Pragmatic Electrical Engineering: Fundamentals

M &C
Mor gan
&Cl aypool
Publishers
Pragmatic Electrical Engineering

Fundamentals
William J. Eccles
SYNTHESIS LECTURES ON DIGITAL CIRCUITS AND SYSTEMS

Mitchell Thornton, Series Editor
Pragmatic Electrical Engineering: Fundamentals
Copyright 2011 by Morgan & Claypool
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any meanselectronic, mechanical, photocopy, recording, or any other except for brief quotations in printed reviews, without the prior permission of the publisher.
Pragmatic Electrical Engineering: Fundamentals William Eccles www.morganclaypool.com
ISBN: 9781608456680 ISBN: 9781608456697
paperback ebook
DOI 10.2200/S00242ED1V01Y201105DCS031
A Publication in the Morgan & Claypool Publishers series SYNTHESIS LECTURES ON DIGITAL CIRCUITS AND SYSTEMS Lecture #31 Series Editor: Mitchell A. Thornton, Southern Methodist University Series ISSN Synthesis Lectures on Digital Circuits and Systems Print 1932-3166 Electronic 1932-3174
Synthesis Lectures on Digital Circuits and Systems

Editor
Mitchell A. Thornton, Southern Methodist University The Synthesis Lectures on Digital Circuits and Systems series is comprised of 50- to 100-page books targeted for audience members with a wide-ranging background. The Lectures include topics that are of interest to students, professionals, and researchers in the area of design and analysis of digital circuits and systems. Each Lecture is self-contained and focuses on the background information required to understand the subject matter and practical case studies that illustrate applications. The format of a Lecture is structured such that each will be devoted to a specic topic in digital circuits and systems rather than a larger overview of several topics such as that found in a comprehensive handbook. The Lectures cover both well-established areas as well as newly developed or emerging material in digital circuits and systems design and analysis.

William Eccles 2011
Introduction to Embedded Systems: Using ANSI C and the Arduino Development Environment
David J. Russell 2010
Arduino Microcontroller: Processing for Everyone! Part II

Steven F. Barrett 2010
Arduino Microcontroller Processing for Everyone! Part I

Digital System Verication: A Combined Formal Methods and Simulation Framework

Lun Li and Mitchell A. Thornton 2010
Progress in Applications of Boolean Functions

Tsutomu Sasao and Jon T. Butler 2009
iv
Embedded Systems Design with the Atmel AVR Microcontroller: Part II

Embedded Systems Design with the Atmel AVR Microcontroller: Part I

Embedded Systems Interfacing for Engineers using the Freescale HCS08 Microcontroller II: Digital and Analog Hardware Interfacing
Douglas H. Summerville 2009
Designing Asynchronous Circuits using NULL Convention Logic (NCL)

Scott C. Smith and Jia Di 2009
Embedded Systems Interfacing for Engineers using the Freescale HCS08 Microcontroller I: Assembly Language Programming
Douglas H.Summerville 2009
Developing Embedded Software using DaVinci & OMAP Technology

B.I. (Raj) Pawate 2009
Mismatch and Noise in Modern IC Processes

Andrew Marshall 2009
Asynchronous Sequential Machine Design and Analysis: A Comprehensive Development of the Design and Analysis of Clock-Independent State Machines and Systems
Richard F. Tinder 2009
An Introduction to Logic Circuit Testing

Parag K. Lala 2008
Pragmatic Power
William J. Eccles 2008
Multiple Valued Logic: Concepts and Representations

D. Michael Miller and Mitchell A. Thornton 2007
Finite State Machine Datapath Design, Optimization, and Implementation

Justin Davis and Robert Reese 2007
Atmel AVR Microcontroller Primer: Programming and Interfacing

Steven F. Barrett and Daniel J. Pack 2007
Pragmatic Logic
PSpice for Filters and Transmission Lines

Paul Tobin 2007
PSpice for Digital Signal Processing

Paul Tobin 2007
PSpice for Analog Communications Engineering

Paul Tobin 2007
PSpice for Digital Communications Engineering

Paul Tobin 2007
PSpice for Circuit Theory and Electronic Devices

Paul Tobin 2007
Pragmatic Circuits: DC and Time Domain

Pragmatic Circuits: Frequency Domain

Pragmatic Circuits: Signals and Filters

High-Speed Digital System Design

Justin Davis 2006
vi
Introduction to Logic Synthesis using Verilog HDL

Robert B.Reese and Mitchell A.Thornton 2006
Microcontrollers Fundamentals for Engineers and Scientists

Steven F. Barrett and Daniel J. Pack 2006

William Eccles
Rose-Hulman Institute of Technology
SYNTHESIS LECTURES ON DIGITAL CIRCUITS AND SYSTEMS #31
M &C
Morgan
& cLaypool publishers
ABSTRACT
Pragmatic Electrical Engineering: Fundamentals introduces the fundamentals of the energy-delivery part of electrical systems. It begins with a study of basic electrical circuits and then focuses on electrical power. Three-phase power systems, transformers, induction motors, and magnetics are the major topics. All of the material in the text is illustrated with completely-worked examples to guide the student to a better understanding of the topics. This short lecture book will be of use at any level of engineering, not just electrical. Its goal is to provide the practicing engineer with a practical, applied look at the energy side of electrical systems. The authors pragmatic and applied style gives a unique and helpful non-idealistic, practical, opinionated introduction to the topic.
KEYWORDS
electrical engineering, power systems, three-phase, transformer, induction motor, magnetics
ix
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv
Basic Stuff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Fundamentals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1.1 Charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1.2 Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1.3 Electrical element . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.1.4 Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.1.5 Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.1.6 Passive sign convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Electrical elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.2.1 Resistor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.2.2 Capacitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.2.3 Inductor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.2.4 Active elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Kirchhoff s laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.3.1 Kirchhoff s Current Law (KCL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.3.2 Kirchhoff s Voltage Law (KVL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Combining elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.4.1 Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.4.2 Parallel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.4.3 Voltage divider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.4.4 Current divider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 1.4.5 Example IVoltage Divider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 1.4.6 Example IIAnother Voltage Divider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 1.4.7 Example IIISolving Using a Voltage Divider . . . . . . . . . . . . . . . . . . . . . . 18 Nodal Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 1.5.1 Example IVNodal Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1.5.2 Steps for Nodal Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 1.5.3 Example VMessier Nodal Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 1.5.4 Example VITwo Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 1.5.5 Example VIIDependent Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.5.6 Example VIIIDeceptively Simple . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Op-amp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6.1 Ideal Op-amp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6.2 Example IXOp-amp Inverter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6.3 Example XOp-amp Non-inverter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6.4 Example XIOp-amp Voltage Follower . . . . . . . . . . . . . . . . . . . . . . . . . . . . Linear implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.7.1 Superposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.7.2 Thvenins Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Formulas and Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
27 28 29 29 31 33 34 34 36 39 41
Power of the Sine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

2.1 Sinusoids and Phasors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1 Differential Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2 Phasors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.3 Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.4 Phasor Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.5 Complex Arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.6 Example ISinusoidal Steady State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.7 Example IIAnother Phasor Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-c Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 A-c Power Quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Example IIIAdding Powers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3 Example IVMotor Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.4 Standard Power Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.5 Example VA Standard Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.6 Example VIAdding S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . More Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Example VIINodal Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2 Example VIIIVII Simplied . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3 Example IXVII Simplied Even More . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.4 Example XPhasor Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.5 Example XITwo Different Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.6 Example XIIA-c Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Formulas and Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 44 46 47 49 51 53 55 56 57 59 61 63 64 66 66 67 68 68 69 70 72 73 75
2.2
2.3
2.4
xi
Three-Phase Power Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

3.1 3.2 Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Three-phase Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Three Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2 Conductor Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.3 Voltages and CurrentsTerminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.4 Y-Y Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.5 Y- Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Power in Three-phase Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 P and Q and S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 Example IP, Q, and S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.3 Example IILine Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Single-phase Equivalent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.1 Generating the Single-phase Equivalent . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.2 Example IIIThree to One . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . More Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.1 Example IV -connected Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.2 Example VY-connected Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.3 Example VI -connected Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.4 Example VIIThree-phase Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.5 Example VIIIPower-factor Correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Formulas and Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 78 79 81 83 83 84 85 85 86 86 87 87 88 90 90 92 93 94 95 95 97
3.3
3.4
3.5
3.6
Transformers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
4.1 Ideal Transformers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 4.1.1 Voltage and Current Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 4.1.2 Power Relationship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 4.1.3 Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 4.1.4 Example ITransformer Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 4.1.5 Example IIApplication Reected . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Real Transformers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 4.2.1 Realistic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 4.2.2 Model with Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 4.2.3 Example IIITransformer in Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
4.2
4.3
xii
4.4
4.5
4.3.1 Open-circuit Test: Rated Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Short-circuit Test: Rated Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.3 Example IVTransformer Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . More Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.1 Example VLoad at Unity Power Factor . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.2 Example VILoad at Lower Power Factor . . . . . . . . . . . . . . . . . . . . . . . . 4.4.3 Example VIIParallel-branch Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.4 Peculiar Transformer? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.5 Example VIIITesting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.6 Example IXTransformer Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.7 Example XTested Transformer in Use . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Formulas and Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
112 112 113 115 115 116 117 118 118 120 121 122 125
Machines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.1 5.2 How they Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Induction Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.1 Motor Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.2 Power Flow Through Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.3 Example IMotor Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.4 Example IIValue of Efciency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Induction Motor Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 Model of the Stator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2 Model of the Rotor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3 Complete Induction Motor Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.4 Example IIIUsing the Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Motor Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1 D-C Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.2 Blocked-rotor Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.3 No-load Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.4 Collecting and Comparing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Choosing a Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.1 Motor Nameplate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.2 Torque versus Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.3 Overmotoring is BAD! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Matching a Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . More Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 132 133 134 135 136 136 136 137 139 140 142 143 143 144 145 145 146 148 149 151 153
5.3
5.4
5.5
5.6 5.7
xiii
5.8
5.7.1 Example IVMotor Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.7.2 Power Factor Correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.7.3 Circuit Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Formulas and Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
153 154 155 157 159
Electromagnetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
6.1 Amperes laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1 Amperes Force Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.2 Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.3 Example IForce between busbars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnetic elds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.1 Magnetic ux density B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.2 Magnetic ux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.3 Magnetic eld intensity H . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.4 Magnetomotive force F . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.5 Connecting all this . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reluctance and inductance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3.1 Reluctance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3.2 Inductance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3.3 Example IIReluctance of a core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3.4 Example IIIInductance for Example II . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3.5 Example IVCurrent for Example III . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3.6 Air gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3.7 Example VAir gap for Example II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Force in the airgap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.1 Force in energy terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.2 Example VIForce in gap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . More example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.1 Example VIIForce in a conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.2 Example VIIIActuator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Formulas and Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 162 163 163 165 165 166 166 167 167 168 168 168 169 170 171 171 172 173 173 175 175 175 178 179 181
6.2
6.3
6.4
6.5
6.6
Authors Biography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Preface
Pragmatic Electrical Engineering: Fundamentals is a collection of introductions to a number of topics from electrical engineering, primarily from the energy side of the eld, topics that you might nd useful in your own work. My goal is to prepare you a little better to deal with electrical questions. That breaks down into three parts: Ability to handle basic electrical problems Knowledge of where your boundaries are and when to get help from other folks like consultants Comprehension so you can listen to what consultants tell you and understand what they are saying. This text is divided into two major topics. The rst two chapters are a review of circuits and the sinusoidal steady state, topics that are basic to understanding electrical circuits and how they are used in the supply of electrical energy. For some folks, this will be a review of the material learned in an EE course, perhaps as part of another major. For some, it will be all new. The last four chapters are about electrical systems that move energy from place to place. We begin with three-phase systems, the bulk movers of energy. Transformers come next because they play a major role in those systems. Next is the induction motor, the most common machine for converting electrical into mechanical energy. We nish with magnetics and in particular, magnets that do mechanical work. A companion text, Pragmatic Electrical Engineering: Systems and Instruments, is an introduction to the use of electricity to do the non-power things. It includes the characterization of physical systems, the basics of closed-loop control systems, simple lters, noise interference, and instrumentation. You may wish to look at that material after you have a rm grasp of the rst two chapters of this Fundamentals text. Pragmatic in the title of these texts has all of the meanings of that word, making these texts a non-idealistic, practical, opinionated look at some topics in electrical engineering. The philosophy folks say that pragmatism is an approach to understanding that starts by nding meaning in the practical side of things. They go on to say that thought can guide our actions and that truth should be tested against the practical consequences of an action. This text is not intended to make you into an electrical engineer, but I hope that youll have a better understanding of the use of electricity in your chosen eld. I think youll nd that you cant get away from some aspect of electricity in your own work, no matter how hard you try! Wm. J. Eccles April 2011
CHAPTER
Basic Stuff
How much physics do you remember fromwhenever? Whether a lot or not much, this chapter starts out with the assumption that not much of that stuff is stuck in your head. But that stuff is important to us here because it forms the basis for all of the circuit work that well be doing. So this chapter may be partly a review for you.
1.1
1.1.1
FUNDAMENTALS
CHARGE
The basic little doojigger that makes electricity work is charge. Hows that for a sloppy introduction? But the statement is true. Charge and the motion of charge are behind this whole electricity show. So it seems appropriate that we should review a little about charge rst. Charge is discrete, which doesnt mean it refrains from repeating rumors. The smallest unit of charge is the electron, which is 1.6 1019 coulombs in size. Charge is measured in coulombs, which, since Coulomb was a person, uses the abbreviation C . Why negative? Somebody, long ago, had to pick directions, Ben Franklin maybe. What somebody picked established the directions of plus and minus and weve followed those choices ever since. But no matter, the electron is negative. The charge on an electron is a very small amount of charge. When we do just about anything useful with charges, we generally involve billions and billions of the little buggers. An exception is the memory in your computer, where the technology uses the presence or absence of a small number of electrons in a memory cell to indicate a binary 1 or 0. Small number is only a few hundred electrons. All of our study of electrical circuits concerns just two things we can do with charges: move them and energize them. Moving them is electrical current. If we push on them, thereby imparting energy to them, we can make them do work for us somewhere down the line (thats a pun!). We give them potential energy, just as we can elevate a brick to give it potential energy. That energy can be returned to us in the form of work later on. Thats enough on charge itself. Its there, we can make it move, we can give it potential energy, and we can make it do work for us. We are more interested, at least when we are talking about circuits, in the motion of charge as currents and their potential for doing work.
1.1.2
CURRENT
Current is charges in motion. The relationship is very simple:
1. BASIC STUFF
i= q(t ) =
dq dt
i( x ) dx + q(0)
0
Current is charge per unit time. Current is measured in amperes, abbreviation A. Positive charge moving from, say, left to right in a conductor is considered positive current. That last statement brings up an additional thought. It says positive charge even though we know that the basic unit of charge is negative. Thats the accident of history as I mentioned before. Ol Ben or somebody had a chance to get it right, but the jelly bread landed on the jelly side. So our convention is that we ignore the electron and consider our charges to be positive.
1.1.3
ELECTRICAL ELEMENT
Lets dene the basic electrical element. Its drawn in Fig. 1.1 and represents a general two-terminal device. It could be anything from a resistor to a refrigerator.The arrow indicates the positive direction of the current. The current that enters the left terminal must come back out of the element via the right terminalthe law of conservation of charge requires this.
Figure 1.1: Electrical element with current labeled
The charges that go in dont have to come back in the same condition, however. We can make those charges do work for us while passing through the electrical element. That work can be as simple as heating the resistor or something as complex as running the motor that cools the frig. When these charges do that work, they come out with less potential energy than they entered with. But our electrical element does not have to be an element that takes work from the charges that ow through it. The element can instead do work on those charges to increase their potential energy. These electrical elements are devices like batteries and generators and solar cells. An analogy to all of this is the wrecking ball, which could represent one unit of charge. The crane raises the ball, which increases the potential energy of the ball. The crane releases the ball, which descends to shatter the concrete below. The ball, upon smashing the concrete, perhaps has not given up all of its potential energy, though. It could, for example, roll off the edge of the concrete and squash a Jag below. Similarly, the charge passing through an electrical element does not necessarily give up all of its potential energy.
1.1. FUNDAMENTALS
1.1.4
VOLTAGE
Voltage is a measure of the potential energy of a charge, the energy available to do work. In fact, we sometimes use potential as a synonym for voltage. In a way, this is a messier concept than current. After all, we can visualize those little charges chugging along in a conductor as if they were water molecules moving through a pipe. This voltage thing is a little harder to visualize. Consider water molecules1 . With a pump, we push harder on them, thereby increasing their potential energy. We measure this potential to do work not in energy units but in pressure units. Likewise, we measure the potential of charges to do work in voltage units. Suppose we have one coulomb of charge that passes through an electrical element and loses one volt of potential as it passes through. That coulomb of charge does one joule of work as it passes through the electrical element. Note the directions in Fig. 1.2. The current arrow, which shows the direction of movement of positive charge, is entering the positive terminal of the element.
Figure 1.2: Electrical element with voltage labeled
A volt is a joule per coulomb. Its abbreviation is V . A typical ashlight runs on 3 V, your computer on 3 to 5 V internally, and your frig on 120 V.
1.1.5
POWER
Power is the rate of doing work. Weve been talking about how charge moves (as current) and loses potential (as voltage) while doing work on an electrical element. Now we need a measure of this. Power is measured in watts, abbreviation W . Consider the following relationships:
watt = joule / second joule = coulomb volt coulomb = ampere second coulomb volt ampere second volt so watt = = second second = volt ampere
So the watt is a volt-ampere. But these dont make clear the signs and directions of these quantities. Well do that with the passive sign convention.
1The water analogy is really pretty poor, since water molecules that go in one end come out the other, while charges hand off
their energy after a very short movement.
1. BASIC STUFF
1.1.6
PASSIVE SIGN CONVENTION
Take a look at Fig. 1.2.The arrow gives the direction of positive current; the plus sign gives the upper end of positive voltage.The current arrow is entering the positive terminal.This arrangement is called the passive sign convention. Why passive? Its because we dene this arrangement as the arrangement where the moving charges give up energy to the element. To say this another way, the charges do work on the element. To say this still another way, the charges deliver power to the element. In other words, (youll get tired of this in a moment!), the element absorbs power and is therefore passive. Now, remember that power is voltage times current and then note that each element in Fig. 1.3 is absorbing power. While the arrows and signs are in various arrangements, if youll pay attention to the signs of voltage and current, power is positive in each case and the element is absorbing power.
Figure 1.3: Absorbing power
Figure 1.4 shows the opposite situationeach of the elements is delivering power via its terminals because the power, the product of voltage and current, is in each case negative. Notice that I have still followed the passive sign convention with my labeling. We are going to consistently follow the passive sign convention except for sources, which we presume to be active. Figure 1.5 shows several elements that are acting as sources, adding energy to the charges and delivering power. Note that the current arrow is leaving the positive terminal.
1.2
ELECTRICAL ELEMENTS
Lots of things can be electrical elements. But we are going to restrict our attention to models of just ve elements. Three are passive in that they can not deliver power on the average. Two are active,
1.2. ELECTRICAL ELEMENTS
Figure 1.4: Delivering power
Figure 1.5: Delivering power
meaning that they generally provide power. The three passive elements are the resistor, the capacitor, and the inductor. The two active elements are the voltage source and the current source, although the current source will turn out to be fairly uncommon.
1. BASIC STUFF
1.2.1
RESISTOR
Figure 1.6 shows the resistor. Since its a passive element, I have followed the passive sign convention in my labeling. The resistors electrical parameter, resistance, is measured in ohms, abbreviated (one of the few electrical quantities that uses a Greek letter). Ohms Law gives the relationship between voltage and current.
R= v i v = Ri
Figure 1.6: Resistor
The resistor always absorbs power, so

p = vi
which can be rearranged using Ohms Law into

p = Ri 2 = v2 R
1.2.2
CAPACITOR
Figure 1.7 shows the capacitor, also a passive element. But this element is different from a resistor
Figure 1.7: Capacitor
it stores charge and hence energy. You can do work on the capacitor by pushing charge into it, thereby increasing the stored energy. That stored energy can be recovered from the element later. The capacitance itself is measured in farads, with the abbreviation F The v-i relationship for the capacitor is
i=C dv dt
1.2. ELECTRICAL ELEMENTS
If we calculate power using this relationship, we get

p = vi = Cv dv dt
If we integrate power with respect to time to get energy, we nd that the energy stored in a capacitor depends on the square of the voltage:
w=
p dt = Cv dt dt = 2 Cv
dv
(The constant of integration is chosen to be 0, representing the fact that, when the capacitor was manufactured, it had zero stored energy.)
1.2.3
INDUCTOR
Figure 1.8 shows the inductor, the third passive element. Like the capacitor, it stores energy, but
Figure 1.8: Inductor
in the form of magnetic energy. Like the capacitor, you can do work on the inductor by pushing current through it, thereby increasing the stored energy. That stored energy can be recovered later. The inductance itself is measured in henries, abbreviated H . The v-i relationship for the inductor is
v=L di dt
If we calculate power using this and then integrate with respect to time to get energy, we nd that energy stored in an inductor depends on the square of the current.
p = vi = Li w= di dt di 1
2
p dt = Li dt dt = 2 Li
1.2.4
ACTIVE ELEMENTS
The active elements are sources of energy. Figure 1.9 shows the two active elements.
1. BASIC STUFF
Figure 1.9: Ideal voltage and current sources
The voltage source provides the stated voltage under every possible condition. In other words, the voltage between the terminals is the value given no matter what happens. This means that the voltage is as stated for any current. Hmmm, since power is voltage times current, this implies an element that can provide unlimited power! We know thats not possible, so we call this source ideal. We cant really have such a source in a circuit, but we usually have a close approximation to it up to some practical limit. Now reread the previous paragraph but change voltage to current and youll have the description of the current source. The current source is less common than the voltage source, but we will see it in our circuits from time to time. There is another form of sources that is useful in, for example, modeling devices like transistors. This form, called a dependent source, uses a diamond instead of a circle symbol. The value of the source (such as Vs in Fig. 1.9) is stated as a dependency such as kix where ix is a current elsewhere in the circuit. Well see these in examples later on.
1.3
KIRCHHOFFS LAWS
If those doojiggers called charges are running around making current, then Kirchhoff is the lawyer, judge, and jury to keep their behavior under control. Kirchhoff s laws tell us the relations among voltages and among currents in a circuit. These two laws are absolutely basic to how things to work in circuits. Note that I said we work with circuits. This is because there is a basic assumption that we must make if we expect Kirchhoff s Laws to hold. This assumption is that our circuits are built with lumped elements. This means that all the properties of a circuit element are concentrated at that elementthey are not spread out over much space. In other words, the resistance property of a resistor that leads to v = Ri is concentrated at a point. If the resistance of the connecting wires is to be considered, that resistance is lumped into the resistor. The underlying physics can best be discussed by looking at the electric and magnetic elds involved, but thats beyond where I want to be doing here. For what we are doing, it is sufcient to
1.3. KIRCHHOFFS LAWS
say that charge never accumulates anywhere in the circuit and that no time-varying magnetic elds link into any loop in the circuit. A corollary to this basic assumption is that, if the frequency of the electrical signal is high enough, our circuit elements no longer can be expected to look like points. That means our circuit elements must be small relative to the wavelength of the signal. OK, enough of that! Assume we have met the requirements of the basic assumption and lets go on with what Kirchhoff discovered. (And got credit forothers working with electricity at the same time probably also found the same facts. The moral is to be the rst to publish!)
1.3.1
KIRCHHOFFS CURRENT LAW (KCL)
Charge doesnt pile up anywhere in a circuit, which requires that a current entering a node (or any closed surface) must be balanced by an equal current leaving. This, with that basic assumption, is conservation of charge: The algebraic sum of the currents leaving any node is zero at every instant of time. Lets look at a several parts of this: Algebraic sum = pay attention to signs as you add the currents Leaving any node = be consistent and choose all currents as leaving so that you dont have to remember which way each one is going Node is the junction of two or more electrical elements Every instant of time = no delay from one place to another in the circuit Figure 1.10 shows several conductors meeting at a node. Kirchhoff s Current Law tells us how to gure out the unknown current. In this equation I am taking leaving currents as positive, which means that I must use a minus sign on two of the currents:
4 + 2 5 ix = 0 ix = 1 A
1.3.2
KIRCHHOFFS VOLTAGE LAW (KVL)
Voltage is potential, the potential to do work. This law is based on conservation of energy. The law says that if you add these potentials around a loop, you have to come up with zero. In other words, you cant gain or lose energy by going around the complete loop:
10
1. BASIC STUFF
Figure 1.10: Currents and KCL
The algebraic sum of the voltages around any closed loop is zero at every instant of time. Notice that the words here are almost the same as for Kirchhoff s Current Law. Note also that any closed loop will do, which says we dont have to follow the actual paths of a circuit. Figure 1.11 is a circuit with a couple of loops. Kirchhoff s Voltage Law tells us how to nd
Figure 1.11: Voltages and KVL
the unknown voltage vx . Ill add the voltages around the left-hand loop by following two simple conventions. First, Ill always go clockwise. Second, Ill use the rst sign I come to as the sign of that voltage. The result is
15 + 8 + vx = 0 vx = 7 V
1.4. COMBINING ELEMENTS
11
1.4
COMBINING ELEMENTS
Well introduce in the next section a formal technique for circuit analysis, nodal analysis, but that technique is usually overkill for circuits we encounter in real life. Many circuits are a simple combination of just a few circuit elements, making the analysis very straightforward. Four such combinations are pretty fundamental: elements in series, elements in parallel, a circuit called a voltage divider, and a similar circuit called a current divider. Im going to walk through the analysis of these four using just resistors and Kirchhoff s two laws.
1.4.1
SERIES
Series means that the elements have the same current through them. This is akin to screwing two pipes together so a stream of water ows through both, one after the other. Elements are in series if the same current ows through both. They are not in series if some of the current leaks out of the path part way along. Figure 1.12 shows resistors in series.
Figure 1.12: Resistors in series
What I just said about current follows directly from KCL. Theres a node (i.e., a junction) between two resistors and KCL says that the current coming in from the left must equal the current going out to the right. If I write Ohms Law for each resistor in Fig. 1.12 and then combine these using KVL, the result describes resistors in series:
v1 = R1i v2 = R2i Combining, v = v1 + v2 = R1i + R2i = ( R1 + R2 ) i so Rseries = R1 + R2
The outcome is simple: Resistors in series add. Figure 1.13 shows series combinations. Beware of resistor combinations where current leaks out. Figure 1.14 shows some that are not in series. In the top circuit, no resistors are in series. In the other circuit, R11 and R12 are in series, as are R14 and R15 . For each of these pairs theres a common current through each member of the pair.
12
1. BASIC STUFF
Figure 1.13: Series
Figure 1.14: Series?
1.4.2
PARALLEL
Parallel means that the element have the same voltage across them. To say this another way, the ends of the elements must meet at two common points. Figure 1.15 shows elements in parallel. Just as I did for series resistors, Ill use KVL to show that the voltages are the same. Then Ill write Ohms Law for each and use KCL to combine the currents.
by KVL, v1 + v2 = 0 so v1 = v2 = v
13
Figure 1.15: Resistors in parallel
OK. We know that both of the resistors have the same voltage v across them. Now Ill nd the current through each of the resistors and then add those two currents:
v v , i2 = R1 R2 v v Combining, i = i1 + i2 = + R1 R2 i1 = 1 1 = v + R1 R2
The result of that is a sort of reverse Ohms Law where i is in terms of v . Turn this equation around and well see the equivalent resistance:
v= 1 i 1 1 + R1 R2 1 = 1 1 + R1 R2
so Rparallel
The result: Resistors in parallel add reciprocally. Figure 1.16 shows parallel combinations. Beware of resistor combinations where the resistors ends are not truly connected to common nodes. Figure 1.17 shows some that are not in parallel. In the upper circuit of Fig. 1.17, R7 is in parallel with the series combination of R8 and R9 . Weve now seen how resistors combine in series and in parallel. Inductors follow the same rules as resistors. Capacitors are opposite (i.e., they add in parallel and add reciprocally in series).
1.4.3
VOLTAGE DIVIDER
The easiest way to describe a voltage divider is to draw it in Fig. 1.18. In words, the voltage vo is a fraction of the source voltage vs . Now lets use Kirchhoff to see what that fraction is. KCL tells us without much thought (I hope!) that the current i is the same current all the way around through all three elements.
14
1. BASIC STUFF
Figure 1.16: Parallel
Figure 1.17: Parallel?
I use KVL to sum the voltages clockwise around the loop and then write the resistor voltages in terms of current using Ohms Law:
vs + v1 + vo = 0 vs + R1i + Roi = 0 Solving, i = vs R1 + Ro
15
Figure 1.18: Voltage divider
From this current I can get the output voltage vo in terms of the input vs :
vo = Roi = vs Ro Ro + R1
The result is that the output voltage vo is a fraction of the source voltage vs , and that fraction is the output resistor Ro divided by the sum of the two resistors Ro + R1 .
1.4.4
CURRENT DIVIDER
The current divider is the voltage divider sort of in reverse. In Fig. 1.19, the output current io is a fraction of the source current is . Kirchhoff will get us that fraction.
Figure 1.19: Current divider
First, I use KVL to show that the two labeled voltages are equal:
v1 + vo = 0 so v1 = vo = v
Next I use this voltage and KCL to relate the voltage to the current is at the node at the top:
is + i1 + io = 0 is + v v + =0 R1 Ro
16
1. BASIC STUFF
I need v in terms of the current and the resistors:

1 1 v + = is R1 Ro Solving, v = is 1 1 + R1 Ro
Finally, I get the value of the output current io in terms of the input current is and the resistors:
io = v 1 Ro = is Ro 1 R1 + 1 Ro R1 = is R1 + Ro
The result is that the output current io is a fraction of the source current is , and that fraction is the other resistor R1 divided by the sum of the two resistors R1 + R0 .
1.4.5
EXAMPLE IVOLTAGE DIVIDER
Figure 1.20 is a voltage divider with several resistors. Our goal is to nd the output voltage vo . The
Figure 1.20: Example I
circuit looks like a voltage divider, but there are two resistors along the top. We should recognize those resistors as being in series, though, so when we combine them, we have just one equivalent resistor at the top. The series combination is
Rtop = 40 + 50 = 90
This voltage divider now has a 60- resistor as the output and a 90the output voltage vo is a fraction of the 6-V input:
vo = 6 60 60 =6 = 2.4 V 60 + Rtop 60 + 90
resistor at the top. So
The result is that vo = 2.4 V .
17
1.4.6
EXAMPLE IIANOTHER VOLTAGE DIVIDER
In Example II in Fig. 1.21, we are to nd the output current ix that ows downward through the 3-k resistor. Lets make this into a voltage-divider problem by labeling the voltage vx as the output voltage as shown in Fig. 1.22.
Figure 1.21: Example II
Figure 1.22: Example II with voltage vx
If Im going to make this into a voltage divider, I need to combine the two output resistors in parallel:
Rright = 2 3 = 1 = 1.2 k 1 1 + 2 3
of 600
Now the circuit is a voltage divider with an output resistor of 1.2 k . The circuit is dividing 27 volts to yield the output voltage vx :
vx = 27 Rright 600 + Rright = 18 V
and a resistor at the top
The voltage vx is directly across both output resistors as shown in Fig. 1.22, so I can nd the current through the 3-k resistor by using Ohms Law:
ix = vx 18 = = 0.006 A = 6 mA 3000 3000
18
1. BASIC STUFF
The output current is 6 mA. Notice that in both of these examples the circuit simplies to a voltage divider and no fancy formal analysis is needed. The following example is again a voltage divider, or really, two voltage dividers in cascade.
1.4.7
EXAMPLE IIISOLVING USING A VOLTAGE DIVIDER
Figure 1.23 shows a more complicated circuit, but it is still one we can solve using just our knowledge of series, parallel, and voltage dividers. We need to be careful, though. It is very wrong to conclude
Figure 1.23: Example III
that the 200:600 ohm pair of resistors is a voltage divider. Why? Because this is a voltage divider with a leak. That means there is current owing from the divider at the junction of the 200- and the 600- resistors. Voltage dividers must not leak. So how can we do this using a voltage divider? First, we need to combine some resistors. The 600- resistor is in parallel with the series combination of the 800- and 400- resistors:
R1 = 600 ( 800 + 400 ) = 1 = 400 1 1 + 600 800 + 400
The resulting circuit is in Fig. 1.24. Notice the voltage labeled v1 . I can use the voltage-divider
Figure 1.24: Example III simplied
relationship to get this value:
1.5. NODAL ANALYSIS
19
v1 = 36
R1 = 24 V R1 + 200
The nal step is to recognize that v1 is the voltage across the 600- resistor (see Fig. 1.25) and it is the voltage feeding the two resistors to the right. Those two resistors are a voltage divider
Figure 1.25: Example IIIsecond divider
that is dividing v1 :
vo = v1 400 =8 V 400 + 800
This second divider doesnt leak. Moreover, it doesnt cause the rst divider to leak because the rst divider was analyzed by including the resistors of the second one. Weve found that vo = 8 V without formal analysis. But now its time to learn a technique for formal analysis.
1.5
NODAL ANALYSIS
Nodal analysis is a systematic way of selecting certain voltages to be the independent variables and then writing equations for them. The key is systematic. Its not hard to choose the right voltages. Once we have those, writing the equations follows a straightforward algorithm. Before we start, Id like you to agree to apply this systematic method in a systematic way. That really means just following the proper steps and not thinking! Youll do best if youll just turn the crank. Now, how about a lesson in topology? I dont think we need to go so far as to argue about how many pieces of donut you get when you make one cut from the hole outward. But a little understanding of branches and trees is probably helpful. Consider the stylized circuit of Fig. 1.26. Two of its characteristics are very important to us, namely, nodes and tree branches. Lets take up nodes rst, since understanding branches needs the node concept. The circuit in Fig. 1.26 has just four nodes. A node is the junction of elements. While the nodes at the top of the circuit are easy to count (1-2-3!), the bottom one can get us into trouble. The whole bottom line is just one node. Three elements join at this node.
20
1. BASIC STUFF
Figure 1.26: Stylized circuit
Remember that dots dont make nodes and nodes dont require dots. The dots just indicate junctions of more than two elements. The node at the bottom is sometimes mistakenly seen as two or three nodes. Now for the tree branches. A branch joins two nodes. The branch is a tree branch if it does not close a loop of branches. A little thought should make us realize that there must be one fewer tree branches than there are nodes. The rst tree branch joins two nodes. Each additional branch adds one more node to the tree. For nodal analysis, we choose all the tree branches to start from a common node. KVL doesnt say we have to follow actual circuit branches. Instead, we make branches directly from this common node to each of the other nodes. Once we have those tree branches, we designate them as the voltage variables for our analysis. For each, we designate that common node as the minus terminal and the opposite end as the plus terminal. Figure 1.27 shows the tree branches of the circuit of Fig. 1.26. Ive added my choice for the names of the voltage variables and their signs.
Figure 1.27: Tree-branch voltages
After doing this once or twice, we begin to realize that we can simplify the appearance of the diagram: I label the common node with a ground symbol and call it my reference node. I can omit the signs because I know the common node, the reference node, is the minus end
1.5. NODAL ANALYSIS
21
I put the names of the voltage variables directly on the nodes, understanding that these are the plus ends. I call these voltages the node voltages Ive done all that in Fig. 1.28.
Figure 1.28: Relabeled with reference node
An example is probably the easiest way to start.
1.5.1
EXAMPLE IVNODAL ANALYSIS
The circuit shown in Fig. 1.29 has ve elements. We could write ve v-i equations for the ve
Figure 1.29: Example IV for analysis
elements, three KCL equations at three nodes, and two KVL equations around the loops. Ten equations in ten unknowns! Nodal analysis reduces this to just three equations. The goal for this circuit is to nd the voltage across the 12-k resistor using nodal analysis. My rst step will be to dene a set of independent node voltages. My choice of voltages is shown in Fig. 1.30 , where I labeled the voltages as I just described a few paragraphs ago. OK, now lets see how to write the three equations. One is pretty obvious, because the voltage v1 is along the branch that contains the 45-V source between the v1 node and the reference node. So our rst of three equations is
v1 = 45
The other two equations come from the application of Kirchhoff s Current Law at the other two labeled nodes. Note that we are going to write current equations, even though this seems strange,
22
1. BASIC STUFF
Figure 1.30: Example IV prepared for nodal analysis
having just dened voltages. But since we know, in terms of the node voltages, the voltage across each element, we can write an expression for the current through that element. Ooops! Wait! How do we know the voltage across each element? I can see that we know some of them. But what about, for example, the voltage across the top resistor on the left. All we know are the voltages at the two ends of this resistor. Lets single out that resistor for study as I show in Fig. 1.31. Ive labeled its voltage vR . The
Figure 1.31: Top left voltages
voltage on the right end is v2 ; the voltage on the left end is v1 . Now write Kirchhoff s Voltage Law around a loop from the reference node to the upper left node to the middle node and back to the reference node:
v1 vR + v2 = 0
Solve this for the labeled voltage across the resistor:

vR = v2 v1
Theres an easy way to state the outcome. Dene the plus end of the resistor voltage as the beginning end and the minus end as the ending end. Then this resistors voltage becomes beginning end minus ending end Now Im nally ready to write the currents leaving the middle node in terms of the node voltages. My goal is to write Kirchhoff s Current Law for this middle node.
1.5. NODAL ANALYSIS
23
I am going to choose currents leaving a node to be positive. I do this so I dont have to think about which terms of my equations should be negative, thereby probably making fewer sign errors. Look at the node labeled v2 . (Figure 1.32 is this node isolated so we can look at the currents
Figure 1.32: Currents at top middle node
leaving it.) There are three currents involved. Each can be found from the voltage across its element divided by the resistance. The three currents are to the left, (v2 v1 )/8 (i.e., beginning end v2 minus ending end v1 ), down to the reference, (v2 0)/12 (i.e., beginning end v2 minus ending end at the reference), and to the right, (v2 v3 )/4 (i.e., beginning end v2 minus ending end v3 ). Combining these into a KCL equation for the node v2 , I get
v2 v1 v2 0 v2 v3 + + =0 8 12 4
Thats one node equation. Finally, consider the node labeled v3 (separated in Fig. 1.33). Two currents are involved:
Figure 1.33: Currents at top right node
24
1. BASIC STUFF
to the left, (v3 v2 )/4 (i.e., beginning end v3 minus ending end v2 ), and down to the reference, (v3 0)/2. Combine these two terms to get the third equation, again paying attention to sign:
v3 v2 v3 0 + =0 4 2
Thats another node equation. Now I have three equations in three unknowns. Dont miss the fact that I wrote them by writing expressions for currents leaving the nodes. In each case, the current is the voltage difference across the element divided by the element value. Here are all three node equations and their solution:
v1 = 45 v2 v1 v2 0 v2 v3 + + =0 8 12 4 v3 v2 v3 0 + =0 4 2 Solutio on: v1 = 45 V, v2 = 15 V, v3 = 5 V
My goal was to nd the voltage across the 12-k resistor. Now, after a whole page of text, I nd that its v2 = 15 V. Was that easier than writing and solving ten equations in ten unknowns? Maybe not, but its a powerful method that is often used by computer programs that simulate circuits (PSpice is an example).
1.5.2
STEPS FOR NODAL ANALYSIS
Here is a list of the steps for analyzing a circuit using the nodal-analysis technique. They rather easily reduce to a straightforward algorithm: 1. Select and label a reference node. Usually there is an obvious reference node at the bottom of the circuit. But the selection makes no difference in the rest of the algorithm. 2. Label all the remaining nodes with voltage variables, plus at the labeled node, minus at the reference. 3. Write a KCL equation for each of the labeled nodes, using the node voltages as the variables. Choose leaving currents to be positive. 4. Solve these equations by any method that works for you. 5. Answer the questions posed about the circuit, since we often want more than just the node voltages. Now lets use this algorithm to solve another problem.
1.5. NODAL ANALYSIS
25
1.5.3
EXAMPLE VMESSIER NODAL ANALYSIS
Consider a little more complicated circuit that has a current source (see Fig. 1.34). Ive already labeled
Figure 1.34: Example V
the reference node (Step 1) and the node voltages (Step 2). Our goal is to nd vo . Step 3 is not really complicated by the current source, because the 5-mA source is merely delivering 5 mA to the node labeled v1 . This time Ill just write the complete set of equations; you should check through them so that you are sure where the terms come from. (Ive written in volts, milliamperes, and kilohms, which are a self-consistent set of units.)
5 + v1 0 v1 v2 v1 v3 + + =0 7.2 8 8 v2 v1 v2 0 v2 v3 + + =0 8 20 8 v3 v1 v3 v2 v3 0 + + =0 8 8 3
Step 4 says to solve these, which yields 18, 10, and 6 volts for the three nodes (left to right). Step 5 says to answer the question, in our case to nd vo , which is the same as v3 :
vo = v3 = 6 V
1.5.4
EXAMPLE VITWO SOURCES
Lets do this with a circuit that is a bit more complicated. The circuit of Fig. 1.35 has two sources, one a current source. I have labeled the reference node and the three remaining nodes. Note that the voltage source establishes the voltage at the top left node as 45 V. We are to nd the voltage across the 2.5-k resistor. When I write these equations, I can save a little writing if I write using volts, kilohms, and milliamperes, which are self-consistent. My equations do not have as many zeroes when I do this:
26
1. BASIC STUFF
Figure 1.35: Example VI

v1 45 v1 0 v1 v2 + + =0 2 3 1.5 v2 v1 v 0 6+ 2 =0 1.5 2.5
The current source appears in the second equation because it is delivering a current to that node. Note that its sign is minus because the current is arriving and I have chosen leaving as positive. The solution of these equations for v2 gives me the result Ive asked for:
v2 = 20.77 V
1.5.5
EXAMPLE VIIDEPENDENT SOURCE
The circuit of Fig. 1.36 is to be solved for the value of the output voltage vo . The circuit contains
Figure 1.36: Example VII
a dependent source, indicated by the diamond. Its voltage is 4000ix , where ix is the current on the right. The dependent source does not change the algorithm, though.
1.5. NODAL ANALYSIS
27
Step 1 of the algorithm says to label a reference node, which I am going to choose at the bottom of the circuit. Step 2 says to label the other nodes with node-voltage variables. Figure 1.36 shows the results of these steps. Step 3 says to write KCL at each of the nodes. But at two of these nodes we already know the voltage. At v1 the voltage is 8 V; at v3 the voltage is 4000ix . We dont write KCL equations for these because we dont know the currents through the voltage sources. That leaves just the node v2 . Three currents are involved: to the left, (v2 v1 )/2000, down, (v2 v3 )/500, and to the right, ix . Because of the dependent source, a constraint equation is also required, since ix is another variable. We get that from the 6-k resistor: ix = (v2 0)/6000. (Note that v2 and vo are the same voltage; I could have saved some writing by using vo as the node voltage instead of v2 .) Here are the three equations plus the constraint:
v1 = 8 v2 v1 v2 v3 + + ix = 0 2000 500 constraint v3 = 4000ix v 0 define ix ix = 2 6000
Step 4 says to solve this set, which yields v1 = 8 V, v2 = 3 V, v3 = 2 V, and ix = 0.5 mA. Step 5 says to answer the question, so we look back to nd that the goal was to get vo . Since this is the same voltage as v2 , the solution is
vo = v2 = 3 V
Why did I write the equations in amperes and ohms rather than a somewhat simpler milliamperes and kilohms? The unit of 4000 on the dependent source is volts per ampere so Ill stick to volts and amperes and ohms.
1.5.6
EXAMPLE VIIIDECEPTIVELY SIMPLE
Heres a deceptively simple problem (Fig. 1.37) that has a twist to it. The circuit in the drawing is already labeled according to Steps 1 and 2. The goal is to nd vo Step 3 wants us to write the KCL equations for the nodes. If I start writing the KCL equation for the node labeled v1 , I notice that there is a 15-mA current arriving, a current of (v1 0)/400 leaving, and a current of. Hmmmm, what is the current through the 6-V source?
28
1. BASIC STUFF
This situation complicates the node equation because we cant nd the current through the voltage source by looking at the voltage source. The voltage source constrains the voltage, but the current through it can be anything. Where, then? Look at the 2-k resistor. The current down through it is (vo 0)/2000. That must be the same current that is leaving the node v1 and going to the right through the voltage source. Aha! So I can write the KCL equation for v1 . But Ive used up vo and I still need another equation. That will come from the fact that the 6-V source establishes the voltage difference between vo and v1 . So the equation vo v1 = 6 becomes my second equation. Here are the two equations for this analysis (written in mA and k ), along with the solution for vo :
15 + v1 0 vo 0 + =0 0.4 2 vo v1 = 6 vo = 10 V
Solution
A voltage source in a position such as the 6-V source is will always establish the difference between two node voltages. Then we have to look beyond the source to nd the current through it.
1.6
OP-AMP
The operational amplier is usually called an op-amp because its real name is too long to say. Figure 1.38 shows an op-amp with all its terminals labeled. The device requires power, shown in the gure as Vcc . But when we draw most circuits, we ignore the power terminals, since they rarely enter our analysis of the circuit around the op-amp. The two inputs are labeled v and v+ . The gain of the amplier is very high, typically of the order of 105 or 106 . If this gain is called A, then the output of the op-amp is the gain times the difference of the inputs:
vo = A(v+ v )
1.6. OP-AMP
29
Figure 1.38: Operational amplier
A is called the open-loop gain, which is the gain straight through the op-amp without any external circuitry attached to it. The output vo will not go beyond Vcc because the amplier saturates. Actually, the output cant quite reach that level but instead falls a little short. In typical semiconductor op-amps like the TL072 and LM747, this shortage is about 1.5 volts.
1.6.1
IDEAL OP-AMP
The current that ows into the inputs of the modern op-amp is very, very tiny, often micro- or nanoamperes. In most of our circuits we dont have currents that small (milliamperes is small for us).Therefore, we can, as a decent approximation, ignore the op-amps input currents when analyzing circuits. Moreover, the open-loop gain A is so large that the input voltage must be minuscule compared with the output voltage. In most of our circuits, the output voltage is in the same range as the rest of the voltages in the circuit. Therefore, the input voltage (the difference between v+ and v ) is essentially zero. So a decent approximation is to consider this voltage difference to be zero. We call this negligible voltage difference a virtual short. It is not really a short circuit because a tiny current does ow from one input terminal to the other. But the voltage between the terminals is zero, as it would be if the terminals were actually shorted together. Finally, the op-amp involves a power supply that is not shown as part of our circuit. Because of this, we cannot calculate the current entering the output of the op-amp. To say this another way, we cant write KCL at the output node of the op-amp until we have nished the rest of our analysis of the circuit. Figure 1.39 shows the ideal op-amp with the decent approximations we have just described.
1.6.2
EXAMPLE IXOP-AMP INVERTER
Lets analyze one circuit, the inverter shown in Fig. 1.40. The op-amp is ideal, or at least we are going to assume that it is, and the power-supply inputs are not labeled.
30
1. BASIC STUFF
Figure 1.39: Ideal op-amp
Figure 1.40: Example IX
The analysis proceeds as follows: Use the assumption that the voltage difference between the + and terminals is 0. Therefore, the voltage at the terminal is the same as at the + terminal, namely, 0 (i.e., ground). The voltage across the 20-k resistor is vo 0. input resistor is vin 0; the voltage across the 40-k feedback
The current owing into the input terminal is 0 because of the ideal assumptions. Applying KCL for currents leaving the minus input terminal requires the sum of three currents:
0 vin 0 vo 0 =0 20 40
which solves to yield

vo = 40 vin = 2 vin 20
The result of our analysis is that the output is a negative copy of the input and is twice the size, provided we dont try to push the output too close to the power-supply voltages.
1.6. OP-AMP
31
This circuit is an inverter that both amplies and inverts the input signal. Figure 1.41 is the general circuit for an inverter that is described by the equation
Figure 1.41: Op-amp inverter
vo =
Rf Ri
vin
1.6.3
EXAMPLE XOP-AMP NON-INVERTER
The circuit of Fig. 1.42 is another important op-amp circuit. Lets see what it does by rst applying the ideal conditions (see Fig. 1.43). Since the voltage between the two input terminals is zero, we can write
vx = vin
Look at the two resistors and notice that they form a voltage divider, although this time its drawn from right to left. The voltage vx is a fraction of the output voltage vo , a fraction that is determined by the two resistors:
vx = vo 2 = vin 2 + 10
So the output as a function of the input is

vo = vin 2 + 10 10 = vin 1 + 2 2
This circuit is called a non-inverter, partly, I guess, because there is no verter in our language. Figure 1.44 shows the general form of the non-inverter. Its output/input equation is
32
1. BASIC STUFF
Figure 1.42: Example X
Figure 1.43: Example X idealized

Rf vo = vin 1 + R1
The non-inverter cannot have a gain of less than 1 because we cant have negative ratios of resistors. Compared with the inverter, this non-inverter has a very important property that is often useful in a circuit. The input to the ideal non-inverter draws no current from the vin source. Thats because the input is connected directly to the plus input of the op-amp, and ideal op-amp inputs require no current. The inverter cannot make that statement. Even though the minus input draws no current in the ideal op-amp, there is still current owing through the resistors. In fact, because the ideal voltage between the terminals is zero, the minus terminal looks like it is grounded. That means the current through the input resistor is vin /Ri .
1.6. OP-AMP
33
Figure 1.44: Op-amp non-inverter
1.6.4
EXAMPLE XIOP-AMP VOLTAGE FOLLOWER
Lets look at one more circuit, this one a special case of the non-inverter. In Fig. 1.44, let the resistance Rf become 0. In other words, replace the resistor Rf with a short circuit. Now the equation for the circuit is
0 vo = vin 1 + R1
The resistor R1 doesnt seem to be doing anything, because in the equation it can be any value (except 0) and have no effect on the result. I can remove R1 entirely, leaving an open circuit in its place. This open circuit makes R1 = . The resultant circuit is the voltage follower shown in Fig. 1.45. The equation that describes it
Figure 1.45: Op-amp voltage follower
is
vo = vin
34
1. BASIC STUFF
Ah, you say, this just sounds like a wire! Why go to the trouble of putting in an op-amp? So I ask you, how much current ows into the input of the op-amp? Ideal? Yes. So the input current is 0. The output current can be anything the op-amp can provide. The voltage follower is used to beef up the output when we have a weak source that cannot provide much current (a sensor in an instrument, for example) [To read more about op-amps, look up operational amplier in any beginning circuit analysis text.]
1.7
LINEAR IMPLICATIONS
All of the circuits weve been studying are linear. They are made up of linear devices whose dening equations are all linear. For example, v = Ri is a linear relationshipdouble the current and the voltage doubles, and so on. Lets demonstrate linearity by adjusting the outcome of Example IV. The original circuit for this example is Fig. 1.46. The outcome of the analysis in Section 1.5.1 was that vo = 5 volts.
Figure 1.46: Example IV again
Suppose we are designing a system and we really wanted this circuit to have an output of just 2 volts. This new voltage is just two-fths of what we got originally. So all we need to do is reduce the source voltage by that factor. The new source = 45 x 2/5 = 18 volts. Suppose we also dont like the general range of values of resistors and would rather have values twice as large. If we simply double all the resistor values, the output voltage will be unchanged. (Note that all the currents in the circuit will be half as large.) Figure 1.47 shows the results of these changes.
1.7.1
SUPERPOSITION
A consequence of linearity is that we can consider the various sources in a circuit independently. For example, in a linear circuit with two sources, we can consider the effects of one of those sources alone with the other source not operating. The only caution in doing this is for us to be clear on what not operating means: A not operating voltage source must produce zero volts, which can be guaranteed only by replacing the voltage source by a short circuit.
1.7. LINEAR IMPLICATIONS
35
Figure 1.47: Example IV adjusted
A not operating current source must produce zero current, which can be guaranteed only by replacing the current source by an open circuit. Ill illustrate this by redoing Example VIII, shown here in Fig. 1.48. I want to nd out what
Figure 1.48: Example VIII
vo is due to each of the sources independently. Ill start by making the 6-V source not operating. Note in Fig. 1.49 that I replaced it with a short circuit. The circuit with only the 15-mA source is now a current divider. (Remember other resistor?) First, nd io :
io = 15 400 = 2.5 mA 400 + 2000
From io , nd vo due to the 15-mA source acting alone:

vo = 2 io = 2 2.5 = 5 V
Now Im going to put the 6-V source back in place and make the 15-mA source not operating. Note in Fig. 1.50 that its been replaced by an open circuit. The 6-V source is pushing the current io clockwise around the circuit:
io = 6 = 2.5 mA 400 + 2000
36
1. BASIC STUFF
Figure 1.49: Example VIII: 15-mA source alone
Figure 1.50: Example VIII: 6-V source alone
From this current I get the value of vo due to the 6-V source acting alone;
vo = 2 io = 5 V
The real test is whether these two outcomes, each due to one source acting alone, add to the value we found for Example VII in Section 1.5.5. Combining these is called superimposing them. Do they check? Thats why this section is titled, Superposition.
1.7.2
THVENINS THEOREM
Lon Charles Thvenin, a French telegraph engineer, published a theorem in 1883 that helps simplify circuits when all we want is what happens to a load (such as a resistance) on the output terminals. He basically said that you can replace a linear source circuit by a two-element Thvenin equivalent and the load will be unable to tell the difference. (For some reason, Thvenin gets the credit, although Helmholz published the same theorem 30 years earlier!) Figure 1.51 is the setup he and I are talking about. Box A must be linear, but it can contain any of our linear circuit elementsvoltage and current sources, resistors, inductors, and capacitors. The sources can even be dependent sources, although we are not going to carry this topic that far.
1.7. LINEAR IMPLICATIONS
37
Notice in Fig. 1.51 that the load voltage and the load current are dened. Thvenin says that the circuit of Fig. 1.52 will produce exactly the same load voltage and load current and he tells us how to nd the right values.
Figure 1.51: Circuit for Thvenin
Figure 1.52: Circuit with Thvenin
The formal statement of Thvenins Theorem is As far as anything that Load B can determine, Linear circuit A can be replaced by a voltage source in series with a resistance, where the voltage source is the voltage measured at the terminals with Load B removed, and where the resistance is the ratio of that voltage to the current measured with the terminals shorted. To say this in equation form:
voc = RThisc
where voc is the voltage measured at the terminals with no Load B attached and where isc is the current measured through the wire shorting the terminals together. Lets show this in an example by reworking Example IV. Figure 1.53 is this example and shows that the 2-k resistor on the right is being considered the load. First, Ill calculate the open-circuit voltage (see Fig. 1.54) by noting that the 4-k resistor has no current owing through it so there can be no voltage across it. Therefore, voc must be the voltage across the 12-k resistor. Ill nd that by using a voltage divider:
voc = 45 12 = 27 V 12 + 8
38
1. BASIC STUFF
Figure 1.53: Example IV for analysis using Thvenin
Figure 1.54: Example IV: open circuit voltage
Next, Ill nd the current owing in the short circuit between the terminals, shown in Fig. 1.55. How about nodal analysis with just one node equation? Dene v as the voltage at the node in the
Figure 1.55: Example IV: short-circuit current
top middle. Put the reference node at the bottom. Now write and solve one node equation:
v 45 v 0 v 0 + + =0 8 12 4 v = 12.27 V
I now use v to nd isc :

isc = v 12.27 = = 3.068 mA 4 4
1.8. SUMMARY
39
Finally, from voc and isc I get RT h :

RTh = voc 27 = = 8.8 k isc 3.068
The resulting circuit is Fig. 1.56. You can answer any question about the 2-k resistor by using
Figure 1.56: Example IV with equivalent source
the circuit of Fig. 1.56, questions like, What is the current through it? What is the voltage across it? How much power is it absorbing? You dont have to go back to the original circuit of Fig. 1.53. But you cannot answer any questions about the original circuit to the left of the terminals. Theres a lot more thats useful about Thvenins Theorem, but what we have here is as far as we are going to go in this text.
1.8
SUMMARY
If you get the feeling that this chapter is packed, it is! Weve covered a large number of topics: Charge, current, voltage, power, and the passive sign convention Resistors, capacitors, inductors, and Ohms Law Ideal sources Kirchhoff s Lawsthe foundation of all this Series and parallel; voltage and current dividers Nodal analysisa formal approach to circuit analysis Op-amps in inverting, non-inverting, and voltage-follower congurations Linear circuits Superposition, which allows us to nd contributions to certain voltages and currents due to individual sources acting alone Thvenin equivalent, which provides the same current and voltage to a load as the original circuit did Now its time to go on to useful applications of circuits.
FORMULAS AND EQUATIONS
41

1. Charge and current
i= dq , q(t ) = dt
i( x ) dx + q(0)
0
2. R, C, and L
v v2 R = , v = Ri, p = vi, p = Ri 2 = i R dv dv dv 1 i = C , p = vi = Cv , w = p dt = Cv dt = Cv 2 dt dt dt 2 di di 1 2 di v = L , p = vi = Li , w = p dt = Li dt = Li dt dt 2 dt
3. Kirchhoff s Laws The algebraic sum of the currents leaving any node is zero at every instant of time. The algebraic sum of the voltages around any closed loop is zero at every instant of time. 4. Series, parallel, voltage divider, current divider
Rseries = R1 + R2 , Rparallel = Rout Rother 1 , vout = vs , iout = is 1 1 Rout + Rother Rother + Rout + R1 R2
5. Nodal analysis (a) Select and label reference node (b) Label remaining node voltages, plus at labeled node, minus at reference (c) Write KCL for each labeled nodecurrents leaving are positive (d) Solve equations (e) Answer question 6. Op-amp inverter, non-inverter, voltage follower
vout = R feedback Rin R feedback , vout = vin vin , vout = vin 1 + R1
7. Thvenin equivalent
vTh = voc , RTh = voc isc
43
CHAPTER
Power of the Sine

Sines and cosines are pretty neat functionsand very useful, too. Neat because they are simple functions that have simple derivatives. Useful because they are what we mean by alternating current, the primary system used to move large blocks of energy from place to place. In this chapter we are going to study two aspects of sinusoids, generally as cosines rather than sines. First, we will consider circuits driven by sinusoids in which we want only the sinusoidal response. Then well see how we can simply our use of sinusoids via phasors. Finally, well work with a-c power to learn how we move energy. To get started, take a look at the sinusoid of Fig. 2.1. This function is drawn as a cosine because
Figure 2.1: v(t) = V0 cos(t + )
it turns out the cosine rather than the sine form is the more useful one for our work. There are some important things to notice: A cosine starts at t = 0 with its positive peak. The period of this cosine is . The frequency of the wave f = 1/ is measured in hertz (Hz). When we write the cosine function, we use the radian frequency , which is 2 f. (It is this 2 that will cause us lots of errors!)
44
2. POWER OF THE SINE
The positive peak of this cosine comes before t = 0, so the phase angle is positive. Notice that the actual offset to the left is /. We say that this is a leading phase angle . Recall from calculus that the derivative of a sine is a cosine. The derivative of a cosine is sine. Well need those in the next section.
2.1
SINUSOIDS AND PHASORS
Sinusoids pop up in a whole array of electrical topics, from power systems to communications. Our study begins with the sinusoidal steady state, which in itself is not a difcult concept. If we break the phrase into its three words, we can readily see what its all about: State is the condition of a system at any instant. Its the values of all the variables in the system at that moment. In this case, though, it has a little broader denition. It means the condition of the system over time as it responds to its inputs. Steady means that we are studying the system only after any start-up transients have died out. Most physical systems, when turned on, exhibit a wiggle of some kind that dies out. The word steady here means that we are ignoring this transient and observing the state of the system only after the transient has disappeared. Sinusoidal is the kind of inputs that we drive our system with. All sources will be simple sine waves with amplitude, frequency, and phase angle. So saying sinusoidal steady state is saying we are studying the non-transient behavior of systems excited by sine waves. Why are sinusoids interesting? Well, for a number of reasons. They are the most common excitation in electrical systems. Mathematically, they are very simple. Also, via Fourier series we can decompose more complicated periodic signals into sums of sine waves. And, just for the record, they are just about the simplest mathematical function to deal with when solving differential equations. Only exponentials are easier to handle. (Hmm, I guess constants would be even easier?)
2.1.1
DIFFERENTIAL EQUATION
Im going to go through the hairy details of solving a simple differential equation that has a sinusoidal forcing function. Youve probably done this in math, so Im going to omit a few of the messy details and just toss in some words at times. My system is described by the following equation, which is giving the current i as a function of time. The driving function is a sinusoid with a frequency of 20 radians per second and a phase angle (an offset on the time axis) of 30 degrees. Since this is a rst-order equation, there must be one initial condition.
di(t ) + 4 i(t ) = 12 cos(20t + 30) dt i(0 ) = 0.2 A
2.1. SINUSOIDS AND PHASORS
45
Perhaps youll remember from math that equations like this one have solutions that can be broken into two parts. The math folks call these two parts the homogeneous and the particular solutions, whose meanings are obvious to just about no one. I prefer to call them the natural or transient response and the forced or steady-state response. My goal here is to solve the equations to nd the steady-state solution only, although in the process Ill develop the complete solution rst. The rst step is to nd the homogeneous/natural/transient solution. Doing this requires us to solve the differential equation with its forcing function, the right-hand side, set to zero. For this step, our equation looks like this:
di(t ) + 4 i(t ) = 0 dt
I recall that all linear differential equations with constant coefcients have a natural solution of the form
in (t ) = Best
Ill plug this solution into our equation and, after a few manipulations, get the natural response in (t):
Bsest + 4 Best = 0 if Best is not 0 everywhere, then s + 4 = 0, and s = 4 so in (t ) = Be4 t
Note the unknown constant B; it cannot be determined until we have the complete solution later. The next step is to nd the particular/forced/steady-state solution. This means solving the equation again but this time with the forcing function on the right-hand side active. I recall from math that I should try a linear combination of the form (note: form) of the forcing function and of all of its derivatives. The beauty of sinusoids is that I need only two forms, sine and cosine. Ill guess that the steady-state solution looks like this:
i f (t ) = C cos(20t + 30) + D sin(20t + 30)
I plug this into the differential equation and determine the coefcients C and D . Since the sine and the cosine are orthogonal functions, I can equate the sine and the cosine terms separately to yield two equations in two unknowns. Skipping the details, I get a steady-state solution of
i f (t ) = 0.1154 cos(20t + 30) + 0.5769 sin(20t + 30)
The next step is to combine the two solutions:
46
i(t ) = 0.1154 cos(20t + 30) + 0.5769 sin(20t + 30) + Be4 t
Finally, I can nd the value of the constant B by evaluating the result at t = 0, using the initial condition given:
i(0 ) = 0.2 = 0.09994 + 0.28845 + B B = 0.1884
The complete solution with all numbers in place is

i(t ) = 0.1154 cos(20t + 30) + 0.5769 sin(20t + 30) 0. .1884 e4 t A
which with a little trigonometric manipulation looks like this:

i(t ) = 0.5883 cos(20t 48.7) 0.1884 e4 t A
It is very easy to see the two parts of the solution, namely, a natural or transient response that dies out and a sinusoidal steady-state response that continues forever. I want just this steady-state solution:
iss (t ) = 0.5883 cos(20t 48.7) A
All I really want when I am working steady-state problems is the steady-state solution. I dont need the transient response because that dies out quickly. Theres got to be a simpler way to get just this one part. [Note: Be careful with degrees and radians. Some software (Maple, for example) and some calculators insist that angles be in radians. Others allow degrees.]
2.1.2
PHASORS
There is a simpler way to do this, though, which makes use of the phasor. The phasor, in a way, is nothing more complicated than the use of Eulers identity to describe sine waves:
e j = cos + j sin
The j in the identity is the square-root of 1, the imaginary. The math folks use i , but for electrical systems, this readily gets confused with the common symbol for current. Ill apply this to my differential equation, writing the forcing function on the right-hand side in exponential form:
di(t ) + 4 i(t ) = 12 e j ( 20 t + 30 ) dt i(0 ) = 0.2 A
47
But you might object by saying, That isnt the same as the original forcing function. Its got an extra term. You started with just a cosine, but now you have a sine as well. But wait! Lets solve the equation rst:
as before, in (t ) = Be4 t so I guess that if (t ) = Ee j ( 20 t + 30 ) j 20 Ee j ( 20 t + 30 ) + 4 Ee j ( 20 t + 30 ) = 12 e j ( 20 t + 30 ) 12 E= = 0.5883e j 78.7 4 + j 20 i(t ) = 0.5883e j 78.7 e j ( 20 t + 30 ) 0.1884 e4 t
Again, the two parts of this solution are visible. (If you get worried about how I converted E , stay tuned.) Since I want only the sinusoidal steady-state, Ill discard the transient to get:
iss (t ) = 0.5883e j ( 20 t 48.7 )
Using Eulers identity, I get

iss (t ) = 0.5883 cos(20t 48.7) + j 0.5883 sin(20t 48 8.7) A
I sure do have an extra term, but there are two really neat aspects of this. First, Im working with an exponential, whose derivatives all have the same exponential form, so the solution process is simplied. Second, although theres a sine term added into the forcing function, it is agged with a j . That means we can pick it out after weve nished the solution and discard it as extraneous. Drop the j term to get the nal result
iss (t ) = 0.5883 cos(20t 48.7) A
which is the same as before.
2.1.3
IMPEDANCE
OK, you say, to solve a circuit for the sinusoidal-steady-state solution, I write the differential equation and solve it as weve done in the past, right? Well, sort of. Thatll work, but theres an easier way. I can simplify my work by developing equations for R , L, and C when they are driven by a general sinusoidal function. Doing this will get rid of differential equations and replace them with algebraic equations. Heres what Im going to do. For each of the three passive circuit elements, resistor, capacitor, and inductor, Im going to take their Ohms Law equation in differential form, apply a general sinusoid using the Euler form, and see what happens. The outcome in each case will be the impedance of that element. Here is the simple one, the resistor:
48
Figure 2.2: Resistor
v(t ) = Ri(t ) i(t ) = Ie j t , so v(t ) = RIe j t = Ve j t ZR = Ve j t =R Ie j t
The capacitor isnt much messier:
Figure 2.3: Capacitor
i(t ) = C
dv(t ) dt dVe j t = j CVe j t = Ie j t dt
v(t ) = Ve j t , so i(t ) = C ZC = Ve j t 1 = Ie j t j C
And nally, the inductor:
Figure 2.4: Inductor
49
v(t ) = L
di(t ) dt dIe j t = j LIe j t = Ve j t dt
i(t ) = Ie j t , so v(t ) = L ZL = Ve j t = j L Ie j t
The outcome is that these impedances look, to the sinusoidal steady-state math at least, just like resistors, albeit a little messier. In other words, by using impedance I can write circuit equations using all the same techniques I have been using for d-c circuits. These equations are not differential equationsthey are algebraic. To solve them means to use just algebra, although that algebra will involve complex numbers containing j . Recapping, these impedances are
Figure 2.5: Impedance
ZR = R ZC = 1 j C Z L = j L
To make this easier, we need one more denition, the phasor transform.
2.1.4
PHASOR TRANSFORM
Our driving functions, forcing functions, sources, whatever you call them, are sine waves. The phasor transform uses Eulers identity to convert these into the exponential form:
A cos ( t + ) Ae j
50
Well use that form to solve what are now algebraic equations that describe the system. Then well use the inverse phasor transform to return to the sine-wave form, a step which includes getting rid of terms that are marked with the j . Its really easier than it sounds. Here are a couple of examples of transforming a sine wave into a phasor:
170 cos ( 377t + 45 ) 170 e j 45 25 cos 10 5 t 80 25 e j 80
But! Be careful! The sine wave must be stated in cosine form! If its in sine form, rewrite it as a cosine while subtracting 90 degrees from its phase angle. Here are a couple of examples:
10 sin (1000t 15 ) = 10 cos (1000t 105 ) 10 e j105 60 sin ( 400t + 30 ) = 60 cos ( 400t 60 ) 60 e j 60
Long ago, when all this stuff was written using a typewriter and making exponents was messy, folks developed another notation for the exponential. It is still the common way to write the outcome of a phasor transformation today. Here are the previous examples written in this magnitude-angle form:
170 cos ( 377t + 45 ) 170 45 25 cos 10 5 t 80 25 80 10 sin (1000t 15 ) 10 105 60 sin ( 400t + 30 ) 60 60 A cos ( t + ) A
Notice that the angles are given in degrees, which is the standard way of doing this. Theres a totally sloppy inconsistency here, though, that we often just slide past. When we write a mathematical cosine function, the quantity in the parentheses must be in radians. The frequency must be in radians per second, not hertz. Multiplying this by time in seconds makes the rst term radians. Yet the second term, the phase angle, is degrees. Apples plus oranges! But we do it all the time and will stick with it. Just be sure to convert the phase angle to radians if you enter the function into Maple or Matlab. Your calculator can be set to handle either degrees or radians. The inverse phasor transform takes the magnitude-angle form and returns to the cosine timefunction form. Doing this requires that we know the frequency, so assume in each of these examples that the frequency is that of the power system, 60 Hz, which is 2 60 = 377 radians per second:
50 76 50 cos ( 377t 76 ) 120 38 120 cos ( 377t + 38 )
Note that the outcome in each case is a cosine with a magnitude and a phase angle as well as a t term. Although the Euler identity would require a sine term as well, we know that we included
51
the extraneous term when we did the transform, so we drop the extraneous term when we do the inverse transform. Whats so good about all this? Without phasors, solving circuits problems would require solving differential equations, even if we wanted only the sinusoidal steady-state results. Phasors allow us to write the sinusoidal steady-state equations as algebraic equations and thereby solve only algebraic equations.
2.1.5
COMPLEX ARITHMETIC
The phasor transform creates a function that has both a magnitude and a phase angle, which we can write either as an exponential or as a phasor. These are complex numbers, which means they have a real part and an imaginary part. A phasor can be written in either of two forms, polar or rectangular. So far, Ive written just polar form. Consider, please, this phasor, written in polar form:
V = 25 e j 30 = 25 30
If I draw this on the complex plane, which has two axes, real and imaginary, I get a point (Fig. 2.6), which has a magnitude and an angle.
Figure 2.6: Polar form
The labels in Fig. 2.7 and some trigonometry give me a clue as to how to convert between forms. That point in rectangular form is gotten by using familiar cosine and sine relationships:
V = 25 cos 30 + j 25 sin 30 = 21.65 + j12.5
The line and angle form is the polar form, written using our angle notation. Figure 2.8 shows the following phasor in rectangular form:
I = 12 + j 20
Converting this phasor to polar form again requires a little trigonometry:
52
Figure 2.7: Polar to rectangular
Figure 2.8: Rectangular to polar

I = 12 2 + 20 2 tan 1 20 12 = 23.32 59.0
Formulas for the conversions can be developed from the drawings:

A = A cos + jA sin B + jC = B 2 + C 2 tan 1 C B
( )
We also dene the real and the imaginary parts of the phasor, which is most easily done in rectangular form:
Re [ X ] = D, Im [ X ] = E (not jE ) X = D + jE
Now for some complex arithmetic. First, addition, which is easier to do in rectangular form. Just add the real and the imaginary parts separately. If the phasors are in polar form, convert them to rectangular form rst.
53
X1 = A + jB, X2 = C + jD
X = X1 + X2 = A + jB + C + jD = ( A + C ) + j ( B + D )
Here are a couple of examples1 :

Xa = X1 + X2 = ( 20 + 36 ) + j ( 5 17 ) = 56 j 22 X 3 = 25 20 = 23.49 + j 8.55 X1 = 20 j 5, X2 = 36 j17
Xb = X1 + X 3 = ( 20 + 23.49 ) + j ( 5 + 8.55 ) = 43.49 + j 3.55 = 43.6 4.7
Subtraction is done the same way.

Xc = X1 X2 = ( 20 36 ) + j ( 5 + 17 ) = 16 + j12
Multiplication is easier to do in polar form. Just multiply the magnitudes and add the angles:
Y1 = A , Y2 = B Y = Y1Y2 = A B ( + )
One multiplication example:

Ya = Y1Y2 = 12 6 ( 60 45 ) = 72 15 Y1 = 12 60, Y2 = 6 45
Division is similardivide the magnitudes and subtract the angles:

Yb = Y1 12 < < 45 = 2105 Y2 = 6 60
For circuits driven by sinusoids where we want only the sinusoidal steady state result, we treat impedances just as we handled resistances. All that we have learned about d-c circuits applies to the sinusoidal steady state. That includes nodal analysis. Be careful to follow convention for expressing the magnitude of a sinusoid in electrical systems. When the sine wave is written in the time domain, use the amplitude of the sine wave. When you are in the phasor domain, use the rms value of the sine wave. (rms = amplitude/ 2)
2.1.6
EXAMPLE ISINUSOIDAL STEADY STATE
The circuit of Fig. 2.9 is driven by a sine wave. We are to nd the steady-state current i(t). Our rst job is to transform the circuit to the phasor domain:
1 You may wish to get a simple calculator that handles complex arithmetic. The Casio fx-115 is under $25 (Walmart, for example),
and its legal for the engineering licensing exams.
54
Figure 2.9: Example I (time domain)
Resistors are resistors The inductor becomes Z = j (20000)(150 103 ) = j 3k The source becomes 12 0 V The current i(t) becomes I Figure 2.10 shows the complete transformation.
Figure 2.10: Example I (phasor domain)
Now comes nodal analysis just as for d-c:

V1 12 V1 0 V1 V2 + + =0 1.4 5 2.4 V2 V1 V2 0 + =0 2.4 j3
To nd the current through the inductor, I need V2 :

V2 = 6.108 49.3 V
The desired current in the phasor domain is
55
I=
V2 = 2.036 40.7 mA j3
Now transform from the phasor current to the time domain, not forgetting to put back (because cosines get actual amplitudes while phasors get RMS):
i(t ) = 2.036 2 cos(20, 000t 40.7) mA = 2.879 cos(20, 000t 40.7) mA
2.1.7
EXAMPLE IIANOTHER PHASOR PROBLEM

resistor.
In the circuit of Fig. 2.11, nd the sinusoidal steady-state voltage vo (t) across the 1-k
Figure 2.11: Example II (time domain)
This is a balanced circuit where the top rail and the bottom rail have the same circuit elements. The inductor with a series resistor represents an actual inductor, where the resistor is the resistance of the inductors wire. First, transform the circuit by calculating the impedances of the inductor and the capacitors:
ZC = 1 1 = 6 jt C j 4 10 0.001 10 <6
)(
= < j 250 Z L = jt L = j 4 10 6 0.2 10 <3 = j 800
)(
The resultant circuit, with the source and vo (t) transformed, is Fig. 2.12. Ive labeled the reference node and the remaining node voltages. Notice that Vo is not one of the node voltages because it is not connected to the reference node. I am going to include in my node equations the equation for nding Vo :
56
Figure 2.12: Example II (phasor domain)

V1 30 V1 V2 V V2 + + 1 =0 j 250 150 + j 800 1000 V2 0 V2 V1 V V1 + 2 =0 + j 250 150 + j 800 1000 Vo = V1 V2
The solution for Vo is

Vo = 41.09 57.1 V
which transforms back to the time domain as

vo (t ) = 58.11 cos( 4 10 6 t + 57.1) V
2.2
A-C POWER
We could treat a-c power using functions of time such as sine and cosine. Figure 2.13 shows a load (i.e., an impedance that is doing work for us) being driven by a voltage v(t) and a current i(t).
Figure 2.13: p(t) = v(t)i(t)
The voltage and the current are sinusoids:
2.2. A-C POWER
57
i(t ) = I peak cos ( 2 ft + i )
v(t ) = Vpeak cos ( 2 ft + v )
If we want to know the power, we calculate the product of these:

p(t ) = v(t )i(t )
If we calculate this product, we get what is known as instantaneous power, the power at any particular instant. But this isnt very useful, because we really need to know how much power is being delivered on the average, and more important, how much energy is being moved (which power companies want us to pay for). From the calculation of p(t), we can derive this average value as well as dene several other quantities that are very useful in dealing with a-c power. All of these new quantities come from looking at p(t) for sinusoidal voltages and currents.
2.2.1
A-C POWER QUANTITIES
The average power comes from averaging the instantaneous power p(t) over one period of the a-c waveform. For our sinusoidal voltage and current in Fig. 2.13, p(t) is
p(t ) = Vpeak cos(2 ft + v ) I peak cos(2 ft + i )
A few trig manipulations yield

p(t ) = Vpeak I peak 2 cos( v i ) + cos ( 4 ft + ( v i ))
If we now average this p(t) over one period, we get Pav :

Pav = Vpeak I peak 2 cos( v i )
Note that Vpeak is the peak value of the cosine wave, a value that makes mathematicians happy because thats the proper way to describe the cosine. But it isnt what we use in a-c power. Instead, we use the rms value. The name, root-mean-square, tells exactly how to compute the value, namely, take the square root of the mean of the square of the function. For a sine wave, this comes out to be Vpeak / 2 = 0.707 Vpeak . Again, note that this is valid for sinusoidal waveforms; other waveforms have different rms values. The average power P in terms of rms values is
P = VRMS I RMS cos ( v i )
58
P has the unit of watts and the unit symbol W . P is sometime referred to as average power or real power, but much more commonly, its just plain power. If someone says simply, power, it means P . Now lets go back to the equation for p(t) and consider the second term, the one with 4 in it. This second term averages to zero because the average value of a sinusoid over any integral number of periods is zero. Yet this second term is still power, even though its average is zero. Its power that ows one way for a quarter of a cycle, then the other way for a quarter of a cycle, and so on. From this second term we dene reactive power Q:
Q= sin ( v i ) 2 = VRMS I RMS sin ( v i ) Vpeak I peak
Q has the unit of volt-ampere-reactive and the unit symbol VAR. Note that Q is volts times amps, which is watts, but we use VAR to distinguish it from the average power P . Q delivers no energy on the average; it represents energy that ows back and forth in the power system. Reactive power Q is arriving at a load for one-quarter of a cycle, leaving for a quarter, and so on. Hence, it represents no net ow of energy, but it accounts for part of the current owing in the system. Because of this, we must include it in power calculations. Q is positive for loads that are primarily inductive, i.e., loads whose impedance has a positive phase angle. Q is negative for loads that are primarily capacitive. Reactive power Q is very important in energy transfer. For example, an induction motor is inductive (duh!) and requires current to supply the magnetic eld that makes the motor work. Therefore, an induction motor requires reactive power Q to operate. To say this another way, we must supply VARs as well as watts to make such a motor run and drive a rotating load. Because of these VARs, the current supplied is larger than that needed for just the average power P . This increased current ows through the wires and transformers of the power system and increases the losses associated with running the motor. Well talk more about this later. The power factor is dened as the cosine of the angle between the voltage and the current:
power factor = cos ( v i )
Power factor, often written pf, is unitless, but you must add a unit to it in the form of either lagging or leading. cos(v i ) is always going to produce a positive power factor, so this lagging-leading distinction is needed. How do you know which to use? A load that is inductive has a current that lags the voltage in time. Hence, the angle of the current i is less than the angle of the voltage v . We always say, The current lags the voltage, and never the other way around. Hence, the power factor is lagging. If the load is capacitive, the current leads the voltage and the power factor is leading. Another power quantity that is useful is the complex power S :
S = P + jQ
2.2. A-C POWER
59
This quantity also needs a unit that is equivalent to watts, so we use the volt-ampere and the unit symbol VA. S is a complex number with a real part of P and an imaginary part of Q. The cosine of its angle is the power factor, lagging if Q is positive. S is also the product of phasor voltage and phasor current:
S = VI *
The asterisk indicates the complex conjugate, which means replacing +j by j in rectangular form or changing the sign of the angle in polar form. The magnitude of S is sometimes useful. It is called the apparent power and is just the product of the voltage and the current without regard to the phase angle between them. The power factor can be dened as the ratio of average power to apparent power:
power factor = P S
The power triangle is an easy way to visualize the relationships among P and Q and S and power factor. The triangle (Fig. 2.14) recognizes that S is complex, with P on the real axis and Q
Figure 2.14: Power triangle
on the imaginary axis. Moving Q to the right to the other end of P gives the triangle. The angle is the power-factor angle, the angle v i . Note that positive Q is upward and represents inductive reactive power. Ill use the triangle to help out with my calculations in the next section.
2.2.2
EXAMPLE IIIADDING POWERS
The system in Fig. 2.15 has two loads whose parameters are stated in terms of power. Find the total power P , the total reactive power Q, the total complex power S , the overall power factor, and the line current I. For the rst load, P1 = 24kW with a power factor of 0.86 lagging. If we look at the triangle and remember that the tangent of the angle is Q/P , then the reactive power is
Q1 = P1 tan cos 1 pf = 24 tan cos 1 0.86 = 14.24 kVAR
60
and the complex power is

S1 = ( 24 + j14.24 ) kVA
Ill do the same thing for the second load, which is given in terms of S2 = 20kVA with a power factor of 0.72 lagging. P /|S | is the power factor, and the triangle tells me that Q and S are related by the sine of the power-factor angle:
P2 = S2 pf2 = 20 0.72 = 14.4 kW Q2 = S2 sin cos 1 pf2 = 13.88 kVAR
Now we use the fact that not only is P conserved, but so also are Q and S . This gives us the totals:
Ptotal = P1 + P2 = 24 + 14.4 = 38.4 kW Qtotal = Q1 + Q2 = 14.24 + 13.88 = 28.12 kVAR Stotal = Ptotal + jQtotal = 38.4 + j 28.12 = 47.6 36.2 kVA P 38.4 = 0.807 lagging pftotal = total = 47.6 Stotal
The current is calculated from S = V I :

* I line =
I line
Stotal 47.6 10 3 36.2 = = 99.2 36.2 A V 480 = 99.2 36.2 A (note the sign)
Ill carry the example one step further by nding the individual load currents.Their magnitudes will come from the equations P = |V ||I |pf and |S | = |V ||I |, and their angles will come from the power factor:
2.2. A-C POWER
61
I1 =
P1 24.0 10 3 = = 58.14 A V pf1 480 0.86 S2 20.0 10 3 = = 41.67 A V 480
I1 = cos 1 pf1 = cos 1 0.86 = 30.7 I2 =
I 2 = cos 1 pf2 = cos 1 0.72 = 43.9 so I1 = 58.14 30.7 A I 2 = 41.67 43.9 A
Note that the angles have to be chosen as negative because both loads have lagging power factors. As a check, lets see if the two currents add to the total current weve already found:
I line = I1 + I 2 = 99.2 36.2 A
Wow! It worked!
2.2.3
EXAMPLE IVMOTOR LOAD
Figure 2.16 shows a two-horsepower motor that is 85% efcient, running on a 240-volt supply with a power factor of 0.78 lagging. It is being fed by wires that each have an impedance of 0.5 + j0.8
Figure 2.16: Example IV
. We are to nd the power delivered to the motor, the line current, the power lost in the line, the total P , Q, and S , the power factor at the generator, the percent voltage regulation, and the percent efciency. Notice that the generator voltage is not given. This is because the motor is designed to operate correctly on 240 volts, so the generator voltage must be larger than that to compensate for the losses in the line. First, we need to put the motors specications into a-c power terms. One horsepower is 746 watts, which is the shaft output of the motor. The motor is 85% efcient, so the motor power input is
62
Pload =
2 746 = 1.755 kW 0.85
the reactive power input to the motor is

Qload = 1.755 tan cos 1 0.78 = 1.408 kVAR
From our knowledge of P comes the line current:

Pload = V I line pfload I line = 1.755 10 3 cos 1 0.78 = 9.375 38.7 A 240 0.78
The loss in the line is due entirely to the resistive part of the line impedance. Resistors dont care about phase angles because the voltage across a resistor has the same phase angle as the current through it, so I -squared-R is the line loss, using just the magnitude of the current:
Pline = 2 Rline I line = 87.9 W
2
I need the generator voltage Vs for a couple of the steps that follow, so I will get it using KVL, calculating 240 plus the voltage across the line impedances:
Vs = 240 + Vline = 240 + (0.5 + j 0.8 + j 0.8 + 0.5 ) ( 9.375 38.7 ) = 256.8 1.3 V
Theres now enough information to get the total Q and S and overall power factor. Heres a neat place to run off the track, though! Its tempting to think of the power factor as the cosine of the angle of the current. But thats wrongits the cosine of the angle between the voltage and the current. Notice that the angle of Vs is not zero! Here are the results, beginning with S = V I :
Ss = ( 256.8 1.3 ) ( 9.375 + 38.7 ) = 2.408 40.0 kVA Qs = Im [ Ss ] = 1.548 kVAR pfs = cos( 40.0) = 0.766 lagging Ps = Re [ Ss ] = 1.844 kW
That leaves percent voltage regulation andoh, whats regulation? The easiest way to look at percent voltage regulation is to think of it as the percentage by which the voltage at the load will rise if the load is disconnected. Our generator voltage Vs is 256.8 volts. If we disconnect the motor, its terminal voltage will rise from 240 to 256.8 volts. Percent voltage regulation is that expressed as a percentage:
2.2. A-C POWER
63
%VR =
Vs Vload 256.8 240 100 = 100 = 7.0% Vload 240
Percent efciency is a little more obvious. Its the percentage of useful power divided by generated power, often stated as %:
% = Pload 1.755 100 = 100 = 95.2% Ps 1.844
How about cost? Lets assume that this motor is part of a plant that operates 16 hours per day, ve days a week, fty weeks a year. Lets also assume that electricity costs 7 c per kilowatt hour plus a charge of $10 per kVA/month for the peak kVA demand during the month. (The meter that measures kilowatt hours also watches for the peak kVA demand using a sliding average with a window several minutes long.) Here are the numbers:
kWhr = (1.755 )(16 )(5 )(50 ) = 7, 020 kWhr/year $ kWhr = ( 7, 020 )(0.07 ) = $491.40 for one year Sload = 1.755 + j1.408 = 2.250 38.7 kVA $ kVAmax = (2.250 )(12 )(10.00 ) = $270.00 per year Total = $491.40 + 270.00 = $761.40 per year
2.2.4
STANDARD POWER PROBLEM
Just about every power problem involves three sections, as in Fig. 2.17. The problem is usually
Figure 2.17: Standard power problem
presented as follows: A load, the part that is doing work for us, stated either in power terms or as impedances The load voltage, which is the voltage the load is designed to operate properly on The line impedance, which is the impedance of the wires delivering the energy, often combined into just one of the two wires
64
The generator, whose voltage must be adjusted to provide the required load voltage Note that this is backwards from the usual circuits problem. Most circuits problems youve been seeing have a generator on the left and then ask something about an element on the right, such as voltage or power absorbed. This standard power problem is different in that we are given the voltage on the right (the load voltage) and generally asked to nd the generator voltage needed to accomplish this. Its sort of like, given the answer, nd the questionJeopardy, perhaps? There arent very many questions to ask, either. Find the line current Iline , the power lost in the line (loss in R ), the voltage drop along the line (across R + j X ), the generator voltage Vgen , the percent efciency, and the percent voltage regulation. There arent many other questions to ask
2.2.5
EXAMPLE VA STANDARD PROBLEM
Lets work the standard power problem shown in Figure 2.18. We are to nd the line current Iline ,
Figure 2.18: Example V
the percent efciency, the generator voltage Vgen , and the percent voltage regulation. Ill start by nding the line current (almost always the starting point) by using the equation for power P :
I line = = Pload cos 1 pf Vload pf 5000
( 460 )( 0.82 )
cos 1 0.82 = 13.26 34.9 A
Theres a very unobvious step hiding in that equation, though. How did the minus sign get attached to the arc cosine? Remember that the cosine of any angle between 90 and +90 is positive, so the arc cosine by itself cannot tell us what sign the angle of the current should have. We have to know which to choose! The choice of sign for the angle depends on whether the circuit is basically inductive or capacitive. Here, it is inductive. That means the current lags the voltage so the angle of the current must be negative.
2.2. A-C POWER
65
Thats where the minus sign comes from. You have to think each time, which sign should I use? Moreover, since most loads are inductive, if you ignore this decision, youll have the wrong answer most of the time. From the line current I can calculate the power lost in the line resistance. I ignore the inductance because it does not dissipate power. Hence, I need only Ri 2 to nd the loss. I do not include the phase angle because the voltage and the current for a resistor have the same phase angle. The power loss is
Plosses = ( 0.5 ) I line
2 2
= ( 0.5 )(13.26 ) = 87.9 W
Now I know both the power that is doing useful work (Pload = 5 kW) and the losses in the system, so I can nd the percent efciency:
% = 100 = 100 Pload Pload + Plosses 5000 = 98.3% 5000 + 87.9
To nd the percent voltage regulation, I need the generator voltage. Ill use Kirchhoff s voltage law to nd Vgen by adding the load voltage (with the phase angle chosen to be 0) to the voltage drop across the line impedance. The phase angle of the current must be included because the line is an impedance:
Vgen = Vload + Zline I line = 460 + ( 0.5 + j 0.8 ) (13.26 34.9 ) = 471.5 0.6 V
Notice the size of the angle of the generator voltagevery small. It doesnt differ much from the angle of the load voltage, which we chose as 0. Thats because the impedance of the line is only a small fraction of the loads impedance. If you get an angle for the generator voltage that is much different from zerosay, greater than 5youve probably made an error. Percent voltage regulation measures how much the voltage at the load changes when the load is removed. Removing the load reduces the line current to zero, so there is no voltage drop across the line impedance. The percent voltage regulation is
%VR = 100 Vgen Vload Vload 471.5 460 = 100 = 2.50% 460
In this example, the efciency is probably unrealistically high, but the voltage regulation is acceptable. Generally, voltage regulation less than 5% is desirable.
66
2.2.6
EXAMPLE VIADDING S
This problem is similar to Example III (Fig. 2.15) but Im going to work it differently. Instead of summing power and reactive power separately, Im going to sum the total power S for each of the loads. After all, if P and Q are both conserved, S must be, also. Note that S is a complex number but |S | is not and is not conserved. Figure 2.19 shows the circuit.
Figure 2.19: Examples VI
Taking a look at the power triangle again, we should see that if we know P and the power factor, the magnitude of S is P /pf and the phase angle of S is the angle whose cosine is the power factor. S for each of the loads is
Smotor = 0.84 = 44.41 32.9 kVA A
( 50 )( 746 ) + cos1 0.84
S fan =
13 10 3 + cos 1 0.75 0.75 = 17.33 41.4 kVA
Sheater = 24 10 3 0 kVA
Summing these gives the total power:

Stotal = 82.37 25.6 kVA
Since S = P + j Q, the desired result is

Stotal = 74.3 + j 35.56 kVA Ptotal = 74.3 kW, Qtotal = 35.56 kVAR
2.3
MORE EXAMPLE
Ill nish this chapter with a few more examples of nodal analysis, phasors, and a-c power.
2.3. MORE EXAMPLE
67
2.3.1
EXAMPLE VIINODAL ANALYSIS

resistor on the right. The reference
In the circuit of Fig. 2.20, nd the power delivered to the 150-
Figure 2.20: Examples VII, VIII, and IX
node is already designated and the node voltages have been labeled. The node equations are straightforward:
v1 24 v1 0 v1 v2 + + 400 650 200 v2 v1 v2 v4 v2 v3 + + 80 200 500 v3 v2 v3 v4 + 80 150 v4 0 v4 v2 v4 v3 + + 150 120 500 =0 =0 =0 =0
When you write node equations in an orderly way, you can look at the equations and see that order. This gives you a check on what you have written. Notice in this set of equations the pattern: In the rst equation, v1 is always positive while the other node voltages are negative. In each equation, the voltage at the node in question is positive and the others are negative. Each term is voltage divided by resistance, so each term is amperes. In the equation for a given node, all adjacent nodes are included. You could start with these equations and draw the circuit. Engineering requires doing correct work, and correct work comes from checking and checking. Solving for all four node voltages yields
v1 = 9.784 v2 = 5.686 v3 = 4.564 v4 = 2.459 V V V V
68
To nd the power delivered to the 150- resistor, I need the voltage across that resistor or the current through it (or both). Ill choose voltage. The voltage across the resistor is v3 v4 :
P150
2 v3 v4 ) ( 2.105 2 = = 29.54 =
150
150
mW
2.3.2
EXAMPLE VIIIVII SIMPLIFIED
Just because a circuit has lots of nodes doesnt mean we have to execute nodal analysis to the letter. The problem asks only about the power to the 150- resistor, so any changes that dont affect that will be OK. One simplication that will reduce the number of equations is to notice that the 120- resistor on the bottom is in series with the 200- resistor at the top. Any current that ows from left to right through the 200- resistor has to return from right to left through the 120- resistor. Since they are in series, I can add them and eliminate the node labeled v4 , making the whole common bottom the reference node. Now the node equations become
v1 24 v1 0 v v + + 1 2 =0 400 650 200 + 120 v2 v1 v 0 v2 v3 + =0 + 2 500 80 200 + 120 v3 v2 v3 0 + =0 80 150
The complete solution is

v1 = 9784 V v2 = 3.228 V v3 = 2.105 V
Now the voltage v3 is the voltage across the 150power is

P150 =
resistor directly to the reference, so the
2 v3 2.105 2 = = 29.54 mW 150 150
2.3.3
EXAMPLE IXVII SIMPLIFIED EVEN MORE
Not only can I combine the 120- and the 200- resistors in series, I can also note that the two resistors on the right form a voltage divider. In other words, I can combine them in series and eliminate the node v3 . Then I can nd the voltage across the 150- resistor using the voltage-divider relationship after I have solved just two node equations:
2.3. MORE EXAMPLE
69
v1 24 v1 0 v v + + 1 2 =0 400 650 200 + 120 v2 v1 v 0 =0 + 2 200 + 120 80 + 150
which yield
v1 = 9.784 V v2 = 3.228 V
The voltage divider gives me the voltage across the 150v150 = v2 P150
resistor and then the power:
150 150 = 3.228 = 2.105 V 150 + 80 230 2.105 2 = 29.54 mW = 150
OK, thats enoughIve beat this example to death!
2.3.4
EXAMPLE XPHASOR DOMAIN
Find the phasor voltage across the 10-F capacitor on the right end of the circuit of Fig. 2.21. Ill
Figure 2.21: Examples X (time domain)
convert the circuit to the phasor domain rst. The impedances are
Z L = jt L = j ( 600 ) 100 10 <3 = j 60 1 1 ZC = = = < j166.7 jt C j ( 600 ) 10 10 <6
The resultant phasor circuit is in Fig. 2.22.
70
Figure 2.22: Example X (phasor domain)
Lets get smart and notice that the elements in the line at the bottom of the circuit are in series with the corresponding elements at the top. So Im going to eliminate the node V2 by combining 50 + j60 with the 50 + j60 at the top. Im also going to combine the two 50- resistors in the right. The one node equation needed is
V1 8 V 0 V1 0 + 1 + =0 50 + j 60 + 50 + j 60 j166.7 50 j166.7 + 50
The solution for the single node voltage is

V1 = 5.888 89.4 V
This voltage is across a voltage divider consisting of the capacitor and the two 50so the voltage across the capacitor itself is
VC = V1 j166.7 = 5.049 120.3 V 50 j166.7 + 50
resistors,
2.3.5
EXAMPLE XITWO DIFFERENT FREQUENCIES
The circuit in Fig. 2.23 is driven by the sum of two different sine waves, one at 1,000 radians per second and the other at 10,000 radians per second. We are to nd the voltage vo (t) as a result of these two voltages acting together. To do this, we need to apply the Principle of Superposition. This states that, in a linear circuit, the voltage across any element is the sum of the voltages at that point due to each source acting alone (Section 2.3.1). What two sources? We can think of the source vs (t) as two sources in series. Well go into the phasor domain to nd out what Vo is for each of these two, then return to the time domain and add them. We cant do them together as phasors because the frequencies are different. Lets start with = 1,000 radians/second. The capacitors impedance is
2.3. MORE EXAMPLE
71
Figure 2.23: Example XI

1 = < j10 k j (1000 ) 0.1 10 <6
ZC =
Now two node equations will produce the voltage Vo :

V1 10 V1 0 V1 Vo + + =0 5 5 j10 Vo V1 Vo 0 + =0 5 j10 Vo = 5.963 63.4 V
Now redo the impedance at = 10,000 radians/second:

ZC = 1 = < j1k j (10000 ) 0.1 10 <6
The nodal result is

V1 10 V1 0 V1 Vo + + =0 5 5 j1 Vo V1 Vo 0 + =0 5 j1 Vo = 0.353 148.0 V
But be careful here! We cannot combine phasor results for different frequencies. Remember that the phasor transform removes from the equations once we have calculated the impedances. Phasor results at different frequencies cant be combined as phasors. Returning the results to the time domain and adding them, while not forgetting to change from rms to peak values, gives us vo (t):
72
vo (t ) = 8.433 cos (1000t 63.4 )
+0.500 cos (10, 000t 148. .0 ) V
This circuit is a lter that sorts out the higher frequency from the lower one. The two frequencies have equal amplitudes at the source but at the output, the 10,000-radians/second signal is reduced to about 6% of the 1,000-radians/second signal.
2.3.6
EXAMPLE XIIA-C POWER
The 230-volt load in Fig. 2.24 is stated as an impedance rather than in power terms. Our job is
Figure 2.24: Example XII
to nd the P and Q and power factor for this load, the percent efciency, and the percent voltage regulation. As usual, we nd the line current rst, this time by dividing the load voltage by the load impedance. I assume the load voltage is 0:
I line = V 230 = = 39.45 31.0 A Z 5 + j3
Now nd the total power S :

* Sload = Vload I line = ( 230 ) ( 39.45 31.0 ) *
= 9.072 + 31 1.0 kVA
Put S in rectangular form to get P and Q:

Sload = 7.78 + j 4.67 kVA Pload = 7.78 kW, Qload = 4.67 kVAR
The power factor comes from the angle of S , which is positive so the power factor is lagging:
2.4. SUMMARY
73
pf = cos 31.0 = 0.857 lagging
The only power loss in the circuit is the resistance of the line, so the power loss is Ri 2 :
Ploss = ( 0.2 ) I line = ( 0.2 )( 39.45 ) = 311 W
2 2
Percent efciency is
% = 100 7.78 10 3 Pload = 100 = 96.2% 7.78 10 3 + 311 Pload + Ploss
To get the percent voltage regulation, we need the generator voltage Vgen , which is the load voltage plus the voltage drop in the line impedance:
Vgen = 230 + ( 0.2 + j 0.25 ) I line = 241.9 1.0 V
Finally, the percent voltage regulation is

%VR = 100 Vgen Vload Vload = 100 241.9 230 = 5.2% 230
2.4
SUMMARY
This chapter and the one before it cover just about everything that well be using throughout the rest of this book: basic circuit laws, nodal analysis, phasors, and a-c power. If at this point any of these topics are giving you trouble, go back and review the material. Getting hold of another basic circuits text might help, too. Nodal analysis rarely deals with circuits as complicated as those in a rst course in circuits. For example, the equivalent circuit of an induction motor can be analyzed with a single node equation and application of the voltage-divider relationship. If you are not comfortable with the procedure (Example IX is a good test), reviewing previous work may be important. Phasors themselves generally do not present problems for studentsbut the complex arithmetic does. Using complex numbers, both in rectangular and in polar forms, is a signicant part of this course. Be sure that you can use your calculator to do these manipulations, including addition, subtraction, multiplication, and division of complex numbers, and conversion between rectangular and polar forms. If you dont know how to do these, nd your calculators instruction manual and learn the procedures. They are basic to the rest of the course. Power in a-c systems is also fundamental to our work because it represents the most common method for delivering energy over a distance. The basics are wrapped up in a few quantities: P , Q, S , and power factor. Knowing how to work with these is basic to a-c power and later, to our study of three-phase systems. Please dont go beyond this chapter thinking, Yup, that stuff is easy, and then nd out that you have trouble with some of it. This is a good time to check out your ability to handle these topics.
75

1. Ohms Law for passive elements
v = Ri, v = L di dv , i=C dt dt
2. Power to resistor, stored energy in inductor and capacitor

p = vi = Ri 2 = v2 1 1 watts, wL = Li 2 joules, wC = Cv 2 joules R 2 2
3. Nodal analysis (a) Dene reference node (b) Label remaining nodes with voltage labels (c) Write Kirchhoff s Current Law at each labeled node except the reference (d) Solve equations to nd node voltages (e) Answer the question posed 4. Eulers formula
e j = cos + j sin where j = 1
5. Phasor transform and inverse

j ( t + ) j v(t ) = A cos ( t + ) = Re Ae V = Ae = A
6. Complex-number conversions
b a + jb = a 2 + b 2 tan 1 , A = A cos + jA sin a
7. Impedances of passive elements

V = RI , V = Z L I = j LI , V = ZC I = 1 I j C
76
8. A-c power
P = V I cos ( v i ) = V I pf watts (W), pf = cos ( v i ) lagging or leading S = VI * = P + jQ volt-amperes (VA) S = V I volt-amperes (VA)
Q = V I sin ( v i ) volt-amperes-reactive (VAR)
9. Power triangle
10. Efciency and voltage regulation

% = 100 Vgenerator Vload Pload P = 100 load , %VR = 100 Pload + Plosses Pgenerator Vload
77
CHAPTER
Three-Phase Power Systems

Got your Ps and Qs straight? We are going to use those and S and power factor heavily in this chapter. Ugh, you think? Yup, just when you though that maybe you could get away from those, here they come again. This time, though, they are coming in threes. In Chapter 2 we studied single-phase a-c systems. The systems have had one generator, two conductors with wire impedances, and a load. We specied the load itself, usually in power terms. We specied the load voltage. Then we answered what are really the standard questions: line current, line loss, voltage drop in the line, percent efciency, generator voltage, and percent voltage regulation. Now we are going to do the same thing, but for three phases instead of one. If you want to say, Ugh, now is the time! Instead of one generator, well have three. Instead of one line current, well have three. And so on. But really, things arent that complicated. We are going to restrict our study to balanced three-phase systems. This means that the three phases are all the same except for 120 and 240 differences in phase angles, so we can analyze just one and have the answers for all three. Everything you know about a-c power for single-phase systems transfers to three-phase systems. So please dont ush from your mind what youve already learned.
3.1
GENERATOR
A single-phase generator is pretty simple, just a magnet spinning between two pole pieces (Fig. 3.1). If that magnet (usually an electromagnet) spins at 3,600 rpm, the voltage generated will be at a frequency of 60 hertz. Figure 3.1 shows the ideal voltage source that represents this single-phase generator. The output of this generator, expressed as a time-domain function, is
va (t ) = Vp cos ( t )
Remember that, most of the time, we use phasors to represent everything we are doing. Our ideal-source phasor voltage is
Va = Vp 2 0
Dont forget to use peak voltage when writing the cosine itself and rms voltage when writing the phasor.
78
3. THREE-PHASE POWER SYSTEMS
Figure 3.1: a. 1-/ 0 generator
Figure 3.1: b. 1-/ 0 ideal source
Generating three-phase voltages makes use of all the free space in the single-phase generator. We add two pairs of poles spaced uniformly around the circumference (Fig. 3.2). When the magnet spins, it produces in the coils three different voltages as shown on the ideal sources of Fig. 3.2:
va (t ) = Vp cos ( t ) vb (t ) = Vp cos ( t 120 )
vc (t ) = Vp cos ( t 240 )
Note that they all have the same peak voltage but differ by 120, corresponding to the 120 spacing of the pole pieces in the generator.
3.2
THREE-PHASE SYSTEMS
Three-phase systems are by far the most common way of delivering any large amount of electric power to customers. In this section, we are going to see what makes up a three-phase system, why we use them, and how we label voltages and currents to avoid getting confused.
3.2. THREE-PHASE SYSTEMS
79
Figure 3.2: a. 3-/ 0 generator
Figure 3.2: b. 3-/ 0 ideal source
3.2.1
THREE PHASES
A three-phase system consists of a three-phase generator, a load that has three distinct sections, and some conductors to join the load to the generator. The circuit of Fig. 3.3 shows the three-phase generator, now expressed with phasor labels:
Va = V 0 Vb = V 120 Vc = V 240
The load in Fig. 3.3 is balanced, which means that all three phases of the load are the same here, a resistance R . (We arent going to study unbalanced systems in this course.)
80
Figure 3.3: Neutral current
The connecting wires in Fig. 3.3 are somewhat unrealistic in that they have zero line impedance. Note that there are four wires. Three are phase wires that connect the three generator voltages to the three loads. The fourth wire is the neutral that connects the common node of the generator to the common node of the load. Using Fig. 3.3 as our circuit, lets calculate the currents in the wires. Im using a fairly common notation here: the currents have two subscripts, the starting node and the ending node. The calculation involves converting the phasors from polar form to rectangular form:
I aA = I bB V V 0 = + j 0 R R V V V = 120 = cos ( 120 ) + j sin ( 120 ) R R R V V = 0.5 j 0.866 R R V V = 240 = cos ( 240 ) + j sin ( 240 ) R R V = 0.5 + j 0.866 R
I cC
But look what happens when we calculate the neutral current. At the load, all three line currents converge at the neutral node N . Kirchhoff s Current Law says the neutral current must be the sum of the three line currents:
I Nn = I aA + I bB + I cC V V V V 0.5 j 0.866 0.5 + j 0.866 R R R R =0 =
81
The neutral current in a balanced three-phase system is zero! So why the wire? Yes, why? We dont need the neutral in balanced three-phase systems, so lets save metal and omit it.
3.2.2
CONDUCTOR REQUIREMENTS
The outcome of this is that a single-phase system requires two conductors, but a three-phase system requires only three. Hmmm, is this perhaps a reason why we use three-phase systems to deliver large amounts of power? I am going to compare the amount of metal needed for conductors for a three-phase system versus a single-phase system. Here are the ground rules: The loads in both systems require the same amount of power, which in this illustration is P watts. The line voltage, which is the voltage between the active conductors, is the same in both systems. That voltage is VS The current density, which is measured in amperes per cross-sectional area of the wire, must be the same in both systems. Figure 3.4 is the single-phase system that supplies a voltage VS to the load P. The current in
Figure 3.4: Copper calculation1/ 0
the single-phase wires is

P = VS I1 so I1 = P VS
Figure 3.5 is the equivalent three-phase system. The load is divided into three parts, each absorbing P/3 watts. The voltage at the ideal generators is VS / 3, which makes the voltage between the wires VS . (The reason for this will be answered later.) The current in each of the three-phase wires is
3P P VS = = I 3 so I 3 = 3 3VS 3 P 3VS
82
Figure 3.5: Copper calculation3/ 0
Now we equate the current densities for the two systems. A1 and A3 are the cross-sectional areas of the conductors in the two systems:
I1 I 3 = A1 A3
Substitute for current and cancel common terms to relate the cross-sectional areas:
P VS P 3VS = A1 A3 A1 = 3A3
The volume of metal required for the two conductors of the single-phase system is
Vol1 = 2 A1C = 2 3A3C
where is the length of one of the conductors. The volume of metal for the three conductors of the three-phase system, spanning the same distance , is
Vol3 = 3A3C
Compute the ratio of these two volumes of metal

Vol3 3A3C = = 0.866 Vol1 2 3A3C
The equivalent three-phase system requires only 86.6% as much metal to deliver the same amount of power at the same voltage between the wires. Thirteen-plus percent savings doesnt sound like a bad deal! There are other reasons for using three-phase that we will encounter later.
83
3.2.3
VOLTAGES AND CURRENTSTERMINOLOGY
Youve already learned that there is some common terminology that everyone uses when talking about a-c power, terms like power and reactive power. Three-phase power is no different. You simply have to learn to use certain terms properly or you will get yourself and anyone you talk with mixed up. The special terms in three-phase systems apply to voltages and currents and where they are observed. We will talk about line currents and line voltages and phase currents and phase voltages. We will use line and phase in a very consistent manner: Line always refers to the wires of the three-phase system. Phase always refers to the load in a three-phase system.
3.2.4
Y-Y CONNECTIONS
Note the four labels for voltages and currents in Fig. 3.6. The generator is connected in a Y cong-
Figure 3.6: 3/ 0 Y-Y connection
uration (as it almost always will be); the load is also connected in a Y conguration. In the drawing are two labeled currents: Iline is the current in the wire that connects the a and A nodes. (Ive labeled just the aA current.) Iphase is the current through the A phase of the load. Notice that the line and phase currents are the same:
I line = I phase
This applies to Y-connected loads. In the drawing of Fig. 3.6 are two labeled voltages:
84
Vline is between the A and B phases; there are two other line voltages but I didnt label them. Vphase is the voltage across one phase of the load. Now write Kirchhoff s Voltage Law around the loop a-n-b-a:
Vline = VS 0 VS 120 = 3 Vphase
The outcome is that the magnitude of the line voltage for Y-connected loads is phase voltage.
3 times the
3.2.5
Y-
CONNECTIONS
Theres another common way to connect a three-phase load, namely, in a delta-shaped conguration. The generator in Fig. 3.7 is still Y connected but the load is now connected.
Figure 3.7: 3/ 0 Y-
connection
Notice that the line and phase voltages are the same:
Vline = Vphase
Applying Kirchhoff s Current Law at the A node makes the line current in that phase equal to the phase current from A to B minus the phase current from C to A. The angles of these are consistent with the three-phase angular steps of 120:
I line = I AB I CA 240 = 3 I phase
The outcome is that the magnitude of the line current for phase current. Heres a summary:
-connected loads is
3 times the
3.3. POWER IN THREE-PHASE SYSTEMS
85
Y connection I line = I phase Vline = 3 Vphase
connection I line = 3 I phase Vline = Vphase
3.3
POWER IN THREE-PHASE SYSTEMS
Now that you have the secret handshake of the three-phase society, you are about to become a full member. The secret word is about to be passed to you. (Whats the handshake, you ask? Understanding line and phase.)
3.3.1
P AND Q AND S
I am going to get us to the secret word by beginning with the Y-Y system of Fig. 3.6. The power absorbed by one phase of this three-phase load must be
P1 = Vphase I phase pf
We learned that when we studied plain ol single-phase a-c power. Since all three phases of this three-phase system have the same load, the total load must absorb
P3 = 3 Vphase I phase pf
Use the Y voltage and current relationships to get this three-phase power equation into line terms:
P3 = 3 Vline I line pf 3
= 3 Vline I line pf
Now well do the same thing for the Yphase of the load must be
system of Fig. 3.7. The power absorbed by one
P1 = Vphase I phase pf
The three-phase load is three times as much:

P3 = 3 Vphase I phase pf
If we use the
voltage and current relationships to get this equation into line terms, we get
P3 = 3 Vline I line pf 3
= 3 Vline I line pf
86
Gosh! The power is the same whether our system is Y-connected or -connected, provided we express the voltages and currents in line terms. Reactive power Q and total power S also come out the same, whether Y or , if we use line terms. Heres a summary of all the three-phase power relationships:
Q = 3 Vline I line sin ( v i )
* S = 3Vline I line = P + jQ
P = 3 Vline I line pf , pf = cos ( v i )
Oh, you are still wondering what the secret word is? Its square-root-of-three. Leave that out of three-phase power equations and youll have wrong results. See how important it is? Now you are a full member of the three-phase society, but you still have an education program ahead of you to become a practicing member.
3.3.2
EXAMPLE IP, Q, AND S
A certain 460-volt, 60-hertz three-phase Y-connected load draws a line current of 65 26 A. Find P, Q, and S for this three-phase load. Before I start, we need a couple of ground rules. First, Voltage in a three-phase system is always stated as the line voltage! A corollary is that the phase voltage is never stated. (Duh!) So the stated voltage of 460 V must be the line voltage for our power formulas:
P = 3 Vline I line pf = 3 ( 460 )( 65 ) cos ( 0 ( 26 )) = 46.55 kW Q = 3 Vline I l ine sin ( 0 ( 26 )) = 22.70 kVAR
Ive chosen 0 as the angle for the specied load voltage. Now lets use P and Q to get S:
S = P + jQ = 46.5 + j 22.70 kVA = 51.79 26 kVA
3.3.3
EXAMPLE IILINE CURRENT
Another three-phase system operates at 8,100 volts and 60 hertz. The -connected load absorbs 285 kW at 0.85 power factor lagging. Find the line current. What do we know here? The power absorbed by the three-phase load. The power factor. The line voltage. Notice that knowing the connection (Y or ) is extraneous information. Start with the equation for P:
3.4. SINGLE-PHASE EQUIVALENT
87
P = 3 Vline I line pf
Compute Iline from that:

I line = = P cos 1 ( pf ) 3 Vline pf 285 10 3 = 23.9 31.8 A 3 ( 8100 )( 0.85 )
Theres something hiding in this example. Remember that we cant get the sign of the angle by just looking at the arc cosine? We have to consciously insert the correct sign into the phase angle. Here, the power factor is lagging so the load must be inductive. That means that the current lags the voltage so its angle must be negative relative to the angle of the voltage.
3.4
SINGLE-PHASE EQUIVALENT
Ah, a mental test for you! In the two examples that Ive just done, in how many ways can you be sure that I am dealing with a three-phase system? I think there are only three: I said three-phase, I used power equations with 3, and I used the words line and phase. Other than those, the examples could have been single-phase questions. In balanced three-phase systems, the three phases are identical except for a rotation of 120 per phase. Therefore, except for the 3, solving the three-phase problem is just like solving an equivalent single-phase problem. Thats where we are headed. It is not hard to convert any balanced three-phase circuit to an equivalent single-phase circuit. Once weve converted the circuit, we can do all calculations using our single-phase abilities, then convert the results so they apply to the three-phase system. For example, if I calculate the power absorbed by a certain single-phase-equivalent load, I just multiply the outcome by 3 to get the result for the original three-phase system. Now we need an algorithm for creating the single-phase equivalent.
3.4.1
GENERATING THE SINGLE-PHASE EQUIVALENT
Before getting to the algorithm, I need to state some ground rules: The three-phase system is balanced, meaning the three segments of the load are identical and the wire impedances in each phase are the same. If a neutral wire is present, its impedance will be dened to be zero since it carries no current. Systems are either Y-Y or Y- . Here are the steps of the algorithm:
88
1. Draw the single-phase equivalent: generator on left, empty box on right, two connecting wires. 2. Divide any voltages by 3. Most commonly, this will be the voltage across the load. 3. In the top wire only, draw and label the wire impedance, generally a series resistor and inductor. The bottom wire represents the neutral, which doesnt carry any current in a balanced threephase system. 4. Write the load description into the empty box: 4a. If the load is stated as power (10 kW at 0.90 lagging, for example), divide the power by 3 and write it into the box. 4b. If the load is stated as power per phase, write that into the box. 4c. If the load is Y-connected and stated as impedance per phase, write that impedance into the box. 4d. If the load is -connected and stated as impedance per phase, divide this impedance by 3 and write the result into the box. 5. Use the resulting single-phase equivalent to solve for whatever quantities are requested. 6. Convert the results back to three phase: 6a. Multiply any voltages by 3. 6b. Multiply the magnitude of any power results by 3. 6c. Line current, efciency, and voltage regulation are unchanged. 7. Answer any other three-phase questions (phase voltages and currents, for example).
3.4.2
EXAMPLE IIITHREE TO ONE
Figure 3.8 is a 480-volt three-phase system with a 25-kW load at 0.86 power factor lagging. Wire impedances are shown. We are to nd the line current, power losses, percent efciency, generator voltage, percent voltage regulation, and generator power factor. Figure 3.9 is its single-phase equivalent. Here are the steps of the algorithm: 1. Figure 3.9 has a generator and a box for the load. 2. The load voltage is 480/ 3 = 277 volts. 3. The wire impedance is inserted into the top wire. 4. Since the load is stated in power terms, Ive divided 25 kW by 3 and written this into the box. 5. Now come all the calculations:
3.4. SINGLE-PHASE EQUIVALENT
89
Figure 3.8: Example IIIthree-phase system
Figure 3.9: Example IIIone-phase equivalent
As a general rule, youll need the line current rst, so calculate it from the power, voltage, and power factor:
I line = = Pload cos 1 ( pf ) Vload pf 8.33 10 3 cos 1 ( 0.86 ) = 34.97 30.7 A ( 277 ) ( 0.86 )
Note the inserted minus sign for the arc cosine, because the power factor is lagging. This is an easy-to-miss step. The only losses are heating of the line resistance. Since for resistors the voltage and the current are in phase, Ill use R|I|2 , where R = 0.2 and |I| = 34.97 A:
Ploss = Rline I line = ( 0.2 )( 34.97 ) = 244.6 W
2 2
Efciency is useful power over useful power plus losses:
90
% = 100
8.33 10 3 Pload = 100 = 97.1% 8.33 10 3 + 244.6 Pload + Ploss
The single-phase generator voltage Van is the load voltage (277) plus the voltage drop along the line. The voltage drop is the line impedance times the current. Phase angles count when you do impedances.
Van = 277 + ( 0.2 + j 0.5 ) ( 34.97 30.7) = 277 + 18.8 37.5 = 292.2 2.2 V
The voltage regulation is

%VR = 100 Van Vload 292.2 277 = 100 = 5.5% Vload 277
Power factor is based on the angle between the voltage and the current. When calculating for the generator, you have to be careful to use the correct angles, namely, the angle of the generator voltage and the angle of the current:
pfgen = cos ( v i ) = cos ( 2.2 ( 30.7 )) = 0.840 laggi ing
Now we resume our regular algorithm: 6. Return the results to the three-phase system: The line current is the same: Iline = 34.97 0.7 A. The power losses are multiplied by 3 to yield the system power loss: 3 244.6 = 734 W. The percent efciency is unchanged: 97.1% The generator voltage must be multiplied by 3 to get the proper three-phase line voltage:
Vgen = 3Van = 3 292.2 2.2 = 506.12.2 V
The percent voltage regulation (5.5%) and the generator power factor (0.840 lagging) are unchanged.
3.5
3.5.1
MORE EXAMPLE
EXAMPLE IV -CONNECTED LOAD
A 4,160-V, 60-Hz, 3 system supplies a 720-kW load at 0.82 lagging power factor through lines with a line impedance of 1.5 + j3.6 /. ( is often written for phase.) Find the line current, the percent efciency, the percent voltage regulation, and the generator power factor, and generator voltage.
3.5. MORE EXAMPLE
91
Figure 3.10: Example IVone-phase equivalent
Figure 3.10 is the single-phase equivalent. The load voltage is 4160/ 3 = 2400 V. The load is 720/3 = 240 kW. The line impedance is for one of the wires. The line current comes from the power, inserting a minus sign into the arc cosine to account for the lagging power factor:
I line = = Pload cos 1 ( pf ) Vload pf 240 10 3 cos 1 ( 0.82 ) = 122.0 34.9 A ( 2400 ) ( 0.82 )
The only loss in the system is heating the wire:

Ploss = Rline I line = (1.5 )(122.0 ) = 22.3 kW
2 2
Note that there is no phase angle in the calculation because for a resistor, voltage and current are in phase. To say this another way, resistors dont know nuthin about phase angles. The percent efciency is
% = 100 Pload 240 = 100 = 91.5% Pload + Ploss 240 + 22.3
The single-phase voltage Van is the sum of the load voltage and the voltage drop across the line impedance:
Van = 2400 + (1.5 + j 3.6 ) (122.0 34.9) = 2400 + 475.8 32.5 = 28135.2 V
This makes the percent voltage regulation

%VR = 100 2813 2400 = 17.2% 2400
The generator power factor, using the angles of the generator voltage and the line current, is
92
pfgen = cos ( 5.2 ( 34.9 )) = 0.765 lagging
Finally, the generator voltage in the original three-phase system is

Vgen = 3Van = 3 28135.2 = 4872 5.2 V
Line current, percent efciency, percent voltage regulation, and generator power factor are the same.
3.5.2
EXAMPLE VY-CONNECTED IMPEDANCE
A 208-V, 60-Hz, 3, Y-connected load has a per-phase impedance of 2 + j1.3 /. The wire impedance is 0.25 + j0.1 /. Find the line current, the power factor of the load, and the percent efciency of the system. Figure 3.11 is the single-phase equivalent. The voltage has been divided by 3, the Y-
Figure 3.11: Example Vone-phase equivalent
connected impedance has been put into the load box without change, and the wire impedance is shown in the top wire only. The line current, in both the single-phase equivalent and the three-phase system, is the load voltage (taken with a phase angle of 0) divided by the load impedance:
I line = 120 = 50.3 33.0 A 2 + j1.3
The power factor of the load comes from the phase angle of the load voltage (chosen as 0) and the phase angle of the line current.
pf = cos ( 33.0 ) = 0.839 lagging
Lagging is appended because the load is inductive (the imaginary part of the impedance is positive). The power absorbed by the single-phase load is
3.5. MORE EXAMPLE
93
Pload = (120 )( 50.3)( 0.839 ) = 5.06 kW
and the losses in the line are

Ploss = ( 0.25 )( 50.3) = 632.5 W
2
which makes the percent efciency

% = 100 5.06 10 3 = 88.9% 5.06 10 3 + 632.5
3.5.3
EXAMPLE VI -CONNECTED IMPEDANCE
The system for this example is the same as that for the previous example (Example V) but the load impedance is connected in fashion. Find the same quantities. Figure 3.12 is the single-phase equivalent. Its the same as before, except that the load
Figure 3.12: Example VIone-phase equivalent
impedance has been divided by 3 because it is The line current has gotten a lot larger:
I line =
-connected.
120 = 150.9 33.0 A 0.667 + j 0.433
The power absorbed per phase is a lot larger, too:

Pload = (120 )(150.9 ) ( cos ( 33.0 )) = 15.19 kW
In fact, this result is exactly three times that for the Y-connected impedance. The line loss is
Ploss = ( 0.25 )(150.9 ) = 5.69 kW
2
so the overall efciency is

% = 100 15.19 = 72.7% 15.19 + 5.69
This is low because the wiring is not large enough to efciently carry the large current.
94
3.5.4
EXAMPLE VIITHREE-PHASE MOTOR
A three-phase 2400-volt, 250-horsepower induction motor is 85% efcient. Its rated power factor is 83%. The line impedance is 1.6 + j0.56 ohms per phase. Find the overall efciency of the system. First, we need to convert the motor output (250 hp) to motor power input. The conversion to watts is 746 W/hp. We divide by the efciency to get the input power:
Pmotor = 250 746 = 219.4 kW 0.85
The single-phase-equivalent load voltage is

Vload = 2400 = 1386 V 3
The power absorbed by one phase of this load is

Pmotor / phase = 219.4 3 = 73.14 kW @0.83 lagging
The single-phase equivalent is Fig. 3.13.
Figure 3.13: Example VIImotor equivalent
We need the line current to get the line loss:

I line = 73.14 10 3 cos 1 ( 0.83) = 63.58 33.9 A (1386 )( 0.83)
The line loss (resistive) is

Ploss / phase = (1.6 )( 63.58 ) = 6.47 kW
2
Hence, the percent efciency of the electrical system is

% = 100 73.14 = 91.9% 73.14 + 6.47
The overall efciency is the product of the electrical and motor efciencies:
overall % = 0.919 0.85 = 78.1%
3.6. SUMMARY
95
3.5.5
EXAMPLE VIIIPOWER-FACTOR CORRECTION
For the motor of Example VII, specify the number of kilovars per phase needed to improve the power factor to 0.9. Figure 3.14 is the power triangle to help us calculate the reactive power.
Figure 3.14: Power factor calculation
Example VII has a reactive power per phase of

Qload / phase = 73.14 tan cos 1 ( 0.83) = 49.15 kVAR
which I got by looking at the triangle and remembering the denitions of tangent and cosine. The reactive power must be reduced to
Qnew / phase = 73.14 tan cos 1 ( 0.9 ) = 35.42 kVAR
to achieve a power factor of 0.9 lagging. The capacitors needed must provide negative reactive power of
QC / phase = 35.42 49.15 = 13.73 kVAR/
3.6
SUMMARY
Why study three-phase power? Because it is so common, the method for almost all delivery of large amounts of energy. It also requires only 86.6% as much metal for a given amount of energy delivery at the same line voltage compared with single-phase power. Terminology is very important. The engineer must use the terms line and phase in order to be understood. Line refers to the connecting wires; phase refers to the load.
96
All power calculations for real power P, reactive power Q, and total power S require a 3 in their equations, along with the magnitudes of the line voltage and the line current. Voltages in three-phase systems are always stated as line voltages, the voltage between the line wires. This is done everywhere in the system. The stated voltage is generally going to be the load voltage, the voltage the load is designed to operate properly on. Since we are restricting our study to balanced three-phase systems, we can reduce any system to its single-phase equivalent, solve for any requested quantities, and then convert back to three phase.
97

1. Current and voltage relationships for Y and
Y connection I line = I phase Vline = 3 Vphase
connections
connection I line = 3 I phase Vline = Vphase
2. Three-phase power equations

P = 3 Vline I line pf , pf = cos ( v i )
* S = 3Vline I line = P + jQ
Q = 3 Vline I line sin ( v i )
3. Three-phase commandment Voltage in a three-phase system is always stated as the line voltage! 4. Three-phase equivalent algorithm (a) Draw equivalent: generator on left, connecting wires, empty box for load on right (b) Divide any voltages by 3 (c) In top wire only, insert wire impedance (generally R and jX) (d) Write load description into empty box i. ii. iii. iv. If load is stated as power, divide by 3 and write into box If load is stated as power per phase, write into box as is If load is Y-connected and stated as impedance per phase, write into box If load is -connected and stated as impedance per phase, divide by 3 and write into box
(e) Use this equivalent to solve for whatever is requested (f ) Convert back to three phase i. Multiply any voltages by 3 ii. Multiply magnitude of any power results by 3 iii. Line current, efciency, and voltage regulation are unchanged
(g) Answer any other three-phase questions
99
CHAPTER
Transformers
Our power systems are based on alternating current. Why? Why not direct current? Well, if we ignore politics and personalities (Edison and the Westinghouse/Tesla teams fought for years), this comes down to two basic technical reasons: transformers and induction motors.This chapter is about transformers; the next is about induction motors. Transformers are pretty remarkable. Theres more than meets the eye here1 . They dont seem to do much, just transform one voltage into another. So what else is there? For one thing, they save metal. Yes, really. They make it possible to move large blocks of power with relatively small conductors. Since power is the product of voltage and current, they save metal by stepping up the voltage. That reduces the current for a given amount of power so smaller wires can do the job. The next time you are driving around, glance about and see how many transformers you can spot. Up on poles, in big switch yards, and so on. They make it practical for devices to run on low voltages, such as toy trains, voltages that are low enough to be non-lethal. They provide a way of automatically regulating voltages in power systems. They give us a way to match a load to a source, a topic beyond this text. Alright, we use transformers in good ways, but arent they wasteful? They do get warm; they require cooling in some manner. Are they efcient? Yes, they are surprisingly efcient, generally at least 95%. So perhaps there really is more than meets the eye here. We are going to look at the ideal transformer rst, one that is 100% efcient. Then well model the real transformer and see how to measure the parameters needed to do this. Finally, well embed transformers into power systems.
4.1
IDEAL TRANSFORMERS
A transformer is very simple. Take an iron core, wrap a couple of coils of wire around it, and you have a transformer. Figure 4.1 shows this. Note the current directions. On the left, the 1 side, the current is labeled into the transformer. On the other side, the 2 side, the current is labeled outward. This is done to imply that there is power moving from left to right. Now and then Ill call the left side the primary and the right side, the secondary. Energy ow is generally from the primary to the secondary.
1 It is left as an exercise for the reader to determine the source and relevance of this quote.
100
4. TRANSFORMERS
Figure 4.1: Transformer
The numbers of turns on the coils is given, an important parameter when describing a transformer and determining how it works. Instead of actually enumerating turns, we often state just the ratio of the turns of the two coils. Figure 4.2 is a cutaway photo of a common pole-mounted transformer.This one, compliments of Cooper Power Systems of Waukesha, Wisconsin, is kind of nondescript, but it is probably in the range of 10 to 25 kVA and a few thousand volts. How does the transformer work? Ill introduce a little about magnetic elds here, but we wont really deal with them until Chapter 6. The basic physics is described by Faradays Law:
v(t ) = d (t ) dt
This law says that a time-varying magnetic ux (t) produces a voltage in a loop of wire. Now apply this to the transformer using the voltages and currents described in Fig. 4.1:
d (t ) dt d (t ) v2 (t ) = n2 dt v1 (t ) = n1
Thats not enough, though, because it doesnt include the currents in the drawing. That comes from Amperes Law, although Ill omit the details here:
(t ) n1i1 (t ) (t ) n2i2 (t )
We are going to deal with a-c power, so its appropriate to switch into phasor notation. Figure 4.3 shows both phasor voltages and currents as well as a schematic way of drawing the transformer. We really dont need the details of how the coils and the core t together.
4.1. IDEAL TRANSFORMERS
101
Figure 4.2: Transformer
4.1.1
VOLTAGE AND CURRENT RELATIONSHIPS
If we take Faradays Law for both sides of the transformer and divide one equation by the other, the derivative of the ux cancels out. Thats because its the same ux all the way around the core. Lets switch to phasor notation at the same time. The result is the relationship between turns and voltage:
V1 n1 = V2 n2
This says, simply, the voltages are related by the turns ratio. For example, if n1 = 1000 and n2 = 500, then V1 will be twice as large as V2 If we take Amperes Law, equate the two equations, and then change to phasor notation, we get the relationship between turns and current:
102
4. TRANSFORMERS
Figure 4.3: Transformer symbol in phasor domain

I1 n2 = I 2 n1
Notice that the order of the subscripts in the turns ratio is opposite that for voltage. For n1 = 1000 and n2 = 500, I1 will be half as large as I2
4.1.2
POWER RELATIONSHIP
Our transformer is ideal, which among other things says that there is no energy lost by placing the transformer in a circuit. Hence, power delivered to the primary (the 1 side) of the transformer must all come out of the secondary (the 2 side). In other words, power is conserved by the ideal transformer. If power is conserved, then total power S must also be conserved. Lets write total power for both sides of the transformer and see what happens:
* (* means complex conjugate) S1 = V1 I1 S = V I*
2 2 2
Now replace V2 and I2 by the equivalents that we just gured out:

n n * * S2 = 2 V1 1 I1 = V1 I1 n1 n2 S2 = S1
Good! Total power is conserved so power passes through the transformer unchanged. What goes in comes out. It may come out with different values of voltage and current, but their product is still the same.
4.1.3
IMPEDANCE
Suppose we connect an impedance Z2 to the secondary side of our transformer. Figure 4.4 shows this connection. When we consider the labeled voltage and current, we can write the impedance as
103
Figure 4.4: Transformer with load

V2 I2
Z2 =
Replace V2 and I2 by their equivalents and we get

n2 V1 2 2 n V n n Z 2 = 1 = 2 1 = 2 Z1 n1 I1 n1 I1 n1 n2
If we turn the equation around, we see that this secondary-side impedance Z2 looks, from the primary side, like it has the value
n Z1 = 1 Z 2 n2
2
We call the process of seeing the impedance through the transformer reecting the impedance through the transformer. Notice in Fig. 4.5 that the impedance of Fig. 4.4 has been drawn on the
Figure 4.5: Load reected
primary side. It and the voltage and the current have been given the same labels but with a prime attached. In Fig. 4.5,
104
4. TRANSFORMERS
V2 =
n1 V2 n2 n = 2 I2 I2 n1
2
n = 1 Z2 Z2 n2
Note the order of all the subscripts!
4.1.4
EXAMPLE ITRANSFORMER APPLICATION
A 240-volt load requires 15 kW at a power factor of 0.84 lagging. The two conductors feeding the load have a combined impedance of 0.1+j0.08 ohms.These are supplied through an ideal transformer with a turns ratio of 8100:240. We are to nd the generator voltage Vin and current Iin , and the power supplied by the generator Pin The system is shown in Fig. 4.6. The 15-kW load requires 240 volts, so this is the typical
Figure 4.6: Example I
a-c power problem where we are given the needs of the load and have to gure out what must be supplied. Although each conductor feeding the load has an impedance, their total impedance is combined into the R-L elements in one conductor. Many of these power problems start with nding the load current Iload . Everything else comes from knowing this current. Recall that P = |V||I| pf, so the line current (with a negative angle because the power factor is lagging) is
I load = 15000
(240)(0.84 )
cos1 0.84 = 74.41 32.9 A
The voltage V2 will be the load voltage plus the voltage drop across the line impedance:
V2 = 240 + ( 0.1 + j 0.08 ) 74.41 32.9 = 240 + 9.535.8 = 249.50.2 V
105
The voltage V2 goes through the transformer by the turns ratio, while the current Iload goes through by the inverse ratio:
Vin = 8100 249.50.2 = 84200.2 V 240 240 I in = 74.41 32.9 = 2.205 32.9 A 8100
Now we have the voltage and the current on the primary side of the transformer, so the power delivered by the generator is
Pin = Vin I in cos ( 0.2 (32.9)) = 15.55 kW
pfin = cos ( 0.2 (32.9)) = 0.838 lagging
So what did the transformer really do? Since in this example the transformer is ideal, all it did was reduce the supply voltage from an 8100-volt system to a 240-volt load. Real transformers do thatalmost. Well learn in the next section how to account for being real, but before that, lets look at another way to get results for this example.
4.1.5
EXAMPLE IIAPPLICATION REFLECTED
We have learned that voltage goes through by the turns ratio, current by the inverse of that ratio, and impedance by that ratio squared. Weve also learned that power is power and is conserved. The circuit of Fig. 4.7 is our previous example reected through the transformer.
Figure 4.7: Example II
The load voltage of 240 volts becomes

= Vload 8100 240 = 8100 V 240
The wiring impedance reects as
106
4. TRANSFORMERS
8100 Rv = 0.1 = 113.9 240 8100 Xv = 0.08 = 91.1 240

2
Now we have a regularpower problem without a transformer. The load current Iload is now the input current Iin :
= I load 15000 cos1 0.84 = 2.205 32.9 A (8100)(0.84 )
The generator voltage is the reected load voltage (8100 V) plus the drop across the reected line impedance:
Vin = 8100 + (113.9 + j 91.1) 2.205 32.9 = 8100 + 321.6 5.8 = 8420 0.2 V
The results are the same as before. Heres a summary of how the ideal transformer reects. This wont change when we model a real transformer: Voltage reects by the turns ratio Current reects by the inverse of the turns ratio Impedance reects by the turns ratio squared
4.2
REAL TRANSFORMERS
Ideal transformers sound like a good idea and they are neat to talk about. Alas, they cant be real. All we have to do is think about what transformers are made of to realize that we are missing some facts. After all, transformer coils are made of wire and wire has resistance. So the practical transformer must have losses due to the heating of that resistance. Also, these coils sure look like inductors, so wed expect practical transformers to exhibit some inductance as well.
4.2.1
REALISTIC MODEL
Thats what Ive drawn in the more-realistic transformer model in Fig. 4.8. This model adds resistances to represent power loss and inductance to represent the effects of coils and iron. This model is good at power-line frequencies, but it is incomplete for high-frequency transformers. Well stick with power here. Lets walk through the six elements of the model of Fig. 4.8: R1 represents the losses caused by the resistance of the primary winding. R2 is the similar resistance of the secondary winding. As current ows through the windings, these resistances account for i2 R heating that transformers exhibit.
4.2. REAL TRANSFORMERS
107
Figure 4.8: Transformer model
X1 and X2 are called leakage reactance. They represent the fact that current owing in the winding creates magnetic ux (the in our original discussion) but not all of that ux is in the iron core. Some of the ux leaks outside the core and makes the winding look partially like an inductance. Rc is core loss. Flux is sinusoidal because our currents and voltages are sinusoidal. This means that the magnetic domains in the core are ipping back and forth in step with the ux. But these domains resist this ipping, so the resistance Rc represents that energy loss.This constant ipping heats the core. Xm is the hardest parameter to describe. It is called the magnetization reactance. We can understand this in terms of the magnetic domains by noting that a ipped domain stores some energy because it took some energy to get it to ip and it will take some energy to get it to ip back. In concept, this sounds like an inductor, where current stores energy in the inductors magnetic eld (1/2Li2 ).
4.2.2
MODEL WITH NUMBERS
When we put realistic numbers into this transformer model, we nd that we can simplify things to some extent. Figure 4.9 has some numbers that came from a commercial 20-kVA, 4160:240-volt transformer.
Figure 4.9: Transformer model with numbers
108
4. TRANSFORMERS
Note the ranges of the numbers. Transformers are obviously designed to be efcient, so the wire resistances must be small to minimize heating. Likewise, we want low core losses, so that resistance should be large. We can simplify this model a bit, though, by reecting the secondary impedance, which is
Z 2 = 0.03 + j 0.016
Lets reect that impedance through the transformer by the turns ratio squared:
4160 v = Z2 ( 0.03 + j 0.016 ) = 9.0 + j 4.8 240
2
The result of this reection is the model of Fig. 4.10.
Figure 4.10: Transformer model with secondary reected
Now its time to get a little tricky. Notice the ranges of the numbers in the model: the series elements are small numbers like 10, while the parallel elements are big numbers like 10s of kilohms. The job of the transformer is to pass power from one side to the other while not wasting a lot inside. Its reasonable to expect that the currents owing downward through the parallel elements are much, much smaller than the power current owing through the series path. Here comes the tricky part. Moving the parallel elements to the right, placing them to the right of the j4.8- inductor, will have only a very small effect on the main current owing through the series elements. Well see how much of an effect this has in an example later. Moving the parallel elements to the right allows us to combine the series elements to produce the simplied model of Fig. 4.11. This is the model well use from here on.
4.2.3
EXAMPLE IIITRANSFORMER IN USE
The 20-kVA, 4160:240-V transformer that weve just been modeled is used in the power circuit of Fig. 4.12 (next page). The generator supplies the transformer through wiring that has a total impedance of 4.6 + j7.1 . The transformer feeds a load through wiring with a total impedance of 12 + j15 m . The load is the rated load for the transformer, namely, 20 kVA at 240 volts. The loads power factor is 0.82 lagging.
4.2. REAL TRANSFORMERS
109
Figure 4.11: Transformer model simplied
We are to nd the percent voltage regulation and the percent efciency for the overall system. Ive said a number of times that we start such problems by nding the load current. But with transformers involved, theres a step before doing that to simplify our work: get rid of the transformer! This example would be a straightforward work-from-right-to-left problem if the transformer didnt complicate things. Weve learned how to reect voltage and current and impedance through a transformer. Lets use what we know to reect the load from the secondary to the primary and eliminate the transformer: The load voltage reects easily, since its the rated voltage. 240 volts across the load becomes 4160 volts. The load current Iload simply gets a prime, Iload , which well calculate later. The wiring impedance reects by the turns ratio squared:
4160 v = Z2 ( 0.012 + j 0.015 ) = 3.6 + j 4.5 240
2
The result is the circuit of Fig. 4.13. Notice that Ive combined all of the series elements (4.6 + j7.1 + 19.5 + j9.0 + 3.6 + j4.5) after shifting the parallel elements to the right. The analysis proceeds from right to left. First, the reected load current:
110
4. TRANSFORMERS
Figure 4.13: Example III reected and combined

20 10 3 cos1 0.82 = 4.81 34.9 A 4160
= I load
The two parallel branches have currents down through them. The voltage across each of the parallel elements is 4160 V, so the current through each is that voltage divided by the branch impedance:
I parallel = 4160 4160 + = 0.128 58.5 A 62 10 3 j 38 10 3
The sum gives the line current:

+ I parallel = 4.81 34.9 + 0.128 58.5 I line = I load = 4.93 35.5 A
The generator voltage is 4160 plus the voltage drop across the series elements:
Vgen = 4160 + ( 27.7 + j 20.6 ) 4.93 35.5 = 43300.04 V
Now watch my hand closely while I perform some magic. We know that voltage regulation depends on knowing what the load voltage will be if we disconnect the load. Until now, disconnecting the load left the circuit with no current, so the voltage at the terminals where the load was will be the generator voltage. But look at the circuit of Fig. 4.13. Notice that disconnecting the load leaves those parallel branches in place. Hence, there is still a current owing. In fact, the voltage across them can be determined by using a voltage divider. Is this really necessary? Didnt I say as I developed the model of the transformer that those parallel elements represent very small currents compared with the power current in the series path? Yup. Which leaves me to wonder whether I can ignore those parallel elements now and just say that the voltage with the load disconnected is the generator voltage. If I can, this no-load voltage is 4330 volts.
4.3. TESTING
111
Lets try some numbers. If you take the circuit of Fig. 4.13 and calculate the voltage across those parallel elements with the load disconnected (i.e., for Iload = 0), it comes out 4326 volts. Hmmm, a 4-volt error on 4,000 volts. Thats 0.1%. Well ignore this. Besides, we generally dont know the parameters of the model to a precision of 0.1% anyhow. While were looking at the parallel elements, notice that the current through them is 0.128 A, which is only 2.6% of the load current. The percent voltage regulation, based on my simplication, is
%VR = 100 4330 4160 = 4.09% 4160
To nd percent efciency, I need to know the losses in the system. There are two places where power is lost as energy is turned into heat, namely, the two resistance elements. Its tempting to argue that the parallel element (62 k ) can be ignored, since weve ignored it before. We cant though, because it represents, in this example, about 30% of the power lost to heat:
Plosses = 27.7 4.932 + 4160 2 = 952 W 62 10 3
Ive never calculated the power absorbed by the load! Ive just used the kVA information, so I need load power:
Pload = ( 20 10 3 ) ( 0.82 ) = 16.4 kW
Now we have enough to get percent efciency:

% = 100 16.4 = 94.5% 16.4 + 0.952
Both percent voltage regulation and percent efciency are reasonable for a power circuit. Voltage regulation better than 5% is considered good, and efciency better than 90% is, too.
4.3
TESTING
Where do the numbers in the transformer model come from? After all, you cant just pry open the case and see a resistance and an inductance. These numbers come from measurements made from the outside. Its not unlike the physician who measures whats going on in your heart by observing from the outside. There are two basic transformer tests that do a good job of divining the values of the four internal elements, the open-circuit test and the short-circuit test. Both are made by measuring current, voltage, and power at the terminals of the transformer.
112
4. TRANSFORMERS
Figure 4.14: Open-circuit transformer test
4.3.1
OPEN-CIRCUIT TEST: RATED VOLTAGE
Lets start with the open-circuit test, whose set-up is shown in Fig. 4.14. Notice that the ammeter and the voltmeter measure the magnitudes of the current and the voltage but dont give us the phase angle. The wattmeter measures the power delivered, and from the three readings we can get all the information we need. In the open-circuit test, the secondary of the transformer is left open (i.e., not connected to anything). The primary is excited by its rated voltage. For example, if the transformer is rated as 4160:240, the primary is excited by 4160 volts. In the open-circuit tests, the current is very small because there is no load on the transformer. Therefore, the series elements have very little effect on the measurements and we are seeing just the parallel branches. From the open-circuit test we have three pieces of data: |Voc |, |Ioc |, and Poc . The calculations hinge on remembering that P = |V||I| cos(v i ). First, nd the angle of the current and then write the current as a phasor:
I oc = cos1 Poc Voc I oc
I oc = I oc I oc = I c jI m
We know the voltage across each of the parallel elements, so we can nd their impedances from the real and the imaginary parts of the current. (The voltage phase is chosen to be 0.) The two parallel elements are
Rc = Voc Ic , Xm = Voc Im
4.3.2
SHORT-CIRCUIT TEST: RATED CURRENT
Now we short-circuit the secondary, which means wiring the terminals together as in Fig. 4.15 below. We apply a small voltage to the primary, just enough to drive rated current into the transformer. This voltage will be much, much smaller than rated voltage.
4.3. TESTING
113
Figure 4.15: Short-circuit transformer test
Because the applied voltage is so small, the parallel elements are only a minor factor and will be ignored. But the series elements are carrying rated current, so we will be able to get their values in the same way as we did for the parallel elements. Again we get the phasor current from our measurements:
I sc = cos1 Psc Vsc I sc
I sc = I sc I sc
The impedance of the series elements must be the applied voltage (taken as 0) divided by the phasor current:
R + jX = Vsc I sc
Calculating the real and imaginary parts of this series impedance yields the values of the two elements:
V V R = Re sc , X = Im sc I sc I sc
4.3.3
EXAMPLE IVTRANSFORMER TEST
A transformer similar to the one Ive used in previous examples is tested using the open-circuit and short-circuit tests. The transformer is a 20-kVA, 4160:240 volts. Rated voltage for the open-circuit test is 4160 V. Rated current for the short-circuit test is 20 kVA/4160=4.81 A. Here are the test outcomes: Voc = 4160V Ioc = 0.155A Poc = 292W Vsc = 121V Isc = 4.81A Psc = 509W
114
4. TRANSFORMERS
From the open-circuit test, the angle of the current is

I oc = cos1 292 = 63.1 4160 ( )(0.155)
so the open-circuit phasor current is

I oc = 0.155 63.1 A
which in rectangular form is

I oc = 0.0702 j 0.1382 A
From this we get the values of the parallel elements:

4160 = 59.3k 0.0702 4160 Xm = = 30.1k 0.1382 Rc =
If you are wondering what happened to a minus sign in the Xm calculation, it disappeared legitimately! If you carry the j with the calculation for Xm , you can see where the sign went:
jX m = 4160 4160 =j j 0.1382 0.1382
Now we use the short-circuit data to get the angle of the short-circuit current
I sc = cos1 509 = 29.0 121 ( )( 4.81)
and then the phasor for this current:

I sc = 4.81 29.0 A
From that current and the voltage (again taken as 0) comes the series impedance:
R + jX = 121 = 25.1629.0 4.81 < 29.0 = 22.0 + j12.2
That separates into the two series elements:

R = 22.0 ,X = 12.2
Ill do another example of this in the next section.
4.4. MORE EXAMPLE
115
4.4
MORE EXAMPLE
Ill nish off this chapter on transformers with six examples that involvetransformers, of course. Theres more than meets the eye.
4.4.1
EXAMPLE VLOAD AT UNITY POWER FACTOR
A 10-kVA, 480:240/120-volt supplies 240 volts to a rated load with unity power factor. Find percent voltage regulation and percent efciency for this arrangement. The transformer and load are shown in Fig. 4.16 with transformer parameters.
Figure 4.16: Examples V, VI, VII, and VIII
What does rated load mean? It means the load requires 10 kVA without specifying the power factor. Since the power factor here is unity, the load power P must be 10 kW. If the power factor were less than unity, P would be less than 10 kW. Im not going to draw a new circuit with the load reected to the primary because I think it is easy to visualize without the drawing. Think of the problem this way: If the transformer rating is 10 kVA and its supplying rated load, that rating is the same on both sides of the transformer. Therefore, I can calculate rated current on the primary side by dividing total power by the primary voltage:
I rated = 10 10 3 = 20.8 A 480
Ill ignore the parallel branches to nd the input voltage:

Vin = 480 + (1.04 + j 0.62 ) 20.80 = 480 + 25.230.8 = 501.8 81.5 V
That gives me percent voltage regulation:

%VR = 100 501.8 480 = 4.5% 480
116
4. TRANSFORMERS
which is good (less than 5%). The only loss in the transformer is the series resistance that accounts for wire heating. The parallel branch has only inductance; the resistance is given as innity (i.e., very large) and can be ignored. So losses are
Plosses = 1.04 20.8 2 = 450 W
which leads to percent efciency:

% = 100 10 = 95.7% 10 + 0.450
Thats a good number, too (over 90%). But lets see what happens to these numbers with a poorer power factor.
4.4.2
EXAMPLE VILOAD AT LOWER POWER FACTOR
The transformer is the same as in the previous example but the load now has a power factor of 0.78 lagging. We are again to nd percent voltage regulation and percent efciency. The load is now smaller. It is still rated 10 kVA, but as the power triangle in Fig. 4.17 shows, the power P is now
Pload = (10 10 3 ) ( 0.78 ) = 7.8 kW
Figure 4.17: Example VIpower triangle
The current still has the same magnitude but now it has a lagging phase angle:
I line = 10000 cos1 0.78 = 20.8 38.7 A 480
The calculation of the input voltage must include that phase angle:
Vin = 480 + (1.04 + j 0.62 ) 20.8 38.7 = 480 + 25.2 7.9 = 505 0.4 V
4.4. MORE EXAMPLE
117
Now the percent voltage regulation is

%VR = 100 505 480 = 5.2% 480
A lower power factor leads to poorer voltage regulation. Losses havent changed:
Plosses = 1.04 20.8 2 = 450 W Pload = 7.8 kW
But because real power P is smaller, efciency is worse:

% = 100 7.8 = 94.5% 7.8 + 0.450
A lower power factor leads to lower efciency. Is ignoring the current through the parallel branches good, bad, or just OK. Lets check.
4.4.3
EXAMPLE VIIPARALLEL-BRANCH CURRENT
Continue with the previous example. The transformer is still rated at 10 kVA, 480:240/120. The load is still 10 kVA at 0.78 lagging power factor. How much current ows through the parallel branches and is it much of a factor in our calculations? Figure 4.18 is the circuit.
We already know the load current reected into the primary:

' I load = 20.8 38.7 A
If I ignore the parallel branch, the line current Iline = Iload . What will Iline become if I include the magnetizing current Im through the j3.5-k inductance? The voltage across this inductance is 480 V, so the magnetizing current is
118
4. TRANSFORMERS
Im =
480 = 0.14 90 A j 3.5 10 3
Adding this to the load current makes the line current larger:
I total = 20.8 38.7 + 0.14 90 = 20.9 39 A
But not by much! 20.9 instead of 20.8 and a phase-angle difference of just 0.3. Thats why we ignore the parallel-branch currents when we calculate the other currents. Note that you cant ignore the losses in the parallel branch, thoughif there are any. In this transformer that resistor is said to be innity, which says the losses are too small to be concerned about.
4.4.4
PECULIAR TRANSFORMER?
The transformer that has been part of the previous three examples might seem peculiar. Take a look at the secondary side of this transformer in Fig. 4.16. Notice that is seems to have an extra terminal labeled N . The other two terminals are labeled A and B . Whats different here is that this arrangement is the common way of supplying power to your home. Lines A and B are the line wires; N is the neutral. From either of the line wires to the neutral, the voltage is 120 V. Between the two line wires, the voltage is 240 V. These supply the typical loads in a home: 120 volts for the usual appliances and lights, 240 volts for the air-conditioner and perhaps the water heater and the kitchen range. The primary voltage in your neighborhood will be much higher than the 480 volts shown in this system, perhaps 4160 V or even higher. While we are still looking at this transformer, lets see what the outcomes of the open-circuit and short-circuit tests would be.
4.4.5
EXAMPLE VIIITESTING
The transformer is still the 10-kVA, 480:240/120 transformer Ive been using. Our job is to conduct the open- and short-circuit tests in reverse via calculations. Figure 4.19 shows the arrangement of this transformer for an open-circuit test. Nothing is connected to the secondary so I didnt even draw the transformer windings. The current Ioc is going to be very small, so we ignore any effects of the series elements. The test requires applying rated voltage, which is 480 V:
Voc = 480 V
The current through the parallel branch is the open-circuit current. That current will be the applied voltage divided by the impedance:
I oc = 480 = 137.1 mA 3.5 10 3
4.4. MORE EXAMPLE
119
Figure 4.19: Open-circuit
Since the resistance of the parallel branch is an open-circuit, it dissipates no power:

Poc = 0
For the short-circuit test, we short out the secondary. This reects as a short circuit to the primary. See Fig. 4.20. The parallel-branch elements are shorted out and have no effect, leaving only
Figure 4.20: Short-circuit
the series elements. [There is some slight-of-hand here. In reality, the parallel elements are not shorted out. They only appear so because they have been slid to the right of some of the series elements. Nevertheless, the currents through them are small enough to be ignored.] Weve already calculated rated current, which we now apply:
I sc = 20.8 A rated
The input voltage is the voltage that the current develops across the series elements:
Vsc = (1.04 + j 0.62 ) ( 20.80) = 25.18 V
120
4. TRANSFORMERS
The power reading for this test comes from the heating of the series resistance:
Psc = 1.04 20.8 2 = 450 W
And those are what the meters would show.
4.4.6
EXAMPLE IXTRANSFORMER TESTS
A 25-kVA, 2300:230-volt transformer is subjected to open- and short-circuit tests. The open-circuit test is done at rated voltage, 2300 V. The short-circuit test is done at rated current, 25kVA/2300 = 10.87 A. The tests produced the following data: Voc = 2300V Ioc = 110mA Poc = 203W First, the open-circuit current as a phasor is
I oc = cos1 203
Vsc = 124V Isc = 10.87A Psc = 1070W
(2300)(0.110)
= 36.6
I oc = 110 36.6 mA = 88.3 j 65.7 mA
From the rectangular form of the open-circuit current we get the parallel elements:
2300 = 26 k 88.3 10 <3 2300 Xm = = 35 k 65.7 10 <3 Rc =
Next comes the short-circuit current as a phasor:

I sc = cos1 1070 = 37.5 124 ( )(10.87)
I sc = 10.87 37. .5 A
Then we get the impedance of the series elements:

Z= 124 = 11.4137.5 10.87 < 37.5 = 9.06 + j 6.94
4.4. MORE EXAMPLE
121
Finally, we get the series elements from the rectangular form of their impedance:
R = 9.06 ,X = 6.94
The resultant model is Fig. 4.21. Now Im going to use this transformer in a circuit.
Figure 4.21: Results of Example IX
4.4.7
EXAMPLE XTESTED TRANSFORMER IN USE
The transformer we just tested, a 25-kVA, 2300:230-volt transformer, is connected to supply an 18-kW load whose power factor is 0.88 lagging. The feeders connecting the load have together an impedance of 18 + j21 m . The feeders supplying the transformer have together an impedance of 2.8+j1.7 . Figure 4.22 shows the complete system.
Figure 4.22: Example X
To begin this, Im going to reect the load and its wiring through the transformer to the primary. The load remains 18 kW at 0.88 lagging. Its voltage becomes 230(2300/230) = 2300 volts. The feeder impedance becomes
(18 + j 21) 10<3
2300 = 1.8 + j 2.1 230
Then Im going to slide the parallel branches out of the way to the right and add all the series elements. The result is Fig. 4.23.
122
4. TRANSFORMERS
Figure 4.23: Example X reected and collected
Start with the current drawn by the load:

I load = 18 10 3 cos1 0.88 = 8.893 28.4 A (2300)(0.88)
From this, nd the voltage drop caused by this current owing through the series elements and add the 2300-V load voltage to get the generator voltage:
Vgen = 2300 + (13.66 + j10.74 ) 8.893 28.4 = 2300 + 154.59.8 = 24520.6 V
I ignored the parallel branches while computing the current, so Ill ignore them again when computing the percent voltage regulation:
%VR = 100 2452 2300 = 6.61% 2300
i2 R
But I cant ignore the losses the parallel branch represents, so my power-loss calculation includes loss in the series resistance and v2 /R loss in the parallel branch:
Plosses = 13.66 8.8932 + 2300 2 26 10 3 = 1080 + 203 = 1283 W
Notice that the loss in the parallel branch is about 16% of the total losses.The percent efciency is
% = 100 18 = 93.3% 18 + 1.283
4.5
SUMMARY
Is there more than meets the eye? A transformer is a pretty simple contraption that performs major service in the a-c power system, modifying voltage and current to t the systems needs while passing power through efciently.
4.5. SUMMARY
123
We can move back and forth through a transformer when we analyze power systems if we remember that voltage goes through as the turns ratio, the current as the inverse, and the impedance as the square. Power is conserved. The real transformer can be modeled by adding just four elements, series resistance and reactance, and parallel resistance and reactance. Their values can be determined by open- and shortcircuit testing. Because of the wide differences between impedance levels of the series and the parallel elements, we can be a bit sloppy in moving the parallel elements and ignoring their currents. Dont jump to the conclusion that this works in other circuits! Our sloppiness works only because of the large disparity between the element values. It doesnt work in general as we will see in a similar model of the induction motor in the next chapter.
125

1. Voltage, current, and impedance ratios
n V1 n1 I1 n2 = , = , n1 I1 = n2 I 2 , Z1 = 1 Z 2 V2 n2 I 2 n1 n2
2
2. Reecting through a transformer Voltage reects by the turns ratio Current reects by the inverse of the turns ratio Impedance reects by the turns ratio squared 3. Transformer tests (a) Open-circuit test
I oc = cos1 Poc Voc I oc Voc Im
I oc = I oc I oc = I c jI m Rc = Voc Ic , Xm =
(b) Short-circuit test

I sc = cos1 Psc Vsc I sc
I sc = I sc I sc R + jX = Vsc I sc
127
CHAPTER
Machines
The world goes around, life goes around, holidays come aroundaround around around. Lots and lots of things we do mechanically go around. Car wheels, merry-go-rounds, lawn-mower blades, they all go around. It seems that the human body is a rarity in that nothing on or in it goes around. Going around seems to be important, so transferring energy into going-around motion must be important. Therefore, transforming electrical energy into rotary motion is also important. Thats what we are going to study in this chapter. Electric motors do the job and they are a major user of electric power. There are lots of different forms of electric motors. Many motors run on direct current, especially those in toys and automobiles. If you count all these little motors, youll probably conclude that they are by far the most popular motors in the world. Most of these, though, are relatively smalltoy-sizedor just large enough to move the sliding door in your van. While there are large d-c motors used throughout industry, these are by far outnumbered by motors that run on alternating current. Common types of a-c motors include single-phase induction motors, three-phase induction motors, and synchronous motors. Most a-c motors, though, are induction motors, which are the simplest motors one can build. Single-phase induction motors are most common in the home, where they power refrigerators, air-conditioners, and ventilating fans. Youve certainly seen and heard those motors running. When it comes to demands of industry, though, three-phase induction motors are it. They are simple, efcient, and reliableand they arent hard to understand, either. One of the most beautiful aspects of the three-phase induction motor is that it has just one moving part, the rotor (unless you get picky and count the balls in the bearings!). Moreover, this rotor is quite easy to manufacture. So lets take a look in this chapter at this simple motor, learn how it works, and most important, learn how to choose a three-phase induction motor for a particular job.
5.1
HOW THEY WORK
Figure 5.1 is an illustration of a simple motor. The coils create a magnetic eld and the rotor, which here is just a bar magnet, reacts to that eld. This simple motor is a two-pole motor, which refers to the two xed coils and the iron they are wound around. An odd number of poles isnt possible, since we always create both a north and a south with the coils. In Fig. 5.1 the rotors north pole is a little off to the right of the xed north pole. That means that theres a force between the two norths that imparts torque to the rotor. Hence, the rotor is going to rotate clockwise. (Remember that likes repel, unlikes attract.)
128
5. MACHINES
Figure 5.1: Clockwise torque
What happens, though, when the moving north pole gets near the xed south pole? Since unlikes attract, it sounds like the rotation will stop with the poles aligned N-S-N-S. Thats what we are approaching in Fig. 5.2. The rotor is about to get stuck.
Figure 5.2: Torque continues
To avoid getting the rotor stuck, lets reverse the xed magnetic eld so that the top pole is S and the bottom is N . Figure 5.3 shows this. If we do this just as the rotor is beginning to get stuck, the rotors momentum will carry it on around. Thats pretty easy to do if the current feeding the poles is a-c.
5.1. HOW THEY WORK
129
Figure 5.3: Field reversal
Now we have torque all the way around, although it isnt smooth torque since itll get stronger as the rotor gets closer to aligning with the poles. But we have a motor and it will make a shaft go around. One way to improve our motor is to go from single-phase alternating current to three-phase. Well see why in a moment, but rst lets look at the structure in Fig. 5.4. It shows three phases and the corresponding poles. Ive left out the rotor for the time being because I want to look at the characteristics of the magnetic eld rst.
Figure 5.4: Three-phase eld
130
5. MACHINES
Lets suppose that each of the pairs of poles creates the same magnetic eld, differing only in the fact that the currents that create them differ in phase. If we assign 0 to the a phase, then the b phase is at 120 and the c phase is at 240. These pole pairs also differ in angular position around the periphery of the stator. If the magnitude of the current in each phase is I , then the magnitude of the magnetic ux created will be kI, where k the appropriate constant:
a = kI cos ( t ) cos b = kI cos ( t 120 ) cos ( 120 )
c = kI cos ( t 240 ) cos ( 240 )
Using
cos a cos b = 0.5 cos(a + b ) + 0.5 cos(a b )
to disassemble this yields

a = kI 0.5 cos ( t + ) + 0.5 cos ( t ) b = kI 0.5 cos ( t + 240 ) + 0.5 cos ( t )
c = kI 0.5 cos ( t + 480 ) + 0.5 cos ( t )
If we now use
cos(a b ) = cos a cos b + sin a sin b
to clean up all the cosines, we get

a = kI 0.5 cos ( t + ) + 0.5 cos ( t ) b = kI 0.25 cos ( t + ) + 0.5 cos ( t ) c = kI 0.25 cos ( t + ) + 0.5 cos ( t )
Adding these three to get the value of the ux due to all three phases working together,
3 total = kI cos ( t ) 2
Umm, oh, so what? Whats important here is that the ux is uniform. In other words, the ux varies only by the angular position around the periphery and it is also in step with the frequency of the supplied power. If you were able to stand at some point inside the eld in Fig. 5.4, you would observe a wave of ux going past you, a wave that would reach its positive peak every 60th of a second (for 60-Hz power). Ah! That means we can put our rotor back and see what happens to it. Figure 5.5 shows the rotor in place. It also shows its place all the time! The rotor wont rotate! It feels the ux wave
5.1. HOW THEY WORK
131
going by, but that wave imparts no net torque, so the rotor just sits there. This motor wont start. If we were to spin the rotor so it was rotating at the same speed as the wave of ux is moving, it will lock in to the wave and follow it. Now its capable of producing torque. This arrangement is called a synchronous motor, which isnt self-starting. Now what? Heres where the inventions of Nikola Tesla enter the story. Tesla invented the polyphase motor in 1889 (U.S. Patent No. 416,194) but he didnt get the rotor into modern form until about a month later (No. 417,794). This rotor is made of laminated layers of magnetic iron with slots in the cylindrical surface for wires. These coils of wire are all short-circuited at both ends. In Fig. 5.6, the wire shown at X goes to the other end of the rotor and then returns as X . It is then connected to X.
Figure 5.5: Add a rotor
Figure 5.6: Add a working rotor
132
5. MACHINES
Whats happening is pretty much the same as happens in a transformer. The changing ux induces a current in the shorted windings. That current is strongest when a winding is directly across the direction of the ux and weakest when that winding is parallel to the ux. Hence, the rotor has a ux of its own that tries to line up with the rotating wave of ux. The result? Torque! Starting torque, too! Credit Mr. Tesla with a major invention, the induction motor. Its much simpler than the d-c motor that would have been required if Thomas Edison had won the a-c/d-c battle and the world ran on d-c.
5.2
INDUCTION MOTOR
Teslas motor was simple, and the modern form of it is even simpler. Figure 5.7 is a cutaway of a modern motor showing the structure of the rotor. The modern rotor can be made by stacking thin sheets of magnetic iron. The sheets have notches around the periphery so that, when they are stacked, the notches form slots. Then molten aluminum is poured into the slots to form the wires. The shorted ends are cast at the same time.
Figure 5.7: Rotor: cutaway view
Figure 5.8 is half of the aluminum casting of the wires and one shorted end. The aluminum at the end also has vanes to stir the air inside the motor to aid in cooling the rotor. This whole structure of shaft, laminated iron, and aluminum conductors constitutes the one moving part in the induction motor.
5.2. INDUCTION MOTOR
133
Figure 5.8: Aluminum rotor
5.2.1
MOTOR SPEED
How fast an induction motor can go is an important question. Im going to assume that we are always dealing with 60-Hz power here; many of the numbers change by a ratio of 50:60 if we are running on 50 Hz as in Europe. The motor that I have been using for illustration has six poles, two for each phase of the three-phase power supply. We say, though, that this is a two-pole motor, stating the poles per phase. The eld rotates at 60 Hz, which is 60 revolutions per second. Thats 3,600 revolutions per minute, the rotational speed of this eld. Since the rotor must have a eld of its own to interact with this rotating eld, there must be current induced in the shorted windings. If the winding rotates exactly in step with the rotating eld, the winding sees no changing ux. No changing ux means no induced current. No induced current means no resultant rotor eld. No rotor eld means no developed torque. So the rotor cannot go at the same speed as the rotating eld. In other words, in this example the rotor cannot go around at 3,600 rpm. So what speed does it go at? A speed a little less than 3,600 rpm. The difference is called slip that is expressed as a fraction:
s= ns nm ns
ns is the synchronous speed, the speed that the eld revolves at. In this example, ns = 3600. Suppose our motor at full load runs at 3,420 rpm, a number that will be stated on the motors nameplate. Then the slip is
s= 3600 3420 = 0.05 3600
We can build induction motors with any number of poles, provided that any is a multiple of 6. The relationship between ns and the number of poles per phase is
134
5. MACHINES
ns = ns =
60 f p 2 3600 for f = 60 Hz p 2
If a motor has 6 poles (6 poles for each of the three phases) and a full-load slip of 4%, then the synchronous speed and the motor speed are
ns = 3600 = 1200 62
nm = 1200 (1 0.04 ) = 1152 rpm
You can always gure out the synchronous speed from the actual speed because the slip is rarely more than 5%. The only synchronous speeds at 60 Hz are 3,600, 1,800, 1,200, 900, 750, 600, and so on, all submultiples of 3,600. If you want something else, you have to change the frequency of the three-phase supply. Thats something done fairly commonly with variable-frequency drives.
5.2.2
POWER FLOW THROUGH MOTOR
Figure 5.9 is a diagram of the ow of power through the induction motor showing what happens as
Figure 5.9: Single-phase equivalent with three-phase power ow
it goes from the input power line to the output shaft. Lets walk through the diagram step by step: At the left is the three-phase power input. The 3 is there, and the voltage is the line voltage. The rst losses are those of the stator, including heating of the wire and the power needed to continually reverse the magnetic domains of the core. The rest of the power is delivered to the air gap and across it to the rotor. This air gap power is related to the developed torque via the synchronous speed of the rotor, not the actual rotor speed:
s =
ns 2 60
5.2. INDUCTION MOTOR
135
The power lost in the rotor is proportional to the slip s . It consists of the wire and the core losses in the rotor. What remains is the developed power delivered to the shaft of the motor. This power is related to the developed torque via the actual speed of the shaft:
m =
nm 2 60
The remaining loss is mechanical, due to windage and bearing losses. The actual output of the motor is everything thats left after accounting for all these losses. Induction motors today typically have efciencies of at least 86%; high-efciency motors are over 92%.
5.2.3
EXAMPLE IMOTOR CALCULATIONS
Specications for a 10-hp totally-enclosed three-phase induction motor in a particular manufacturers catalog show volts = 208-230/460 V, current = 28.0-25.0/12.5 A, power factor 83.5%, NEMA efciency 88.5%, full-load speed 1,740 rpm. Calculate the following: Power out: Motors are rated for output power, so Pout = 10 hp746 W/hp = 7.46 kW. Power in: The list of currents matches the voltages, so Pin = 3 460 12.5 0.835 = 8.316 kW. Percent efciency: % = 100(7.46/8.316) = 89.7%, a little better than the manufacturers stated value. Number of poles: The full-load speed is 1,740 rpm, which is just under 1,800 rpm. So the motor has four poles because ns = 3600/(p/2) = 1800 for p = 4. Slip: s = 100(1800 1740)/1800 = 3.33%. Now walk through the diagram of Fig. 5.9 from right to left to nd the various power values: Developed power, assuming windage if windage and bearing losses are 200 W: Pdevel = 7.46 + 0.2 = 7.66 kW. Air-gap power: Pairgap = 7.66/(1 s) = 7.66/(1 0.333) = 7.924 kW. Rotor loss: Protor = Pairgap Pdevel = 7.924 7.66 = 264 W. Stator loss: Pstator = Pin Pairgap = 8.316 7.924 = 392 W. Output torque: m = (1740/60)2 = 182.2 radians/second, so Tout = Pout /m = 40.94 Nm.
136
5. MACHINES
5.2.4
EXAMPLE IIVALUE OF EFFICIENCY
While looking for 10-hp motors in the manufacturers catalog (which is where I found the motor in the previous example), I noticed another motor with the same specications but with a rated efciency of 91.7%. Catalog data show volts = 230/460 V, current = 25.0/12.5 A, power factor 82%, full-load speed 1,765 rpm. Lets check the efciency of this motor: Pout = 10 746 = 7.460 kW as before. Pin = 3 460 12.5 0.82 = 8.167 kW. % = 100(7.46/8.167) = 91.3%, close to the stated value. How much will it cost to run each motor at full load 16 hours per day, ve days per week, 50 weeks per year? Suppose the plant has a marginal (top-end) electric rate of 4.5| c/kWh (i.e., the cost of adding 1 kWh on top of all the other loads in the plant). Motor I: Cost 1 = 0.045 8.316 16 5 50 = $1, 500 per year. Motor II: Cost2 = 0.045 8.167 16 5 50 = $1, 470 per year. Motor I lists for $875, motor II for $1,266, a difference of $391. It will take a decade to amortize the added cost of the more efcient motor. The choice of which motor will have to be left to the suits in the front ofce.
5.3
INDUCTION MOTOR MODEL
Weve just seen how energy ows through the induction motor, which is the energy model of the motor (Fig. 5.9). But in understanding the induction motor or any other complex electrical device, it is often helpful to have a circuit model as well. The circuit model has to take into account the various places where energy is lost. One place is obvious, the conductors, leading to i2 R heat loss. The magnetic core is not loss-free either, because it takes energy to ip the magnetic domains back and forth sixty times a second. This loss also appears as heat. Our model also has to include the reality that magnetic ux is not completely conned to the iron. Some, not much but some escapes, which is modeled as inductance. The iron core also stores some energy in the form of a magnetic eld, which is also modeled as inductance. Our model is going to be a single-phase equivalent just as we set up a single-phase equivalent of a balanced three-phase system.This means that our voltages, always stated initially as line voltages, must be divided by 3.
5.3.1
MODEL OF THE STATOR
The model of the stator is pretty straightforward (Fig. 5.10). The resistor R1 in the series path is the conductor loss. The reactance X1 is there to model the ux that does not stay completely in the
5.3. INDUCTION MOTOR MODEL
137
Figure 5.10: Single-phase stator equivalent
iron core. Two parallel branches represent the core itself. Rc models the losses in the core due to energy needed to ip the magnetic domains. Xm models the reactance that makes the core look like an inductor. What goes into this stator model is the line voltage, divided by 3. What comes out is a voltage that is transformed across the air gap to the rotor. Note the subscripts. The 1 refers to the stator; 2 refers to the rotor. We need to reect the elements of the rotor across the air gap to the stator so that our model is complete as viewed from the terminals of the motor itself.
5.3.2
MODEL OF THE ROTOR
Well, this should be simple, but theres a complication that requires a little explanation. The stator is a 60-Hz model, meaning that the frequency of operation is 60 Hz. Umm, yah, so? Thats not true for the rotor. If the rotor is stopped dead, it is operating at 60 Hz because the coupling from the stator to the rotor is very much like the coupling in a transformer. A stopped rotor isnt too useful, though. When it is running, its frequency is reduced by the slip s . If the rotor is running exactly at the synchronous speed (such as 3,600 rpm for a two-pole motor), s= 0 and hence the frequency in the rotor is 0 Hz. In my example in Sect. 5.2.3, the slip s = 3.33%, so the rotor frequency is 0.0333 60 = 2 Hz. Thats a long way from 60! The effect of this is shown in Fig. 5.11. The rotor equivalent has copper loss, although the conductor is usually aluminum. It also has leakage reactance, but with an added s . Why the s ? I want the models elements all to have their 60-Hz values. In other words, I want X2 to be the value wed observe at 60 Hz if we could get in there and measure it. By multiplying X2 by s , I get the actual value of the leakage reactance in the rotor when its running. The input voltage to the rotor model is shown with an s as well because this is the actual voltage on the rotor itself after being transformed across the air gap. Ill reect all this across the air gap. To start this, I need to calculate the actual rotor current I2 in Fig. 5.11:
138
5. MACHINES
Figure 5.11: Rotor equivalent

sE2 E2 = R2 + jsX2 R2 + jX 2 s
I2 =
Ill use the second part of this equation to draw a new model of the rotor. Figure 5.12 is this new model. But we arent done yet. This model has the right elements but its missing a big piece.
Figure 5.12: Rotor equivalent adjusted
Whats missing? Well, doesnt this motor deliver energy via its rotating shaft? Where is this power ow represented in the model? That shaft power is actually there in the model but it is hard to see. So the nal task in developing an equivalent circuit of the induction motor is to separate the losses in the rotor from the power delivered to the shaft. The power delivered to the resistor in the rotor model of Fig. 5.12 is the power delivered across the air gap:
Pairgap = I 2
2
R2 s
139
In the rotor, though, the only loss is due to the resistance R2 that models the conductor loss. That loss should be
PCu = I 2 R2
2
A certain amount of power (Pairgap ) crosses the air gap. Some of that is lost as heat (PCu ). The rest must be the developed power to the shaft:
1 s 2 Pdevel = Pairgap PCu = I 2 R2 s
Figure 5.13 is our nal rotor model. Ive separated the resistor R2 /s into two parts. R2 models the conductor loss; the variable resistor (the part with s) models the power to the shaft.
Figure 5.13: Rotor with rotation power
Consider two different values of the slip s : s = 1: The rotor is stopped and delivering no power. The variable resistor has a value of zero ohms, a dead short, not capable of absorbing power. s = 0: The rotor is running exactly at synchronous speed and the rotor can deliver no power. The variable resistor has a value of , an open circuit, not capable of absorbing power. Now lets take the nal step and combine the models of the stator (Fig. 5.10) and the rotor (Fig. 5.13).
5.3.3
COMPLETE INDUCTION MOTOR MODEL
The rotor model of Fig. 5.13 is as seen from the air gap. Now I have to refer that model to the stator as I connect the two models together. This means labeling the elements of the rotor model as referred or reected by adding primes to their labels. Figure 5.14 is the nished model. If you have already studied the transformer, you know that the two parallel elements can be moved to the left or right in the transformer model because their values are much, much larger that
140
5. MACHINES
Figure 5.14: Complete single-phase equivalent with rotor reected
the series elements. Moving the parallel elements has a very small effect on the currents, only 1 or 2 percent or so. This is not true for the induction motor. The parallel elements play a much larger part in the representation of the motor, so they have to stay where they are shown in Fig. 5.14. If you have not studied transformers, ignore what I just said.
5.3.4
EXAMPLE IIIUSING THE MODEL
The equivalent circuit of a particular induction motor is shown in Fig. 5.15. (How we got this is a subject for a later section.) This motor is a 460-V, three-phase, 50-hp, 60-Hz, 4-pole induction motor. Its full-load speed is 1,770 rpm. From the model we are to calculate the line current I1 , the input power with power factor, the percent efciency, and the starting current.
Figure 5.15: Example III: 50-hp motor equivalent
141
First comes the full-load slip

s= 1800 1770 = 0.01667 1800
so we can calculate the value of the shaft resistor:

Rshaft = 0.0767 (1 0.01667 ) = 4.5244 0.01667
Ill ignore bearing and windage losses and nd the one -phase equivalent power out per phase:
Pout = 50 746 = 12.43 kW 3
so the current I2 (with a chosen angle of 0) becomes

' I2 =
Pout 12.43 10 3 = = 52.420 A Rshaft 4.5244
Now nd the rotor voltage as seen from the stator:

' E2 = I 2 (0.0767 + j 0.6829 + 4.4244 ) = 243.838.4 V
The voltage E2 is across both of the parallel elements, so the current through those elements is
I parallel = E2 E2 + = 20.53 80.4 A j11.88 569.8
and the input current I1 is the sum:

' I1 = I 2 + I parallel = 59.41 19.9 A
Finally, we can get the single-phase input voltage

V1 = E2 + I1 (0.0881 + j 0.4865 ) = 263.213.4 V
and from that the input power factor, remembering to use the angle between the voltage V1 and the current I1 :
pf = cos (13.4 (19.9 )) = 0.836 lagging
The input power for one phase is

Pin / phase = V1 I1 pf = 263.2 59.41 0.836 = 13.07 kW
142
5. MACHINES
The total three-phase power is 3 times that:

Pin = 3Pin / phasse = 3 13.07 = 39.21 kW
Whew! The overall efciency (ignoring bearing and windage losses) is

% = 100 50 746 = 95.1% 39.21 10 3
One check on these calculations is to see how the input voltage, stated as a line voltage, compares with the motors stated specications. V1 in the single-phase model is 263.2 volts. This times 3 gives a line voltage of 455.9 volts, compared with the specication of 460 volts. To nd the starting current, we note that s = 1 so the shaft resistor is 0 , a dead short. If I nd the input impedance of the entire circuit, then I can get the starting current by dividing the supply voltage by that impedance. The input impedance is
Zin = (0.0881 + j 0.4865 ) + 569.8 j11.88 (0.0767 + j 0.6829 = 1.143482.1
If we assume that the supply voltage stays at the level we calculated for V1 (but choosing a phase angle of 0), the starting current is
I start = V1 263.2 = = 230.2 82.1 A Zin 1.143482.1
with a power factor of

pfstart = cos(82.1) = 0.137 lagging
The starting current is about four times the full-load running current, which means that the supply voltage probably drops until the motor gets rolling and the current decreases to the running current. With some motors, reducing the starting current by reducing the line voltage temporarily avoids tripping circuit breakers. We will see later how this model and these calculations stack up against the actual measured data for this motor.
5.4
MOTOR TESTS
Where did the data for the example I just nished come from? Very simply, from the manufacturers data sheets. The motor is a modern, high-efciency motor rated at 460 V, 50 hp, 1,770 rpm full load. Suppose there are no available test data for such a motor. How then to obtain the numbers? There are two speeds at which its rather easy to conduct tests. They yield pretty good information. Nothing other than three-phase metering is required, plus a place to fasten down the motor.
5.4. MOTOR TESTS
143
The two test speeds are 0 rpm and no-load rpm. The rst is done using a blocked rotor, meaning that the motor is run with the shaft rmly blocked so it cannot turn. The other iswell, sort of obvious: running with no load. A third test is a simple d-c test to determine the d-c resistance of the stator. I have a set of data from these tests on the 50-hp motor weve been using for an example, so lets see what these data yield for an equivalent circuit.
5.4.1
D-C TEST
The d-c test could be a simple ohmmeter test, but most portable ohmmeters dont read low resistance values very precisely. If we apply a voltage across two of the Y-connected three-phase terminals of the motor and observe the current, well get double the value of the resistance R1 . Its double because we are seeing into two legs of the three-phase Y. At d-c, impedances of the inductances in the model are 0. Sometimes the terminals for the individual windings are accessible. The d-c data are
Vdc = 5.0 V I dc = 28.4 A
Now we calculate the resistance and divide by 2 to get R1 :

Rdc = Vdc 5.0 = = 0.1761 I dc 28.4 R1 = Rdc 2 = 0.1761 2 = 0.0880
5.4.2
BLOCKED-ROTOR TEST
When we block the rotor and apply power, the motor will behave like we are starting it. If we apply full voltage, the current will be large and the motor windings would overheat quickly. Instead of full voltage, well do this test at full rated current. At rated current with a blocked rotor, the motor voltage will be low compared with the normal running voltage. This means the parallel Rc and Xm will not be much of a factor in the current. The result is that we will essentially see just R1 , X1 , R2 , and X2 in series. (See Fig. 5.16 on the previous page.) As we go through these tests, we must keep in mind that the tests are being done with three phase power and meters but our model is the single-phase equivalent. That means a 3 now and then. Rated current for this motor, according to the manufacturers specications, is 58.0 A. When we block the rotor and run the current up to 58.0 A, we get the following (three-phase) results:
V = 114.8 V , P = 1589 W , pf = 13.78% lagging
I reduce the voltage to the single-phase equivalent, calculate the angle of the current from the power factor, and divide the total power by 3:
144
5. MACHINES
Figure 5.16: Equivalent during blocked-rotor test

Vph = 114.8 3 = 66.28 V
I ph = 58.0 cos 1 pf = 58.0 82.08 A Pph = 1589 3 = 529.7 W
From these single-phase-equivalent numbers, I calculate the input impedance, whose real part is R1 + R2 and whose imaginary part is X1 + X2 :
Zblock =
' 2
Vph I ph
66.28 = 1.14382.08 58.0 82.08
R1 + R = Re [ Zblock ] = 0.1575
' = Im [ Zblock ] = 1.132 X1 + X2
From this information and the d-c value of R1 , I can get R2 . The values of X1 and X2 are assumed to be equal since we have no data to separate them:
' R2 = 0.1575 R1 = 0.1575 0.088 = 0.0695 ' X1 = X2 = 1.132 2 = 0.566
5.4.3
NO-LOAD TEST
During the no-load test, the rotor will be revolving at almost zero slip. In fact, sometimes this test is done with a second motor set up to spin the test machines rotor at exactly synchronous speed. During this test, the motor draws very low current, so the series elements of the model can be neglected. The data from the test will see only the two parallel elements. (See Fig. 5.17.) The no-load data, run at the rated voltage of 460 V, yield
I = 21.5 A, P = 464.6 W , pf = 2.71% lagging
5.5. CHOOSING A MOTOR
145
Figure 5.17: Equivalent during no-load test
For the single-phase model, the data become

Vph = 460.0 3 = 265.6 V I ph = 21.5 cos1 ( pf ) = 21.5 88.45 A Pph = 464.6 3 = 154.9 W
Rc and Xm will come from the real and the imaginary parts of the applied voltage divided by the appropriate part of the current:
Rc = Xm = Vph 265.6 = = 456.7 Re I 0 .5816 ph Vph 265.6 = = 12.36 I ph Im 21.49
5.4.4
COLLECTING AND COMPARING
The single-phase equivalent in Fig. 5.18 (next page) shows the values of the elements that have come from our tests. Also shown are the values from the manufacturers more extensive tests. You be the judge! One way to judge this is to compare data on the motor as calculated using the two sets of values with the data the manufacturer obtained by more extensive tests. These comparisons are in Fig. 5.19.
5.5
CHOOSING A MOTOR
What do you do if you are called upon to replace a motor or to specify a proper motor for a job? It would be nice to know something about three things: what the motors nameplate tells you, how torque and speed are related, and how a new motor needs to be sized. Well tackle all three of these in this section.
146
5. MACHINES
Figure 5.18: Text book tests compared with manufacturers tests
Figure 5.19: Comparison of motor data for three models
5.5.1
MOTOR NAMEPLATE
The nameplate in Fig. 5.20 contains the information required by NEMA, the National Electrical Manufacturers Association. Its their standard way of giving you the data you need to properly use or replace that motor. Heres what each of the entries tells us. Ill highlight the ones that are particularly important when you consider a replacement for this motor: Model is the makers catalog number Frame is the tof the motor such as the mounting holes, shaft diameter, keyway, shaft height, etc. A new motor with the same frame code should t if it also has the same electrical characteristics
147
Figure 5.20: A motors nameplate redrawn
Type is a makers designation Des (Design) is a code: see Section 5.5.2 Phase says this is a three-phase motor Code rates the locked-rotor reactive power. Here, H means 7.18.0 kVA/hp Ins Cl is the temperature rating of the insulation on the wire of the coils. Class B is rated to 180 C Encl is the type of motor enclosure. This motor is Drip Proof. Duty here means that the motor can operate continuously as long as other specications on the nameplate are followed (such as ambient temperature) Max Amb is the maximum air temperature around the motor if the motor is to operate at full load continuously Shaft End Brg and Opp End Brg are standard bearing numbers, useful if bearings need replacement Volt is the required three-phase voltage. This motor is wired for two different voltages, 208-230 and 460 volts. 208-230 means any voltage in that range Amp is the rated current at each of the stated voltages RPM is the full-load speed of the motor. This motor must be a four-pole motor that operates with a full-load slip of 2.78%
148
5. MACHINES
Hz is 60 hertz, although many 60-Hz motors can be operated using variable-frequency drives. You have to read the makers specications to nd out NEMA Nom Eff is the rated efciency of a large population of motors of the same design. Today, 88.5% is considered minimally energy efcient. Nom pf is the power factor at rated full load Max Cap kVAR is the amount capacitance stated in kVARs that will raise the power factor to unity HP is the rated full-load horsepower SF is the service factor. If the motor is operated at no more than the maximum ambient temperature, the continuously-running load may exceed the rated load by 15% Corr Amp is the rated currents if the power factor is raised to unity
5.5.2
TORQUE VERSUS SPEED
When you match a motor to a load, the torque-speed curve can tell you a great deal about making the right choice. While there are lots of different curves designated by the NEMA Des code, only a few are common. Figure 5.21 shows four common curves. These are plotted against percent slip, so full synchronous speed is on the right and zero speed is on the left. The torque axis is percent of rated full load. Since Design B is probably the most common, lets look at some important spots on the Design B curve: The normal operating range of torque is on the far right with only a small amount of slip. That portion of the curve looks pretty straight, and we are going to assume that it is. The peak of the curve is called the breakdown torque. If the torque takes the motor over the top, its speed drops very rapidly, but it takes over 200% of rated torque to get there. The valley is the pull-up torque, which is the minimum torque the motor will produce as the speed runs up from starting to normal operation. The torque at 100% slip, which is zero speed, is the starting torque that overcomes the inertia of the load as it starts and also unsticks bearings. Why the different curves? Which design we choose is based on the torque requirements of our intended load. Here are a few of the characteristics of these four designs: Design A has the highest breakdown torque. The curve in the normal operating range is very steep so speed doesnt vary as much with varying load as with Design B. Starting current is often higher.
149
Figure 5.21: NEMA induction motor torque curves
Design B is the common garden-variety motor. It has modest starting current and usually can be started directly from the power line without reducing the voltage. Design C has high starting torque and moderate starting current, but it is generally not as efcient as A and B. Design D has high starting torque but doesnt require a high starting current. It can operate over a wide range of slip (up to 10% at full load in Fig. 5.21).
5.5.3
OVERMOTORING IS BAD!
Yes, bad! To show you why, Im going to throw at you four motor curves. In each of these, I am including only the right-most slip from 0 to 20%. All of these curves are for a motor that delivers full-load torque at a slip of 3%. First, lets look at this motors torque-speed curve in Fig. 5.22. The full-load (100%) rated torque for this motor is 410 N-m. Note that the curve, in the region of normal operation, is pretty much a straight line. Suppose we have decided we will use this motor to drive a load that requires a torque of 275 N-m. We have made this choice because, everyone knows you should be conservative and run these motors easy.
150
5. MACHINES
Figure 5.22: Torque and Power factor
The load on this motor is now just about two-thirds of its rated load, so it is going to operate at 2% slip. That sounds pretty good, right? After all, we arent stressing the motor as much and everyone knows it will last longer. Butovermotoring is expensive! What we are doing here is overmotoring, choosing a motor larger than what we need, assuming that this is a good thing. Look what happens to the power factor in Fig. 5.22. This graph shows that the power factor decreases (i.e., gets worse) as slip decreases. The graph shows that at 2% slip the motor will operate with a power factor of about 0.8. In other words, we are going to pay extra for poorer power factor. The message? Dont overmotor! Now consider efciency. Motors are generally designed to operate at their peak efciency at full load. The curve in Fig. 5.23 shows that the peak efciency of 91% is achieved at full rated slip. But weve chosen to run light at 2% slip. From the graph, it looks like the efciency drops to about 85%. In other words, we are going to operate at a lower efciency, which means more money for the same job. Get the message? Dont overmotor! Finally, the graph of line current (Fig. 5.23) shows we dont come out ahead there either. The oversized motor does not draw rated line current because it is running below capacity. While the
5.6. MATCHING A LOAD
151
Figure 5.23: % efciency and Line current
current is lower than rated (on the graph is looks like about 80 A), the current hasnt decreased by a full 2/3. Here we are reinforcing the message from the efciency graph, namely, that it takes extra power for a given load to drive it with an oversized motor. Lets hear the cheer once more: Dont overmotor! When we select a motor for a given load, we should select a motor that matches that load while using the motor at its full rating. Hence, a load that requires, say, 500 N-m of torque should be driven by a motor that is designed to deliver 500 N-m as its full load. But what if I cant nd such a motor? Should I go high or low? Many motors have a service factor of 1.15, so a motor that can deliver at full load 440 N-m wont be overloaded by a load of 500 N-m (440 1.15 = 506). The proper choice would be the 440 N-m motor.
5.6
MATCHING A LOAD
Now that we have learned not to choose a motor that is too large for the job, its time to try an example of choosing a motor for a particular job. The process is easy to state: Choose a motor whose rated output torque equals the loads torque demand. Gee! Thats pretty simple! But as usual, the problems are really in the details. One obvious way to avoid these details is to base our decision on power. Shouldnt we just nd the motor in the
152
5. MACHINES
catalog whose horsepower rating is larger than the loads horsepower requirement? No! As well see in this example, that can lead to the wrong decision. Our example load is a centrifugal fan that requires a torque of 70 N-m at 1,800 rpm and 100 N-m at start. We are to select a motor to properly drive this load. The only other thing we know about the fan is that the power required is proportional to the cube of the shaft speed. The simple way to nd the right motor appears to be simply nding the horsepower required by the fan and then picking the next larger motor in the manufacturers catalog. So lets do that. Power is the radian speed times torque, so
f =
1800 2 = 188.5 radians/second 60 Pf = 70 f = 13.20 kW W = 17.7 hp
A look at our favorite makers catalog shows us a three-phase, 60-Hz induction motor rated at 20 hp. The next smaller one is rated at only 15 hp. Both have a service factor of 1.15, so the 15-hp motor can be loaded to 15 1.15 = 17.3 hp. I guess our choice is limited to the 20-hp motor. OK, lets check a little further into the 20-hp motor. Its listed as 1,800 rpm, 208-230/460 V, 91% efciency, 1,750 rpm full load. Full-load torque is 60 lb-ft, which translates into 81 N-m. Locked-rotor torque (and hence starting torque) is 115 lb-ft, which is 156 N-m. That sounds just ne. But not so fast! Could the 15-hp motor actually do the job? Nah, you say, 17.7 hp is just too much over the 17.3 hp limit. Consider, though, that the operating speed is going to be less than 1,800 rpm, so the actual power required by the fan will be less than 17.7 hp. Maybe the 15-hp motor will work. We start by nding the motor constant km for the 15-hp motor that relates torque to shaft speed, or more important, that relates torque to slip. This relationship for induction motors is linear, so we can write
Tm = km sm
where km is the slope of the torque-slip curve at rated slip. The makers catalog says that this 15-hp motor produces full-load torque of 61 N-m at 1,750 rpm, so we can nd the motor constant km :
61 = km 1800 1750 , km = 2.20 10 3 N-m 1800
Now we do the same think for the fan, nding the constant kf that relates fan torque to speed. Since power for this fan is proportional to the cube of the speed, the torque must be proportional to the square of the speed. (Remember that torque is power divided by speed.) So we write fan torque as a function of fan speed squared:
T f = k f n2 f
5.7. MORE EXAMPLES
153
and then solve at rated torque and speed for the fan constant kf :
70 = 1800 2 k f , k f = 2.16 10 5 N-m/rpm 2
Equate these two torque equations because the torque required by the fan must be produced by the motor:
T f = k f n2 f = Tm = k m sm
The left side is in rpm but the right is in slip, so well change slip to rpm:
3 2.16 10 5 n 2 f = 2.20 10
1800 n f 1800
Simplifying this yields a quadratic equation, which we solve for shaft speed:
4 8 n2 f + 5.658 10 n f 1.019 10 = 0,
n f = 1, 747 rpm
Hmmm, thats just barely smaller than the motors full-load speed of 1,750 rpm. I wonder how much horsepower is being developed? Motor torque is
Tm = 2.20 10 3 1800 1747 = 64.8 N-m 1800
so motor power is
1747 Pm = 2 64.8 = 11.85 kW = 15.9 hp 60
Wow! Thats well under the 1.15 limit of 17.3 hp! We can use the 15-hp motor and we wont be overmotoring. The catalog specs for this motor are 15 hp, 1,800 rpm, 208-230/460 V, 91% efciency, 1,750 rpm full load. Locked-rotor (starting) torque is 91 lb-ft, which is 123 N-m. Right choice! And 20% less expensive, too.
5.7
5.7.1
MORE EXAMPLES
EXAMPLE IVMOTOR CALCULATIONS
A 5-hp, 3-phase, 60-Hz motor has a full-load speed of 1,165 rpm. It is rated at 230/460 volts, 14.0/7.0 amperes full load, 75.5% power factor. Rotational losses are estimated to be 175 W. Find the following: Number of poles and slip. The next larger synchronous speed above 1,165 is 1,200 rpm:
154
5. MACHINES
ns =
3600 = 1200, so p = 6 poles p 2 1200 1165 = 2.92% 1200
s = 100
Torque output. First we nd the speed in radians per second and the output power in watts (746 watts per horsepower). Then we can nd the torque:
1165 2 = 122.0 radians/second 60 Pout = (5 )( 746 ) = 3.73 kW
s =
Tout =
Pout 3.73 10 3 = = 30.57 N-m 122.0 s
Power input and percent efciency. Well use the 460-volt rating and remember this is threephase power:
Pin =
( 3) ( 460)( 7.0)(0.755) = 4.21 kW

3.73 Pout = 88.6% = 100 4.21 Pin
% = 100
Power lost in the rotor and in the stator. Work backward through the power-ow diagram of Fig. 5.9:
Pdevel = Pout + Protation = 3.73 + 0.175 = 3.905 kW Pairgap = Protor Pdevel 3.905 10 3 = = 4.023 kW 1 s 1 0.0292 = sPairgap = (0.0292 )( 4.023 10 3 ) = 117 W
Pstator = Pin Pairgap = 4.21 4.023 = 187 W
5.7.2
POWER FACTOR CORRECTION
For the motor just studied, specify the capacitance per phase (in kVARs) needed to improve the power factor to 90%. Then nd the magnitude of the line current with the added capacitance. The following calculations are on a per-phase (single-phase equivalent) basis, First we need the input power and the input VARs for one phase:
460 ( 7.0 )(0.755 ) = 1.404 kW 3 Qin = Pin tan(cos1 0.755 ) = 1.219 kVAR Pin =
5.7. MORE EXAMPLES
155
Now nd the input VARs for a power factor of 90%:

Qnew = Pin tan ( cos1 0.9 ) = 0.680 kVAR
This will yield the VARs of capacitance per phase needed to make this improvement:
Qcap = Qnew Qin = 0.680 1.219 = 0.539 kVAR
We need 0.539 kVARs of capacitors per phase to make the power-factor 90%. (About 20 F at 460 V for each phase.) The magnitude of the total power S per phase will give us the magnitude of the current:
S = Pin + jQnew = 1.404 + j 0.680 = 1.560 kVA =V I = 460 I , I = 5.87 A 3
Thats an 16% reduction in current, which will reduce wiring losses. (The equation because this is for a single-phase equivalent calculation.)
3 is in the S
5.7.3
CIRCUIT ANALYSIS
A single-phase equivalent circuit for a 5-hp, 460-V, 3-phase, 60-Hz induction motor is shown in Fig. 5.24. The motors full-load speed is 1,165 rpm. Determine the input power, the power lost in
Figure 5.24: Single-phase equivalent for Example VI
the stator, the power lost in the rotor, the developed power, the rotational power lost, and the output power. All these results are to be three-phase results. Ill calculate the value of the output resistor rst and then write one node equation at E2 to get E2 and the currents I1 and I2 . From these I can answer all the power questions. Ill carry my results to the end of this section before I multiply by 3. The value of the output resistor is
156
5. MACHINES
s= Rout
1200 < 1165 = 0.0292 1200 R' 1.34 (1 < 0.0292 ) = 44.55 = 2 (1 < s ) = s 0.0292
Only one node equation is needed, written in terms of the air-gap voltage E2:
E2 266 E E E2 + 2 + 2 + =0 0.45 + j 3.1 490 j 85 1.34 + j 8 + 44.55
Solving this yields

E2 = 250.9 3.5 V
and from this come the currents:

' I2 =
E2 = 5.39 13.4 A 1.34 + j 8 + 44.55 E E ' I1 = I 2 + 2 + 2 = 6.99 37.2 A 490 j 85
The power input is

Pin = Vin I1 cos(37.2) = (266 )(6.99 )cos(37.2) = 1481 W
The power lost in the stator involves both the series resistor representing the copper loss (0.45 ) and the parallel resistor representing the core loss (490 ):
Pstator = R1 I1 +
2
E2
490 250.9 2 = 150 W 490
= (0.45 )(6.99 )2 +
The power lost in the rotor is represented by the copper-loss resistor (1.34
' ' Protor = R2 I2 = (1.34 )(5.39 )2 = 38.9 W 2
):
The developed power, which is the output before subtracting windage and bearing losses, is calculated using the output resistor value:
' Pdevel = Rout I 2 = ( 44.55 )(5.39 )2 = 1295 W 2
5.8. SUMMARY
157
This motor is rated at 5 hp, so the output power per phase is

Pout = 5 746 = 1243 W 3
The difference between the developed power and the output power is the rotational loss:
Protation = Pdevel Pout = 1295 1243 = 52 W
These results are all for one phase. The results for the full three-phase motor are three times these:
Pin = 4443 W Pstator = 450 W Protor = 117 W Pdevel = 3885 W Protation = 156 W Pout = 3730 W
5.8
SUMMARY
A summary probably should review the major topics of the chapter, which means making a list of them and commenting on the list. But that might obscure several important points: Induction motors are very simple devices, provide smooth, non-pulsating torque, and have just one moving part. Motor specications are well standardized so that replacing a motor is generally straightforward. Overmotoring, the use of a larger-than-needed motor, is not a good idea. The concept of a safety factor doesnt hold here. Choose the motor that can drive the load at the full rating of the motor. If you need a motor to drive a load, match motor and load by equating the loads torque-vsspeed equation with the motors equation. As one oil company used to say, Happy motoring!
159

1. Induction-motor slip
s= nsync nshaft nsync
2. Synchronous speed
nsync = 60 f 3600 = for = 60 Hz p 2 p 2
3. Radian/second speed
=
n 2 60
4. Single-phase equivalent with three-phase power ow
161
CHAPTER
Electromagnetics
Electromagnetics is a very different topic, very different from just about anything else youve studied before. New and different variables. Different units of measure. Unusual formulas and numerical constants. This is not unlike being a uent speaker of both English and, say, German, then landing in China and being told to speak, read, and write Chinese immediately. OK, its not really that bad! An engineering topic that is most similar to electromagnetics is uids. Both areas use much of the same math, and in fact some of the math constructs used in electromagnetics came from the math for uids. (Curl is an example.) The quantities we are going to deal with dont lend themselves to feel, either. If we are studying uids, for example, we can visualize, feel, and touch uids. Pressure makes some intuitive sense. When you got into electricity, quantities became more abstract. You cant see voltage, although you canand dont want tofeel it. Current makes some intuitive sense. So why are we studying electromagnetics? The simple reason is that it is the basis for just about every form of conversion of energy between mechanical and electrical forms. Motors, solenoids, generators, relaysthese all apply electromagnetic principles. Thats why its appropriate that we learn something about these principles. But not too much! If you really want to learn a lot, take an EE course! Weve got a different purpose here, namely, to see fundamentally how these things work and to associate numbers with what is going on. One shortcut we are going to take is to study electromagnetic phenomena in their worst case forms. For example, when we study the force between conductors in the next section, we are going to align the conductors so we get the maximum possible force. This means that well use a scalar product in the formula rather than the integrals of dot and cross products. This simplies our work, which is to nd the largest possible force that could be applied to the wires involved. Keep in mind the adage, when you are up to your tail in alligators, its hard to remember that you were sent to drain the swamp. When you are buried in a mass of detail, look back to see what the original goal was. To help keep track of the alligators, the last page of this chapter is a summary of all the formulas and equations that well study.
6.1
AMPERES LAWS
Andr-Marie Ampre, working almost 200 years ago, established the relationships between electricity and magnetism. In only about a decade he developed the laws that are the basis for electrodynamics
162
6. ELECTROMAGNETICS
and much more. We are going to use Amperes Force Law in this section to determine forces between conductors. In Section 6.2.4, Amperes Circuital Law will lead us to magnetomotive force.
6.1.1
AMPERES FORCE LAW
Amperes Force Law describes the force between conductors that are carrying current. As I said before, Im restricting this to the worse case. For this Law, I have the two conductors parallel to one another so that the force is maximum. While this isnt general, most of our magnetic systems involve conductors that are arranged for maximum force. Figure 6.1 shows the basic structure. The conductor on the left is carrying a current, owing upward, of I1 amperes. The conductor is innite in both directions, which is impractical, but as long as the loop is closed a long way away, innite is OK.
Figure 6.1: Amperes Force Law
The second conductor in the drawing is carrying a current, owing downward, of I2 amperes. We are considering only a part of this conductor, a segment of length . This conductor is a distance r from the rst one. Ampre says theres a force F acting outward between these conductors. The relationship is
F=
+ I1 I 2 C newtons(N) 2/ r
Its not surprising that the force depends on the strengths of the two currents. It also should not be surprising that the force diminishes as the distance between the conductors r increases. Also, if we omit the length of the second conductor, the equation becomes the force per unit length, which would be N/m. Most of the formulas we are going to learn about involve a physical constant of some kind, a constant that makes the units work out properly, generally by describing some property of the
6.1. AMPERES LAWS
163
material we are working with or in. In this equation, is the magic frigger factor that describes the magnetic characteristic of the material.The larger the , the more easily is the material magnetizable.
6.1.2
PERMEABILITY
This has a proper name: permeability. If we are working in free space,

+ 0 4/ = 10 <7 newtons/ampere 2 (N/A2 )
In fact, just about everything is free space as far as magnetics are concerned. For our purposes, the only materials that have a permeability other than are iron and iron-bearing metals plus a few others such as cobalt, nickel, and rare-earth alloys. The differences in permeabilities can be stark. Some alloys of iron, for example, have permeabilities hundreds of times that of free space, a fact that we already made use of when we studied motors and transformers. We often state the permeability of a magnetic material by relating it to free space through relative permeability. This is the multiplier to get the actual permeability. For example, if a certain iron alloy has a relative permeability r = 900, we mean that its actual permeability is 9000 . Free space, then, must have a relative permeability r = 1.
6.1.3
EXAMPLE IFORCE BETWEEN BUSBARS
Busbars are copper or aluminum bars that are arranged to carry large currents. They are often used where round, insulated conductors would be too large. They are generally mounted on insulators and enclosed in a bus duct that provides support and safety isolation. Two busbars are mounted parallel to each other as shown in Fig. 6.21 . The bars are held in place by a bolt through both bars into an insulating block (see Fig. 6.3). The bolt is insulated from the metal bars by insulating sleeves. The bars are held apart by a bushing on the bolt. Bolts are installed at one-foot spacing along the busbars in the bus duct. These bars are to be built to carry 2,500 A d-c. We are to nd the force on the bolts. Lets see. The currents are in opposite directions if this is to be a complete circuit. Hence, the forces are attempting to separate the bars, which puts tension on the bolts. Ill nd the separating force per bolt, so = 12. But where is the current? For d-c, the current will be uniform throughout the bar. If I am to do this 100% right, Ill dene a small thread of current i and then use the calculus to get the force. However, Im lazy and instead will consider that all the current is concentrated dead-center in the cross section of the bars. This means that the radius r in Amperes Law is 1 + 1/4 + 1/4 = 11/2. Note that these dimensions need to be metric, where 1 = 2.54x102 m.
1The dimensions of the busbars needed have been taken from Table 5, Standard T1.311, Alliance for Telecommunications Industry
Solutions.
164
6. ELECTROMAGNETICS
Figure 6.2: Busbars
Figure 6.3: Busbar support

4 10 7 ( 2500 )
F=
2 (1.5 2.54 10 2 )
(2500) (12 2.54 102 )
= 10 N/ft
Thats the force attempting to stretch one bolt. How much is this in pounds? Divide by 9.8 m/s2 and divide by 2.2 lb/kg:
F = 10 9.8 2.2 = 2.25 lb/bolt
That sure isnt much! But wait! What about the possibility of electrical failures that could cause huge fault currents? One EE estimates the fault current could be as high as 100,000 A before circuit breakers clear it. Since the force depends on the square of the current, the fault could produce a force of
6.2. MAGNETIC FIELDS
165
Ffault = 2.25
100000 2 = 3600 lb/bolt 2500 2
Quite a different outcome.
6.2
MAGNETIC FIELDS
Think about this force between the conductors for a moment. Notice that it is in a way like gravity. With gravity, I hold a ball near the ceiling, let go, and it drops to the oor in a predictable manner. If instead I hold the ball near the oor, it still heads for the oor. Likewise with the second wire. I hold it somewhere and a force moves it in a predictable manner. I hold it somewhere else and there is still a predictable force. So whats happening around the innite wire? Theres something there, much like there is gravity in the room. That something is a magnetic eld. The current I1 has somehow lled the space around itself with a eld that can interact with another current. Take a look at Fig. 6.4. Its a simple representation of this force eld that seems to be out there around the rst conductor. If I place the second conductor anywhere on the dotted circle, a distance r from the rst conductor, I will feel a force pushing outward on that conductor. Move the conductor out to 2r and Ill feel half as much force.
Figure 6.4: Force eld
6.2.1
MAGNETIC FLUX DENSITY B
The eld created by the rst conductor is independent of the second conductor. Its strength depends on the intensity of the current I1 and the distance r from the rst conductor. Theres no second conductor as far as this eld is concerned. This eld is described as the B eld, more properly called the magnetic ux density. It depends on I1 and r :
B=
+ I1 webers/m 2 (Wb/m 2 ) 2/ r
166
6. ELECTROMAGNETICS
Ill use the webers-per-square-meter unit, but there is an ofcial unit for this, the tesla, abbreviated T . Take note of some of the properties of this eld: Its called a density so it has area in its unit Its caused by the moving charges of a current appears, so it depends on how magnetizable the material isa certain I1 will produce a larger ux density B in iron than in air It has a direction that can be determined by the right-hand rulehold your right hand so your thumb points in the direction of the current I1 and your nger will curl around the wire in the positive B direction
6.2.2
MAGNETIC FLUX
What happens if we multiply B by an areajust in general? Since B is webers/m2 , the result must be something in webers (Wb). That something is ux, just plain ux, no density involved. The symbol is a capital :
\=
0 B>dA webers(Wb)
A
But this is more complicated than I said we were going to get. Lets reduce this to the worst case by stating that the ux density B is uniform and perpendicular to the area. Then the formula for the magnetic ux reduces to a scalar:
= BA webers (Wb)
Total ux is ux density B through an area.
6.2.3
MAGNETIC FIELD INTENSITY H
Now we have a ux that is created by a current, but this ux depends on the material. We need something akin to an ideal voltage source that provides the push to create this eld in the material, a push that is independent of the material. If its to be independent of the material, it cant have a permeability in the equation. We dene the magnetic eld intensity H as the drive that creates the ux B :
H= B I ampere-turns = 1 (A-turns/m) 2 r meter
Turns? OK, thats not really a unit, but we use it here as a count, a multiplier. Very often we are dealing with a coil of wire in the magnetic structure. Hence, the current goes past lots of
6.2. MAGNETIC FIELDS
167
times. For example, if a coil of 100 turns carries a current of 50 mA, the number of ampere-turns is (100)(50x103 ) = 5 ampere-turns. This coil will act the same as a single wire loop carrying 5 amperes.
6.2.4
MAGNETOMOTIVE FORCE F
Magnetomotive force (mmf) is the electromotive force (emf) of magnetic systems. Just as a voltage source provides emf for an electric circuit, current in a wire provides mmf for a magnetic circuit. The formal denition of mmf is
F=
0 H >d Campere-turns(A-turns) x
This is the integral around a closed path that encloses the current. But again, Ill reduce this to the worst case:
F = H C ampere-turns(A-turns)
Since this is in ampere-turns and those turns constitute a coil in the magnetic structure, the mmf also turns out to be
F = Ni ampere-turns(A-turns)
These come from Amperes Circuital Law, but we wont go into detail here.
6.2.5
CONNECTING ALL THIS
Lets work from the end backward to see how all of this connects. Weve really seen two different sides of this electromagnetics stuff. One is the force created between current-carrying conductors. Amperes Law tells us what the force is, which we use to consider the mechanical stresses on conductors and their supports and enclosures. The other side is the eld relationships. These well use to see how force is created in an air gap in a magnetic structure and while we are at it, how inductance comes about. Work backward through this section: A current owing through the wires of a coil creates a magnetomotive force, mmf, F, that provides the magnetic eld. From the mmf, F, and the length of the magnetic path, we can get the magnetic eld intensity H = Ni/ , where is the length of the path. The magnetic ux density B = H, where is either = 4 x107 for air or a much larger number for magnetic materials. Finally, the total ux = BA, where A is the cross sectional area of the magnetic path.
We will use some of this in the next section to develop the magnetic concept of reluctance and from there, inductance. Our goal with these is to be able to nd the force in an air gap in a magnetic structure, which will be useful for converting electrical energy into mechanical energy.
168
6. ELECTROMAGNETICS
6.3
RELUCTANCE AND INDUCTANCE
Reluctance is a strange-sounding term that sort of conveys the same idea as resistance. Grammatically, they arent all that far apart. I resist going to the pub tonight because I have to study. I am reluctant to go to the pub.
6.3.1
RELUCTANCE
Lets play with the equations we already have to get to the concept of reluctance. First, we relate magnetomotive force to magnetic eld intensity and then to magnetic ux density, all of this in some magnetic structure that has a magnetic path length (a closed path) of :
F = HC =
B C +
Now move from ux density to ux by giving the magnetic path a cross-sectional area A:
B= A
Gather terms and rearrange things a bit:

F=
C \ A C = \ + A +
Interesting result? I think so! Look at what it says. Theres magnetomotive force on the left and magnetic ux on the right. They are connected by a term that is entirely dependent on the dimensions of the magnetic structure and what it is made from. We give this term a name, reluctance, which is
= C ampere-turns/weber +A
Reluctance is proportional to the length of the magnetic path and inversely proportional to its cross-sectional area. The larger the permeability , the smaller the reluctance. Now we can write the Ohms Law for magnetic circuits:
F = \
6.3.2
INDUCTANCE
Reluctance has wrapped up into one quantity all the physical parameters of the magnetic structure except one. Theres nothing in what we have just done that says anything about a current or a wire or a coil, yet weve implied all along that amperes and ampere-turns are part of the picture. What happens when we wind a coil of wire around our magnetic structure? Weve just written Ohms Law for magnetic circuits. This gives us the magnetomotive force in terms of the ux and the reluctance. Magnetomotive force is also ampere-turns:
6.3. RELUCTANCE AND INDUCTANCE

F = Ni = \
169
Inductance is dened as ux per unit current. Since we often have coils of wire instead of just a single wire, each turn of the wire in the coil provides inductance, so the number of turns N is included in the denition:
L=N henries (H) i
Replacing the ux the physical setup:
using the Ohms Law equation yields inductance entirely in terms of

Ni N 2 = i
L=N
Heres the complete result:

L= N2 +A = N2 henries(H) C
6.3.3
EXAMPLE IIRELUCTANCE OF A CORE
Find the reluctance of the core shown in Fig. 6.5.The cores cross-section is square and all dimensions are given in centimeters. The core is made of an alloy whose relative permeability is 900.
Figure 6.5: Magnetic core
The reluctance in terms of the dimensions is

=
r
C
0
Hmmm, three of the numbers needed are easy to get because we know the relative permeability, we know the permeability of free space, and we can easily calculate the cross-sectional area:
170
6. ELECTROMAGNETICS
+ r = 900 +/ 10 <7 N/A 2 0 = 4

A = (0.01)2 = 10 <4 m 2
But whats the path length ? Theres no obvious single path length. If we choose to go around the outside of the core, we get one result; if we choose to go around the inside, we get another. We need a good estimate. For relatively small cross-sections, we usually choose the mean path length, which we dene as the middle of the road. The dotted path in Fig. 6.6 is along the center line of the legs of the
Figure 6.6: Mean path length
core. This gives a path length of

<2 C = 2 ( 20 < 0.5 < 0.5 ) + 2 (10 < 0.5 < 0.5 ) 10
= 0.56 m
The reluctance of this structure is

= 0.56 (900) ( 4 107 )(104 )
= 4.952 10 6 A-turns/Wb b
If you are bothered by the question of what does this number mean, just take it as a number that represents a physical property. Its like resistance, which represents a physical property of a resistor.
6.3.4
EXAMPLE IIIINDUCTANCE FOR EXAMPLE II

500 2 = 50.5 mH 4.952 10 6
Find the inductance of the coil of 500 turns wound on one leg of the structure of Fig. 6.5.
L=
6.3. RELUCTANCE AND INDUCTANCE
171
It doesnt make any difference where we wind the wire as long as all the loops go around the core in the same direction. But we usually wind the coil with the turns tightly together as well as tightly against the core. That tends to reduce ux leakage, which we assumed didnt exist in the ideal transformer of Chapter 4.
6.3.5
EXAMPLE IVCURRENT FOR EXAMPLE III
A current of 50 milliamperes is applied to the coil of Example III. Find the total ux in the core, the ux density B , and the magnetic eld intensity H Ill start with Ohms Law for magnetic circuits since I know the magnetomotive force F and the reluctance R
F = Ni = ( 500 ) 50 10
<3
) = 25 A-turns
\=
25 = 5.049 10 <6 Wb 4.952 10 6
The ux density B eld comes from knowing the cross-sectional area:

B= 5.049 10 6 = = 50.49 10 3 Wb/m 2 4 A 10
The magnetic eld intensity H relates to B through the permeability:

H= B 50.49 10 3 = = 44.64 A-turns ( 900 ) ( 4 10 7 )
That answers all the questions, but we have a way to check this result. Remember that the magnetomotive force F is related to the magnetic eld intensity H and the path length, so we can check to see if we get the right number of ampere-turns from H :
F = H C = ( 44.64 )( 0.56 ) = 25 A-turns
Good! Thats what we started with.
6.3.6
AIR GAPS
Suppose the iron core is not a complete loop. What happens if there is a gap in that loop, a gap we call an air gap? Well, the permeability of that air gap is certainly a lot smaller than the iron so lets see, the reluctance should be a lot larger. (The is in the denominator of R.) Consider this: reluctance is kind of like resistance and the Ohms Law that we have for magnetic structures looks like the real Ohms Law. So it sounds like reluctance of an iron path with an air gap would be like resistors in series. We nd the reluctance of the iron and the reluctance of the air gap and add them to get the reluctance of the structure.
172
6. ELECTROMAGNETICS
Figure 6.7: Magnetic circuit
This idea is shown in Fig. 6.7, where resistors are the analogs for reluctances. In this circuit, the following applies:
F = \ ( i + a )
To say this another way, reluctances around a magnetic path add just as do resistors in series.
6.3.7
EXAMPLE VAIR GAP FOR EXAMPLE II
The iron structure and coil of the previous examples has a thin air gap cut into one leg (Fig. 6.8). This gap is made using a thin saw that leaves a kerf of just 0.1 cm. What happens to the reluctance
Figure 6.8: Core with air gap and winding
and the inductance of this structure? The path length in the iron is a tiny bit shorter, which we could almost ignore, considering that we are already approximating the path length. But lets not ignore it:
<2 Ci = 2 ( 20 < 0.5 < 0.5 ) + 2 (10 < 0.5 < 0.5 ) < 0.1 10 = 0.559 m
Thus, the reluctance in the iron part of the path is
6.4. FORCE IN THE AIRGAP
173
i =
0.559 (900) ( 4 107 )(104 )
= 4.943 10 6 A-turns/ /Wb
The length of the air gap is 0.1 cm= 103 m, so the reluctance of the gap is
a = 10 3 (1) ( 4 107 )(104 )
= 7.958 10 6 A-turns/Wb
That tiny gap has almost twice the reluctance of the entire iron path! Notice that the relative permeability of air is 1. Now add these two to get the total reluctance of this magnetic structure:
= 4.943 10 6 + 7.958 10 6 = 12.90 10 6 A-turns/Wb
The inductance is now

L= 500 2 = 19.4 mH 12.90 10 6
That small gap sliced through the iron cut the inductance by more than 60%. In general, we dont want air gaps when we are building inductors and devices like transformers. But as well see in the next section, it is the air gap where we can get our magnetic structure to do work for us.
6.4
FORCE IN THE AIRGAP
Take a look at the core in Fig. 6.8 (previous page). Its not hard to imagine that there is a force that is trying to close that air gap. Why? Just think about magnets that youve played with as a kid. You found that in certain positions they attract each other. You also learned, I am sure, that opposite poles attract, like poles repel. Thats the situation in an air gap. The current in the coil produces ux in the iron. One side of the gap is going to be a north pole, the other, a south. (Which is which isnt important, only that they are opposite.) So the two poles attract, trying to close the gap.
6.4.1
FORCE IN ENERGY TERMS
Theres a force F here, trying to close the gap. Suppose it succeeds by a small change in the gap, x. Then there is a change in the mechanical energy:
Wmech = F x
174
6. ELECTROMAGNETICS
where x is the spacing of the gap. The sign is negative because the energy is being decreased. This change in mechanical energy must be the same as a reduction in the energy in the magnetic eld:
W field = Wmech = F x
The force attempting to close the gap is

F = W field x
Notice that the magnitude of the force increases as the elds energy increases. Now lets switch to the calculus to nish the job:
F = dW field dx
We should be able to recall from electric circuits that the energy stored in an inductor is
W field = 1 2 Li 2
Use this in the derivative to nd the force while keeping the current i constant:
F = d ( 0.5 Li 2 ) dx = (0.5i 2 ) dL ( x ) dx
Yes, that looks funny, but only the inductance itself can be affected by the gap length x as we saw in Example V. We now need the inductance in terms of the dimensions of the air gap:
L= N 2 N 2 A = x
Ive used for the length of the gap x . Differentiate L(x) with respect to x (remembering that d(hi/ho) = ho dee hi minus hi dee ho over hoho):
dL ( x ) dx =
AN 2 x2
Finishing this up by combining this result with the force equation and using relative permeability gives us the force in an air gap:
+ AN 2 F = (<0.5i 2 ) < x2
+ + A ( Ni ) newtons (N) = r 0 2 2x
6.5. MORE EXAMPLE
175
6.4.2
EXAMPLE VIFORCE IN GAP
Whats the force trying to close the gap in Fig. 6.8 in Example V if we continue with the coil current of 50 mA that we used before?
F=
(1) ( 4 107 )(104 ) (500 50 103 ) 2 (2) (0.1 102 )
= 39.27 10 3 newtons (N)
Dividing by 9.8 m/s2 converts this to 4.00 grams, not a large force. But notice that we have control over a number of parameters in the equation for force. Both current and the number of turns are squared, so doubling the current, for example, quadruples the force in the gap. The gap spacing is also squared and in the denominator, so making the gap half as large quadruples the force.
6.5
6.5.1
MORE EXAMPLE
EXAMPLE VIIFORCE IN A CONDUIT
Two 60-Hz conductors are to be placed in a run of EMT, which stands for electrical metallic tubing. Its often called thinwall because it is much lighter than pipe-style conduit. The conductors are to be able to carry 100 amperes. We are to choose the proper conductors and then nd the force between the conductors under both full-load conditions and when the wires are carrying maximum fault current.
Figure 6.9: NEC Table 310.16: Ampacity
176
6. ELECTROMAGNETICS
The rst place to check is the National Electrical Code, which species the proper conductor sizes for various currents and other conditions. Table 310.16 in the Code species the ampacity for a given size of wire. Wire sizes in the United States are given by American Wire Gaugethe smaller the AWG number, the larger the wire. A check with a local electrical goods supplier nds that they dont stock very many types of wire and that the best choice is THHN-insulated, stranded copper wire. (THHN stands for Thermoplastic High Heat-resistant Nylon-coated.) The table (Fig. 6.9) shows that we need AWG #2 wire, which is good to 130 A. While AWG #3 would do, its not very commonly available. The next question is how large a conduit is needed for these wires. The National Electrical Code also species this based on lling a conduit to no more than about half full so there is sufcient space for cooling. Table C.1(A) of the Code, a part of which is Fig. 6.10, shows that two #2 THHN conductors need at least 1 EMT.
Figure 6.10: NEC Table C.1(A): Max. conductors in EMT
Now that weve selected the conductors, we need their dimensions if we are to calculate the force between them. A table on the Phelps-Dodge web site (Fig. 6.11) shows these dimensions. From this table, THHN AWG #2 has a diameter of 9.63 mm. Lets assume the conductors are lying together inside the EMT. If they are touching, the distance between their centers will be one diameter. Amperes Force Law gives the result:
F=
( 4/ 10<7 ) 100 2 100 2 1 + I1 I2C = () 2/ r (2/ ) (9.63 10<3 )
= 0.415 N/m
If we divide by 9.8 m/ss to get kilograms, then multiply by 2.2 lbf/kg to get pounds, and nally by 16 to get ounces, this result is just 1.5 ounces! Hardly worth worrying about.
6.5. MORE EXAMPLE
177
Figure 6.11: Phelps Dodge wire dimensions
Note in the calculation that I have used 100 2 to get the peak value rather than the rms value of the current and hence the peak force. The permeability is that of free space. This result is for a meter of EMT. What happens if there is a sudden fault on the circuit and the current rises to a peak fault current of 10,000 A? How much force is now pressing the wires apart? Lets assume the force is large enough to press the wires outward against the sides of the EMT. To nd the separation of the wires under this condition, we need the inner diameter of 1 EMT. That comes from Table 4 of the National Electrical Code, shown partially in Fig. 6.12. The table shows that, for a trade size of 1, the internal diameter is 26.6 mm. So if the wires of diameter 9.63 mm are pressed against the side of a tube 26.6 mm in diamter, the centers of the wires must be 26.6 9.63 = 16.97 mm apart. Ill use this in Amperes Force Law with a peak current of 10,000 A:
F=
( 4/ 10<7 ) (10000) (10000)(1) + I1 I2C = 2/ r (2/ ) (16.97 10<3 )
= 1179 N/m
178
6. ELECTROMAGNETICS
Figure 6.12: NEC Table 4: Dimensions of EMT
Converting this to pounds force yields 265 lbf outward against the sides of the EMT for each meter of tubing. If that force were spread uniformly around the circumference, it would be equivalent to only about 2 lbf/in 2
6.5.2
EXAMPLE VIIIACTUATOR
The device shown in Fig. 6.13 is an actuator. It is designed to pull a 1-kg mass upward 3 mm.
Figure 6.13: Actuator
6.6. SUMMARY
179
The magnetic structure will have a very high permeability compared with air, so we are not really interested in the structures dimensions. We need to know only the size of the air gap (length and cross-section). Our job is to nd the number of turns of wire for the coil if the coil current is to be 1 A. The available magnetomotive force is
F = Ni = 1N A-turns
The area of the air gap is

2 10 2 2 4 Agap = = 10 m 2
2
Using the equation we derived in Section 6.4.1 for an air gap gives a result in terms of the number of turns on the coil:
F=
(1) ( 4 107 )(104 ) N 2 2 (2) ( 3 103 )
= 21.93 10 6 N 2 newtons
The force necessary to just move the mass is

F = 1 9.8 newtons
Equating the required force to the available force and solving for the number of turns:
9.8 = 21.93 10 6 N 2 N= 9.8 = 668 turns 21.93 10 6
A check of wire tables shows that #24 AWG copper wire can handle 1 A. If the wire has a heavy lm insulation, 668 turns can be wound in a coil six layers deep and 2.5 long. Notice that the force equation contains the term Ni. If we want to use just half the current, we need twice the turns, and so on.
6.6
SUMMARY
Weve just scratched the surface of the eld of electromagnetics (no pun intended), probably going just far enough to convince you that youd rather not scratch any further! But I think weve gone far enough to be able to understand the common magnetic devices that an engineer can encounter. Current-carrying conductors have a force between them, a force that is of interest when we build supporting structures for conductors that carry large currents. Most mechanical devices that use magnetism to transform energy from one form to another do this via changes in air gaps in magnetic structures or via changes in the relative positions of current-carrying coils. Weve seen a little of how this energy arises.
181

1. Amperes Force LawForce per length between two conductors r meters apart and carrying currents I1 and I2 . Force is a separating force if currents are in opposite directions.
F=
+ I1 I 2 C newtons(N) 2/ r
2. Permeability of free spaceMagnetizability of materialr is permeability relative to

+ 0 4/ = 10 <7 newtons/ampere 2 (N/A2 )
3. Magnetic ux densityAt a distance r from a conductor carrying current I1 . Direction can be determined by the right-hand rule: thumb in current direction, curl of ngers in ux density direction
B=
+ I1 webers/m 2 (Wb/m 2 ) 2/ r
4. Magnetic uxTotal ux through an area A when the ux-density vector is normal to the surface area A. This assumes that the ux density B is uniform over the area
= BA webers (Wb)
5. Magnetic eld intensitySource of a magnetic eld from a current ow I1

H= B I ampere-turns = 1 (A-turns/m) 2 r meter
6. Magnetomotive forceAkin to electromotive force, which is voltage in an electric circuit

F = H C = Ni ampere-turns(A-turns)
7. ReluctanceResistance of a magnetic structure of length and cross-section A

= C ampere-turns/weber +A
8. Ohms Law for magnetic circuits

F = Ni = \
182
9. InductanceCoil of N turns on a structure with reluctance R

L= N2 +A = N2 henries(H) C
10. Force in an air gapMagnetomotive force is Ni, gap has length x and cross-section A
+ r + 0 A ( Ni ) newtons(N) 2x2
2
F=
183
Authors Biography
WILLIAM ECCLES
Bill Eccles has been Professor of Electrical and Computer Engineering at Rose-Hulman Institute of Technology since 1990 (except for one year at Oklahoma State). He retired in 1990 as Distinguished Professor Emeritus after 25 years at the University of South Carolina. He founded the Department of Computer Science at that university, and served at one time or another as head of four different departments, Computer Science, Mathematics and Computer Science, and Electrical and Computer Engineering, all at South Carolina, and Electrical and Computer Engineering at Rose-Hulman. Most of his teaching has been in circuits and in microprocessor systems. He has published Microprocessor Systems: A 16-Bit Approach (Addison-Wesley, 1985) and numerous monographs on circuits, systems, microprocessor programming, and digital logic design. In this Synthesis Lectures in Digital Circuits and Systems series, Bill has published several texts in this Pragmatic series, all to introduce electrical topics to non-electrical engineers. Bill and his wife Trish have two children and four grandchildren. Bill is also a conductor (appropriate for an electrical engineer) on the Whitewater Valley Railroad, a tourist line in Connersville, Indiana. He is a Registered Professional Engineer and an amateur radio operator.

Pragmatic Electrical Engineering: Fundamentals

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Pragmatic Electrical Engineering: Fundamentals

Transféré par

Droits d'auteur :

Formats disponibles

M &C