the reader can choose to make use of the green pages without disturbing the continuity of the study. To mention a few numerical details, this book contains more than seven hundred figures, a hundred photos and more than fifty fully worked problems. The purpose of the book is to explain the principles and applications of power electronics. Electronic switches and converters are studied in volume 1 and drive technology and positioning systems are dealt with in volume 2. The largest part of this book is distilled from more than 40 years of lessons, talks and projects. The most important source of information is my students, especially the few hundred of whom I was the mentor I guided during their thesis for Master of Applied Engineering Sciences.

These I quided in wich I remain thankful and indebted to them.

To my publisher Peter Laroy of Academia Press I wish to express my thanks for many years of pleasant cooperation. Our thanks also goes out to Paul Fogarty from Rotterdam University for the accurate English translation. I would also like to thank Prof. dr. ir. Bernard Baeyens of the Ibague University (Colombia) for correcting and improving the Spanish technical vocabulary. Last but not least, we have to thank the advertisers. As a result of their support, we have been able tot minimize the recommended retail price (RRP) of our textbook.

In conclusion we wish the readers of this book a fruitful study.

Blankenberge, Belgium, September 2015

Jean.Pollefliet@telenet.be

With thanks to the advertisers:

Heidenhain LEM Maxon Motor Siemens

p. 18.26 p. 17.24 p. 20.34 p. 19.45

TABLE OF CONTENTS

16. ELECTRIC MACHINES

16.1 16.20 16.37 16.67 16.72

17. DRIVE SYSTEMS

1. Static transformers 2. DC commutator machines 3. Three-phase asynchronous motors 4. Synchronous machines 5. Small appliance motors

1. History 2. Control theory 3. Types of drive systems 4. Electronic drive technology 5. Useful mechanical formulas 6. Moment (of torque) and power of a motor 7. Run out test to determine moment of inertia of a load 8. Numeric examples

17.1 17.2 17.6 17.7 17.13 17.21 17.22 17.23

18. CURRENT-, ANGULAR POSITION-, SPEED TRANSDUCERS 1. Current sensors 2. Angular position sensors ( 2.7 Speed sensors: p. 18.26)

18.2 18.14

19. SPEED and (or) TORQUE CONTROL of a DC-MOTOR A. DC-MOTOR supplied from an AC POWER GRID 1. Control of an independently excited motor 2. M-n curves 3. Regulated single quadrant drive 4. Two quadrant and four quadrant operation 5. Functional control diagram of a single quadrant drive 6. Optimising controllers 7. Numeric example19-5

19.2 19.6 19.12 19.18 19.30 19.34 19.44

B. DC-MOTOR supplied from a DC SOURCE 8. Chopper controlled drive 9. Chopper control of a series motor 9.1 Traction service 9.2 Line filter

19.46 19.53 19.53 19.55

20. SPEED- and (or) TORQUE-control of THREE-PHASE ASYNCHRONOUS MOTOR 1. Three-phase asynchronous motor 2. Electronic control of an induction motor 3. Scalar regulation of induction motor speed 4. Slip control 5. Scalar frequency converters 6. Indirect frequency converter of the VSI-type 7. Vector control 8. Microelectronics with power electronics 9. Softstarters 10. Indirect frequency converters with a current DC link (CSI)

20.2 20.8 20.10 20.10 20.12 20.14 20.36 20.80 20.83 20.86

21. ELECTRONIC CONTROL of appliance motors, switched reluctance motor, synchronous three-phase motor, induction servo motor

1. Appliance motors Universal motor Single phase induction motor 2. Switched reluctance motor 3. Synchronous reluctance motor 4. Synchronous AC motor 5. Asynchronous induction servomotor

21.1 21.1 21.4 21.6 21.16 21.20 21.20

22. ELECTRICAL POSITIONING SYSTEMS 1. Servomechanisms 2. Electrical positioning systems. Definitions 3. Position control with DC servomotor 4. Position control with brushless DC motor 5. Position control with stepper motor 6. Position control with AC servomotor 7. Position control with linear motor 8. Computer guided motion control 9. An integrated system (SIMOTION from Siemens)

23. e-MOBILITY

VOCABULARY KEYWORD INDEX VOLUME 1: Power electronics

1. Renewable energy 2. Electrical traction 3. Electrical automobiles 4. Electrical boats 5. Electrical bicycles

. Semiconductor switches: diodes, transistors, thyristors . Electronic power converters: DC and AC controllers , choppers, SMPS, inverters . Applications of power electronics . Computer simulations

22.2 22.4 22.7 22.19 22.26 22.46 22.49 22.53 22.54

23.2 23.3 23.11 23.16 23.19

VO.1 I.1

vii

PRINCIPAL SYMBOLS α transistor current gain α firing angle thyristor (rad) β conduction angle thyristor (rad) B magnetic flux density (T = Wb/m²) AC alternating current DC direct current δ duty ratio (%) e instantaneous e.m.f. (V) E RMS-value elektromotive force (e.m.f.) (V) / DC-e.m.f. (V) E electric field intensity (V/m) E on ,Eoff   energy dissipation during transistor switching “on” and “off” respectively (J) f frequency (Hz) Φ flux per pole DC-machine / rotating air gap flux induction motor (Wb) ΦS1 flux one stator winding of an induction motor (Wb) gfs transconductance (Siemens / mho) gm transconductance coefficient (Siemens / mho) H magnetic field intensity (A/m) hFE current gain common emitter connection i / î instantaneous current (A) / peak value of sinusoidal current (A) i 0 / I0  output current of a circuit (A) iµ magnetizing current (A) îµ peak value of magnetizing current (A) I AV average value of a semiconductor current (A) I RMS / I r.m.s. value of current (A) / DC-current (A) J (polar) moment of inertia (kgm²) L b load self inductance Lo magnetizing inductance (transformer / induction motor) (H) L Laplace transform µ0 permeability of free space (4.π.10-7H/m) µr relative permeability M momentum (of torque) (Nm) Mem electromagnetic momentum( of torque) (Nm) M J accelerating or decelerating momentum (of torque) due to inertia (Nm) Mmax = M po   peak value momentum (of torque) induction motor (Nm) Mt total momentum of load torque (mechanical load ML  + static friction torque MF  + windage torques M W ...) (Nm)

viii

reluctance (A/Wb) Fµ  magnetomotive force (m.m.f.) (Aw) NSe equivalent sinusoidal (stator) winding induction motor n motor speed (r.p.m. or rad/s) nS synchronous speed (rotating stator field) induction motor (r.p.m.) nR speed rotating rotor field induction motor (r.p.m.) η efficiency of operation (%) P DC-power (W) / average power (W) Pe eddy current loss density (W/m³) Ph hysteresis loss density (W/m³) p number of pole pairs DC-machine p number of pole pairs stator winding induction motor σR.L0 leakage inductance rotor induction motor σS.L0 leakage inductance stator induction motor R b load resistance s Laplace operator T time period (s) T temperature (°C ; K) ton time to switch on a power semiconductor (switch) (µs; ns) toff time to switch off a power semiconductor (switch) (µs; ns) tON time that the power semiconductor is conducting (ON-state) (µs ; ms) tOFF time that the power semiconductor is blocking (OFF-state) (µs ; ms) td delay time (to switch a transistor on) (µs ; ns) tf fall time during switching off transistor (µs ; ns) tr rise time during switching on transistor (µs ; ns) ts storage time (to build off the space charge in a BJT) (µs ; ns) τ time constant (s) v instantaneous voltage (V) v 0 / V0  output voltage of a circuit (V) v s/V s supply voltage (V) ^        v  peak value sinusoidal voltage (V) V voltage (DC, average, ...) (V) V L/ VF  line voltage / phase voltage in a three-phase system (V) V RMS root mean square voltage (V) V di (dc-) average voltage for ideal rectifier (V) V diα (dc-) average voltage for ideal controlled rectifier with firing angle α (V) v speed (m/s) W energy (J) ω angular frequency (rad/s)

Rµ 

16.1

ELECTRIC MACHINES

ELECTRIC MACHINES 16.1 TRANSFORMERS CONTENTS 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12

Transformer at no-load Transformer with load Vector diagrams Impedance transformation Magnetizing induction Leakage inductance Energy losses Equivalent diagram No-load and short circuit test Nominal values of a transformer Three-phase transformers Types of transformers

1.1 Transformer at no load The simplest form of a single phase static transformer consists of a ferromagnetic circuit of Sisteel plates upon which two separate windings have been placed (fig. 16-1). The primary coil p has N p windings and the secondary s has N s windings. As long as no load is connected to the secondary we refer to the unloaded transformer or transformer at no-load. We now connect the voltage     ^  . sin ωt to the primary. A primary sinusoidal current flows, resulting in a sinusoidal flux in v =  v   p Vp  _____________ the core. The self inductance with respect to this flux Φ 0 is L0 .Primary current: ≈   _____________       2 2 2 Rp   + ω     . L0     laminated sheets

primary coil

√

secondary coil

Ø0

core of (Si-)steel Fig. 16-1: Single phase transformer, no load

vs = es

16.2

ELECTRIC MACHINES The primary current cannot yet be exactly determined since we first need to take the leakage reac p  V _____ tance in the transformer into account. If we neglect the resistance R p of the coil, then Iµ  ≈  ω .L     = 0 magnetizing current. This wattless current lags 90° behind V p (fig. 16-2) and creates the flux Np  . Iµ  l ______  Φ0  = _____   R      . Here R µ  =  µ µ  is the reluctance of the magnetic circuit, with l being the aver A   0 r µ age length of the field lines and A being the cross-sectional area of the core. Iµ  is calculated later when we takeR pand the leakage flux into account.

. is in phase with I   and π/2 behind v  = v  . sin ω.t → Φ = Φ   . sin (ωt − π/2) Flux Φ   → . produces iron losses (see number 1.7.2): P  ≈V  . I  ; I is in phase with V   .   I    +  I    =  I   = no-load current, lags almost π/2 behind V   . induces an emf e = N .  ddtΦ  in each coil − ) Primary: . e = N .  Φ . d sin(dt ω.t   = N  . ω . Φ     . sin ω.t . effective value: E =  N . ω2 . Φ (16-1) . E is in phase with V  and is the self induced emf of the primary.

^ p        __________ __  0    √

^ π __    0   . d ( sin ( ω.t −  2 )) Φ __________________

. e = N    . effective value: s

  ____0        ^ π __    0    2  ________________

→        →        →   µ   v   n

Secondary:

^  p

= Ns . ω . Φ    0   . sin ω.t dt    

Es =

^ NS  . ω . Φ    0   __________ __        √ 2 

A secondary emf is produced with

(16-2)

the same frequency as the applied primary

voltage. The primary emf E p is eliminated by the applied voltage V p . If we ignore the voltage losses: E p = Vp = the applied primary voltage. p  E ____  ω.L       The magnetizing current can be written as: Iµ  = (16-3) 0

Ep ω

φn

In Im

(Vp)

φn

Ø0

Fig. 16-2: Vector diagram of no-load transformer

In Im

Ø0

16.3

ELECTRIC MACHINES E  Np  __p ___ From (16-1) and (16-2) follows:  E    = transformer ratio =  N s    s V  N p __p ___ At no-load Ep ≈ Vp and Es = Vs so that V       =  N    = k = turns ratio s s

(16-4)

^ p . ω . Φ  N   0  

2 . π . f      ^ ^ __     ___    In addition: E p  =  _________  =N p .  ______   . Φ    0   = 4.44 . Np  . f . Φ    0   (16-5) √ 2    √ 2    By neglecting the losses, the flux is directly proportional to the primary voltage (assuming that the

frequency is constant).

1.2 Transformer with load 1.2.1 Secondary and primary currents When a load is connected to the secondary, then a current Is flows. The power consumed by the secondary load is drawn from the net by the primary, which means that the current Ip is larger than the no-load currentI n. With V p constant, E p ≈ Vp = constant. From (16-3) and (16-5), it follows that Iµ  and Φ 0 are practically unchanged. Constant flux means unchanged iron losses (I v) so that the current In also does not change. In other words, the secondary current Is and the primary current Ip produce the same flux Φ0 asI nat no-load (see fig. 16-3a).

→ →   I  s      I  s                                       → → → → → ___ ____ = Np   .  I  n     →   I  p     −   k      =   I  n    →   I  p    =   I  n    +   k            → From (16-6) the vectorial construction of   I  p  in fig. 16-3b follows.

→ → Np  .  I  p    − Ns  .  I  s   

(16-6)

Since with a good transformer In is quite small with respect to Ip , we find from (16-6) that        → Ip    I  s           → 1 ____ ___ __   I  p    ≈   k     or:   I    ≈  k = s

Ns  ___     

(16-7)

From (16-7) and (16-4) it follows, by approximation:

Ø1

Ø2 Np

p  N Vp  Is  __ ___ ___      =       =     =

ω Is

φ1 (Ø1 - Ø2 = Ø0)

(a)

φ2

Is/k

Fig. 16-3: Vector diagram of loaded transformer

Ø0

(b)

(16-8)

16.4

ELECTRIC MACHINES 1.2.2 Leakage flux The current Ip produces a flux Φ P  which is mostly enclosed (Φ1 ) within the core. A small part Φ pl  is not coupled with the secondary coil, so that Φ P  = Φ 1 + Φ pl  . We refer to Φ pl  as the primary leakage flux. On the secondary side the currentI sproduces a flux Φ S which for the most part (Φ 2) flows through the primary and a small component Φ sl  (leakage flux) that is not linked to the primary: Φ S= Φ 2 + Φ sl  . Np . Ip 

Ns . Is 

We therefore have: Φ P  = _____   R       in phase with Ip andΦ S= _____  R       in phase with Is .

In fig. 16-4, the instantaneous currents and voltages are drawn. Here we see that Φ 1 andΦ 2oppose each other, so that the resulting flux Φ 0 = Φ 1 - Φ2  = flux which was considered at no-load ( = no-load flux !). This resulting flux Φ 0 is practically constant for every load and includes the emf’s EP  andE Sas already shown. Ø1

ip Ø p = Ø1 + Øpl

is Np

Ø2

Øpl

Øs = Ø2 +Øsl

Ø = Ø – Ø2 ) 0 1

(

Fig. 16-4: Fluxes in the transformer

Øsl

1.3 Transformer vector diagram Primary:

N . (  Φ    )

p pl __    Leakage flux Φpl    it’s practically proportional to I p : ________    =s p. Ip  √ 2 

s p = coefficient of primary self inductance with respect to leakage flux Φpl 

. The leakage flux produces an emf in the primary coil: e  = N .   = s .   . If i =  i  . sin ω.t , then e  = s .  i  .   =s . ω .  i  . sin ( ω.t +   ) e  . is a sinusoidal emf , which leadsI  by 90° . effective value:E  = ω . s . I 

^ p

d ( sin ω.t ) ^ _________     p p  dt

^ p

d ( Φ  ) d i  ___p   p dt    dt  

pl ______

π __ 2

Summary: Primary:

1. Induced counter-EMFE Pthat leads Φ 0 by 90°

2. Induced counter-EMF Epl  = ω .sp . IP  which leads IP  by 90°

3. Voltage dropI P.R Pin phase with IP 

4. If we include the voltage lossesI P.R Pand ω . s P.I Pwe find:       →

→

  V  P  =   E  P  +   E  pl   +   I  P  . RP 

(16-9)

16.5

ELECTRIC MACHINES Secondary: 1. Flux Φ 0 produces an emf E S which leads Φ 0 by 90° 2. FluxΦ slproduces an emf E sl  = ω . ss . IS which leadsI Sby 90° 3. Voltage drop IS . RS in phase with IS        →

→

4. Terminal voltage V S is formed by:   V  S  =  E  S  − ω . ss .  I  S  −   I  S  . RS  (16-10)

With what we have considered up to now, we can create a diagram in fig. 16-5 of a loaded transformer. Øpl

Ø 0 Iv

Iµ L0

Ø sl

Rs Zb

Fig. 16-5: Transformer with losses and secondary load

With the help of (16-9) and (16-10) we now construct fig. 16-6. ω . sp . Ip ω

Is IpRp Vp

Is/k

ω . ss . Is

φp

Is/k

Es I R s s Vs φs

In Iµ

Ø0

Fig. 16-6: Vector diagram of transformer with inductive load

Remark From the figures 16-2 and 16-6 we see that the displacement angle between primary current and        →

→

voltage decreases from almost 90° (at no-load) to a value determined by  I  n  and   I  S  .

16.6

ELECTRIC MACHINES 1.4 Impedance transformation 1.4.1 Transformation formula Fig. 16-7a shows an ideal transformer, loaded with a series R-L-C circuit. An ideal transformer is a transformer without losses. Ip

R’ = k2 . RS

RS Vp

L’ = k2 . LS

C’ = CS / k2

(a) Fig. 16-7: Ideal transformer, loaded with a series R-L-C circuit: impedance transformation t dis  __ 1 __ Secondary: v s = Rs  . is  + L s.  dt   +  C        ip . dt   s

∫  0

(b)

∫

p 1 Application of (16-8) gives: vp = k  .R s.i p+k  .L s.  ___  + k 2 . ___      i p. dt

For an ideal transformer, the secondary load can be represented as an equivalent circuit seen from the primary side (fig. 16-7b), as long as: R’ = k 2 . Rs  ; L’ = k 2 . Ls  ; C’ = CS / k 2 . More generally: an ideal transformer with secondary impedance Z sec   . may be seen as a

(  ) Np 

primary impedance:

  . Z  ( Ω ) Z’ prim. = ______ sec. Ns  

(16-11)

From (16-11) it follows that we can transform a primary impedance to an equivalent secondary impedance:

(

)

s       . Z  Z’ sec. = _______ prim. ( Ω ) N    p

(16-12)

1.4.2 Numeric example 16-1: 1. An electrical oven is supplied with 46 volt and has a power of 4 kW. The supply network is 230V-50 Hz. If we had an ideal transformer available, what is then: a) the transformation ratio b) the primary and secondary current c) impedance seen from the 230 V-50Hz net?

16.7

ELECTRIC MACHINES Solution: NP 

230

a) k = ___  N   = ___  46   = 5 S

4000

b) secondary current:I S= ___  V   = ____   46     = 86.95A S

IS  86.95 primary current: IP  = __     = _____   5     = 17.39A

Transformed impedance seen from the source: Z prim.   =k 2. Zsec.   = 5   x 0.529 = 13.23Ω

Proof: Z prim.xI P= 13.23 x 17.39 = 230V !!

If maximum power transfer is required from the generator to the consumer, then the consumers impedance should be the complex conjugate value of the generator impedance. We have a power amplifier with an output resistance of 48Ω and wish to connect a loudspeaker with the following characteristics: 30 W - 4Ω . Maximum power transfer is possible by placing an impedance transformer between amplifier _____

k VS  46 c) load impedance: ZS = __  I   = _____  86.95   = 0.529 + j.0 Ω S

√

___

NP  prim. 48 and loudspeaker. The turns ratio should be k = ___  N   = ____  Z       = ___   4     = 3.46. S

√

sec.

1.5 Magnetizing inductance

NP  .I µ

With a magnetising currentI µthe magnetic field strength in the core is H = _____   l      and the k magnetic induction is B =µ 0.µ r. H so that the flux in the core of the transformer is: µ0 . µr  .A k

Φ0 = B . Ak = ________   l      . Np . Iµ 

. Φ  (Wb): no-load flux ≈ resulting flux (Φ − Φ ) with load . B (Wb/m²): magnetic induction in the core . µ  : relative permeability of core material . µ  : = 4 . π . 10  H/m . l (m): average length of field line in the core . A (m²): cross-sectional area of core . I  (A): magnetising current of the transformer. k

Whereby:

−7

If we call L 0 the self inductance of the primary with respect to the flux Φ 0 in the core, then we may µr . µ0 . Ak 

write: Np  . Φ0  = L0 . Iµ  so that: Np .Φ0 = ________   l     .  Np    .I µ= L0 . Iµ  2

from which follows:

µr . µ0  . Ak 

L0 = N p . ________   l        2

(H)

(16-13)

16.8

ELECTRIC MACHINES Numeric example 16-2: 1. A ring core transformer (fig. 16-8) consists of: average radius 60 mm; cross-section of torus 45 mm; µr =1600.

Core:

Insulation layer: 1 mm thick

Primary:

3 layers: respectively 201, 189 and 140 windings AWG 18. Each layer is separated by 1 mm thick insulation.

Insulation layer: 4 mm thick

Secondary:

two layers: 50 and 22 windings AWG 10, separated by 1 mm of insulation

2. Extract from winding wire table (AWG = American wire gauge) diameter (with insulation) in mm min. max.

AWG

resistance (per 100 m) Ω

admitted current (on base of 2A/mm²) A

2.64

2.69

0.3276

10.38

1.08

1.11

2.095

1.624

insulation

60 mm

primary

secondary

core ø 45 mm

1mm 1,11

ø 45 mm

(a)

(b)

Fig. 16-8: Ring core transformer (a) cross-section (b) core and windings

Question: 1. Resistance of primary and secondary coil 2. Magnetising inductance Solution: 1.1 Primary resistance LAYER 1: length of one winding: π x 0.04811 = 0.15114 m

total length:

1,11 1 1,11

201 x 0.15114 = 30.38 m

2,69 1mm 2,69 4mm

ELECTRIC MACHINES LAYER 2: length of one winding: total length: LAYER 3: length of one winding: total length: Total length primary winding: 86.32 m

16.9

π · 0.05233 = 0.1644 m 189 · 0.1644 = 31.07 m π · 0.05655 = 0.1776 m 140 · 0.1776 = 24.87 m

86.32

Resistance: R p  = _____   100    · 2.095 = 1.8 Ω

1.2 Secondary resistance LAYER 1: length of one winding: π · 0.06835 = 0.2147 m total length: 50 · 0.2147 = 10.73 m LAYER 2: length of one winding: π · 0.07537 = 0.238 m total length: 22 · 0.238 = 5.234 m Total length of secondary winding: 15.96 m

15.96

Resistance: Rs  = _____   100     · 0.3276 = 0.0523 Ω

2. Magnetizing inductance µ0 · µr · Ak 

−7

4 · π · 10  · 1600 · π ·0.0225 

L 0= Np    · _________   l      = 5 30 2· ________________________   2 · π    · 0.06     = 2.38 H 2

1.6 Leakage inductance From the viewpoint of voltage loss leakage inductance is undesirable. Transformers are therefore constructed to minimise the leakage fluxes. Fig. 16-9 shows, for example, how a coaxial implementation of primary and secondary coils minimises the leakage reactance by minimising the distance between consecutive coils. On the other hand, possible short circuit currents are limited by the leakage reactance, which can form a protection for the transformer. In practice distribution transformers are constructed with sufficient leakage reactance, so that short-circuit current is limited to 8 or 10 times the full load current. In electronic power supplies ring core transformers are frequently used. Due to the construction method they have a minimum leakage reactance. Electronic technicians talk about “hard” transformers since large variations in the load coupled with low leakage inductance can produce large current spikes. These varying load conditions occur for example during commutation of one rectifier element to another on the secondary side of three-phase transformers.

16.10

ELECTRIC MACHINES

Ø0

llp Øls

Ølp

Øls lls

Ø 0

secondary

Als

primary

Alp

Ø0

core

Fig. 16-9: Leakage fluxes by a coaxially wound transformer

To determine the leakage inductance, we consider the primary leakage flux (the same reasoning is valid for the secondary side). We can not make an accurate calculation since the cross-sectional area through which the flux flows can not be accurately determined. It is possible to make an approximate calculation. If the cross-sectional area where in the leakage flux flows is A lp  and the average length of the field line isl lpthen similar to expression (16-13), it may be written as: lp A   2 ___ Llp   = sP  = NP   · µ0  ·  l    

(16-14)

The field lines of the leakage flux complete their circuit through the air (µr = 1) instead of through the ferromagnetic core (µr ), which explains the difference with expression (16-13). Numeric example 16-3: We reuse the data of numeric example 16-2 ensure the possible parts of the leakage fluxes in fig. 16-10a and fig. 16-10b.

23,5 22,5

Øls Alp

50 x 2,69

Ø lp

ø 4 mm

Als

23,5 22,5

CORE ø 45

primary

28,83 32,83

Ølp Fig. 16-10a: Primary leakage flux of transformer in fig. 16-8a

Øls Fig. 16-10b: Secondary leakage flux fig. 16-8b

16.11

ELECTRIC MACHINES Primary leakage inductance −7 2 2 −6 µ0  ·Alp   4 · π · 10  · π · ( 23.5  − 22.5  ) .·10  2 _______ sp  = N p    ·  l      = 530 2·  _______________________________             = 228 µH −3 201 · 1.11 · 10 

Secondary leakage inductance −7

µ0 · Als 

−6

4 · π ·10  · π · ( 32.83  − 28.83  ) · 10 

ss  = Ns   · ______   = 72 2 ·  ________________________________         = 37.42 µH −3  l      2

50 · 2.69 · 10 

If we realise that the magnetising inductance for this transformer is 2.38 H then we see that the leakage inductance is indeed minimal. It is clear that the path of the leakage fluxes depends upon the practical implementation of the transformer windings. The present numeric example gives us a rough idea of the relative magnitude of the leakage inductance.

1.7 Energy losses 1.7.1. Copper losses In the primary and secondary windings energy losses occur. IfR Pand R S are the respective resis2 2 tances of the windings then the losses may be written as R P  . IP   andR S. IS   . The sum of both is the total energy loss. This is referred to as the copper losses of the transformer.

1.7.2 Iron losses In ferromagnetic materials, subjected to a varying magnetic field, hysteresis losses occur: n     ^

Ph = kh . f .B     

W/kg

(16-15)

whereby: k h = material constant of the ferromagnetic material used in relation to hysteresis losses. f = frequency (Hz)     ^  B  = amplitude of the magnetic induction (T = Wb/m²) n = empirical constant for the magnetic material (1 < n <3). Since the magnetic circuit of a transformer is constructed from metal plates, hysteresis losses occur. To limit these losses, it is desirable that the material constant be as small as possible. A possibility in this case is an iron alloy using silicon (e.g. 3% silicon). If the core was made from solid iron, then considerable eddy currents would occur. These can be dramatically limited by making the magnetic circuit from plates which are insulated from each other and the surface of which is in the direction of the flux. As result of this the path of the eddy currents is limited. The eddy current losses P w  can be determined with a formula in the following form:     ^2

    Pw  = kw  . δ 2.f 2. B

W/kg

(16-16)

Here in : k w = material constant with respect to the eddy current losses δ = plate thickness in mm. By adding silicon the electrical resistance is also increased as a result of which the eddy current losses are reduced. According to the last formula, it is advantageous to have the plates as thin as possible. Typical plate thickness lies between 0.3 and 1mm for 50 Hz operation. The plates can be 0.02 mm for high frequencies. For band wound cores thicknesses of 0.003 to 0.3mm are possible.

16.12

ELECTRIC MACHINES In many applications it is possible that non-sinusoidal waveforms occur. The eddy current losses VRMS  

are proportional to the square of the form factor a = ____   V      so that: AV

kw      ^2 2 Pw  = ____   2    · a  ·δ 2·f 2· B      1.11 

(16-17)

1.11 = form factor of sinusoidal voltage a = form factor of the actual voltage. Hysteresis and eddy current losses form the iron losses. They are sometimes called the constant losses of the transformer since they do not depend upon the load but only the magnetic induction. The magnetic induction only depends upon the applied voltage. The following table provides an idea of the iron losses with plates between 0.2 and 0.5 mm thick, and a frequency of 50 Hz with an induction of 1 Tesla.

Material

Losses in W/kg

commercial iron

5 ... 10

Si-Fe, warm rolled

1 ... 3

Si-Fe, cold rolled and crystal orientated

0.3 ... 0.6

50% Ni-Fe

0.2

approximately 65% Ni-Fe

0.06

Fig. 16-11 shows the total iron losses at 50 Hz for toroidal band wound cores of 0.3mm (data for cold rolled 3% Si-Fe cores). In fig. 16-12, we see the influence of the frequency on the total iron losses for the same material. Such cores are used for power transformers, impulse transformers, welding transformers, line transformers, etc... .

Fe-losses

(W/kg)

f = 50Hz

5kHz

2kHz 400Hz

0,36 0,1

0,1

0,01

0,87

0,001 0,01

0,1

Fig. 16-11: Iron losses as a function of induction

(T)

0,001 0,01

0,1

(T)

Fig. 16-12: Iron losses with the frequency as a parameter