University of Central Florida University of Central Florida

STARS STARS

Retrospective Theses and Dissertations

1977

Using Ebers-Moll Equations to Evaluate the Nonlinear Distortion in Using Ebers-Moll Equations to Evaluate the Nonlinear Distortion in

Bipolar Transistor Amplifiers Bipolar Transistor Amplifiers

Amanollah Khosrovi University of Central Florida

Part of the Engineering Commons

Find similar works at: https://stars.library.ucf.edu/rtd

University of Central Florida Libraries http://library.ucf.edu

This Masters Thesis (Open Access) is brought to you for free and open access by STARS. It has been accepted for

inclusion in Retrospective Theses and Dissertations by an authorized administrator of STARS. For more information,

please contact STARS@ucf.edu.

STARS Citation STARS Citation Khosrovi, Amanollah, "Using Ebers-Moll Equations to Evaluate the Nonlinear Distortion in Bipolar Transistor Amplifiers" (1977). Retrospective Theses and Dissertations. 346. https://stars.library.ucf.edu/rtd/346

USING EBERS-MOLL EQUATIONS TO EVALUATE THE NONLINEAR DISTORTION IN BIPOLAR TRANSISTOR AMPLIFIERS

BY

AMANOLLAH KHOSROVI B.S., ~ well Technological Institute, 1974

Lowell, Massachusetts

RESEARCH REPORT

a · al fulfillment of the requirements e o Master of Science in Engineering G duate Studies Program of the o lege of Engineering of

or "da Technological University

Orlando, Florida 1977

ABSTRACT

USING EBERS-MOLL EQUATIONS TO EVALUATE THE

NO I EAR DISTORTION IN BIPOLAR TRANSISTOR AMPLIFIERS

by

Amanollah Khosrovi

e ers- oll model , which is applicable to static

o a ic cond"tions , is used as a basis for

0 ,

0 0

a ~ p l e ethod for the evaluation of harmonic

enerated in bipolar transistor amplifiers .

eq ations are transformed into the desired

ac aurin Series expansion . A computer

•tte to provide numerical results of the

t ese predictions are compared to measured

·al es.

ACKNOWLEDGEMENTS

I _.is to acknowledge Dr . R. L. Walker for his

i ce d advisement in reviewing this research

r e 0

I 0 i to extend thanks to my committee members

• • E i c on and Dr . B. E . IVlathews . I am especially

0 • E . . ·a heMS for his assistance and

e co 0 au e entire school program .

iii

TABLE OF CONTENTS

c 0 lENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... . . iii T TS ••••••••••••••••••••••••.•••.••••.••.• i v

I . 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

0 PROBLE . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 17

TI-lE 'ffiTHOD •••••••.••••••••••••.•••• 18

o Basic Equations ............ 18 for Collector Current ... 22

LE •• · • · • · • • · • • • · • • • • • • • • • · · • · · • • • • 35

0 0 ERRORS •••••••••••••••.••••••• 47

CLUSI ON ••••••••••••••••••••.•••.•. 49

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... . .so D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -54

0 , •................•................... 57

iv

a

0

I. INTRODUCTION

Distortion In Transistor Amplifiers

I an ideal amplifier , the amplified output voltage

0

e ersion of the input voltage without any

ap . Ho ever, in practical amplifiers

difference in the waveshapes of

o ol a· e and the applied input

ere ce is called distortion. The

e ca sed either by the transistor itself

a d c1rc "tor by both . An important

o ·sto tion is the so- called "non­

o " ·h·ch ·esults from a non- linear

o · en u e instantaneous values of the

he amplified output current . Such a

1 uwed generally by the non- linearity of

c racteristics .

In c ses here distortion is present it is con-

en nt o express the response of an amplifier using a

pe iod·c ·nput waveform. A periodic input waveform may

be e·pressed in terms of its Fourier Series components in

the 10rm:

e

=

2

n=oo

e1 = E0 + 2:: E~ Sin(nw t + q, n)

n=1

.pl ·· tu e of n h harmonic in val ts .

( 1. 1)

E = . c. co ponent of the input voltage (in volts) .

~ = w =

=

0 (

=

0

1 of the nth harmonic .

requenc of the input signal ei .

c o the input signal ei .

e s ideal , the amplitude of all

appear at the input are multiplied by

n A, and the phase angles are

p oportional to their frequencies.

pl"fier , the amplified output wave-

po n ) be expressed in the form :

e0 = A Em1 Sin( \U t + 'f + cp1 )

+ A Em2 Sin(2 UJt + 2 'f + <P 2>

+ A Em) Sin() uJt + J 'f + c.p3) + (1 . 2)

Let W t + \f' = U.J t '.

hus ,

J

+ AEm2 sin ( 2 W t • + cp 2 J

+ A Em3 Sin ( 3 Wt ' + <fl 3 ) + ... (1 . J)

ro . (l.J) it can be seen that the amplified

t o e0 as the same waveshape as the applied

, /w,

,

ol e ei ' the only difference being an increased

e a t · e delay of the wave by an amount

0 1 there is no distortion .

o the amplifier differs from the

~s p ese t . Different types of distor­

in an amplifier , either separately

eo 1 , are :

plitude distortion

uenc distortion

3· P se distortion .

t ion

ase of amplitude distortion the voltage

the amplifier varies with the amplitude of the

n u , ·.e ., the amplifier output waveform has a

non-l ·ne relationship with the applied input voltage.

An e. ample of a characteristic which results in ampli -

tude distortion is shown in Fig . 1 . 1 .

2. Frequency distortion

Frequency distortion occurs when the voltage gai n

4

c

Linear portion

input voltage

i . 1.1.Input-output characteristics of an Ideal

d Pra t cal Amplifier.

5

i h he equency of the applied input voltage .

T , the input oltage as given by Eq . (1 . 1) , ex-

pe ences a di ~ ent amplification factor dependent on

e e

e

0

J .

e r disto tion is primarily caused by the

nc of eac i e ele .. ents in the amplifier circuit .

e t

o a o

0

0

0

0

1 , plitude and frequency distortions may

c

e

o sl and may cause the output wave­

from the input waveform .

:as stated that for an ideal ampli ­

s o ld be either zero or propor-

e e c ~. If an amplifier has a phase

oportional to the frequency , phase 0

e distortion, like frequency

esults from the frequency depen­

cte istics caused by reactive elements

ssociated with the amplifier .

Sou~ces of Distortion

The sources of distortion in transistor ampli ­

r· rs a listed below:

1 . Non-constant spacing between constant- current

curves , particularly along the load line .

6

2 . on-l·near input resistance.

J . Too-l e a signal so that clipping occurs from

satur ation or cut off .

o em·n of the bias point with variation in

·ure, which causes clipping of large­

si al inputs.

t o de d- ase connection the constant-

,e e s are normally nearly equally spaced

e u

ol

0

e

0

e

e a

0

~·~~o-L.l. The presence of distortion

tra; sister amplifier by dra~::ing

g a plot of Ic versus IE . If the

~ ·ne up to the saturation region ,

o t · a f om non-equal spacing .

i tance of the transistor amplifier

n -dependent, distortion is present .

li ie~ this ·ill result in a distorted

_ d the output current will be an ampli -

o th"s distorted wave .

Di o ·an caused by clipping occurs when the

1 c· uses the amplifier to swing into saturation or

into the cut-off region . Both saturation and cut- off

will c u e the output to be clipped . If the allowable

po~. · e dissipation permits , the clipping may be remedied

by moving the operating point farther from the origin

(of the raph).

7

I he signal operates close to saturation or the

c - o f re ·on , distortion may be caused by the bias

poi s J. ing because of a change in ambient temperature.

a

0

h"rt rna· ause the signal to move into either the

0 he cut-off region .

eristics of the grounded emitter amplifier

o o as uniformly spaced as the grounded

e rounded emitter has inherently more

o be cor~ected to a certain extent

o ~ source resistance . Since the

o ·nput resistance is non- linear , as was

0

o ounded base amplifier , the source

ce osen to compensate for this non- linearity

o o o e e inherent distortion mentioned

o considerations are opposing ,

an optimum value of source resistance to

d·s~ortion . The clipping and bias sta­

pects of the grounded emitter are identical

o the grounded base.

Graphical Method For Calculating Distortion

The output current of a transistor amplifier may

be e.pressed by a Fourier Series of the form :

io

)

= Io

= I

=

=

0

00

+2 =

+ ff

e

c

8

2 r 0k Cos k w t

I 1 o s \.U t + I O 2 Cos 2 Wt ... ) ( 1 . 4)

e a e value , in amps , of i 0 when a . c .

is impressed at the input .

0 h . .

.armon~c ~n amps .

f" .al position of the a . c . load line

se the follo,ing procedure to deter-

1 . 2 sho san output character­

o . The intersections of the a . c .

ed a.c . load line (due to distor-

·stic curve corresponding to

, de er ·ne the quiescent operating point

Point A is the intersection of the

d . c . load line with the d . c . load line .

i t e harmonics higher than the fourth are

ne li · le , the output current can be written as :

+ r 03 Cos J uJ t + r 04 Cos 4 w t) .

9

p e ons for the five unknown coefficients , lOA '

r 01 , 02 , r03 , and r04 may be obtained by evaluating the

o e a

o. 0

c 00

0 0

0 . '

ol

Io

I a~

Iop

e dif erent points on the a . c . load line .

=

=

=

he a . c . input signal is assumed to be

inp ~ bias to be xc . The input signal is

+ 2 A X Cos w t c

0 i of t e time axis so that it

e val- e of the input signal .

w t as 0 , TTt J , lT/2 , J , andTI;

s o x are : xc + 2 ~x , xc + l:l x ,

- 2 6 The corresponding values

ed b lomax , l ao< , lop , lop , and

ted in Fig . (1 .2) .

ese values for UJ t and i 0 into Eq . (1 . 5) ;

i e s1multaneous equations may be written :

1oA + f2<1o1 + 1o2 + 1oJ + 1o4>

1oA + fi< 1o1 - tro2 - iiOJ - -l1o4>

1oA + f2< - 1o2 + 1o4>

00

0 .

10

x -AX c

The generalized output characteristics

~1) th d . c . load line , (2) the a . c . load line ,

( ) the . c . load line sho ing the shift due to

d. to t"on .

o e

11

Q5 = 10A + i2(- 101 - 2102 + 1oJ -~!04)

1om·n = 1oA + 12(-!01 + 102 - 10J + 1o4)

e e a o s ; the expressions for harmonic

d ·e ob ained as follo~s:

1 1omin) 1

+ Icy.g ) (1 . 6) = .., Io + + J<root. . ax

~-:- -( Io . ) 1 - IO{J ) (1 .7) = + J<ro~ J 0 n

~ ( 1omin) 1 (1 . 8) = 2 1op 0 a:·

Io . ) 1 Iop ) (1 .9 ) 0 = - -(I nu.n J ()o(

+ Io . ) - 1 + Ia/9 ) = J<Ia~ nun

1 + 2 1ap (1 .10)

ous harmonics are usually given as a per­

rna itude of the fundamental term . Thus,

e c n e harmonic for second , third , and fourth

h rmon c are defined by:

0

0

he e :

e - o 1

= I

I = - ol.

12

DJ = !02 X 100% 1o1

D = 1o4

X 10()% . 1o1

Ebers-Moll Equations

- all ations a e for low frequency

.e ra sistor and are derived 0 0

1 0 1 io . For pnp transistors ,

i ons are gi- en by equations 1 .11 and

( qv. I

- 1) - o( RICS (e CB KT - 1)

(1 . 11)

(eq .EB/ KT _ 1 ) + I (eqvCB/KT _ 1) ES CS

(1 . 12)

IES = Emitter saturation current when collector is

shorted to the base .

Ics = Collector saturation current when emitter is

1J

horted to the base .

O( F = Fo ~' rd short-circuit current gain .

D( = de rsed short- circuit current gain .

J .

fo constants depend on the diffusion

, t d"ffusion lengths , the equilibrium

oncentration , the area and the base

o s ade in deri ~ing these equations

0 ~

e s o is operated at low frequency , so

e · . ,e elements can be neglected .

o e terminal currents which change

c r iers stored in the transistor

c ed.

ndence of the base- width on junction

o t e s neglected , i . e ., base- width is

constant .

he transistor is subject to low- level

injections only .

5. Ohmic voltage drops at the contacts and in the

neutral regions are neglected .

Equations (1 . 11) and (1 . 12) apply to pnp transi s­

tors . The corresponding relationships for an npn tran-

14

i o e iren by:

0

lie

0

I

- ,

0

-q -qv I = -IES (e ~B/KT - 1) + o< RICS (e CB KT - 1 )

(1 . 13)

-q ( - qv I = o( I S ( e EB KT - 1) - Ics ( e CB KT - 1)

( 1 . 14)

l ,eq Y.ations, .'hich describe the large­

e i ·cs of the idealized transistor

a s ple and useful interpretation in

model that uses two idealized ex-

is oael is shown in Fig . 1.J . The

i erpretation in terms of the inter-

~ e ransistor . The emitter and col-

a eac be resol-ed into two components .

1 o ··s into the idealized diode is the

or -carrier injection at that junc-

o e component , which is provided by the

o· c , is the consequence of minority- carrier

c on · he other junction and transport across the

b Thus , the first component of emitter current

qv I I S (e KT - 1) results from diode action at the

emitter junction . qvCBIKT

The second current - ol RICS ( e

- 1) is the consequence of diode action at the collector

(

+ •

15

B

all static model for a pnp

-

I- I (eqv/ KT - 1) - s

•

i . 1 . (b) . Idealized pn- junction Diode symbol .

c

16

j ct · o d exis s because a fraction of that diode

en , c<R R' is transported across the base to the

e i t e , he e i contributes to the total emitter

r nt .

0

0 I

e ba·e

0

0

o(

e c

e

II. T T ,1E T OF THE PROBLEM

p· ical ethod, described in Chapter I , to

e o ic content in the output of a bi-

s·o amplifier is very laborious and cumber­

. ~ · a , to determine the a . c . component of

t (co on emitter connection) assumptions

o ~ a .p e , Vbe is assumed negligible .

1 errors in the calculations .

~e normally inaccurate .

e o ~ect is to investigate a method

e ea sonable accuracy the harmon-

0 . .put val tage . The basis for

e s- all equations .

e ions the four parameters o( F '

C.;:, ( nom as the forward short- circuit

e erse short-circuit current- gain , emitter

ent rhen emitter is shorted to the base ,

1 ) must be kno' ·n.

These fou parameters are normally measured from

th ~ tran i to characteristic curves or taken from the

ve e d ta supplied by the manufacturer .

III. D I TION OF THE METHOD

A. Derivation of Basic Equation

) . 2 ) '

= - o<. c

0 0

0

(

0

ers- moll equations given by Eqs .

o tains :

qv / - 1)- ~RIGS (e CB KT - 1)

(J .1)

KT - 1) + Ics (eqvCB/KT - 1) .

(J .2)

o o an a plifier , the emitter-

·ased and the collector- base

d i sed :

c = ( C., VBE) <- 0 . 1 volt .

t 00 empe tu e 4~T/q = 0 . 1 volt .

hu , the term eqvCB/ KT in Eqs . () . 1} and (J .2) can

be con idered ne 0 li ibly small compared to unity . Hence

qs . ( . 1) and (J .2) reduce to :

19

= q '::B/ KT s (e - 1 > + ~ R1cs (J . J)

i = - o( I S ( e q ~B/KT - 1) - lcs . () .4)

o po· t on , onl the collector current

. ) . i 1 e studied , and the corres-

e •

0 . . • J . se . ·rch off ' s voltage

0 0

+

= (J .S)

. ( J . 5) and qj KT ~ L in Eq . ( J . 4) ,

0 0

(J . 6)

Now for transistor working in the normal active r egion

the ollowin equation holds :

B B

. • •

1 E

c

E

20

plifier circuit .

qv / Ico(e CB KT- 1)

--

Fig . ) . 2 . Equivalent Ebers-Moll model of Fig . J . l .

21

e o ~ . (J .6) yields :

o( F

- 1

0 :

A 1 = - o( ES

k ~ Lico

R' LvBB + D( 2 F s

nd k ~ LRs(1 - o( F)

) o( F

B.

22

i = () . 10)

Harmonic Equation For Collector Current

0 e e armonic content , Eq . (J . 10) will

a n Series Analysis . This

ppendix A. In this work only

e 0 1 CS ill be considered , as the

0 .. 0 ~at the higher order terms are

- L + k i ) o S 3 C can be expanded as :

1 + (k2 - Lv5 + kJiC)

(k2 - Lvs + k3iC)2 + 2

2LKJ

23

sic + 2k2kJiC k2 LJVJ + kJiJ + 2 s J c

6 2

2 - L

=

+ JL2v2(k s 2 + k~ic) 6

s> - 6k2Lv5k3

i 0 6

. () . 10) , one obtains :

kJ sic + k2kJic + ~

or , in terms of the descending powers of ic :

+ (

+ 1

0 0

24

1 3 - Lk1k3v5 + k1k2kJ

2 2

+ L k1k

2v8 - k1k2k

3Lv5 - 1) ic 2

2 L2 2k + 1 2 + s 1

2

J k LJVJ 2

+ k1k2 1 s k1Lk2v8

2 6 6 2

2 s

- 1cs) 0 () . 11) =

e aclaurin Series it is neces-

e ic successively with respect to

ppenidix A. The technique is as

·n i· in ~q . (J.11) ith respect to v5 yields :

25

) die

(-L ( . ) 1 ~ s a 5

+ ~c

( -

2 1 2 s = 0 (J . 12)

op t ng point , v5=o , ic=Ic (in this work

o in pain "11 be established) . So to obtain

di lu of the derivative d C at the operating point , vs

{J . 12);

v =0 s

substituting v5=o and ic=Ic in Eq .

26

2 die L =0

~ + 2J{J + k1k2kJ

- 1) 3 1 2 dv3 s

) ,., + ( - 1 2kJL ) Ic + { - Lk1

)

2

= 1 , and collecting terms con-=0

27

+ Lk1 Lk k

2]

+ Lk k + l 2 1 2 2

0

!2 + c LklkJIC + k1k2k3LIC

!2 + 2 2 c (klkJ + k1k2kJ) Ic

k1k~L} + 2

{J . lJ) 2

-1J .. ? ...... + 1k2k

3 .)

e

= = constant coefficient of the =

s o as indicated in Appendix A.

= o c rent (d . c . ) at operating point ,

o s ants given by Eqs . {J .?) , (J .8) , and

( J .

s· e differentiation must be taken of Eq .

(J . ) o e A2 , the harmonic coefficient of the

n Se ies e~pansion .

Di ntiating once again yields ,

v

[·

J

2 +

28

d. k LJ - + - ) + L2k1 - 1 2vs

d s

= 0

op

nd u titut·

ing point , i . e ., when v5=0 ,

d2. ~c

2 dv5 v =0 s

= A2

yields ,

29

s e final equation for A2 :

- (

2 L · 2Ic 1 2 - L 1 - k1k

{J .15)

h

=

e

oll o

d2. ~c

d 2

s=o

JO

= a coefficient of the Maclaurin

o as indicated in Appendix A, and Ic =

(d . c . ) at the operating point .

e i lds c . ( J . 16) :

d2 ' di d. 2 ~- + ~-c 2i _Q + ( ~c)

d .. C dv dv s s s

[ dJ. d2. d2. die 1 ~c + 2ic

l.c + vs

l.c vs2ie J

dv2 2 dv5 dv5 s dv5

die d2. die 2 d2. ~e

2 + 4ie ~e

+ 2v52- 2 + (- ) 2 dv5 dv5 dv5 dv5

die die ] dJ. l.e

(k1kJ + kl k2k) + 4-- + dvJ dv5 dv5 s

( - 3 )

J

J J =

31

(3 .16)

= ) , 5=0 (at the operating point) ,

i el ds Eq . ( 3 . 1 7 ) .

f> - (k1k~ + k1k2k~) (JA1A2)

+ A~) + )Lk1k~2 - JA1L2

k1kJ

+ kL3

(3 . 17)

0 ed· 0

oe . '

i

+ vs

32

d4. ~c

4 = 4 dv5

, the Maclaurin Series

dJ. die d2. d2. +

~c 2ie 2

~e ic

~e

3 -+ 2 2 d s dv5 dv8 dv8

c die die dJ. die 2 ~e --+ 2i -

dvJ + 6( - )

d s C dv dv8 s s

di dJ. +·4i _Q

J.e c e dv dvJ

s s

dJ. d. d2. d2. d. dJ. c I ~e ~c ~e J.e ~e 1e

+ -+ dv2 2 +- J J dv8 dv8 dv8 dv8 dv8 s

d' dJi d2. d2. 3[ d4. ~e J.e k 1Lk) . J.e

+ 2 2 2 ) - 2 vs21e 4 d s d s dv5 dv8 dv8

d). die dJ. dJ. c

2ic J.e

+ 6ic J.e

2-+ dvJ dvJ dv5 dv8 s s

+ 0

dlc

s

33

d2. d2. d. d3· di~ , ~c ~c ,. 1 c 1 c

_.J + ovs 2 2 + ovs - J d-s

d2. c

d-2 s

- 1)

+

dv5 dv5 dv5 dv5

] die d2ic

+ (k1k3 + k1k2kJ + 12 - 2 dv5 dv5

d4. ~c

+ (-Lk1k3

) dv4

s

2 21kJ [ d4.

? l.c 4 + s dv5

d . d3· 1 c 1 c --+ ) =0 d s dv~

[ vs d4. dJ. l.c l.c

4 + dvJ dv5 s

d3· d2. l.c l.c 2v5 +

dv3 2 s dv5

= A4 , v5 = o and ic = Ic (at the

o t n point) and rearranging the terms yields

. {J .18):

6

)4

J ( 2 6 2 2 2 J C .1 J + 3IcA2 + A1A2) - (k1kJ + k1k2kJ)

( ? 2

1 ~J + J 2 + 4k1Lk) (ICAJ + JA1A2 ) + 4AJ (LklkJ)

1 · J + (k1k2kJ ) 4AJ

= ----------------------------------------------

2 I

().18)

o e ~ . (A.4) from the Appendix A,

=

Co

().19)

= - alue of the input signal. The values of

e co i t A1

, A2

, A3

, and A4 are evaluated from

o · alues of the Ebers-Moll parameters ( o< F , o< R, ,

, nd Ic ); the harmonics can be determined from Eq.

( J .19).

0

0

0

IV. N MERICAL EXAMPLE

shovn in Figure 4 .1 . To check the

i entally measured values of VBB ' RB '

. e e substituted in Equation (J .19)

t , and a program was run on the

e program is shown in Appendix B.

es of c<. R, o( F, IES , and Ics were found

0 . •

pter III, for the normal active

to nt Ic is given by :

a d in"tion ,

[VeE = canst . 0 l normal gain region

= o( F ll ~ F 1- ~ F

( 4 .1 )

J6

=10KJ1. B

= · .535

• . 1 . The basic amplifier circuit •

37

, {J F can be found by noting the average spacing

o he o 1-emi tter output curves in the normal region .

e 1" ' o( F can be calculated from Equation (4 . 1) .

e p n of the common-emitter output curves in

o c i e region as found using a curve tracer .

1 e sur,ements ere taken the average value

0 a o p ed and was used in the experiment .

(J = 11.5

= fJF

1 = 0 . 92 + (A F

e ermined by noting the average

0 o -e "tter output curves in the inverse

0

[ CE = const . ) 0 l = ~1 __ 1_o(_R

·n erse-gain region

ut he , for a more accurate result , the emitter and

collector leads were interchanged (collector grounded) ,

and by noting the average spacing of output curves in the

0

s 10

)8

on of the inverted connection , the value of ~ R

d as :

o(. R =

1- ci.. R 1.5

saturation current parameters , IES

e e ed most accurately by direct

o ransi stor . The technique is to let

0 , · ·c· reduces the Ebers- Moll equation

es o Ic or different values of VCB were

o T n Ics as determined by plotting a graph of

Ic us VCB for VCB~0 . 1 , and by extrapolating back

To determine the corresponding reverse parameter

ICS ' on setting VcB=O the emitter current becomes:

39

qVEJ!KT IE = IES (e - 1)

A en ioned in chapter III , the four parameters

ollo ·ng equation.

(4 . 2)

, ·e need to know only either Ics or

e i e ·.as performed to determine Ics . Then

) .as plotted and the value of Ics was

o t intercept of the ordinate . Then using

( , I~s as calculated .

= 8.9J X 10- 9 A

Th e.:perimental results , which were obtained

u in H lett-Packard Model J02A wave analyzer (Fig . 4 . 2)

are shown in Table 4 . 1(a) and 4 .1(b) . The results using

40

t . 0.

0 r-i J

0

.J .4

vCB in volt-..

or

or

Fi0

• 4.1. The graph log of Ic versus vcB ·

HP

0

2A

Sig

nal

Gen

erat

or

c·

cu

Und

e T

W

ave

An

aly

zer

Fig

. 4

.2.

Har

mon

ic D

isto

rtio

n M

easu

rem

ent.

(

CE

0 c

0 c

3rd H

m

onic

s

v ee

in

(d.e

.)

eel

n in

er-

in

in p

er-

rms

vo

lts

in v

olt

s o

lt

rms

e n

ta

e rm

s ee

nta

ge

vo

lts

of

V eel

vo

lts

of

V

1 ee

+=

-l\

)

0.9

212

28.0

11

.6

0.5

8 4

. 96%

0

.084

O

.?J%

1.2

783

28.1

5 1

.14

6.?4

%

0.2

J 1.

J4%

)

0 0

vs

Ic

( . c

.)

01

0 J

d m

onic

in

( u

) u

) (p

e k

· lu

e)

VRL

v v

rms

vo

lts

Ic =

RL

1c1

=

i2

c2

=

Ic

.J =

ce

x

2 R

~

\.....

)

0.9

212

2.1

0 rnA

1

.10

rnA

0.0

54 r

nA

0.0

08

rnA

1.2

783

2.0

9 rn

A 1

.60

rnA

0.1

1 rn

A 0

.022

rnA

44

the measured para~e~ ers i n t he co mputer program a re

sho~n in Table 4.2. h co mparison of th e harmonic ralues

1s ;iven in Ta bl e 4.~.

0 Q

_ I

0

0 c

s 3

rmon

ic

in

pea

k

(d.c

.)

0 c

-rn

A in

rnA

in

per-

rms

val

ue

(Pe

k c

nta

e

(Pea

k ce

nta

ge

vo

lts

of

V8

in A

mp

Ic1

n

rnA

v

a t

) 0

Ic1

v

alu

e)

of

rc1

+=-

\.1\

0.9

212

1.)

02

6

2.J

9

1.2

7

0.0

64

5.

06%

0

.006

4 0

.51%

1.2

783

1.8

090

2.J

9 1

. 71

0.1

30

7.4

8%

0.0

17

1.0

9%

0

J

3

46

TABLE 4 . .)

S - 0 THE RESULTS OF THE EXPERIMENTAL

D COMPUTER METHOD

c

( . .

l.C

. c. )

onic

d armonic

d Harmonic

Experimental

ethod

2.100 mA

1.100 mA

0 . 054 mA

0 . 008 mA

2.09 mA

1.60 mA

0 . 11 mA

0 . 022 mA

Computer

Method

2 . .)941 mA

1.2700 rnA

0 . 0640 rnA

0.0064 mA

2.)941 mA

1.707 mA

0.1J rnA

0.017 mA

EST! TED SOURCE OF ERRORS

~~~~~~~~~o~d: The Hewlett-Packard Model J02A wave

e

0

·o 0

e

fo determining the harmonics has a

0 of less than 1%. The sensitivity of the

0 JO ,. v to JOO volts. Since the harmonic

sing the Ebers- ~oll equation , and

0 e not expected to approach the

t the experimental results have , these

e a ill be considered the standard for

_c rae of this method depends on (1)

o e bers-1oll model, (2) accuracy of the

e or measuring the voltages), and (J)

o e ssumptions and approximations made in

o he expression for harmonic evaluation .

o sl s~ated in the derivation of the Ebers­

the effect of the space-charge-layer

. on o( F, o( R, IES and Ics has been ignored .

cond , ~ F and o( R were considered to be independent of

u r nt ~n the model. In actual practice this is not the

e. So these assumptions will cause some error in the

results. Experience has shown that these errors are of

48

1 o o na nitude of 5%.

In de · in the computer method two approximations

e e (1) VCB~0 . 1 volt , where for the most part ,

0

a d- of this voltage is much greater than 0 . 1

e ~a can be considered small . (2) The expan­

o p ( 2 - L S + KJiC) was terminated after the 4th

e

e

0

0

circumstances , this could introduce

o·e er , previous experimental work

oi as and load networks are linear

ics higher than the Jrd do not

t changes in the waveshape . In

o e literature and experience with this

0 e o der of 10~ can be expected

e

0

·I. SU~~RY ArD CONCLUSION

bas·c objective of this work was to obtain a

o e 1 a in5 the harmonic distortion generated in

0

-ansistor amplifiers . Ebers- Moll equations

e o

0 f(

e - o 1

0

~ ... - ,- ···-1 .

basic equations for developing this

tions ··ere then transformed into a

.e ~ claurin Series expansion .

o · e .alue of harmonics that will be

en applied input signal , knowing the

~ I s and Ics (the four parameters of

a · ~ s), can be determined .

a 2N12J4 transistor , the four

ed e~perimentally , and harmonics

the derived method . The computer

·n in ppendix B. An experiment wa s

onics v·ere measured applying the two

. 1 the esul s of the two methods (computer and

· ent 1) were compared as shown in table 4 . J .

n u.wary , considering the experimental errors and

th assu ptions made to derive the computer method , the

results a e in excellent agreement .

( ) =

= ) :

PPENDIX A

e e sa function Z=f(x) , it can be expanded

) + ~ t ( 0 ) + X 2 f" ( Q) 2!

n :r(n) (0) + • • • + X n!

e e i 1 es are taken at point x=O .

is a ction ic=f{v8) , {Eq . J . 10) , we can

)

o bout the operating point (v8=0 ,

v =0 s v =0 s

v =0 s (A . J)

but ·0

= C + ic(d .c. component+ a . c . component) but

Ic=f(O) in Eq . (A . J) . So Eq . (A . J) reduces to

( . .

0 =

here

51

omponent)

v =0 s

di and A for _Q

f &a J 4 dv5

. , 0 0 ins:

=

o e s are retained .

0 w

1 + Cos 2 UJ t 2

v =0 s

v =0 s

v =0 s

v =0 s

=

0

52

0 J V-J t = Cos 3 t + )Cos t 4

Co 4 \U t = Cos 4 w t + 8Cos2 w t - 1 8

= Cos 4\IJt + 4cos 2UJt + 3 8

al e in above equation for ic yields :

A2

: A vJ o w + ~- + 4 Cos 2Wt + ~ Cos J W t

0

A 4 m

w t + 24x8 3 ·

ou th and higher order terms (terms

os 4 UJ t and higher) yields :

(A .4)

53

Co e · c · ents of Cos w t , Cos 2 UJ t and Cos 3 wt in

P e eses represent fundamental , second , and third

ic espectively . The last term (in the brackets)

l e e ectified a . c . component . Hence , the first ,

0 i d 1armonics are given by :

J = A1v + AJ;m

4 +

A4vm c 2 = 48

J m

cJ =

0 0 =

APPENDIX B

I I e p o r mming data were taken as follows :

L = L = I = )8 .46 volts- 1 (assumed)

= · 5.35 ol (Fig . 4 .1)

= = 10600 ohms (Fig . 4 .1)

= = -9 ( a e J?)

= = 1 -9 amperes (page J9)

0 = 0 = 33 " 10-9 amperes (page .39)

= = J .6 .. 10-9 amperes (page J9)

. J = c = 0 t i d from eq . () . 10) for V =0 . s

D D

55

PROGRAM FOR EVALUATION OF HARMONIC I TORSION IN TRANSISTOR AMPLIFIER

) ) .

I

-. OI*RSt AF / AF

. 0 =2 .O*R) / 27 . 0

3· )/27 . 0) . / J . O)

_(52 )**(1.0/ J .0))*( - 1 .0)

C/ 2.0+(1 . 0+T2)*(1 . 0+TJ*QC)

/ 2.0+(1 . 0+T2))+(T1*TJ)*(1 . 0+T2

1 J ' ) () . L*( C*AL(2)+AL(1)*AL(1) - TJ*{J . O* ) ( )+AL(1)*AL(1)*AL(1)) - J .O*AL(1)*AL(2)

- ( . ) .0 ABL*T1*TJ*(AL(2) - AL(1)*ABL+AL(2)* ( J .O) )

I D . L~TJ*TJ*{QC*AL(J)+J . O*AL(1)*AL(2)) -T1 * ( J -' . 0) * ( AL ( J) * C * AL ( 1 ) *4 . 0+ 6 . 0* AL ( 1 ) * AL ( 1 ) *

{. ) . 0* *A ( 2) *AL( 2)) -T1 *TJ*TJ*( 1 . 0+T2) *( 4 . 0* (1) · A ( )+ . O*AL(2)*AL(2))+ABL*T1*TJ*(4 . 0*AL(J)

- . 0 ABL*AL(2)+4 . 0*AL{J)*T2) A ( )= N t AD1 005 =1 , V =V (J) _IC=1 . 0*AL(1)*VM(J)+AL{J)*({VM(J)**J . 0) / 8 . 0)

BIC=(AL(2)*(VM(J))*(VM(J)) / 4 .0)+AL(4)*((VM(J)) **4 . 0)/ 48 . 0

CIC=AL{J)*((VM(J))**J . 0) / 24 . 0 DIC={(((VM(J))**2 .0)*AL(2)) / 4 . 0+(AL(4))*((VM(J)**4 . 0)/

64 . 0)-2 . )94

56

5

= .JO

. - 0 . - 02 , - 0 . - OJ, = 0 . - 0

( . ) = . 3 1

= . JO

-o .- . = =

. ) = .J

3

5

LIST OF REFERENCES

, Paul ., and Zemanian , Armen H. t o ·cs. McGraw- Hill Electrical and

Electronic ~ngineering Series . New York : ·-ill Book Company , Inc ., 1961 .

o . , • S. Principles of Radio Engineering . New o craw-Hill Book Company , Inc ., 19)6 .

. , and Searle , Campbell L. Electronic r~nc~ples Physics, Models, and Circuits . New

John Wiley & Sons, Inc., 1969 .

, o , d Halkias , Christos c. Integrated ~lec~ronics: Analog and Digital Circuits and S stems New York: McGraw- Hill Book Company ,

72.

, 1 P ., and Riddle , Robert L · ransistor Physics and Circuits . 2nd Ed .

lewood Cliffs N J.: Prentice- Hall , Inc .

oothroyd , A. R.; Angleo , E . J ., , Pa E . ; and Pederson , Donald 0 .

ementary Circuit Properties of Transistors . ol. J. New York: John Wiley & Sons Inc .