This action might not be possible to undo. Are you sure you want to continue?
INTERMEDIATE
ELECTRONICS
R. J. MADDOCK
C. Eng., M.I.E.R.E.
Lecturer in E/ectronics,
Department in Electrical Engineering,
Southampton College of Technology
Book I
Springer Science+Business Media, LLC
ISBN 9781489958617
ISBN 9781489958594 (eBook)
DOI 10.1007/9781489958594
©
Springer Science+Business Media New York 1969
Originally published by Butterworth & Co. (Publishers) Ltd. in 1969.
Softcoverreprint of the hardcover Ist edition 1969
Suggested U.D.C. number: 621·37/·38
Library of Congress Catalog Card Nurober 6918297
PREFACE
Many textbooks adequately cover the development of small signal
equivalent circuits for electronic devices. Students frequently
experience difficulty in applying such circuits to the solution of the
various configurations found in practice. In this volume I have
demonstrated step by step procedures by which the required solutions
may be obtained by the use of such equivalent circuits.
The first chapter shows clearly the necessity for equivalent circuit
methods and also discusses the d.c. biasing requirements and circuits
for valve and transistor amplifiers. A chapter follows covering the
fundamentals of network analysis required for the remainder of the
book. Valve and transistor equivalent circuits are then introduced.
and detailed methods of application are given with many fully
worked examples. Later chapters extend these methods to a discussion on the effects of feedback, the solution of practical feedback
amplifiers and to the analysis and design of a range of sinusoidal
oscillators. A final chapter introduces the modifications required
when considering high frequency operation. The required high
frequency equivalent circuits are given together with sample
calculations.
I have included worked solutions and problems with answers in
all chapters. The parameters chosen for the various equivalent
circuits are those most commonly available from manufacturers'
published data. British Standard symbols and units are used throughout except where very recent changes may be unfamiliar to the
majority of students.
Both valve and transistor circuits are discussed as I feel that for
many students the valve provides a useful teaching medium, but
in the later chapters, the stress is, in general, on transistor circuits.
The physical operation of the devices has not been covered, as
adequate Iiterature is available elsewhere. A list of works for further
reading is provided at the end of the book.
I have written this book as a result ofthe experience gained through
teaching students at Higher National Diploma and Certificate level
but I feel that it should also be useful for students at the early stages
of degree courses and the final year of technician courses.
V
PREFACE
I would like to offer my thanks to those members of the staff and
students of Southampton College of Technology who have helped
in the preparation of this book and to Miss L. Lavender and Mrs. R.
Huntingford for their assistance in typing the manuscript.
R. M.
vi
CONTENTS
Preface
v
1. Graphical Analysis of Electronic Circuits .
1
2. Fundamentals of Network Analysis .
42
3. Low Frequency, Small Signal Equivalent Circuits for Valves
and Transistors .
94
4. Use of the Valve Equivalent Circuit.
112
5. Use of Transistor Equivalent Circuits
140
6. The Theory of Feedback Amplifiers.
181
7. Practical Feedback Amplifiers .
220
8. Oscillators
255
9. Modifications to Equivalent Circuits for High Frequency
Operation .
291
Appendix 1. Use of Logarithmic Units
312
Works for Further Reading
315
Index
317
vii
transistors. for example. D. a capacitor.1. thermionic valves.c. Also large changes in temperature or humidity may well affect this model. CHARACTERISTICS The first description ofvalves and transistors that will be considered are the socalled d. Any chosen description will be suitable only for a limited set of conditions. The characteristics of a resistor . The difficulty in designing or analysing such circuits lies in their inherent nonlinearity. These characteristics are graphs showing how the various direct voltages applied to the device +I V (a) (b) I Figure 1.1 GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS Electronic engineering is principally concerned with the behaviour of electrical circuits or networks containing various nonlinear and active devices. The physical explanation of their behaviour will not in general be discussed in this book since this aspect is more than adequately covered by other authors.C. This representation holds only if the peak voltage does not exceed the breakdown value for the capacitor. characteristics. may be shown as a pure capacitance shunted by a resistance representing the dielectric loss. This concept is not peculiar to these electronic devices. We must first consider suitable methods of describing the behaviour of a particular device. transducers such as photocells and many other devices. These include the various types of rectifier.
In this case the parameter could be the ambient temperature.1. t1= [ 7oooc fa=40°C ta=20°C . The parameter again is VGlK· 2 .3b). the ambient temperature is constant thus collector current (/c) is plotted against the collector emitter voltage (VcE) using the base current (/B) as the parameter (Figure 1. as shown in Figure 1. With multielectrode devices such as pentode valves.2.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS change the direct currents ftowing in it. For the transistor collector characteristics.____ 650°C 6QQ°C V Vv=.z=::.2a and b.3a).QOC I (a) (b) I Figure !. Diode characteristics. Similarly with thermionic and semiconductor diodes the cathode and ambient temperature respectively could be taken as the parameters. while the anode current (h). The anode characteristics of a pentode are quoted for constant filament voltage (VF) constant screen voltage (VG2K) and constant suppressor cathode voltage (VGaK). other potentials or currents will have to be kept constant for a particular set of characteristics. The resulting family of characteristics are shown in Figure 1.+ V +V ta= 20°C fa=f. A variable that is fixed is known as a parameter of the characteristics. (a) Thermionic. lf there are more than two variables. is measured for different values of anodecathode voltage (VAK) taking the gridcathode voltage (VGK) as the parameter. A simple example of this would be the variation of current ftowing in a resistor with variation of applied voltage. the values of one are plotted as a second is changed with all others held constant. A 'family' of characteristics is thus obtained for different fixed values of VGK (Figure 1. (b) pn junction With threeelectrode devices such as triode valves and transistors the cathode or ambient temperatures are assumed to be constant and the parameters of the characteristics are taken as one of the applied voltages or currents. For example the anode characteristics of a triode valve are shown with constant cathode temperature (filament voltage).:=~:.
operating point for (a) triode valve and (b) transistor given by Ic 2 mA. For example a particular operating point for a triode valve could be given as VAK 150 V.C.2 V. h 50 flA or Ic 2 mA.D. In practice it should not lie outside the shaded area shown. Thus we can say the characteristics provide one restriction 3 . Active device characteristics. (Figure 1.3. or h 3 mA. the operating point is completely defined by any two of the three. since no information is given in the particular set of characteristics. (b) Common emitter transistor D.4b).2 V (see Figure 1. (a) Triode valve. h 3 mA. VaK. Since these three variables are not independent.C. Permissihle range of d. Foratransistor it might be VcE• volts (a) (b) Figure 1. VaK. VcE 3 V etc.A 5 10 15 VcE· volts (a) (b) Figure 1.4. CHARACTERISTICS <l: 10 E ~5 OIJ. The operating point can lie anywhere within the range given by the characteristics.c. Operating Point The operating point is given by the values of the two variables and the parameter corresponding to a particular point on the characteristics.4a). or as VAK 150 V.
will be considered in detail before proceeding to basic transistor circuits. Under these conditions.C. (h) Load line resulting from the circuit shown in (a) VHT· The grid will be maintained at a potential negative with respect to cathode by a second battery of VB volts as shown in Figure 1. VAK or h is fixed then the other two can have only one possible pair of values.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS on the possible range of values of the operating point.5b having intercepts if h = 0 VAK = VHT if Thus in this circuit the operating point may lie only at any point on the straight line. 4 . These.C.c. therefore.5a. supply valtage JA VHT RL rVHT ~ 0 (a) VAK VHT (b) Figure 1. For example. Load Lines The basic problems concerned with graphical solutions may be most easily understood by a consideration of triode valve circuits. An equation can now be written relating VAK and h.hRL This is a straight line law as shown in Figure 1. if one of the three variables. circuit and Ioad for valve. This line is known as a Ioad line since the slope 1/RL is governed by the Ioad resistor RL. consider the case of the triode valve connected in series with a resistive Ioad RL and a d.5a we can say that the operating point is restricted by (a) the valve characteristics and (b) the Ioad line. VAK = VHT. In the circuit of Figure 1. D. This Ioad line is a function of the circuit only and if the valve were replaced by another device such as a lamp or transistor it would be unchanged.5. VGK. First. (a) D. Further restrictions will be provided by the electrical circuit into which the device is connected.
CHARACTERISTJCS if VaK is fixed at 0 V. '0 J . the operating point must lie on the zero volt characteristic.::. and the load line is plotted as shown. lt must also lie on the load line. volts Figure 1.6 connected in the circuit of Figure 1.C. 10 . (() <i J E IARK + 2 ~ 5 200 100 VAK. Consider the valve having the characteristics shown in Figure 1. Triode characteristics with Ioad and bias lines The intercepts of the load lines are found: VAK = VHT = 300 V h =0 If If VHT J A = .5a with VHT 300 V.D. Reading from the graph we find the 5 .::..= 15mA Rr.. to say 2 V. and the only possible operating point now lies at the intersection between the zero volt characteristic and the load line. and RL 20k0. and thus VaK.6. Voltage Amplification The operating point will now be fixed by setting the VB supply.
of the circuit is 30.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS operating point is VAK 156 V. Thus a change of + 1 V in Vax results in a change of (156.156 . The value of VAK changes to 289 V. If an alternating voltage generator of es volts is now connected in series with the VB battery. Since a valve is a nonlinear device. we cannot expect the voltage gain to be the same for all changes of Vax.rads 0 31fJ2 211V 2rr 1/) 1 0 > 4 (b) (a) Figure 1. To investigate this we shall now change Vax from 2 V to 8 V.(2) = 133 6 = 22·17 This result is of the same order as the previous one and is only reduced as the spacing between the characteristics becomes less for higher values of VAK· A. for this particular case we can say that the voltage amplification. the operating point moves to VAK 126 V. but with alternating voltages. the instantaneous value of Vgk will be given by Vgk = 2 + Es sin wt where Es is the peak value of the alternating voltage and w the angular frequency. AMPLIFICATION For many electronic applications we are not concerned with changes of direct voltage. and in general ßVout ßVAK 30 Av = ßVtn = ßVax = +1 = . h 8·7 mA. Thus Av = +289.30 Nonlinearity So.126) = 30 Von VAK· This effect is known as voltage amplification Av.7. It is assumed that the value of wt. h 7·2 mA. Input and output waveforms for simple triode amplifier 6 · .C. a change of + 1 V. Now suppose Vax is changed from 2 V to 1 V.8 . or the voltage gain.
c. The peaks and zeros of the waveform should be sufficient to indicate the degree of distortion obtained in this case. In practice the use of batteries is inconvenient and alternative methods are provided.6. The corresponding values of Vak can be obtained from the graph in Figure 1.8.h(RL 7 + RJ{) (1. For most amplifiers. Biasing In the circuit discussed above.c. AMPLIFICATION w is such that the effect of any stray capacitance or inductance may be neglected. 180° out of phase with Vgk and amplified by a factor of about 28. the cathode bias circuit shown in Figure 1.8 is used. This bias is essential since if it were not present. supply. In addition the given characteristics show no information for the positive grid region and the variation of operating point could not be predicted. This output waveform is not quite identical in form to the input waveform. Initially neglecting the a. The resultant Vak is nearly sinusoidal. This would cause grid current to flow with possible valve damage. This is the result of the nonlinearity of the valve characteristics and is therefore known as nonlinear distortion.hRK VHT .7b.1) .C. source es. the equation for the circuit is VAK = = VHT . Triode amplifier with cathode bias would cause the grid to become positive with respect to the cathode. The !arger the signal amplitude the greater the distortion and vice versa.hRL .7a. If Es is 2 V the waveform of Vgk is shown in Figure 1.by making VB= 4 V and Es= 4 sin wt. and are shown in Figure 1. the mean value of VGK or the grid bias was obtained by the use of a separate d.A. positive half cycles of the signal h Figure 1. The reader can investigate this for himself.
c. VaK = 0·25 V. To evaluate the amplification of this circuit we must find how the operating point moves when the a. h must lie on this line. The second equation is therefore (1. for 3 V h = 12 mA. The valve has characteristics shown in Figure 1. Example 1.2. operating point and voltage amplification.8 with values VHT 300 V. The second point cannot be plotted since we have no characteristic for VaK = 0·25 V.20h V. From equation 1.1. Since the operating point is not known the grid bias voltage hRK cannot be found directly. Determine the d. The intercepts are now VHT and VHT/(RL + RK). (Note the bias line is not quite straight and must in generat be plotted from more than two points.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS Also the voltage at the grid with respect to cathode. h = 8 mA. RK 250 n.1 relates VAK to h and can be plotted on the anode characteristics to give a load line as before. VaK = VEK = VKE where VKE is the voltage at the cathode with respect to earth. if h = 0. These four points are shown plotted with the characteristics and the resultant d. RL 19·75 kO. Any given value of h would result in a known grid bias.c. Such values of grid bias can be plotted against h on the anode characteristics for any given value of h. Ra has no effect on VaK since grid current is assumed to be zero.h(19·75 + 0·25) V = 300. bias line has been drawn. Taking equation 1. Thus Ra maintains the grid at d. bias line and the load line intersect is the only operating point satisfying both equations 1.2) Equation 1. This gives the same load line as before.6.1. and once again the operating point VAK. earth potential.c.1 and 1. h 7·4 mA. The point at which this d. If h = 1 mA. VaK = 0. But we have a characteristic for VaK = 1 V and the corresponding h may be found +1 VaK RK lA===4mA 0·25 Similarly for 2 V.c. source e8 is connected between 8 . A numerical example will illustrate this method.2. VAK = 300 . Consider the circuit in Figure 1.c.) The operating point of the valve in this circuit may now be read off the graph VAK152 V.
The input signal to the valve is Vgk and Vgk is the difference between es and ia. The voltage between grid and cathode is now given by Vgk = Ia.RK. and The extremes of VAE are 109 194 + 0·25 and Av X 5·3 83·9 = _4 = = 195·3 V 20·975 Thus the addition of RK to the circuit can provide the required grid bias but it also results in reduction of the overall amplification Va.k/Vgk = 85/2·925 = 29 which compares very closely with the result previously obtained (28).3) Vgk = hRK + 2 to Vgk = (1.Es 4 Strictly the output voltage should be given by the voltage at the anode with respect to earth. 2 V is used as in the previous example then es moves between + 2 V and 2 V.C. The resulting grid cathode voltage must vary from (1.4) hRK. The explanation in this case.. This is given by VAE = VAK + VKE = VAK + hRK + 0·25 X 9·6 = 111·4 V.6 and the intersections with the load line give the limits of the variation of the operating point.3 and 1. This is due to an effect known as negative feedback which will be discussed in a later chapter. bias line by +2 V and 2 V respectively.8). ßVAK 85 Voltage gam Av = . Taking peak to peak values Vgk = 4 .5·3) 2·925 V Valve amplification = Va. h = 9·6mA h = 5·3 mA . These are shown in Figure 1.2 Equations 1.= 21·25. however. VAK es = 2.1·075 = . AMPLIFICATION grid and earth (Figure 1.A.0·25(9·6 = 4.c. From the graph these are found to be = es= +2.4 represent two further bias lines which may be obtained by shifting the original d. is simple.Rk +Es sin wt If the same value of Es.= . 9 . VAK = 194 V. 109V. D..e/es.
but a simple example will illustrate the operation.1 above. or very 10 . so using a figure of 1 000 Hz. The resultant .VaK would be 4. voltage iaRK.uF 7T The operating voltage of the capacitor is only 2 V and miniature capacitors. Vak now becomes 3·8925 X va1ve amplification = 3·8925 = 115V X 29 This is in fact greater than the value obtained wit.=:::= 6 . load line. voltage can be eliminated without change in the d. Provided the signal frequency is not too low a capacitor may be selected having reactance very much less than RK. say 25 il. The exact analysis will be left for a later chapter. If the a. The cause of the reduction was the a.uF 20 = .h battery bias (113·5). but in this case we have neglected the effect of nonlinearity of characteristics.iaZk. and iaZk can be made negligibly small compared with es.c. Rk was 250 Q. are readily available should lower frequency operation be required. If Xe was.iaRK = es . Vgk becomes es . A.6.c.c. 6 V working at 100 .uF or more. the bias point VaK will be unchanged. 1 Xe= 27TfC' 1 C= 27TfXe 106 = 27710325 .GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS Bias Decoupling At this stage we shall assume that a reduction in amplification is a disadvantage and we shall see how the circuit gain can be made equal to the valve gain.c.0 = es. while the a. Finally the value of a suitable capacitor depends on the frequency of es. hRK. Considering Example 1.c. Load Lines Up to this point we have found that the operating point can move only up and down the d. Vgk will be es . then the iaZk voltage would be only a tenth of its previous value 1·075 V.C. If this decoupling capacitor CKis connected in parallel with RK.0·1075 = 3·8925. This is true for d.c.
T.c. The form of the graphical solution + 300 V + . AMPLIFICATION low frequency changes of Vgk.8 becomes Vgk =es. relationships are VAK = VHT .c. are related only by the Ioad resistance RL.7) (1. equation 1. the d. For the cathode bias circuit of Figure 1. battery : 200 V _. operating point must then move along an a. The slope of the a. H.c. Circuits having different a. value. Equation 1.c.9 the a. load lines are very different to the d.h(RL VGK = + RK) (1. Ievel with respect to VHT· The instantaneous amplitude of the a. load line of slope 1/RL (compared with slope 1/(RL + RK) for the d.c. quantities.c.9 gives the a. ia and Vak. Ioad was 20 kQ and the a. RK has been short circuited by CK.9.c.ia(RL Vgk = laRK + es + RK) (1. For the circuits shown in Figure 1. lf however the signal amplitude is reduced to zero the operating point must return to its d. (a) Decoupled cathode bias..c. Ioad was 19·75 kil.C.c. load lines.c. Ioads.7 must become Vak = (1. load line is not changed by the signal amplitude so it must pass through the d.5) (1. (b) RC coupling for next stage 11 . and equation 1.9) VHT. and d. operating point. relationships may also be written Vak = VHT. It would be difficult to differentiate between these two load lines on the graph so this effect is negligible. The a.. a.c.8.iaRL since to a.c.c.c.A.c.c._ (b) (a) Figure 1.c. In the last example the d.8) If CK is included and the frequency suchthat XcK is approximately zero. variations in VAK together with the d. load line).6) hRK In the absence of CK.c.
perhaps. operating point is unchanged. Ioad line.c. draw a d. resistance of the 1 Slope. Ioad line occurs when the Ioad is reactive.c. A. operating point will move up and down the a.10.c. The first circuit shows the same situation as in the last example but the component values are such that the slopes of the a.c.. To simplify this problem.c.. Ioad line parallel to this and passing through the operating point. Construct the a..9b an RC coupling network is used to apply the amplified output to. Ioad line for the a.!. we shall imagine a Ioad 12 .. The operating point moves up and down the a.c. in this case 40 V. In the circuit of Figure 1.9 HT battery is negligible. RL Figure 1. and d. Thus the reciprocal of the slope of the a. since the a. Assuming any convenient value for VHT.10. Load Lines with Reactive Loads A further complication of the a. The calculation of d.c. These are in parallel to a.c.C. Ioad line is RLRL' RL RL' + = 20 X 5 20 X 5 = _ 4 kQ The dotted line in Figure l.C.c.c. load line Slope. but the a.c.C.c.c. D. load lines are now very different.JOb shows a convenient method of constructing the a. both RL and RL' are connected between anode and earth.c. proceeding as follows. Once again the a. Ioad value of 4 kO. load line according to the instantaneous value of the applied Vgk.RL•RK A. load lines for the circuits shown in Figure 1. and thus to a. and the peak to peak value of the alternating output voltage can be read off the graph.c.c.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS is shown in Figure 1. a second valve.c. Ioad line according to the applied instantaneous value of Vgk. load is now the parallel combination of RL and RL'. and a.
anode voltage must be zero when the instantaneous ia is a maximum or minimum. The a.lla and b.c. (a) Triode valve with reactive load.A.c.c.c. (A capacitive Ioad would result in rotation in the opposite direction. We shall further assume that an alternating Vgk will cause an alternating ia.c. Thus for these values of ia.C. When however the a. operating point thus moves around the circle in the direction shown.c.c. Remernhering that in an inductor the current lags the voltage by 90°. In Figure l. The circuit and characteristics are shown in Figure l. but it can be seen that the only single continuous line joining the four points is an ellipse or circle depending on w. Points could be calculated for intermediate values.ll. This gives us the two further points (3) and (4). component of Vak must be maximum or minimum. In practice the Ioad would not be purely inductive and the procedure 13 . the corresponding Vak is VnT giving points (1) and (2) on the graph. component of ia passes through zero.) The peak values of Vgk required to produce this Ioad Iine may now be read from the graph and the voltage gain calculated. the a. operating point is found by the battery bias Vn. (b) Construction of the resulting ellipticalload Iine a. In this circuit the d. The a. The values of Vak at these points is given by VnT ± ia peak X wL where wL is the coil reactance at the signal frequency.Jlb the narrow ellipse represents a lower frequency. Ioad line is vertical (RL = 0) and the d. AMPLIFICATION consisting of a coil having inductance but negligible resistance. the instantaneous value of the (a) (b) wt Figure J. Vgk will cause ia to vary with time as shown.
operating point for a particular circuit. The common emitter transistor amplifier. The results obtained are as accurate as these methods permit and demonstrate clearly such effects as distortion due to nonlinearity. bias circ11it and load. The commonest form of transistor amplifier is very similar to the valve amplifiers discussed above.c. These methods also enable the d. or the circuit for a required operating point. (a) D.12. (b) The output and (c) the input characteristics 14 . The circuit is shown in Figure 1.12a and the forms of the output and input (a) A VcE' V (b) Figure 1.GRAPIDCAL ANALYSIS OF ELECTRONIC CIRCUITS for constructing the load line is so complex as to make it rarely useful. Summarizing: we have found that simple valve circuits not involving reactive components may readily be analysed using graphical methods. TRANSISTOR CIRCUITS Weshall now see how far these methods can be applied to simple transistor circuits. tobe found.C.
These pairs of values may then be plotted on the input characteristics.VBE + E sin wt with maximum and minimum values of.e. on both input and output characteristics. signal e8 now applied to the base as shown will result in an input voltage Vbe = . The input and output currents 1B and I c flow out of the base and collector respectively. But the parameter of the input characteristics depends upon V cE.V cc and V cc/RL. the operating point. The waveform is then further modified by the nonlinearity of the output characteristic.Vcc + IBRB This leads to a further straight line with intercepts at VBE = . A dynamic input characteristic may be plotted for a particular circuit by reading the V cE values corresponding to lB values along the Ioad line. If JB were known the operating point would now be defined. input and output voltages. The circuit in Figure 1. An a. i. The intersection of the line and this dynamic input characteristic then gives the values of VBE and lB which determine the operating point. Now looking at the input circuit we obtain VBE = .c.12a is said to be in the common emitter configuration since the input and output voltages VBE and V cE are measured with respect to the emitter. These in turn lead to the variation of V cE from the output characteristics.VnE ± E. The input characteristics show variations of lB with applied VBE using VcE as the parameter. in the conventional direction for pnp transistors. The corresponding range of IB can be determined from the input characteristics using the dynamic curve for the circuit. The output characteristics show variations of Ic with changes of VcE and with lB as the parameter. As before we may write an equation for the circuit VcE = .TRANSISTOR CIRCUITS characteristics shown in Figure 1.12a. With a transistor. the base current waveform is distorted by the nonlinear input characteristic. A numerical example of this type of calculation will now be given for the circuit in Figure 1.Vcc and lB = V cc/ RB. This procedure is somewhat involved and is only necessary when calculating maximum power output and distortion for power amplifiers. and input and output currents.Vcc + IcRL giving a Ioad line with intercepts at . four variables must be investigated. 15 . It is interesting to note that for a sinusoidal input voltage.12b and c respectively.
. Load line 8±~ HtH ++l lii · l± m= +t+++ ~ ~  0 1 2 3 r8 :::BO mA i i Load line 60mA 40mA 50 m 10 Vce• V Estimated characteristics 20 30 0 mA 20 mA 1111 1 11 111 111111 i 111 11 II tHW 1111111 I 111111111 1111 30 mA 9:5n.. 3oV.I~ V8 E .. " Cl I t' ' ... 30V. 111 "'"' ~~THHffiB crr.' ..13.rr. "' ' . ' f±. ~~ oa 9·5mA BmA 6mA 4mA 2mA OmA . ..0 10 20 30 40 50 11 HH~ 1k Q. 70 R 1m FITiffi+ffi11Ti ilTITTITT"I ...mV · ' rTT... 11111111111111111111111 1111111 ffrH70 rnÄ Figure 1. The characteristics for Example 1. rn 601 111 1111 I IIIIJJ 111 111 11 111 EI Dyn a m 1c m p~t charactenst1c ...2 <1:  · ~ ~ c . ......
A simple transistor amplifier is operated from a d. Comment on the effect of the characteristic nonlinearity in each case. operating point. If IBis 70 mA.c.[ß ~ 41 mA which can be plotted on the input curve for VcE = 10 V. Similarlywhen VcEis 10 V.13a and b.13a. Thus the input 'Ioad' line may be plottedas shown in Figure 1. so the slope of the line must be used. operating point. determine the d. source is 3 V peak with 100 n internal resistance.c.13b.20 V and 30 V are IB 17 mA and IB 0 mA respectively and these may be inserted in the same way.TRANSISTOR CIRCUITS Example 1. Using this we can now plot th!! dynamic input characteristic as follows. If the transistor characteristics are those shown in Figure 1. and (b) the a. Following the procedure outlined above we must first insert the Ioad line on the output characteristic. d/B = 1\~ = 1 mA.c. and the base bias resistor RB is 1 k!l. Find also the a. If dVBE is 1 volt. The Ioad resistor RL is 9·5 n. source is 40 mV peak with zero internal resistance.c.5 = 3·15 A This Ioad line is shownon Figure 1. the operating values of VcE and Ic may be found from Figure 1. supply of 30 V. output voltage if (a) the a.2. The resulting dynamic curve is shown in Figure 1.d/B = 29 mA. The intersection between this line and the dynamic input characteristic gives the d. The intercepts for the input 'Ioad' line are =0 IB = 0 VBE IB = VBE Vcc RB = = 30V 1 k!l Vcc = 30 mA = 30V The second point cannot be plotted. VcE ~ 0·5 and this point may be placed on the inputcurvefor VcE = 0·5. IB = 30 mA . Therefore at VBE 1 volt. From Figure 1.c.13b VcE = 14·5 V 17 /c = 1·65 A . Two further points for VcE.13a VBE = 560 mV IB = 29·5 mA ~ 30 mA Now since /Bis known.c. The two intercepts are given by: lf If /c =0 VcE VcE = 0 Ic = = = Vcc Vcc RL 30V 30 = 9.13a.
c.13b. a measure of the distortion is given by the ratio of these two results .c. The resulting /b values obtained from the dynamic input characteristic are Vbe 520 mV 30mA 20mA 560 mV 600 mV 560 mV 70mA 30mA The peak values of the /b waveform are 30 . base current is given by 18 . Thus the peak a.c. the d.c. VBE varies between 520 mV and 600 mV in a sinusoidal manner. Vbe Vce = 520 mV 560 mV 600 mV 560 mV 18·5 V 14·5 V 0·5 V 14·5 V The peak values of the Vce waveform are 18·5 . 10r = 40 10 = 4 The corresponding limits of the variation of Vce can now be found from Figure 1. output voltage required is 18 V peak to peak.14·5 = 4 V and 14·5 .30 = 40 mA This represents a considerable degree of distortion resulting from the curvature of the input characteristic. source of 40 mV peak and zero internal resistance is connected. :. and the a. The a.c. D1 d"Jstortwn c h. Rtn is given by Rtn = VBE IB = 560mV 30 mA = 18 n This is very much less than the source resistance and may therefore be neg1ected. Rtn· At the operating point.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS When the a. In the second case the source has an internal resistance of 100 n.0·5 = 14 V The corresponding distortion factor D2 14 = 4 = 3·5 This is less than that obtained for the /b waveform since the output characteristic nonlinearity acts in the opposite sense to that of the input characteristic. base current will be determined by this and the transistor input resistance. For comparison purposes.20 = 10 mA and 70 .
the a.VBE 145 kQ = 5·8 x 103 A 145 p. therefore lB === Vcc/RB.14b determine (a) the d.c. (and a. and the slope of the input characteristic at this point is ß. This result is less than before and in this case the negative half cycle is the larger. From this result it seems likely that a suitable value of source resistance.c. For small signa1 amplification. say 20 Q.c.TRANSISTOR CIRCUITS es/100 = 30 mA. could result in the input and output nonlinear effects cancelling each other and leaving a sinusoidal output.c. VBE is very much less than the d. (c) the current gain. The corresponding distortion factor Da= 14·8/11·2 = 1·33. Vbe = ibRtn These approximations enable the selection of a suitable operating point and an estimate of the required input voltage to be made without difficulty. input resistance of the transistor Rtn is taken as the slope of the input characteristic at the operating point.14·5 = 14·8 V and 14·5.c.m. If it is connected in the circuit shown in Figure 1. and (d) the maximum output signal for negligible distortion. calculations are usually simplified by making certain practically valid assumptions.14a operates with VBE at 0·2 volts.s. Firstly. (b) the output voltage and voltage gain for an input signal of 28·28 mV r. operating point.c.Vcc = 6 V and Ic = RL = 1 = 6 mA (a) The d.VBE 10 mV.3. First the d. The a.) Ioad Iine can be constructed with intercepts at Vcc 6 VcE = .3 V 14·5 V The peak values ofthe Vce waveform are now 29·3. supply voltage Vcc. Example 1. MB 5 pA.c. The resulting variation in lb values and the corresponding Vce values are Ib Vce OmA 30mA 29·3 V 14·5 V 60mA 30mA 3.3·3 = 11·2 V. output voltage in case (b) is 26 V peak to peak. 19 = 40 fl A . operating point is now determined by lB which may be found from Vcc . A transistor having the output characteristics shown in Figure 1. Secondly. the phase of the distortion has therefore been reversed..
a.!Ä 2 3 4 5 6 VcE• V (a) (b) Figure 1. mput reststance = 5 ttA = 2 u An r.1·5)/2y2 = 0·885 mA.m. The current gain A1 is therefore 0·885 mA/14·14 ttA = 62·5. lOmV kr.s. base current ib = 28·28 mV 2kQ = 14·14 uA r = y2 X 14·14 ttA = 20ttA and the peak to peak value = 20 x 2 = 40 ttA.s.s. ic of (2 .m.3 From the characteristics the operating point is now VcE 2·9 V. the voltage gain Av = 850 mV/28·28 mV = 30. the a . input resistance and is given by ~VBE ~h . 20 . The r.s. (c) The value of ic varies from 1·5 to 4 mA giving an r. Ic 3 mA.m.JÄ ÜJ. (b) The slope of the input characteristic is the a.c.2·0)/2y2 = 0·85 V. As VBE goes positive lB is reduced and VcE goes negative. Thus as with simple valve amplification we get phase reversal and Av = 30. The peak value of ib Thus from the characteristics.c.m.c. .GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS 6V <! E 1· 5 tlc~~. The characteristics and circuit for Example 1. input voltage of 28·28 mV will result in an r. output voltage = (4·2.20 j. The corresponding values of Vce are 2 V and 4·2 V respectively. operating point moves between ib = 60 ttA and ib = 20 ttA.14.
IE flowing out at the collector. Vce of 2·9 .s.)IE flowing out of the base. These relationships can be simply shown and remernbered by the diagram shown in Figure 1.10) IE = Ic +In (1.TRANSISTOR CIRCUITS (d) The peak output voltage must not cause VcE to exceed 6 or the transistorwill cut off.)IE .· Figure 1.a.m. Figure 1. Fundamental current relationships for the transistor (1a)0 biased collector base junction will have a reverse leakage current flowing conventionally from base to collector._. In addition the reverse Figure 1.a. A peak a.12) h = (1 . . This current is known as /co or lcBo since it is the value of Ic when JE is zero. The maximum output signal is therefore 2·7/v2 = 1·9 Vr.Ico 21 .11) lc = a.16. VcE cannot fall below 0·2 V from the characteristics.0·2 will not cause distortion due to bottoming or cut off. Waveforms for Example 1.16.c.15 shows the distortion that would arise with this circuit if the peak ib was increased to 60 pA. transistor action results in a.3 \ .15. base and emitter currents in a transistor. Before we can discuss these we must examine the relationships between the collector.IE + lco (1. In practice the simple biasing arrangement used in the last two examples has a number of disadvantages. If a current IE flows into the emitter lead. By inspection of this diagram the three basic relationships may be found: (1. and (I .
we are not concerned with JE.rx Now applying equation 1.rx)IE = Ico IE = Ico 1 .(1. is known as Ico' (or sometimes lcEo as opposed to lcBo for the common base circuit). Ic = = rxlco + Ico rxlco + Ico  1 _ rx 1oc rxlco Ico =1a: This current.1 0. From equation 1.11. The values of Ico vary from 1 mA for a high power germanium transistor to 1 flA for a small signal germanium transistor or to 10 nA for small signal silicon transistor. the va1ue of Ic when IB is zero.rx Now since rx is nearly equal to unity having typical values from 0·96 to 0·995. Firstlet us find the value of Ic when lB is zero. Ic. From equation 1. To find the value of /c when IB is not zero we must rearrange equation 1. Thus . Equate to 1.11. Typically the corresponding value of Ico' would range from 0·1 A to 1 flA respectively.14) 1rx (1. Ico Ico = .16) .GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS In common emitter circuits.13) 1 . (I .11 to obtain /E and equate to equation 1.rx) = rxh and rx Ico Ico + Ico rx Ico lc=h+1rx 22 (1. but we do wish to know how Ic varies with IB.Ico IE=(:1.12. Ico' is very much larger than /co.10 /c rx/c + IB = Ic  + rxh = /c  Ic(l .
Example 1. lco 1 ttA and VBE 0·2 V.17.0·2)V 970 Kil ' = 1 1 _ oc ttA = lco = 10·1 ttA 1 oc 1 . now suppose such an OC 75 was connected into the circuit shown in Figure 1.17.4.I+ oc' 1 = ttA = 91 ttA 190 91 23 ttA .17) = oc'/B + lco' (1.oc (1. 10V 5kfi 970k!l Figure 1.16 becomes lc oc =1 .19) lco' = 1 + oc' (1.17. Simplebias circuit for Example 1. Typical values for an OC 75 transistor are a' 90.TRANSISTOR CIRCUITS The second term in the result is lco' and for convenience we shall put and equation 1.13 and 1. These lead to: oc' IX= 1 lco + oc' (1.20) We are now in a position to discuss the disadvantages of our simple transistor biasing arrangement.4 IB = (10.18) oc' Other useful relationships can be obtained by rearranging equations 1. A numerical example will adequately illustrate these points.
1 000 .18. Taking these extreme values we get Ico' Ic VcE 70 71. Unfortunately all OC 75 transistorsarenot typical and the specified range of r~.. Summarizing.c. The effect of temperature on the output characteristics 24 . the output VcE would have been severely distorted by cut off in the first case and bottoming in the second. with a peak output voltage of 4·8 V. operating point without distortion. r~..uA Now VcE = 10 + (5 X 1) = 5 V and since VcE may vary between 10 (cut off) and 0·2 (bottoming) this allows maximum variation of a. The reverse biased leakage currents Ico (collector base) and lEo (emitter base) are the result of the intrinsic properties of the semi'@ /.. Alternatively if the input signallevel had been maintained.'@ 30°C /.'h + Ico' = 90 X 10·1 + 91 = = 1mA.' unless the signal level is severely restricted.' 130.QOC _ _ lO~C _ leb 2ooc Vcc VcE Figure 1.GRAPillCAL ANALYSIS OF ELECTRONIC CIRCUITS Ic = a.' Temperature Effects A second disadvantage isthat transistors are temperature sensitive. the simple bias circuit cannot be designed to suit all transistors with the normal commercial range of r~.uA 1·444mA 2·78V Thus to prevent distortion the output signal must be reduced to 3·9 V peak with r~.' 70 or 2·08 V peak with r~.QOC .uA 0·778 mA 6·1 V 130 131 .' is from 70 to 130..
As can be seen from Figure 1. The effect on the input characteristics is shown in Figure 1. and T is the absolute temperature in °Kelvin. The effect on the output characteristics would be very similar tothat shown in Figure 1. Again the operating point would move towards bottoming. In was maintained at a constant value.Ico IB = (1 . if temperature rises by 10°C. at 20°C the permissible peak a. In this case. An alternative approach might be to maintain VBE constant. When VBE is very small Ico will have a significant effect but when VBE is larger we can neglect Ico as Figure 1.Ico (1. IB is constant and Ico' doubles as /co doubles.c. Thus for any particular value of VBE. Both currents approximately double for each 10°C increase in temperature. Vce is approximately Vcc/3 while at 40°C it has been reduced to zero. compared with the exponential term. In the simple circuit discussed above.22) In this expression a change of 10°C will have 1ittle effect on Twhich will be of the order of 300°K.18. 2 25 . the characteristic corresponding to IB = 40 ttA moves up with Ico'. The operating point must lie on the load line and it therefore moves towards the bottomed or saturated condition with increase in temperature. Now IB = IE(l . In Example 1.TRANSISTOR CIRCUITS conductor material.oc)/Eo[exp (KVBE/T) .19. lc = oc'h + Ico'.21) where K is a constant.4.1] .oc) .18.1] (1. lEo and thus In doubles.19. lt can be shown that the emitter current can be expressed in terms of the emitter base voltage by IE = /Eo[exp (KVBE/T) . The effect of temperature on the input characteristics we can one.
IERE oc)/E . lcoRC20°C) Figure 1. d V+ lcoR' w1t mtercepts at V + IcoR. The 1oad lines are drawn on this for the same temperatures. The base is supplied from a source of medium resistance instead of one having high resistance (tending to constant current) or zero resistance giving constant voltage.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS The commonest form of bias circuit emp1oys a compromise aimed at stabi1izing the emitter current and therefore the collector current regard1ess of changes of transistor (hence oc' and required VBE) or temperature.24) This equation represents a '1oad 1ine' on the VEB/IE characteristics.lco V' ..IE[RE + (1  oc)R'] + lcoR' (1. Graphical analysis of stabilization 26 .23) hR' . Figure 1. an at RE+ R'(l _ rxf Figure 1. . Stabilization of operating point The simp1est form of this is shown in Figure 1.21 shows the VEB/IE characteristics for 20°C and 30°C..20. The V". .20.h.. .21. The equation for this 1oop is given by = V' IB = (1  VBE but VBE = (1.
01: + RE) . Thus the conditions for a stable emitter current are. V' much greater than VEB and R' small.18 V' .25) V'= VccR2 and that R1 (1. Potential divider. emitter resistor stabilization In practice V' is obtained from a potential divider circuit across the main Vcc supply. 01: 27 + RE) lco' ( R . This may then be rewritten V' .26) + R2 The operating point VcE.22. Also if the line could be made nearly horizontal the change in IE could be reduced.TRANSISTOR CIRCUITS resulting change of /E is shown by the intersections marked.22. .VBE = IcRE +~ (Ic 01: Ico')(R' lc [R . + RE( 1 + 01: ')] = .25 and 1. as shown in Figure 1. RE large. ( I'+18 ) Ja I' R2 Figure I.VBE = IBR' + (In + Ic)RE then collecting terms and substituting for IB from equation 1. thus if R' is made small this movement will be limited.23. Ic can be found by calculating V' and R' from equations 1. The movement of the Ioad line is due entirely to IcoR'. By Thevenin's theorem (see Chapter 2) it can be shown that R' = R1R2 R1 + R2 (1.26 and substituting in equation 1. Vcc R.
27 giving R' +RE K = R' + RE(1 +IX') by assuming VBE ~V' and that V'.29) K = dleo' die dleo S= Krx' die = diX' (1. . For good stability K is small and in the limit when R' K = I 1 + IX 1 F or the warst case R' ?> RE and K = 1.VBE = constant.28) An alternative approach to the stabilizing action of the circuit can be seen by consideration of the circuit (Figure 1.30) (1.32) ~RE. Any change in /e due to temperature or IX' changes.GRAPHJCAL ANALYSIS OF ELECTRONJC CIRCUITS From which Ie = IX'(V' . For R' to be small. IERE. Suppose /e tends to rise. the voltage across RE.31) K may be found directly from equation 1.27) VeE may then be found from VeE =  Vee ~. also cause a change in JE.VBE) Ieo'(R' +RE) R' + RE(l +IX')+ R' + RE(1 +IX') (1.Vee + IeRL + lERE + Ie(RL +RE) (1. These are die (1. will also rise. Stability Factors Three measures of circuit stability are often encountered. reducing VBE· This in turn reduces /B tending to maintain Ie and IE at their original values. 28 (1.22). R1 and R2 must be small making I' much greater than /B. Thus the valtage between base and earth is approximately V' and nearly constant.
33) l +MtX' RE M= RE+R' where (1. S tend to 1 ~ die = IX dleo 1 _ IX = dleo . Design Considerations The limitations on RE being large is that the volt drop IERE must be supplied by the supply battery Vee (equation 1. and typical values of R1 and R2 will be used in the following examples. This will be discussed in later chapters. RL +RE is given by 13mVA = 3 kQ. Finally to find Krx· we must take equation 1. Thus for a 6 V battery and operating point VeE 3 V. If the Ioad is 2 kQ. S tends to one and d/e = dleo.28). On rearranging and differentiating S can be found R' +RE (1·35) s = RE+ R '(1.TRANSISTOR CIRCUITS K is also sometimes written (1.IX) In this case if RE~ R'.IX'M (1 + M1X') 2 29 [V' .27 and substitute from equations 1.24 and substituting for IE in terms of Ie and Ieo. Je 1 mA.VBE) IX'leo lc = (R' + RE)(1 + MIX') + 1 + MIX' (1. If R' ~ RE.3 6) Note Ieo' ~ tX'/co le Krx' die = 1 + MtX' = diX' = IX' (1 [V'R' + VBE J RE + lco + MIX') .33 and 1.VBE R' +RE J + leo .34 IX'(V'. RE can only be I kQ. The minimum size of R' is determined by the minimum permissible input impedance to the amplifier. For higher current transistors the values will of course be much lower.34) S may be found in a similar way from equation 1.
VBE) + RE)(1 + Ma. Collector feedback.'lco + 1 + Ma. any increase in Ic increases the voltage across RL. reducing h and thus /c.' (R' a.') a. Ka.' J From equations 1.23.32. This is shown in Figure 1. For large changes. emitter resistor stabilization The equation may be written in the same way Vcc. Figure 1.'.36 Klc Ka.VBE = = (/c + IB)RL + IBRB + IERE Ic(RL + RE) + h(RL + RE + RB) Now by making the same Substitutions as before.'(V'. 30 . 1.' 1 = 1 + Ma.38) In this circuit.35) may be used therefore for large changes in /c as weil as small ones. This reduces the voltage across RB.· however is the slope of a curved relationship and equation 1.36 whichever is most convenient. values should be inserted in equations 1. The formulae (equations 1.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS This may be rewritten Ka.37) K and S are the slopes of approximately straight line relationships.23.' 1 [ X a.33.33 and 1.·=. IX (1.37 should only be used for small changes in a. and differentiating we find (1. 1.27 or 1. Another stabilizing circuit is occasionally encountered where the base resistor is connected to the collector.
s.c. The reverse collector leakage current Ico is quoted as 2 pA at 20°C and the amplifier may be used in the temperature range 20oC to 50°C. such a. Example 1. operating points.c.23) the stabilization is performed by d. circuits will be discussed in later chapters.24.TRANSISTOR CIRCUITS Bias Decoupling In both stabilizing circuits (Figures 1. currents could be fed back through RB. Weshall now consider some numerical examples on the operating conditions for a given circuit and on the design of a suitable circuit to meet required operating conditions. The exact mechanism of these a.C. but in addition a.c. voltage across RE.24b the same applies. Figure 1.25 employs a transistor having rx' in the range 5095. Determine the range of possible d. The single stage amplifier shown in Figure 1. (a) (b) Figure 1. feedback will reduce the gain of the stage. First we must find the range of d. As with the triode valve amplifier.c. operating point at 20°C due to the spread of rx'.c. So if XcB ~ RB12.5. 31 . decoupling circuits In Figure 1. In Figure 1. output current. and hence the maximum permissible r.24 shows how this is done.22 and 1.m.c. Assurne VBE is 200mV.c. A. This behaves in the same way as the RK in the valve amplifier. negative feedback. Decoupling must be used to overcome this effect.24a if XcE ~RE there can be no a. currents can flow to earth.
25.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS 10V 2kfi 1kfi Vcc• volts (a) (b) Figure 1. Circuit and graph for Example 1.VBE) + Ico'(R' R ' + RE(l + Cl. 32 .') + RE) 10 + 10 = 2·323 V I X 7·67 kO Using equation 1..27 lc = _50 lc  Ic = X 2·125 + 0·102 X 8·67 A m 7·67 + 51 1·825 mA _ 95 X 2·125 + 0·192 X 8·67 A m 7·67 + 96 Ic Ic = 1·95 mA Increase in ambient temperature can only increase Ic...' 50. Ico' = = (1 + CJ.')Ico 0·102 mA = 0·002(51) mA Ico' = 0·002(96) = 0·192 mA Next. Cl.5 For T1. so maximum I c will occur if T2 is used at 50°C. to find Je we need V' and R' for this circuit. V I = 10 33 X 33 10 R = lO + 33 = CJ.'(V' .
5·525 = 4·475V.c. We must now consider the effect of the decoupling capacitor.0·2 = 3·6 V. Maximum VcE = 10 + 1·825(1 = 5·525 V For Tz at 50°C.29 and 1. Vce has a maximum value of 3·6 V. 1oad is only 2 kil. Thus to allow for all possibilities. Vce must not exceed 3·811 .28.c.c. lco = 0·002 X 2 X 2 X 2 mA and = 0·016mA ß/co = 0·014 mA ßlco' = (1 + oc')!:l. cut off will occur if peak a./co = 96 X 0·014 mA = 1·34mA From equations 1.c. 1·34 mA = 1·95 + 0·113 mA = 2·063 mA For T1 at 20°C. peak a. With T1. If Xe~ RE at signa1 frequencies. the peak a. + 2) V Minimum VcE = 10 + 2·063( 1 + 2) V = 3·811 V To avoid distortion due to bottoming VcE must not become less than VBE· To avoid distortion due to cut off VcE cannot be greater than Vcc.TRANSISTOR CIRCUITS For Tz at 50°C. the a.32 K = 6./c ßlco' R' +RE = R' + RE(1 + oc') 8·67 = 7·67 + 96 = 0·084 ß/ce = K!:l.Ico' = 0·084 x = 0·113mA K Maximum /c From equation 1. To avoid bottoming with Tz. We can 33 . Vce exceeds 10.
has Ico 2 pA and oc0·98. RL 4 '.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS therefore draw an a. Ioad line (lk!l. The peak a.) \ \ <! E .S.3 V.c.u \ \ \ ~ 2 0·3 1·0 3·0 Vcc• volts (a ) (b) Figure 1. Circuit and graph for Example 1.6 34 6 .26.c.M. the maximum r. signal current of 2 mA is to produce an R. If the minimum VcE to avoid bottoming distortion is 300 mV. The available d. output voltage of 1·414 V.c. = RL\1'2 = 2 y'2 mA = 1·27 mA Example 1. design a suitable bias circuit. Ioad line will occur at a lower value of VcE· To find the value that would cause cutoff distortion we can say ßVcE = 2 kQ ßlc and ßVcE = 2kQ = X 2·06 mA 4·12 V Our peak a. supply is 6 V. output current is given by Vpeak 3·6 lr.s. A.m.. A transistor whose operating point at normal room temperature is tobe VcE .s.m. Cutoff on the a. Ic 2 mA.load line of 2 kQ passing through the operating point as shown in Figure 1.C..25b. Vce is only 3·6 V so there is still no risk of cutoff distortion..c.c. and the transistor is subjected to a possible temperature rise of 40°C.6. Assurne that Ico doubles for each 10°C rise in temperature and that any emitter resistor will be suitably decoupled at the signal frequency of I kHz. Finally.
VcE + peak a. To determine K. and the 300 mV bottoming line are also shown. so the minimum instantaneous Vce is 3 + 2 = 1 V.c.34 K I = l + MIX' where M = RE 35 RE + R' IX and I X ' . The a. RL is determined by the output conditions from Vac iac = RL 1·414 =2 = X 0·707 l kO. ßVcE ßlc = RL +RE = 0·7 1. we must find ß/c. 3 + 2 + 0·3 = 0·7 V.c.c. Rt.26a. the operating VcE is 3 V and the peak signal Vce is 2 V.c.c.TRANSISTOR CIRCUITS We shall use the common form of circuit shown in Figure 1.Ico = 30 f1A .26b.\ u/co K = I 30 == 1IX ß/c ßlco' = A 1500f1 466 1 500 = 0 "311 Using equations 1.98 = 2 mA But IE = Therefore RE = 2 mA = 500 n IV Now since IE = Ic we can draw the d.5 mA (Note this is a d. = 3 + lcRL + IERE V IERE = 3 .2 = 1 V 6 So Ic . the a. Rz and CE..load line. RE. First we find the permissible ßVcE: at normal room temperature. signal voltage.Ico 1·998 IX = 0 .c.1IX . Ioad line as shown in Figure 1. Vce + Vbottoming. change with temperature) = 466 11A ßlco = 2 4 /co . RE may now be found since VcE is given. Now VcE must not fall below 0·3 V (bottoming) so permissible shift of VcE is given by ßVcE = d.33 and 1. The problern is to determine the values of RL.
hR' .' and ) 2·22 1 = 49 = 0·0453 =RE(~ R' = 49 I) = 2I·IRE = 10·5 kQ To find the values of R1 and R2.23 VBE = Now IB = Taking iE = Ic V' . = 39 ttA V'= VBE 0·5 X 2 10·5 X 0·039 = 0· 3 + 1 + 0·41 = l· 71 V + R2 R1 1·7I So and 6R2 V'= But = R1 6R2 + R2 6 R1+R2= 1.IERE ' 2. ttF 36 .GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS Wehave a.' = 0·98 1 . we use the Thevenin equivalent circuit and equation I. for adequate decoupling X cE ~RE at 1 kHz.7 IR2 R1R2 R1 10' 5 = 6R2/I·11 = 3·5 and R1 = 37 kQ and R2 = 6 R1 = 14·7 kQ 1·71 .0·98 1 (1 K M = a..O·I 49 + + Ic.Ico' (. I XcE Let Thus XcE = = 27Tl03C so IO Q C 1 C = 27Tl03XcE 50 IQ6 = 27T 104 F = :..1 Finally.
or (b) an H. How is the gain for case (b) modified if the bias VGK is changed to 0·5 V? Ans. operating conditions may be found. Using the circuits described in Example 1. We have investigated suitable circuits to give the correct d. in this chapter we have seen how simple valve and transistor circuits may be analysed by graphical methods. supply of 300 V and Ioad of 60 kil.c. Find the d. (a) 25·5. Example 1. and in the case oftransistors we have seen how this operating point may be stabilized against changes of temperature and transistor. signal 0·5 V peak in each case.c. EXAMPLES Example 1. if with (a) VGK is 0·5 V and with (b) h is 1·5 mA. supply of 100 V and a resistive load of 8 kil. II I 7S 100 I 12S ISO 117S 200 122S 2SO 127S 300 0 2S SO 0 0 for Vax (volts) O·S 3·3 1·0 6·6 3·3 0·2S 0 1·3 0·5 3·3 1·7 6·1 3·S 9·S 13·S 6·S 9·7S 0 0·7 0·2 2·0 1·0 41 2·S 6·3 4·4 10·0 6·9 10·2 0 0·4 O·IS 1·4 0·6 2·8 1·5 4·85 7·4 2·9 5·0 0 0·2 0·1 0·75 0·4 1·6 0·9 3·2 2·0 5·3 3·S 8·0 S·75 8·S 0·15 O·S 0·2 1·2S 0·7 2·3 l·S 4 ·0 2·8 6·1 4·5 6·S 0 0·3 0·1 0·7S 1·7 0·4 1·0 3·0 2·0 4 ·6 33 0·1 1·2 2·3 VAK (volts) fA(mA) 1·0 J·S 2·0 2·S 3·0 3·5 4·0 4 ·S 10·3 14·0 6·1 9·S J3·S 17·S S·O 5·S 17·S J3·S 6·0 6·5 7·0 13·6 10·4 7·6 10·6 0·4 Ans.7. these will make the solution exceedingly difficult if not impossible.8. 37 . but as with valves.7. (b) 208 V. determine the voltage gain. We have found the limitations of these methods.EXAMPLES In practice a 100 p. Summarizing. conditions.T. Assurne an a. operating point in each case.c.c. A triode valve having the characteristics given below is connected in series with (a) an H. 5 V. Wehave not considered the effect of coupling circuits or reactive Ioads on the graphical solution of transistor amplifiers. and we have seen how the d. 4·7 mA.F capacitor would be used allowing the amplifier to be used at lower frequencies. (b) 24 or 45·5.T. (a) 63 V.
30·5. Find the output voltage and voltage gain if (a) es = 1 sin wt and (b) e8 = 1·5 sin wt. (a) 43. Ans. 32. 0·65. of 250 V. 21·5. The circuit described in Example 1. (b) 10 ill. Circuit for Example 1. A single stage valve amp1ifier employs a resistive load of 4 kU. The characteristics are suchthat the 200 V H. h 6 mA. 250V H. Ans.9.27.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS Example 1. A triode valve having the characteristics given for Example 1. 166 Q. RK 250 Q and an H. (b) VAK 150 V.27. h 9·75 mA. Find the new value of vo and Av if a capacitor CK is connected in parallel with RK such that its reactance at the signal frequency is negligible.T.13.27 is now connected in the circuit shown in Figure 1. A triode valve using the characteristics given for Example 1. (a) 24·8 kU.10.11 Figure 1. 154 Q. Circuit for Example 1. Example 1. (a) 61. Determine the voltage gain.T.7 is connected in series with an anode Ioad resistor RL and a cathode resistor RK and an H. Ans. (b) 96.28.7 is connected in the cathode follower circuit shown in Figure 1. 790 Q.28. Find the values of RL and RK ifthe required operating point is (a) VAK 100 V. + H. 21·5.c. grid bias is provided by an RK of 1k0. Example 1. Ans.T. The d. supply of300 V. If the triode valve shown in Figure 1.T + 175V I· Figure 1.T. results in an operating 38 . 19·2. (b) 64·5.11. Example 1.9 employs RL 17·15 kU.12 Example 1.12. calculate the voltage gain and the terminal input impedance.
15. A single stage amplifier is operated with battery bias and an H.EXAMPLES anode current of 20 mA. A triode having the characteristics given for Example 1. Ans. (b) the maximum peak a.c. Determine the operating point on the input and output characteristics taking RL and RB as 500 n and 47 kO respective1y and Vcc as 6 V. Ans. Assuming that anode current excursions to less than 5 mA d.c. anode voltage when RK is adequately decoupled. Either positive VaK or h less than 0·4 mA results in excessive distortion.16.T. T. Example 1. (a) 15 mA. Ans. (b) 11 V low h. Hence estimate the voltage gain and phase shift. (a) with the circuit as described and (b) if an externalload of 5·7 kO is coupled to the anode through a capacitor of negligible reactance. (b) 60 V.12a. calculate (a) the maximum peak a. The anode characteristic for VaK 0 V is approximately linear having an ra of 15 kO. is assumed tobe purely inductive.14. I0 I VBE(mV) IB (ttA) for VcE (volts) 0 1·5 0 I 150 I 00 22 3 90 18 200 345 60 135 235 360 1 14 10 50 117 42 105 213 332 198 313 39 190 300 3·0 4·5 6·0 8 39 200 I 250 50 98 300 . The anode Ioad. If the frequency is 5kHz. (a) 45 V positive VaK.c. draw the operating Ioad line for a peak alternating anode current of7·75 mA. 30 L 315°. VcE 2·7 V. a coil of 0·2 H.7 is operated with a bias voltage of 1· 5 V and an H. supply of 150 V. Example 1. Calcu1ate new values for RL and RB to change the operating point to Ic 8 mA. Ievel will result in excessive distortion. 60 V. Find the maximum peak alternating anode voltage and state the limiting factor. A transistor having the characteristics given be1ow is connected in the circuit shown in Figure 1. Example 1. anode current and hence the peak anode voltage. of 200 V resulting an anode current of 2·5 mA through the 40 kO Ioad.
lB and JE. 204 mV. RE 500 n. andRE 1 kil. RL is 3·3 kil. RB 140 kil. Jco.' 160. Jco'. Je. a 4 kil load. 1 V. 30 ttA. and a 40 . Example 1.20. Ans. A transistor has a stable bias condition fixed by the collector feedback emitter resistor circuit shown in Figure 1. Example 1.c. 0·97 and Jco 4 ttA. If VBE is taken as 0·3 V. and the current gain.21. Ans. An alternating signal of 50 mV peak is applied to the base ofthe simple amplifier circuit in Example 1. Ans. calculate a. Example 1. 1·24 ttA. 35. If the transistor has a. and JB.GRAPHICAL ANALYSIS OF ELECTRONIC CIRCUITS .' 45. 0·99 and Jco 3 ttA is connected in the potential divider emitter resistor bias circuit shown in Figure 1. the voltage gain..'.16. Calculate a. 133 ttA. calculate the operating point and the stability factor K. and Jco' 0·2 mA. If JE is 2 mA.18. 36. Ans. R2 22 kil. RL 1·8 kil. 488 Q. 5·03 mA. A transistor operating at 5 mA is known to have a.22. 123 ttA. and the supp1y battery 10 V. The d. supply is 12 V and the circuit components are R1 33 kil. Example 1. Example 1.23.17. 56 ttA. Example 1. 0·46. Calculate the input impedance. 413 Q. A transistor having a. The design requirement forasinglestage transistor amplifier include a collector current of 1 mA. A certain transistor is found to have a. 1·944 mA. 3·93 mA. 5·1 mA. calculate the operating point and the stability factor K. Ans. 32·3. 4·5 V. 3·43 V. 0·994.VcE (Volts) Je (mA) for In (JtA) 1 o 1 0·2 0 0·4 11·0 6·0 0·2 1·2 0·2 1·4 0·2 1·5 0·2 1·9 80 120 2·4 3·6 28 29 4·3 3·9 5·7 160 200 4·8 6·0 5·6 6·9 5·8 7·3 7·6 9·5 240 74 8·4 8·8 11·5 0 40 4·1 Ans. 1·45 mA. Jco' 0·1 mA and VBE 0·2 V. 0·125.19. 28·8 kil.22.
The available power supplyis 12 Vand the silicon transistor has rx' 120 and negligible Ico. 70 Q. the shunt primary reactance is very much greater than the reflected Ioad impedance.23. 41 . supply is 12 V and the selected operating point is VcE 8 V. Rz12 kQ and Vcc is 6 V. Example 1. At the signal frequency. A transistor amplifier uses potential divider emitter resistor bias with adequate decoupling.m. an Vcc 12V Figure 1. 1 V r. determine: (a) the remaining bias components if the maximum temperature is 50°C (b) Ifthe effect of distortion is ignored. 1·11 V. RE 1 kQ. R140 kQ.s. The transformer has 2:1 turns ratio and 0·5 Q primary resistance. Ans. 2 kQ. 45·6 kQ. The power transistor shown in Figure 1. and further. The available d. rx' 150. 195 Q. 1 V.29. 10 ftF. Circuit for Example 1. calculate the values ofthe remaining components if a potential divider emitter resistor circuit is to be used. the maximum output power. Assuming VBE tobe 0·7 V. 3·6 W. and at 20°C. Ans. Ifthe transistor has negligible Ico and VBE of 0·2 V.c. calculate the maximum peak alternating output voltage (a) if rx' is 50. It is required to amplify signals widely differing in amplitude in the frequency range100Hz to 10kHz. (c) the approximate input voltage to obtain this output. Assuming the transistor to have VBE 0·2 V. 139 kQ.24. that it will bottarn at VcE 0·4 V. and (b) if rx' is 150.29 has a maximum collector dissipation 16·4 W. 1! Q. The components are RL 1 kQ.24 Ico of 500 flA. Ic 2 A.EXAMPLES stability K of 0·05. Example 1. Ans.
Section 3 will state some additional theorems which frequently simplify analysis and Section 4 will explain the analysis of a common form of network. admittances and generators. the four terminal or two port network. t.c.e. any signa1 to be amplified was supplied from an a. together with impedances or admittances. I= VY (2. In addition. Section 1 FUNDAMENTALS Ohm's law states that the current I flowing in an electrical circuit is directly proportional to the electrical pressure or voltage V applied to the circuit.2 FUNDAMENTALS OF NETWORK ANALYSIS Equivalent circuits for electronic devices will be shown to consist of simple current or voltage generators. we found that such devices could only operate correctly if they were connected to suitable 1oad and bias circuits consisting of impedances or admittances. we must be ab1e to analyse complex networks of impedances. Thus if we are to analyse these equivalent circuits. Correct application of Ohm's law demonstrates a nurober of important relationships which will be shown in the first section of this chapter. These methods and the solution of the resultant equations will be considered in Section 2 of this Chapter. having dimension mhos. The basic rules for such analysis are fortunately simple and will probab1y be familiar to the reader. The constant of proportionality is known as the circuit admittance Y.1) An alternative way of expressing this is: the potential difference V across an electrical circuit is directly proportional to the current I 42 . current or voltage source together with its associated impedance or admittance. Kirchhoff's laws lead to the two most important tools for network solutions. mesh analysis and nodal analysis. These are Ohm's law and Kirchhoff's laws. In Chapter 1.
For steady state alternating or sinusoidal currents and voltages.1. (2. In this case. the impedance and admittance are complex and become Z=R+jX (2. we must consider the sense of measurement of voltage and current. circuit shown in Figure 2. Circuit notation for direct currents and voltages R y 43 . in general I 1 B ::. i.c.FUNDAMENTALS flowing through the circuit. Consider first the simple d. zero.2) V=IZ By inspection we can see that V 1 Z=1= y or I 1 (2.3) Y=v=z For direct currents and voltages the impedance of a circuit is the resistance R and the admittance is the conductance G.1. Note. having dimension ohms.i: X and G =F R If a circuit has Z = R + jX.4) Y= G + jB (2.5) where X and B are known as the circuit reactance and susceptance respectively. B = 1 R jX R2 + X2 1 Z=  B X = R2 + X2 (2.e. the constant of proportionality is known as the circuit impedance Z. If the voltmeter and ammeter are centre zero and connected with the (I) X Figure 2. I Y= R + jX R G=R2+X2 and onIy ·r R lS.6) x·1 Notation Before we can proceed to further relationships.
In terms of this current i we can say that iZ2 va = iZa = +jwLi = i v2 = = iR ji = wC 44 (:J) . Consider the part of a circuit shown in Figure 2.2. the meter would read . This rule applies to alternating quantities in exactly the same way. If the sense of voltage measurement was also reversed we should find Vyx = . The instantaueaus value of the current will be i = isin (wt ± 4>) where 4> gives the phase with respect to some unknown reference. lf however. Conventional current flow is from positive to negative so the sense of measurement of +I is as shown by the arrow (I). Circuit notation for alternating · currents and voltages is flowing. the ammeter connections and thus the (I) arrow were reversed. given that current Figure 2. The voltmeter measures the voltage at x with respect to y and will measure +V= +IR= +Evolts in the sense shown by the second arrow (V).2. the voltmeter will read +E volts and the ammeter +I amps where I= (E/R) amps.I amps and we could say that I= (E/R). This example Ieads to the deduction of a useful rule: The potential difference that is produced by a current flowing in an impedance will be positive ( + IZ) if the sense of the potential difference is taken in the opposite direction to the chosen sense of current flow.Vxy and Vyx =IR= (E/R) X R = E as would be expected from the circuit.FUNDAMENTALS OF NETWORK ANALYSIS positive and negative terminals as shown.
7) eZz = Zt + 22 Potential division between series impedances is in the direct (2. . z2 Iv. Series and parallel circuit arrangements Total impedance ZT Also and = ~l = Zt + Zz.3a e = v' + vz = iZt + iZz = i(Zt + Zz) et V' :Jz. rewording the rule: In terms of a specified current. e Vt = 1Z1 = 21 + Zz X Zt . f V2 i! J Jl' (b) (a) ef ~ ~ :]'' y2 lv. Use of Ohm's Law Now applying Ohm's law to a nurober of simple seriesandparallel circuits.FUNDAMENTALS So for greater clarity.8) ratio of the impedance values. 45 . From Figure 2.3. Impedance in series may be simply added. r v2 'I~ (c) (d) J J!' Figure 2. a volt drop due to an impedancecurrent product will be positive only if taken in the opposite sense or direction to that of the current. vz = zZ2 (2.
3c e = v1 + vz = . 46 .9) X Y1 iYz = ==yl + Yz Current division between parallel admittances is in the direct ratio of the admittance values. = eY1 = Also h and iz = eYz i + Y1 Yz (2.3b i = h + iz = eY1 + eYz + = e(Y1 Yz) Total admittance Admittances in parallel may be simply added.i + i = i . The duality of a VZI system with an IYV system will become more obvious with further examples. = 1 Y1 1 YT 1 =E + + 1 Yz 1 Yz Total admittance of a nurober of admittances in series is given by the reciprocal of the sum of the reciprocals of the individual admittances.7 and 2. (2.FUNDAMENTALS OF NETWORK ANALYSIS Now referring to Figure 2.10 with 2. The similarity between the results in a series impedance system and a parallel admittance system is known as duality.+ Y1 Yz Y1 Yz (1 1) Total admittance YT i =. Consider Figure 2.9 and 2.11) This rule is frequently simplified to YT or YT = = yly2 1. .8. y 1 + y 2 10r two a nuttances m senes Y1Y2Ya y 1 y 2 + y 2 Ya + Ya y 1 for three in series etc." d . (2.10) The "reader should now compare 2. .
1 + . Potential division between series admittance is in the (2.1 47 . Total impedance of a number of impedances in parallel is given by the reciprocal of the sum of the reciprocals (2. Finally. Circuit for Example 2. .1..13) of the individual impedances.n. and Z1Z2 iZz = Z1 (Z1 + Z2) = Z1 + Zz 12 iZ1 = Z1 + Zz Current division between parallel impedances is in the (2. 2 = e (~1 + ~z} i Total impedance e =i = ZT 1 Z1 1 1 or 1 zl + Zz ZT = 1 1 + Zz and ZT = Z1Z2 zl + Zz etc.FUNDAMENTALS Also V! i = Y1 = e YT Y1 vz = Similarly Y1 Yz e Y1 X Y1 + Yz = i Yz = eYT Yz = = e Yz Y1 + Yz eY1 Y1 + Yz :. Example 2.4.12) inverse ratio of the admittance values. In (a) the branches are given in their admittance values while in (b) impedances %. Now from Figure 2.3d = ii + iz = .14) inverse ratio of the impedance values An example will illustrate the use of the above rules. Figures 2. 11 e = Z1 = iZT Z1 i . 4 mho ·1 ·1 io io (b) (a) Figure 2.4(a) and (b) show the same circuit.
therefore i4 = 4v amps ia is found by calculating i2 and dividing i2 between the 3 mho and 1 mho branches (rule 2. the 4 mho branch is in parallel with (the 2 mho branch in series with the 3 mho and 1 mho branches in parallel). the value of io in terms of i. First consider circuit (a). to find io from circuit (a).10). YT  4 + 2 2(3 + (3++1)1)  4 + ~6  1 53 mhos For circuit (b) we must use rules 2.11.13 since all branches are quoted as impedances.. Now multiplying numerator and denominator by t. and the potential difference v' across the 1 Q branch in terms of V. . io is the sum of the currents flowing in the 3 mho and 4 mho branches.= v amp 6 3+1 ia =  i0 = v + 4v = Sv amp 48 . 12 = v [2 X (3 + 1)] 8 2 + 3 + 1 = 6 v amp 8v 3 X .FUNDAMENTALS OF NETWORK ANALYSIS are given. the same result as that found when working in admittances.7 and 2. Applying rules 2. io = i4 + ia The 4 mho branch is connected directly across v.9 and 2. Using both (a) and (b) determine the input admittance ifv. Now.
FUNDAMENTALS
To find io from the impedance circuit (b) we follow the same procedure using rules 2.7, 2.13 and 2.14.
As before,
io = i4 + ia
V
}
V
=1+1 txlx!+1
2+!+1
=
4v
+
V
1
4
2 X 3
+
1
a
6v
= 4v +  = 5v amp
4+2
Finally, to find v' from each circuit, we can ignore the 4 mho
branch and note that v' is the potential across the 3 mho and 1 mho
branches in parallel.
For circuit (a) using rules 2.9, 2.11 and 2.12
v' = v
X
2
2
V
+ 3 + 1 = 3 volts
and for circuit (b) using rules 2.7, 2.8 and 2.13,
V
X!
V
=   = volts
i+!
3
Section 2
From Example 2.1, it can be seen that simple circuit problems
may be solved by the rules based on Ohm's law. There are three
disadvantages to this approach. Firstly, with more involved circuits,
the resulting expressions become exceedingly unwieldly and the
possibility of a mistake increases. Secondly, if several different
unknowns are required, as in Example 2.1, a separate solution is
required for each. Finally, if the circuit involves more than one
generator, the combined effect cannot be determined. Two most
49
FUNDAMENTALS OF NETWORK ANALYSIS
important methods of circuit analysis are based on Kirchhoff's law.
These are really only common sense and will be explained in these
terms.
KIRCHHOFF'S LAWS
1. The Current Law
In everyday language this would be defined as 'That which goes
in must come out'. Consider the part circuit shown in Figure 2.5a.
is
is
(a)
(b)
Figure 2.5. Summation of currents at a node
This shows a circuit node or junction between a number of branches.
It is not shown as a capacitor and can therefore store no charge.
Thus as a quantity of charge enters the node, an equal quantity
must leave it. But moving charge is current, so we can say: the sum
of the currents entering a node must be equal to the sum of the
currents leaving that node. In Figure 2.5a this relationship is given by
h
+ ia + i4 =
i2
+ is
An alternative definition is: the algebraic sum of the currents
entering a node is equal to zero.
Again referring to Figure 2.5a, currents i2 and is leaving the node
are equivalent to currents  i2 and  is entering the node as in
Figure 2.5b. From this alternative definition,
h
+ ia + i4 
i2  is
=0
which Ieads to the same result as that obtained by the first definition.
A useful analogy is traffic flow at a roundabout, where unless
there is an accident or breakdown, the total number of vehicles
entering the roundabout is equal to the total number of vehicles
leaving it.
50
KIRCHHOFF'S LAWS
2. The Voltage Law
For this law, in everyday language, we could say that however
far one falls in altitude, one must climb by the same amount to
reach the original starting point.
A particular point in an electrical circuit can possess only a
single value of potential or voltage. It might be at earth or zero
potential or it might be at say, 5 000 V but it cannot be bothat the
same time. lf we start at such a point (at say + 10 V) and move
araund the circuit, we might climb to a high positive valtage or fall
to negative voltage, but when we return to the starting point, it
must be at + 10 V.
Change in potential may either be due to currents .flowing in
impedances (/Z volt drops) or due to generators or sources of e.m.f.
Thus we can say that for any closed loop in an electrical circuit,
the sum of the rises in potential due to generators, must be equal to
the sum of the falls in potential due to IZ volt drops. Consider
Figure 2.6a.
(b}
(öl}
Figure 2.6. Summation of voltages around a loop or mesh
Starting from point X and travelling in a clockwise direction, we
shall first sum the changes in potential due to the generators. We
climb through V1, fall through Vz and climb again through Va.
Thus the sum of the rises in potential is V1  Vz + Va. To find
the sense of the /Z volt drops, we need to know the sense or direction
of I. This could only be found if values for V1, Vz and Va were given.
We therefore guess the direction and then if on inserting values, we
obtain a negative answer, we can change the direction for the solution.
In this case we guess that I flows in a clockwise direction as shown.
Now to obtain positive /Z products, the potential must be measured
in the opposite direction. The positive sense of these /Z products
have also been inserted in Figure 2.6a. Now starting from point X
51
FUNDAMENTALS OF NETWORK ANALYSIS
we can add the changes in potential until we return to the starting
point where we must return to our starting potential, i.e. the total
change in potential is zero.
+ V1 
IR1  V2  IR2  !Ra+ Va  IR4
=0
or the algebraic sum of the potential difference taken around a closed
loop is zero.
Alternatively we can say that the sum of the potential rises due to
generators is equal to the /Z volt drops when taken around a closed
loop in the same direction.
For the circuit of Figure 2.6a, this would be given by
V1  V2
+ Va =
l(R1
+ R2 + Ra + R4)
from which I could be obtained if numerical values were given.
In practice the currents in the different branches may have
different values as shown in the a.c. circuit in Figure 2.6b. The method
of writing the equations is exactly the same for a .c., and the solution
will give the value of the unknown currents in the sense shown and
their phase with respect to the given a.c. generators. The positive
sense of IZ drops has been shown and the equation may be written
e1
+ R1(h i2) 
(R2
+ jwL2)i2
(wd2) (i2 ia) e2 = 0
Further examples of writing loop or mesh equations and junction
or nodal equations will be found when we see how Kirchhoff's
laws are applied to mesh and nodal analysis.
MESH ANALYSIS
For mesh analysis, all branches are shown as impedances and all
sources are voltage generators. Unknown currents are selected
(i.e. named h, i2, etc.) and their sense indicated. The loop or mesh
equations are written and solved giving all the branch currents in
terms of the source voltages and branch impedances. This technique
will be best understood by considering some examples.
Example 2.2. Using mesh analysis determine the current flowing
in the 2 0 branch in the circuit shown in Figure 2.7.
The first problern is to insert the unknown currents. Any notation
can be used provided Kirchhoff's current law is obeyed at the junctions. One possible set of unknowns can be inserted as follows.
Let the two batteries have currents h and /2 as shown and Iet the
52
KIRCHHOFF'S LAWS
required current in the 2 Q branch be lx. Now, following the current
law, the 4 Q branch must carry (h lx) and the 5 Q branch
(h + h  lx).
This is not the simplest method but we shall first obtain the
required solution using these unknown currents. There are six
I,
6V
lz
c
2ß
2V
Figure 2.7. Circuit for Example 2.2
possible loops for which equations can be written, but since there
are only three unknown currents only three equations are required.
The equations for meshes A, B and C are respective1y
+ 5(h + h  lx)
5(h + h  lx)  4(/2 
A
6 = 3h
B
0 = 
C
2
=
+ 2/x
2/x /2
Collecting terms
6 = 8h
0
lx)
=
+ 5h 
5h  9/2
2 = 2/x
5/x
(2.15)
+ 11/x
(2.16)
+h
(2.17)
One method of so1ving these equations is by substitution as
follows
From 2.17
(2.18)
substitute in 2.16
SI2
=
18
+ 18/x + 1llx
18
h=5+
or
29
5
(2.19)
Ix
substituting for h and h from 2.19 and 2.18 in equation 2.15
6
144
232
=  5 +5
lx
53
+ 10 
10lx  5lx
FUNDAMENTALS OF NETWORK ANALYSIS
Collecting terms, 6
+ 5144 
10
= lx (232
5 
15 )
24·8 = 31·4/x
simplifying,
24·8
lx = 31 .5 = 0·79 A
and
Maxwell's Circulating Currents
Now since any notation for the unknown currents may be used,
we should investigate the quiekest and most convenient method.
This is known as the Maxwell's circulating current rule. The circuit
for Examp1e 2.2 is redrawn as Figure 2.8.
6V
2V
Figure 2.8. Use of Maxwell's circulating currents
Maxwell's circulating currents (h, i2 and ia) are shown. lt is assumed
that each c1osed 1oop has a current associated only with that 1oop.
Where a branch is peculiar to a loop, the branch current is equal to
the loop current. If, however, a branch is common to two loops
(the 5 n branch for example), the branch current is equal to the
difference between the two loop currents.
Thus the 5 0 branch current is (h  i2) fiowing down or (i2  h)
fiowing up. Similarly the 2 0 branch current is (i2  ia) fiowing
down or (ia  i2) fiowing up.
So writing the equation for the h loop we find
6 = 3h + 5(it i2)
and for the i2 and ia loops respectively,
0
2
=
+ 4i2 + 2(iz 
S(i2  h)
= 2(i3 i2) + i3
Collecting terms and rearranging
6 = 8h 5i2
0
=
2=
5h
+ 11 i2  2i3
2i2 + 3i3
54
i3)
(2.20)
(2.21)
(2.22)
This is replaced by the constants. Loop current x branch impedances = h(3 + 5). sum the e. in the direction of the loop current and equate to the loop current times all the loop branches. minus any adjacent loop current.5i2 which is the same as equation 2.. Further application of this rule will appear in the next example. A better method for the solution of a number of simultaneous equations is the applications of determinants.f. Equation becomes 6 = 8h . we shall apply the method in full detail so that the reader can regard this method as a useful aid to the solution of circuit problems.m. For the theory behind this method the reader is referred fo any good mathematics textbook.f. In this case ß = 0 2 3 5 11 2 8 5 0 To obtain the value of one of the unknown currents. Here. times the common branch sharing that current. Adjacent current x shared branch = Si2. Following this for the first loop: The e.20 above.m. For any loop. The determinant ß of the network is made up from the coefficients of the unknowns. Thus for ßt ßt = 6 0 2 5 11 2 0 2 3 Similarly for i2 and ia we should require ß2 and ßa ß2 = 8 5 0 6 0 2 0 2 3 and Now ßa = ia = 55 8 5 5 11 0 2 ßa tl 6 0 2 . we write a second determinant ßt which is the same as ß except for the h column. h. in the direction of the loop current is +6 V.KIRCHHOFF'S LAWS These equations could have been obtained more simply by applying the following rule.
%~ :: :: Cl C2 = Ca 01 I b2 ~ ba._ ~ bi I 02 ~ oa ~ I+ I Ci 02 ~ oa ~ I Note in each case if a horizontal and a vertical line are drawn through the principal coefficient.75 = 157 0 8 6 5 0 2 = 8(0 . Now applying this to equations 2.FUNDAMENTALS OF NETWORK ANALYSIS Now all we have to do is to find the numerical values of the determinants.6( 15 .0) + 0 0 2 3 =58 ßa= 8 5 5 11 0 2 6 0 = 8(22.20. Alternatively.21 and 2. the remaining second order determinant or cofactor consists of the remaining terms.22 for Example 2.4) . we may expand on the first column.ia = 58 124 = .0) + 6(10. This process is known as expanding the determinants. The expansion of a second order determinant is as follows: I %~ :: I = Oib2  bio2 Fora third order determinant the expression is as follows: o b 1i Ci b022 c2 obaa Ca = 02 I b2 C2 I I ba Ca  02 bi Ci ba Ca I+ oa I bt Ci b2 c2 I This is known as expansion on the first row.0) 2 = 66 58 i2 = 157 A.=0·79A 157 56 + 66 A 157 . ß= 8 5 0 5 11 2 0 2 3 =8(334)+5(150)+0 = 232.2.0) + 5(10. 2. Unknown 66 and ia = 157 A h = i2 .
1 A (4 . as is the case in the next example.3.j) . for equation 2. S(cos 53° + j sin 53°) = 5(0·6 + j0·8) V = 3. Determine the current supplied by the 6 V generator in the circuit shown in Figure 2. Circuit for Example 2.9.f. but with practice it is possible to write down the values for the determinant directly from the original equations. expressing all impedances in the complex (a + jb) form.24.24) .(4.j3) Now collect terms and converting generator e.j4)i2 Evaluating the e. 6/0° = it(3 + j2 + 4 . lt 3 I I (6 + jO) (4 .j4) 57 (2.j) {3j4) (7j4) = .f.j) = 5/53° h(4.j) + i2(4.j4 V Now solving by determinants .m.j)i2 (2.9. Figure 2. Example 2.23) 5(cos 53° + j sin 53°) = (4 .m.KIRCHHOFF'S LAWS This method may appear longer.s to (a + jb) form 6 + jO = (7 + j)h.j)h + (7 .j) (7 .(7_+_j_)(4j).j + 1 + 2.3 First we insert the currents as before and then write the normal mesh equations.i2(4 . In addition it is the only convenient method if the coefficients of the unknowns are complex.
j4) .10.j) A . Vc.FUNDAMENTALS OF NETWORK ANALYSIS expanding the determinants .(3 + j4)(4 .j24. l1 = y/{262 + 372) ctan1 H y/{382 + 132) /tan1 HA = J2048 = /550. 11 = (6 + j0)(7 .49 42.jl3 Converting the numerator and denominator into the R/0 form: .16 + 1 + j4 + j4 A +4  _ 26.12.j4).4.j) {7 + j)(7 .jl6 + j3 j28 + j7 . The dual system known as nodal analysis is based on Kirchhoff's current law.10.38. NODAL ANALYSIS Mesh analysis was developed by the use of Kirchhoff's voltage law.j)(4.{4. Circuit suitable for nodal analysis circuits consisting of current generators and admittances. Consider the circuit shown in Figure 2. In practice we usually express the voltage at a 58 . This circuit has three nodes each of which will have a potential or voltage VA.{19o)A 1484 0·686/36° A Thus the current is lagging the 6 V generator voltage by 36°. We solve for unknown node voltages (as opposed to unknown mesh currents) in terms of current generators and the circuit admittances.j37 A . VB. Since it is a dual we shall expect to use this method on c Figure 2. Many other examples of mesh analysiswill occur in Iater chapters.
VB)Yz and at node B [z =(VB.4. Foreach node. In this case we shall Iet node C be at earth. Circuit for Example 2.O)Y1 +(VA. Determine the current flowing in the 5 mho branch of the circuit shown in Figure 2. We shall now apply Kirchhoff's current law in the following manner. !1 + [z = VA(Yl + Yz). equate the currents entering the node to that node valtage times the sum of all admittances connected to the node. minus each adjacent node valtage times the connecting branch admittance.4 This circuit has four nodes. we can see that these nodal equations could be found in another way. Thus at node A.VA)Yz Collecting up terms. D Figure 2.11.e. and we shall consider node D to be at zero potential. For node A.11.4Vc + VB(1 + 1 + 5) 59 1 Vc . 0 = = VA(l 1 VA + 2 + 4)  1 VB . V c = 0. Currents entering a node from current generators may be equated to currents leaving a node through admittance branches. h + [z =(VA. Example 2.YAYz + VB(Yz + Ya + Y4) From this.NODAL ANALYSIS circuit node with respect to earth or zero. 3.VBYz lz = .O)(Ya + Y4) +(VB. This is of course the dual ofthe rule for formation of the mesh analysis equations on page 55. i.2 For node B.
Section 3 In this section a nurober of useful theorems will be stated and demonstrated without academic proof. In general if a circuit has mainly parallel branches.1(8. 7 1 4 1 0 1 4 3 8 V VB= 7 1 4 1 7 1 4 1 8 Expanding 7(0 + 3).4).1) + 1( 8 . nodal analysis will be quicker.0) VB = 7(56 .4Vc 0 = VA+ 7VB. I= 5 X 0·17 = 0·85 A Other examples of nodal analysis will appear in later chapters. 1= 1VA. 2 + 1 = 4VA.4) .4(3. Such conversions may be accomplished by the use ofThevenin's theorem and Norton's theorem which will be discussed in the next section. Examples using each will also be given.4(1 + 28) V 21 + 12 + 12 VB= 392.VB. and hence the nurober of equations. It may be necessary to convert valtage generators to current generators or vice versa. If there are more series branches mesh analysis is best.1VB+ Vc(3 + 1 + 4) Rewriting.VB+ 8Vc By determinants. the other in terms of an admittance network.116 45 = 264 = 0 "17 V The current flowing in the 5 mho branch is given by VY. 60 . In some instances. THE SUPERPOSITION THEOREM This may be stated in two forms.12. use of the better method will result in reducing the number of unknowns.Vc 3 = 4VA. one in terms of an impedance network.FUNDAMENTALS OF NETWORK ANALYSIS For node C.
we remove the 12 V battery leaving only its internal impedance of 6 Q. 6 l=h= 6 5x6x5+6A 2 +5+6 36 36 = 22 + 30 =52 A By the Superposition theorem. the current flowing in one brauch is equal to the sum of the currents flowing in that brauch due to each generator taken separately with all other generators replaced by their internal impedances. 24 + 10 = 24 52 A Now taking the 6 V battery alone.7. Now for this circuit l=h= 12 2 2x5x2+5 6 +2+5 (Rules 2. I= h +h = + 36 24 52 61 = 1155 A . Circuit for Example 2. Check this answer by use of mesh analysis.12.THE SUPERPOSITION THEOREM In any linear network of impedances and generators.) = 42 Evaluating. By use of the superposition theorem calculate the current flowing in the 5 Q branch ofthe circuit shown in Figure 2. Example 2.0. for the whole circuit. 2.12.12 and 2.(l 12V 6V Figure 2. Now applying the same rules.13 on pages 45 and 47.5 Taking the 12 V battery alone.5. 6. 2. we remove the 6 V battery leaving only its internal impedance of 2 Q.
12. h + ?iz 84.FUNDAMENTALS OF NETWORK ANALYSIS Now checking this result by mesh analysis and using the unknown currents h and iz shown in Figure 2. Figure 2. Circuit for Example 2. By the use of the Superposition theorem calculate the potential across the branch YL in the circuit in Figure 2.13.6.13. Example 2. 1 2 X 1 6 +2+1 1 = 20 V V due to h alone = V1 = = Vz = V due to /z alone 3 2 x 6 1 +2+6 18 Vz = .20 V 62 1 X 2 +IV X 2 +6V 6 .6 Check the answer by use of nodal analysis. The dual form of this theorem may be stated: In any network of admittances and current generators the potential across one branch is equal to the sum of the potentials across that branch due to each generator taken separately with all others replaced by their internal admittances.iz = H A as was found by superposition. 12 = 11h.5iz 6 = 5h By determinants.30 = 77  iz = 25 54 = 52 A 66 + 60 52 6 =52 A But the required current I = h .
at node 2. 63 .14. Thevenin's theorem The twoterminal equivalent is shown on the right. Figure 2. Now v = v1 + v2 = 17 20 v Now checking by nodal analysis: at node I. Thevenin's theorem states that any twoterminal network of generators and impedances may be replaced by a single valtage generator in series with a single impedance.THEVENIN'S THEOREM Note V2 is negative following from the direction of h and the required sense of V. the network is shownasabox with two terminals.14. In Figure 2. The components of the equivalent are found as follows: Vofc is the valtage measured across the terminals of the network when no Ioad is connected.= v 20 20 which is the same result as that found using the Superposition theorem.. 3 = 2V1 + 3V2 3+6 9 v1 = 244 = 20 and 24 + 2 26 V2==20 20 Now v = v1  926 17 v2 = .
Now. 1l! 1 I= 3 = 1l. 3V Figure 2. The truth of this may be demonstrated by a simple example.15b. Determine the value of RL that will carry a current of i A in the circuit shown in Figure 2.+ RL = 10 + 7RL = RL 15 10 + 7RL A 15 1= = (45  7 10) 45 Q = 5n Checking by basic methods 1 I = 3= 1 2 3 5RL + 5 + RL 15 X 5 5 + RL A 3 = ::10~+:2::::R::L+::5:=R:L 64 .15.FUNDAMENTALS OF NETWORK ANALYSIS Z1n is the impedance measured between the terminals when all internal generators are suppressed or replaced by their internal impedances.7.15. Example 2. Circuit for Example 2. Now 3 5 15 2 X5 10 X Votc =2+5=7 Zin = 2 + 5 = 7n V The equivalent circuit is now shown in Figure 2.7 First break the circuit at points XX and apply Thevenin's theorem to the lefthand half of the circuit.
8. j 2. . i26 V= (0·77 + J·0·154) V 65 2~ V .~i(5 + j) V = (26 + j26 = 20 26 10 +4 j2 . write the mesh equation for the closed loop: 2 .16.5 .j3)i . Figure 2. 5 + j) = ~~ (5 + 1  = ~~ (1 . Example 2.0.jt·54) n j12 r\ u .J j3) .c. 5 = 1 +J+ = 1+j J + (2 .1 .(4 Ztn + j2)(1 .j l=.v'2/45° 2. example.j and Now = i(5  j) = i(5. 1.THEVENIN'S THEOREM + 10 = 45 Q RL = 5 n as before. (1 .4  + 6 + j2. and hence find the powerthat it could supply to a Ioad of (3 + j2) 0.8 Applying Thevenin's theorem. Determine the equivalent generator for the circuit shown in Figure 2.j)(5 = (2·31 . 5.16.j20) + J.j3)(1 .j) V . .J Votc = \1'2/45° + (1 .j) . 7RL Now consider a more difficult a.J j5 + j) n To find Votc. Circuit for Example 2.
The components of this equiva1ent are found as follows: fstc is the current that would flow in a shortcircuit connected across the terminals. the load current may be found.17 the network is shown as a box and the Norton O:=IT ls/c Figure 2.FUNDAMENTALS OF NETWORK ANALYSIS When the resulting Thevenin equivalent is connected to the load of (3 + j2) 0. i A Ytn = i + i.= l!i mhos fstc = 66 . Y1n is therefore the reciprocal of the Thevenin equivalent Ztn· Example 2. Repeat example 2. note that with a short circuit across XX in Figure 2. the 1oad power is given by 0·772 + 0·1542 = 5·312 + 0·462 X 3 W 0·612 X 3 28·61 = 0·0645 w = or P= 65mW Norton's theorem is similar to Thevenin's theorem except the equiva1ent is expressed as a current generator in parallel with an admittance.7 using Norton's theorem. Ytn is the admittance measured between the terminals with all generators suppressed.5·31 + j0·46 Since power can be dissipated on1y in resistance. 1 0·77 + j0·154 = 3 + j2 + 2·31 .15. . no current will flow in the 5 n resistor.9.17. In Figure 2. Norton's theorem equiva1ent circuit is shown on the right. To determine fstc.j1·54 A _ 0·77 + j0·154 A .
current divides in the direct ratio of parallel admittances.j3 10(1 .j6  + 1 + j _ 2(1 .j2) .j) 3 J·2 X lJ 5 J. 1 1 3.j2) 10(1 .j2 Now YL = ZL = 3 + j2 = _1_3_ mhos 4 h = 3.8 using Norton's theorem. h = 4(3.j) A When the load is connected.30 . Repeat Example 2.j).j3) + (4 + j2) 4 + 6 .j63 67 + 65  j13 A .j2) + 13(5 .j2) A .j) + j2 = + j2) A 4 10(1 .j2 10(1.4 2 . o.j20 _ 4(3.16 1 Yln = 4 + j2 + 1  1 j3 = (I . For the circuit shown in Figure 2.20 .j30 .j = l0(1 _ j) mhos __ 2_ + j2 lsjc.75.j3) + (1 + j)(4 1 .+ 10(1  A j) Multiplying numerator and denominator by 130(1 .j) 4(3.NORTON'S THEOREM Since current divides in the direct ratio of parallel admittances ! 1 X YL 3 = YL + l0 !YL + :lö = tYL YL = 7 32ö ! _ ! = 45 _ 10 mhos 1 35 RL===5!2 and 7 YL Example 2.j 12 + j2 mhos 5.j)(3 .10.j) +4  2 + j4 10(1 .
e. h. 1~/1 h2 f Y1 Network f V2 V ~ Figure 2. Z Parameters Let h and /2 be the independent variables. This approach is similar to the description of twoterminal networks by means of Thevenin's and Norton's theorems. there are six possible sets of parameters. but not always. For any particular set ofparameters. or other networks may be connected.25) V1 = hP + l2Q V2 = hR 68 + ]zS (2. A fourterminal network The currents and voltages at the two ports are conventionally taken in the directions shown. common. Since any two may be taken as independent. The diagrammatic form of this is shown in Figure 2. the same result as that obtained using Thevenin's theorem) Section 4 FOURTERMINAL NETWORKS A common form of network. Ioads. occurring frequently in electronic circuits. Such networks have two pairs of terminals or two ports to which sources.FUNDAMENTALS OF NETWORK ANALYSIS lhl 2 RL 16(9 + 4) X 3 = 752 + 632 w Load power= = 65 mW (i. There are four variables V1. Terminals 1' and 2' are frequently.26) .18. two ofthese variables are considered as being independent while the other two are dependent.18. One description of such networks is in terms of network parameters. This allows for each porttobe taken as either input or output. In general we may write two equations (2. V2 and ]z associated with the network. is known as a fourterminal network.
The parameters of a particular network may be found by writing the mesh equations and rearranging so that the parameters can be found by inspection. i.29 and 2. lz Figure 2.28 with equations 2. Zu relates V1 to h.27 and 2.2' are open circuit. we can see that the Z parameters of the circuit shown in Figure 2. By comparing equations 2.29) (2.FOUR·TERMINAL NETWORKS To show that such equations are possible.28) which have the form of equations 2.26 above.30) The double suffix notation indicates which pair of variables is related by the particular parameter.e.30 must be true for all values of the independent variables. they must be true for either h or /2 equal to zero.19. Solution of a fourterminal network by mesh analysis Dimensionally the parameters of the network P.19 are given by A more general method of determining the parameters for a network is as follows. This can only occur if terminals 2.19. Q. Since equations 2. consider the simple circuit shown in Figure 2. R and S must be impedances since in each case the parameter multiplied by current results in voltage.30 above.27) (2. Suppose 12 is zero.29 and 2. I.25 and 2. Applying Kirchhoff's laws we can see that + Z2) + hZ2 V2 = hZ2 + hZ2 V1 = h(Z1 (2. The equations may therefore be written V1 = hZu + /2Z12 (2. 69 . In each case the first number is given by the dependent and the second by the independent variable. Z12 relates V1 to h etc.
Find the Z parameters of the circuit shown in Figure 2.1' be open circuit making h zero. h Figure 2.FUNDAMENTALS OF NETWORK ANALYSIS Equations 2.11. we obtain Z12 = v11 and /2 I!=O Z22 = v21 h I!=O Example 2.29 and 2.20. Applying the rules based on Ohm's law 3(1 + 2) v1 = h x 3 + 1 + 2 = 1!h v1 Zn=h Also Now Iet h V2 = ft =0 v2 = h Also V1 X 3 I 12=0 = 3 + 1+2 1! Q X 2 = ft + 22(1 + 1++3)3J= 3!/2 Z22 = 3tn x [2 =h X 2 2 + 3 + 1 X 3 = /2 z12 = 1n 70 .11 Let /2 = 0. . if we let terminals 1.30 now become From which Zn and Z21 may Zn = v11 h be defined. and 12=0 Z21 = v21 h 12=0 Similarly.20. Circuit for Exarnple 2.
so Z12 must be shown as a voltage generator of hZ12 volts. Zn can be shown simply as an impedance.22.22. 1~ Figure 2.21. A passive network is one containing no elements such as valves or transistors.29 V1 = hZn + hZ12 Since this represents the sum of two voltages.21. I. A loaded fourterminal network 71 . lz Figure 2.FOURTERMINAL NETWORKS Note Z12 = Z21· This is always true for a passive network. The resulting equivalent circuit is shown in Figure 2. /z does not flow in this part of the circuit. As h flows into this part of the circuit. General Solutions in Terms of Z Parameters A common problern is to find the input impedance and voltage gain of a network when a load ZL is connected to one port as shown in Figure 2. The Z parameter equivalent circuit No additional information is given by this equivalent circuit but its use sometimes makes the formation of complete network equations much easier. we can see that the input side of our equivalent circuit must contain two components. From equation 2. It is frequently convenient to show the Z parameter equations in the form of an equivalent circuit. Similarly Z22 can be shown as an impedance in the 'output circuit but Z21 must appear as a voltage generator of Z21h volts.
which are obtained by connecting an impedance Zs across termirrals 1. V2 = hZL Equation 2.Z21Z12 lt=~~~~~~~= The input impedance given by Vt = Zu(Z22 + ZL) .32 and 2.30 now becomes hZL = ftZ21 + hZ22 (2. There are two further general solutions.Z21Z12 Z22 + zL Ztn Z21Z12 Z22 + ZL = Zu  (2.1 '.32) o.31.Z21Z12 Also /2=~=~ Zu(Z22 From which the transfer admittance h Z21 Vt = Zu(Z22 Now since V2 = + ZL)  Z21Z12 (2. = ltZu + hZ12 0 = ftZ21 + h(Z22 + ZL) Vt Solving for h by determinants V1(Z22 + ZL).31) Rewriting equation 2.33) hZL.FUNDAMENTALS OF NETWORK ANALYSIS From the sense of V2 and h we can say.33 are known as the general solutions for the network in terms of the Z parameters.34) Equations 2. The equations now become + Zs) + hZ12 V2 = ItZ21 + hZ22 0 = h(Zn 72 .29 and rearranging equation 2.0 Zu(Z22 + ZL) .Z21V1 + ZL) . the voltage gain Avis given by v2 zLz21 = Vt Zu(Z22 + ZL) .Z21Z12 (2.
12. A certain electronic device is represented by the equivalent circuit shown in Figure 2. The results that should be obtained are (2. 400. n v2 = 5ovl + 5h 73 . (b) the voltage gain (V2/Vl). and the volt drop across the 5 resistor.0.23. when loaded with 400 Q and (c) the output impedance (V2/I2) if the source impedance is 5 n. Figure 2. Let h = 0. To find the Z parameters we apply the standard technique. VI = h(20 + 5) V Zn = vh1 I 12=0 = 25 n There will be no potential difference across the 100 Q resistor (/2 = 0). Circuit for Example 2. 50 VI.36) Example 2. first Ietting h = 0 and then h = 0. h flows through the 20 n and 5 n resistances only.FOURTERMINAL NETWORKS The reader should check that these are correct and solve for the output impedance V2/I2 and the reverse voltage gain V1/V2.35) and (2. and the load is disconnected. so V2 is the sum of the generated voltage.23. Since the output terminals are open circuit.12 Determine the Z parameters for the device and hence find (a) the input impedance (V1/h) when the output is loaded with 400 Q. Writing equations by inspection then leads to the required parameters.
For the remaining solutions we need only apply equations 2. In this case Ia flows through the 100 Q and 5 Q resistors. 2.34 and 2. (a) Z12Z21 Zln =Zn.32.FUNDAMENTALS OF NETWORK ANALYSIS v1 = But . 25h V2 =50 X 25h v2l h and Z21 =  12=0 + 5h = 1 245 n Now Iet h = 0. and there is no potential difference across the 20 n resistor. Thus and and V2 = 105/2 .5ov1 = 105/2 .35. Z21 = 1 245 Q and Z22 = 145 n.50 X 5h v2l =1450 h h=O Z22= Summarizing Zu = 25 il. thus v2 = VI 400 251 ~4~55 +5 = 39·5 (c) The output impedance of a circuit will be discussed in detail in later chapters.z22 + ZL = 25 5 X 1 245 I45 = 49·4 n + 400 (b) Voltage gain V2 VI= Zn(Z22 ZLZ2I + ZL)  Z21Z12 400 X I 245 I45) + I 245 X 5 = 25(400 This last expression is conveniently simplified by dividing numerator and denominator by I 245. but simply it is the impedance of the Thevenin 74 ... Z12 = 5 il.
The use of the different types will then be illustrated in this and later chapters. .38) V:~ be zero. We shall investigate one of these in detail. VI and h. VI and [z. The other possible choices. Sinillarly R. In this case . h and Vz. When the parameters are mixed in this fashion. relating [z to · V2. Our two equations must have the form + V2Q 1z = hR + v2s VI= hP Notice in this case our four parameters cannot have the same dimensions. VI and Vz. is a number and S. Now let h and Vz be the independent variables. These were obtained by selecting h and lz as the independent variables. we shall step directly to the final result for the other sets. and Vz and [z each lead to a separate set of parameters. and noting that the results are very similar to those of the Z parameters. While P relates VI to h and is therefore an impedance. they are known as a hybrid set of parameters and are given the symbol h.FOURTERMINAL NETWORKS equivalent generator determined at the output termirrals of a circuit in the absence of a load. Our equations thus become VI = hhu + V2hiz lz = hh2I + V2h22 To define our parameters we can now let either h or leading to the following relationships [zl fi h2I =  Ve=O 75 (2. a current ratio.Zu+ Zs Q = _ 145 + 5 x 1245 0 25 + 5 = 630 h Parameters So far in this section we have considered only the Z parameters of a fourterminal network. Q relates VI to V2 and is simply a number. Output unpedance = Vz ZziZiz 12 = Zzz.37) (2. is an admittance.
13. Note /z is the current fiowing in the short circuit. Circuit for Example 2. Now hz1 = ~~ =!4 h Vz=O For h12 and hzz we must opencircuit the input to make h = 0.24). first Iet Vz equal zero. this is simple current division between parallel conductances. Zu is not the same as hu since Zu is an open circuit parameter (/z = 0) while hu is a short circuit parameter (Vz = 0). First determine hin terms of Vz by finding the total conductance at the output terminals. In this case.FUNDAMENTALS OF NETWORK ANALYSIS Note that hzz is the reciprocal of Zzz since both are determined with the input open circuit (h = 0). This implies a short circuit across the output terminals. For hu we must determine hin terms of V1 and since the 7 mho conductance is short circuited h hu = V1(6 = vll h + 2) A Vz=O =! n 8 For hz1 we must find h in terms of h.24. Following the same method as we used for Z parameter calculations.13. Thus /z = h X 2 2 1 + 6 = .24. hzz h I 2 X6 = 8! mho = Vz = 7 +II=O 2+6 76 . Example 2. Figure 2.4h A The minus is required since the parameter convention requires that lz fiows into the netw0rk (see Figure 2. Determine the h parameter equivalent circuit of the network shown in Figure 2.
V1 = Vz X 2 2 +6V hlz = Vz vll r =o =!4 1 Note once again the passive network leads to the same numerical value for h21 and h12 but the current convention results in h12 = hz1 As with Z parameters. The second equation is the sum of two currents so our equivalent must have two parallel components. Figure 2. it is often convenient to use an equivalent circuit. Rewriting the h parameter equations V1 = hhu + V2h12 /z = hh21 + Vzh22 We can see that the first equation is similar to the Z parameter equations in that it is the sum of two voltages. The current V2h22 will flow in an admittance of hz2 mho when Vz is applied across it. The h parameter equivalent circuit General Solutions in terms of h parameters If the four terminal network is loaded. Since this part of our equivalent circuit will carry h. The other voltage Vzh12 must be produced by a voltage generator. In this case the second equation 77 .25. we have potential division across series conductances so. the voltage huh will appear across an impedance hn Q.25. The second current h21h can be provided only by a current generator in parallel with the admittance h22· The resulting equivalent circuit is shown in Figure 2. general solutions may be derived in terms of the h parameters.FOURTERMINAL NETWORKS Finally to find V1 in terms of Vz.
we can find. = reverse current gam = h22(hn + Zs) _ h21 h12 (2. We have now derived two sets of parameters.43) and h .h21h12 From which (2. Z and h.40) Solving by determinants for h. The only difference is the result for which each equation is true. so the required Substitution is h = writing and rearranging the equations. 78 . h and h. Bach set of equations has exactly the same form.36.35 and 2. by taking h. V2 as our independent variables.43 and 2.41) Solving for V2 From which (2.41.hn + Zs (2. Since the steps for developing the y and g parameters are precisely the same we shall summarize only the important results. I = VI(h22 + h) hn(h22 + h) . The reader should now compare equations 2. 2.39) + V2(h22 + YL) (2. 2. hl2 I. = output admlttance = h21h12 h22 . 2. . 2.. we obtain V1 v2 y L· Now re = hhn + V2h12 0 = hh21 (2.42.42 and 2.32.44) Equations 2.43 are important since they will be very useful for the solution of transistor circuits.42) Similarly by putting V1 I~ V2 = hZs.41.44 with equations 2. 2.FUNDAMENTALS OF NETWORK ANALYSIS is for h.33.
FOUR·TERMINAL NETWORKS Y Parameters Independent variables V1.rrent gatn = .51) 79 . The y parameter equivalent circuit General solutions: .Yu + Ys .Y21 Y12 /2 yl2y21 Output admtttance = V2 = Y22 . h Reverse current gatn h = (2 49) · (2. (2. h h Cu.= . Equations: V2.26.46) Conditions for defining parametersshort circuit input or output. Definitions: hl Yu = V1 hl Vz = O Y21 =V 1 V2 =0 mho mho Equivalent circuit: Figure 2. +Y21h Yu( Y22 + YL) . h Input admtttance = Vt y = u  yl2y21 y 22 + YL (2.45) (2.50) Y12Ys y 22(Yu + Ys) _ y 21 y 12 (2.47) From which .
h = V1 = gu  /z h Current gatn = . g21 = V 1 12=0 rat10 Equivalent circuit: Figure 2.= g12g21 +ZL g22 . The g parameter equivalent circuit General solutions: . one further rule should be discussed.s . Output tmpedance =.53) Conditions for defining parametersshort circuit input or open circuit output. This is simply achieved by drawing the 80 . VI Reverse voltage gam = T7 = ( +g12 y.56) (2.54) (2 55) · (2.) r2 gzz gu s .L2 + ZL)  g12g21 = gz2.57) Before applying these results to some examples. h Equations: h = V1gu + hg12 V1g21 + hg22 v2 = (2.52) (2.g21 gu(g22 ~ .27. Definitions: · v21 .gugn~2 + y. Input admtttance . It may be necessary to convert from one set of parameters to another.g21g12 (2.FUNDAMENTALS OF NETWORK ANALYSIS g Parameters Independent variables: V1.
/2 Iv = 1(1 = v2 y22 = 0·85 mmho 1 0 r 0·5) = 0·25 r 1 r 1 r 0.Ia. Circuit for Example 2. Calcu1ating yz2 in the same way as yu.14. To determine the y parameters.28.28 determine (a) the y parameters and (b) by conversion the g parameters. For the network shown in Figure 2. Using 4kn I.5 = (0·25 2] 2:1A X r 0·2) mmho y21 = 0·45 mmho Now put V1 = 0.FOURTERMINAL NETWORKS equivalent circuit for the available parameters. Example 2. Converting to admittances we have 0·25 mmho in parallel with [0·5 mmho in series with (1 mmho in parallel with 1 mmho)] /1 I 0·5(1 r 1) = V1 v2 =o = 0 ' 25 r 0·5 r 1 r 1 yu = 0·65 mmho yu Also the current in the short circuit is . we must Iet either V1 or Vz be zero by assuming a short circuit across the required pair of terminals. First Iet Vz = 0.14 each form calculate the current gain when the network is loaded with an impedance of 1 kQ. Figure 2. so h y21 and = 0·25Vl 1 I =h V1 v2=o 0·5 X V1 [ 2. writing the equations for the required parameters and solving by conventional methods.5 mmho y22 81 .
This eliminates the y21 generator. J. 0·85 0 65 mmho mmho Figure 2.14 To find the g parameters we must Iet either V1 or ]z be zero by either short circuiting the input terminals or open circuiting the output terminals. let V1 = 0. Thus Vz = (0·45 X I03Vl 0·85 X I03 v2 1 and V1 12=0 = g21 = 0·53 We can now express the y12 generator current in terms of V1.29.FUNDAMENTALS OF NETWORK ANALYSIS Also lt = 0·25V1 + y12 = 1v:12 I V1=0 vl X 1·5] 0·5 [ 11 + 1·5 X BA = 0·45 mmho The required equivalent circuit is shown in Figure 2. Let [z = 0.0·45 X 103 X 0·53V1 = gu = (0·65.0·258) X 103 mho = 0·412mmho To obtain g12 and gzz.29. and write an equation for ft. Now gz2 = Vzl [z Vl=O = 0·85 82 1 X 103 = 1·178 kQ . U oder these conditions the whole of the y21 generator current must fl.ow in the admittance yzz. Solution for Example 2. ft = 0·65 and hl  Vt 12=0 X 1Q3Vl.
At= Working in mmho. Series input. input and output impedances or admittances may be found using whichever parameters are available.FOURTERMINAL NETWORKS Since we have a short circuit on the input terminals all the y12 generator current flows as fi. 83 .49 and 2. First using y parameters Current gain. Parallel input and output. At= 0·65(1 yu(y22 y 21 YL YL) . series output.0·45 X 0·45 === 0·45 Now using equation 2.55 g21 + ZL)  At=~~~~ gu(g22 . parallel output.55.0·412 X I03(1 g21g12 + 0·53 1178) X 103 + 10·53 ~~~~~~~ X 0·53 === 0·45 Thus properties of fourterminal networks such as voltage and current gain. g12 = hh I (0·45 X 103) X 0. Interconnection of Fourterminal Networks We shall now see how these parameters are of use when two or more fourtermina1 networks are interconnected in various configura· tions. We shall see that if the individual networks are described by the appropriate parameters. These configurations are Series input and output. and Parallel input. the combined network parameters will be the sum of the separate network parameters.85 X1 10_3 V!=O = = 0·53 To find the current gain using each set of parameters we can apply formulae from equations 2.y21y12 + 0·45 X 1 + 0·85).
Example 2. Z22 3 Q. and the overall Z parameters are given by the sum of the individual Z parameters. First we must find the Z parameters of the subnetwork shown in Figure 2. I. 2 Network A tVz' Z parameters 12' I{' Network 8 V.FUNDAMENTALS OF NETWORK ANALYSIS Figure 2. series output.3la.''f I2 Z parameters jvi' Figure 2. and z21 = z12 2 n is connected in series with a 4 n resistor as shown in Figure 2. I . Network A has Z' parameters and network B has Z" parameters. 1 I1 V{~ v.15. Two fourterminal networks connected in series Inspection of the circuit shows that = h' = h" 12 = 12' = 12" h and + V1" = hZn' + hZu' + hZn" + hZ12" V2 = V2' + V2" = hZ21' + hZ22' + hZ21" + hZ22" Also V1 = V1' and Collecting terms + Zn") + h(Zu' + Z12 ") V2 = h(Z21' + Z21 ") + h(Z22' + Z22 ") V1 = h(Zn' But these are the equations for the combined network.30. A network having Z parameters Zn 5 n.3lb. Determine the overall Z parameters and hence calculate the output impedance if the network is supplied from a source of internal impedance 3 n.30 shows two networks interconnected series input. 84 .
Z21Z12 36 Zout = z22 . parallel output.Zn + Zs = 7 .35. Network A has y' parameters. by inspection Zu= Z22 = 40 Z12 =Vtl  h h0 Z12 = 40 but and since the network is symmetrical Z21 Now the overall Z parameters are: +4 = 2+4 = = 4 0.32 shows two networks connected parallel input. With this connection we can see that: Also I' and h = + h" = /2' + h" = = ft' + V2y12' + Vtyu" + V2y12" Vty21' + V2y22' + V1y21" + V2y22" Vtyn' 85 . apply equation 2.FOURTERMINAL NETWORKS (b) (a) Figure 2.15 Since these are open circuit parameters.9 + 3 0 =40 Parallel Parallel Figure 2.31. Circuit for Example 2. + 4 = 60 3 +4 = 7o Zn= 5 90 Z12 = 2 Z21 = 60 z22 = To find the output impedance. and network B has y" parameters.
Series Parallel Figure 2. Two fourterminal networks connected series input. J Figure 2.33 shows two networks connected series input and parallel output. In this case fi =h' =hw 86 . and the overall y parameters are given by the sums of the individual y parameters.32.FUNDAMENTALS OF NETWORK ANALYSIS Figure 2. Two fourterminal networks connected in parallel so collecting terms and h = V1(yu' h = V1(Y21' + yu") + V2(Y12' + Y12") + y21 ") + V2(y22' + y22 ") But these are the y parameter equations for the combined network.33. parallel output Network A has h' parameters and network B has h" parameters.
16. These are the fundamentals without which he cannot proceed to the topics to be discussed throughout the book. Ans. are equally essential. Parallel Series This is exactly the reverse connection for the previous case. The theorems in Section 3. particularly those of Thevenin and N orton. YT Y. No further examples will be given at this stage since practical applications of this work will not becorne apparent until later chapters.1 with the circuit rnodified by interchanging the 4 mho conductance with the 1 rnho conductance. EXAMPLES Example 2. The derivation of this is left to the reader as further practice. 87 . io Ht. v' i v. Repeat Exarnple 2. SUMMARY Summarizing the work of this chapter. mho. and provided the g parameters for the individual networks are used then the overall parameters for the cornbined network are given by the sums of the individual g parameters. The work on interconnected networks provides an alternative approach to the solution offeedback problems but is not essential for any ofthe work in later chapters.SUMMARY Also and + vl" = hhu' + V2hl2' + hhu + V2h12 /2 = h' + h" = hh21' + V2h22' + hh21" + V2h22" vl = Vl' II 11 Collecting terrns + hu ") + V2(h12' + h12 ") !t(h21' + hz1 ") + Vz(hz2' + hz2 ") V1 = h(hu' h = Thus for the cornbined network the overall h parameters are given by the sum of the individual h parameters. the reader should ensure that he is proficient in the use of the rnethods discussed in the first and second sections. The generat background frorn Section 4 is very useful and although problems rnay be solved directly with the derived general solutions we shall also work from first principles in rnany cases. using the parameters only to construct an equivalent circuit.
and (c) the current i ' . Figure 2. Circuit for Example 2. The admittance circuit shown in Figure 2. determine v' 1.36. (b) 0·445i. Circuit for Example 2. Example 2. and (c) the valtage v' in terms af the inputvaltage v.34.36.17 (a) the input impedance v/1.19. Ans. Example 2. 0·756 V.FUNDAMENTALS OF NETWORK ANALYSIS Example 2.35 is 2mho Figure 2. (c) 0·727v. (a) 39·6 mV. Far the circuit shawn in Figure 2.35. (c) 31·2 mA. Figure 2. Ans.18 supplied fram the current source i af 0·2 A.17. (b) the valtage ratia v0 fv. (b) 0·526. (a) 1·63 Q. Ans.34. Using mesh analysis calculate the valtage v' acrass the 3 ahm resistar shown in Figure 2. Circuit for Example 2. (b) the current io in terms af the input current i.19 88 .ß.18. Calculate (a) the input valtage v.
determine the potential at the node marked X using mesh analysis. Repeat Example 2.!1. V Figure 2.EXAMPLES Example 2. (6+jO)Vt rv i2 Figure 2. 0·25. 2·2 V. If the circuit shown in Figure 2. and using nodal analysis.22 Example 2.24.21.9 is modified by changing the centre brauch to (2 . and by reversing the connections to the 5 V generator.20 Example 2. I.23.37.38. calculate the resulting current from the 6 V generator.j2) n. Example 2. Circuit for Example 2.37.21 using nodal ana1ysis. calcu1ate the current i2 in the circuit shown in Figure 2. Circuit for Example 2. voltage sources to current sources.19 by converting impedances to admittances. Example 2.!1. Using mesh analysis. 1 L 7o 18' A. Ans. For the circuit shown in Figure 2.20.22. Ans. 8 + j26 Ans. 4 89 . mA.38.25. Repeat Example 2. Repeat Examp1e 2. 37 j2k.20 using nodal analysis. Example 2.
FUNDAMENTALS OF NETWORK ANALYSIS Example 2. and determine the node voltage V2 using nodal analysis. Repeat Example 2.28. to be connected to these termirrals suchthat the current supplied to it is 1 A .26 admittancecurrent generator form. Example 2.27.21 using Thevenin's theorem. By repeated applications of Thevenin's theorem. Figure 2.. Convert the circuit shown in Figure 2..5~]R.. Example 2. Example 2. determine the components of the equivalent Thevenin generator seen at the termirrals T.39 to the Figure 2. Repeat Example 2.39. 0·737 L 121 ° 26'. Circuit for Example 2..26. 1·35 Q.40. T' on the circuit shown in Figure 2. Example 2. 90 .29. Ans. Circuit for Example 2. Repeat Example 2.40. Ans.21 using the Superposition theorem..3 using Thevenin's theorem.30.28 Hence calculate the value of R 1. T .
Determine the h parameters of the network shown in Figure 2.32. Ans. Using Norton's theorem. 1·29. Ans. Repeat Example 2.33.34.28 using Norton's theorem. 1·9 X I02. Example 2.20 is modified by interchanging the 3 Q and the I Q branches. calculate the output current io in terms of the source current i. hl2 = h21 = u. 73 .35.43. h22 H mho. Circuit for Example 2. Figure 2. determine the new z parameters. Example 2.42. Circuit for Example 2. 4. The circuit shown in Figure 2. 14·3 Q. zu i n. Ans. Example 2.42. Figure 2. Zln lJ!l n.umho.41. hu ~l n. Ans. Repeat Example 2.33 of 106 rad/sec. Z22 3t n.43.26 using Thevenin's theorem.36. Circuit for Example 2.0. Determine the h parameters for the circuit in Figure 2.41 is the equivalent circuit for part of an amplifier to be used at an angular frequency 1 OOOpF 4kfl. 0·44 L 2° 30'i.EXAMPLES Example 2. 91 .31.36 Example 2. Z21 = Zl2 t n. Hence find the input impedance V1/lr when the output is loaded with 4 n.35 Figure 2. Example 2. If the circuit shown in Figure 2.
31. 5 000 n. Ans. h21 75. 125. is loaded with 4 kQ and driven by a source es of internal impedance 600 Q. Determine (a) the y parameters. . The T network shown in Figure 2. Calculate the value of es if the Ioad voltage is to be 3 V. Example 2. h12 103 . 37 n. 393 Q. determine the terminal input and output impedances and the voltage gain V2/ V1.38 hence calculate the input impedance v1/ h and current gain i2/ h when it is loaded with 2 kil. 375 X 103. Z12 10 n. h22 Example 2. Ans.41. 50 X 106. h21 50. 40 mV. 2 X }04.umho. Repeat Example 2.umho.37. is connected in series with a I 00 n resistor as in Figure 2 . Circuit for Example 2. Z21 92 .2· 3.40. (b) from the results of (a). Z22 500 n.38 using g parameters. 600. Example 2. If the combination is loaded with I 000 Q and supplied from a source of impedance 200 n. A network having hu 1 000 !2. Z1n 1·25 kQ. 5 000. y. A fourterminal network having Zn 100 n.44. 5.FUNDAMENTALS OF NETWORK ANALYSIS Example 2. Example 2. Ans. 2·39 kQ. (c) from the results of (b) the z parameters. A certain active device has the following h parameters: hn 500 Q. 45 X 10 3.44 represents an active device. 2 X 103. the g parameters. Find also the overall output impedance of the circuit including the Ioad. Z1n. 3 000.38.39. 5·7. z. 1 944 Q. Determine the y parameters for the device and Figure 2.umho. 8 X 103. g. 0·04. 2 X 106. A 1 54. 23 . h22 200 . 200 . Checkthese results by finding the input impedance in each case when loaded with 2 kil. Ans. 0·15. h12 103 .
(Networks in parallel. h12 I0. h21 250.43 networks. Figure 2.46 shows two interconnected fourterminal ·. 1 2000 Jj 1 500fi I I I I I I I I  I il 50fi ~~ 10on I I I I L~ Figure 2.45. 9·8 mV.45 is known to have 100kfi 2kfi Figure 2. 93 .42 hu 2 000 0. 157.) Ans. Compare the current ratio lz/h with and without the 100 kO resistor connected as shown.4 . Circuit for Example 2. Example 2.EXAMPLES Example 2. Circuit for Example 2. hzz 300 p.42. 38.46. Calculate the appropriate parameters for each network and hence find the voltage Vo. The network shown in Figure 2.mho.43. Ans.
(c) A.1.3 LOW FREQUENCY. Iet us first consider a diode having the characteristics shown in Figure 3. i. (b) Diode circuit for forward and reverse bias. (a) Junction diode characteristics. First consider the circuit shown in Figure 3. the diode is forward biased.$"·~:~ 6 I 200mV 1 (ft Al (c) (a) Figure 3. If the switch is in position B.e.C. circuit with forward biased diode In order that we may understand the idea of an equivalent circuit.2 . SMALL SIGNAL EQUIVALENT CIRCUITS FOR VALVES AND TRANSISTORS In Chapter 1 we found that the a..lb. and an alternative approach is desirable. the first 94 .1a. This approach is to replace the active device by an equivalent circuit and then to analyse the resulting arrangement using the network methods discussed in Chapter 2.c.4 R 200 300 . For more complex circuits these methods become exceedingly difficult and time consuming. operation of valve and transistor circuits could be investigated by graphical methods. +J(mA) V (vol t s) 20 10 5  +V (mV) .
on the other hand. This circuit will impose a load line on the characteristics as shown. If the a. the load line will move between the two dotted lines shown. Note the required condition for the use of such an equivalent circuit is that: Rr ~ R ~ Rr where Rr and Rr are the reverse and forward bias diode resistance respectively. supp1y voltage is 200 mV.c.lc.c. Thus in this circuit our diode equivalent could be a switch. Here the diode resistance is given by V 20 ==20MQ R I06 Now this value is very much greater than R and will determine the circuit current. giving an operating point of 110 mV. Here suppose R is 10 Q and the d.c.INTRODUCTION quadrant on the characteristics. An alternative situation is shown in Figure 3. as in Figure 3. Also the diode resistance is negligible compared with R which therefore determines the circuit current. resistance which is 95 .c.C. (a) D. Diode equivalent circuits. The approximate resistance of the diode is given by V 200 X 103 I = 20 X I03 = 10 Q If the resistance of R is much greater than this.2a. the switch is moved to A. generator has a peak value of 20 m V. is approximately zero. the Variation in total series resistance due to the nonlinearity ofthe characteristics will be negligible. lf. the diode is now reverse biased and operates in the third quadrant of the characteristics. 0·3 ftA. say 300 n. (b) Small signal a. Thus the a.2. for low frequencies and (c) for high frequencies for the opposite case. 7·5 mA. open for an applied voltage of one polarity and closed + ·~ 7[1 ~ + (a) (b) l 7[1 0·01 JLH (c) Figure 3. this current. By comparison with the first case when the circuit current was 20 mA. Over the range of resulting operating points the characteristic is approximately a straight line.
LOW FREQUENCY.3 shows a set of linear hiVAK or anode characteristics and a corresponding hiVaK or mutual characteristic. A further term in the chapter heading is low frequency.c. operating point. The inverse slope of the anode characteristics AC/ABis the first constant ra. SMALL SIGNAL EQUIVALENT CIRCUITS given by Vacliac can be obtained from the slope of the characteristic at the d.3. The slope of the mutual characteristic EO I DO is the 96 . Note that theseadditional components are present at all frequencies but their effect may be neglected at frequencies where Xe~ 7 0 and XL~70. bias current is 7 mA. These are known as ra. the anode slope resistance. Figure 3. operating point Q. voltage and current variations are sufficiently small so that over the operating region the characteristics can be assumed linear. and provided the a.c.c. In considering valve and transistor equivalents in this chapter we shall ignore the effects of shunt capacitance and series inductance which will be considered in a later chapter. A Ioad line of slope . At this point the value of the resistance is 105 X 103 15 x 103 = 7 0 Thus for this particular case the a.c. and p the amplification factor. Small signal for one condition may mean a few millivolts while another willlead to a reasonable approximation with signals of hundreds of volts amplitude.c. equivalent circuit is a 7 0 resistor. series inductance. gm the mutual conductance or transconductance.1IRL is shown passing through the d . We can call the 7 0 resistor a small signal equivalent circuit for the diode. At very high frequency the equivalent circuit becomes that shown in Figure 3. signal is sufficiently small so that the characteristic may be assumed linear. This is necessary since a diode by its construction will also have shunt capacitance and to a very small degree.c. The relationships between the various electrode voltages and currents can now be specified in terms of valve 'constants'.2c. provided the d. First let us imagine we have a valve with linear characteristics as shown in Figure 3. VALVE EQUIVALENT CIRCUITS To find a small signal equivalent circuit for a valve we must consider the characteristics to be linear over the operating region. The term small signal implies that the a.
AB .VALVE EQUIVALENT CIRCUITS 0 Anode characteristics _. i.2) r _ a  g and ra. is not constant. =~Vgk Ma ~ O (3.4) ~Ia = QB. ) h ( QB and ~ Vgk = gm mutual c aractenshc and AC AB= ra. X ~Vgk ~Vak I = f1..Load line 0 0 Figure 3. llVak~o m.~Vgk 6.c.3.~Ia I (3. a change of ~ Vgk on Ioad results in ~Ia = QA. But (3.. ~Ia = gm~ Vgk 97 ~Vak Ta. = ~Vgk AB Rewriting equation 3. (Note this is not a geometrical relationship on Figure 3. 'Ideal' triode characteristics for development of a.3) To find an equivalent circuit in terms of these constants and RL only we must find an expression for ~ Vak/ ~ Vgk on Ioad. equivalent circuits mutual conductance gm.. .4. when Ia.e.1) .3.) These three constants are related since ~Vak I (3.V8k ~ o ~Vak =~Ia ~Ia. From the graph. The ratio of change in anode voltage to change in grid voltage for constant anode current FQ/ Vgk is !' the amplification factor.. X gm ~Ia.
Vgk Ö./aRL = Ö.Vak Ö.Vgk = f1./a ( 1 A and u/a gmÖ. (b) Constant current equivalent 98 .Vgk = "'71 + RL fal But = Ö. SMALL SIGNAL EQUIVALENT CIRCUITS Ö. If direction is taken into account: Voltage gain Av = fl.Vgk Putting gmra Voltage gain = Ö. result in the voltage gain given by equation 3·5 or 3·6. (a) Constant valtage equivalent.. I a 1 I I I I I I Vs ___ j t.1: ra gmraRLÖ.LOW FREQUENCY. when loaded with a resistor RL.6) Any equivalent circuit for a valve must._____.r.4. fa + RL ttRL = ra + R L Note this equation does not show the expected phase reversal.Vak ß/a on Ioad Now = RL RL ß/a = groß Vgk . Small signal equivalent circuits for loaded triode valve. This is because magnitude of changes have been considered.Ö.VgkRL R 1 + .:~~ ~ ~~v:k (b) (a) Figure 3.5) + RL or (3.RL ra (3.Vak gmÖ. r../a ra So + ~:) = gmÖ.
These two equivalent circuits satisfy the requirements and we shall consider an alternative approach for obtaining them.gmraRL v. the valtage at the grid with respect to cathode is the same as the input signal voltage v8 • Applying mesh analysis to this circuit we obtain: = ia(ra .c.5. + RL Note that putting gmra = fl makes these two results the same. As before Vgk = v8.c. changes at a d. The characteristics shown in Figure 3. fWsRL = laRL = '_.5. The alternative circuit in Figure 3. Triode characteristics for application of Thevenin's and Norton's theorems examining the characteristics with reference to a.. Vgk..= A _ .. + RL ". By Figure 3.=ra + RL (lRL ra. so Av Vo = Vs = gmva = and and Vo = Vo ':= (~ + ~J gmvara.VALVE EQUIVALENT CIRCUITS Consider the circuit shown in Figure 3.5 are those of a valve having the d.4b can be solved by nodal analysis.ra. The section enclosed by the dashed line represents the valve. operating condition indicated by point 0. Thevenin or Norton equivalent circuit for the valve.RL ra. 99 .c.fWs • Ia and Vo Thus the valtage gain = + RL) fWs ra + RL . + RL which is identical to the expression in equation 3.4a. operating point we can obtain an a.c.:.
If a circuit is open then the current is zero.. the operating point can only move along the line CD on our characteristics.6a and by comparison with Figure 3. In this case we are concerned with a. condition. 100 .3. any linear circuit may be represented by a voltage generator in series with an impedance.c. Applying Norton's theorem in the same way. The open circuit output voltage is Vak and given by Vak = . This means that the a. See equation 3. SMALL SIGNAL EQUIVALENT CIRCUITS According to Thevenin's theorem. Vak g vgkl a .1 Vgk .6..e. The voltage generator has a value equal to the open circuit output voltage..LOW FREQUENCY.= ra Ia So our Thevenin equivalent is that shown in Figure 3. voltage must be zero. the current is constant and the operating point can only move along the line AB.. so if the alternating current ia is zero. Thus to a.c. i.4a. hence the resistance to be measured is that of the anode characteristic line passing through point 0. Vgk must be zero. The minus sign arises since a positive change in VGK causes a negative change in VAK· The Thevenin series resistance is that measured with all generators suppressed. our current generator is the current that would fiow in a short circuit. g Y= . Now from equation 3./lVgk. (a) Thevenin equivalent and (b) Norton equivalent shown in Figure 3.c.4a we can see that the valve equivalent is a gm Vgk .6a while the loaded circuit equivalent is shown in Figure 3.!_ ra Vak Vgkt (a) (b) Figure 3. Triode small signal equivalent circuits.
gmVgk. for each configuration we can obtain z. input Vgk is (See equation 3. by comparison with Figure 3. but since one terminal is usually common to both input and output connections. Again. 101 . i. or y parameter equivalent circuits and other possibilities are known as T and 7T equivalents. when Vgk is zero. h. and the input and output currents are the base and collector currents. it is 1/ra. Under these circumstances the applied voltages are measured at the base and collector with respect to the emitter.fva.TRANSISTOR EQUIVALENT CIRCUITS Now the current variation ia.6b is the equivalent for the unloaded valve. Thus we can expect to find at least three different equivalent circuits representing the transistor.6b. There are in fact many more possibilities. The application of these valve equivalents to more complex circuits will be discussed in Chapter 4.) The parallel admittance with generators suppressed is given by ia.c. resulting from an a. Alternative connections also used are common base and common collector. Since the common emitter circuit is the most important we shall investigate this in full. This is because these are the parameters that can most easily be measured. h Parameter Equivalent Circuits For low frequencies the most popular equivalents are those based on the h parameter equations.4b we can see that Figure 3.2. TRANSISTOR EQUIVALENT CIRCUITS The transistor is a threeterminal device. With the common collector configuration the voltages are measured at the base and emitter with respect to collector and the currents are the base and emitter currents. g. This approach is more useful than the graphical approach used first. For common base the voltages are measured at emitter and collector with respect to base and the currents are emitter and collector currents. since it can be easily extended to the case of the transistor. it may be treated as a fourterminal network. Separate measurements may be made for each configuration or one set can be measured and the other two found by network manipulation. In practice the emitter is most frequently used as this common point and the transistor is then said to be in the common emitter configuration. The resulting Norton equivalent is shown in Figure 3.e.
SMALL SIGNAL EQUIVALENT CIRCUITS Figure 3. operating point. value. fand o for input. reverse. Now remembering the definitions for the parameters obtained in Chapter 2 we can write hte = v_be lb hre = ic 7 lb I ''ce=O I l!ce=O Remernher that all these voltages and currents are a.7 shows the transistor connected in the common emitter configuration as a fourterminal network. Similarly ib = 0 indicates that the base current is fixed at the required d.c.LOW FREQUENCY.8) In each case the e stands for emitter and the i. Thus the statement Vce = 0 means that the collector voltage is fixed at its d.7. quantities measured about a particular d. 102 .c.c. 'BlackBox' representation of transistor in the common emitter configuration Now replacing the variables by those specifically applicable to this transistor configuration Vbe ic = = + Vceh12 ibh21 + Vceh22 ibhll Finally since these parameters refer only to a common emitter connected transistor we can introduce a new notation for the parameters (3.7) Vbe = ibhte + Vcehre ic = ibhre + Vcehoe (3. forward and output respectively. For common base and collector these would become htb and htc etc. r. First writing the general h parameter equations: V1 = /2 = + V2h12 hh21 + V2h22 hhu Figure 3.c. Ievel.
If ib is zero we can move only along the constant IB line on Figure 3. This parameter is very small. (a) Input and (b) Output characteristics characteristic with the output short circuit to a.8b. it is the short circuit input resistance. and as we shall see in later chapters. Typical values are of the order of 1 000 Q.8a. we can see that is the direct ratio of ic and ib when Vce is maintained at a constant Ievel. the output characteristic. i.8a.J Constant Vce (vce =0) Vce Constant la (ib = 0) ic Vbe t I 10 I Constant lib Constant Vce Vae Ia (ib=Ol Vce ( Vce= 0) (a) (b) Figure 3. The relative Vce and Vbe under these conditions is shown and hre is referred to as the reverse transfer parameter.8.c.e. Transistor characteristics for determination of h parameters. may frequently be neglected.TRANSISTOR EQUIVALENT CIRCUITS Weshall now relate these parameters to the approximate characteristics as shown in Figure 3. typically I03.8. 103 hre .. h1e is therefore the slope of the input la . hre ic = : I le Vce = O Referring to Figure 3. Now considering each of our definitions in turn: hle = V~e lb I Vce=O If Vce is zero the operating point can move only along the input characteristic on Figure 3.
In commercial transistors the value of hre will be in the range 10500 depending upon type and application.10. hre.umhos.9 but with appropriate notation. But it is usually more convenient to convert. In either case the resulting h parameter equivalent circuit will have the same form as that shown in Figure 3 . The hybrid parameter transistor equivalent circuit for common emitter connection Common collector and common base parameters may be obtained in the same way from the appropriate sets of characteristics if these are available.8b we can see that this represents the slope of the output characteristic. SMALL SIGNAL EQUIVALENT CIRCUITS This is the most important parameter and is known as the short circuit current gain. The T Equivalent Circuit Another equivalent circuit often encountered is the equivalent T. To complete this section we should Iook at the resulting h parameter equivalent circuit as shown in Figure 3. an output resistance of 10 kQ. Using conventional currents.9. These manipulations will be shown in Chapter 5. This circuit is based on the common base configuration and is sometimes thought to represent the physical structure ofthe transistor. hoe is therefore an admittance and it is known as the open circuit output admittance. b e Figure 3. using network methods from the available set of parameters to those required. i.9. both lB and Ic flow out of the transistor (pnp).e. A typical value for hoe is 100 . The ratio of the two. 104 . It is shown in Figure 3. is therefore positive. Looking at Figure 3. so changes in these directions are both negative when considering the fourterminal convention.LOW FREQUENCY.
11.11.TRANSISTOR EQUIVALENT CIRCUITS c e . The T equivalent for a transistor in the common base configuration Typical values for these parameters are re 50.10). (c) From fundamentals. ib b  otie rc 8 ic ib c b (a) (b) b c (e) Figure 3. Thus ocierc generator may be split into two components.10. Conversion from common base T to common emitter T equivalent circuits (a) Shows the T circuit turned so that the emitter is common. ie may be expressed as sum of ic and ib (equation 1. 105 . Tb I 000 Q.. Figure 3. rc I MQ. and oc0·99. (b) Thevenin's theorem has been applied to branch AB. Conversion to the common emitter form is achieved by the steps shown in Figure 3.
(b) Open circuit input By writing nodal equations we can solve for the voltage V in terms of ib and thus put: ib + oc'ib = V ( re1 + rcoc') (Since the sfc puts re and rc/oc' in parallel.12. the conversion is completed oc oc (equation 1.m. SMALL SIGNAL EQUIVALENT CIRCUITS (d) Since ic fiows in brauch AB. Conversion to the h parameter and vice versa may be achieved by applying the h parameter definitions to the T equivalent circuit. it is convenient to reverse its direction..) = ('c + oc're) rerc V ib rerc(l + oc') (rc + oc're) Now from the typical values we can see that rc ~ oc're.oc) ~ !.17).f. First for hte and hre we must let Vce be zero. An example of this use of the circuit will be given in Chapter 5. Tc !'•• /rx' r. Knowledge of typical values allows valid approximations leading to simple conversion factors. (e) Norton's theorem applied tobrauch AB. 106 . the direction of the current generator. and with it. Also putting 1 oc = oc' and (1 . Circuits for determining h parameters from T parameters. The minus sign gives the correct polarity for this e. (a) Short circuit output.12a. the ocicrc generator may be rep1aced by a resistor of ocrc n. ib je e (a) (b) Figure 3.oc).LOW FREQUENCY. The required circuit is shown in Figure 3. and the two resulting series resistors have been added to make rc(1 . (f) Since ib is the input current.
(3. hre Tc==hoe hoe (3.12) Equations 3.10. rx. Since ib + must comp1ete its circuit to the input terminals. 3. 3.' = hre (3.' Tc I __T_e_~ ib=O  Tc Te+.9) + The current (1 rx.')ib fiows in Te.12 give the conversion factors for the h parameters in terms of the T parameters. By inspection hoe =~ Vce Also _ h re Vbe Vce I ib=O 1 ..11 and 3. 107 .' }>Te and we can assume that (1 rx. From equation 3. the short circuit ic must be the remainder.  rx.'re Tc (3. h re and ic =: I lb Vce=O .13) From equation 3.14) hreTc hrehre hre Te===rx. = rx.16) From equation 3.')ib fiows into Te and Tc/rx. rx.=== Te(1 lb + rx.12b. rx.'.10) For hoe and hre the condition is open circuit input. The opposite conversions can be found using these results.' in parallel. This makes ib zero and therefore eliminates the current generator from our T equivalent. hoe===oc' = Tc Te+ rx.11.11) (3.9.10. But Tc/rx.' hrehoe hoe (3.TRANSISTOR EQUIVALENT CJRCUITS So neglecting oc'Te and cancelling the Tc terms we obtain: V .') (3. The resu1t is shown in FiguTe 3..12.
A triode valve has the characteristics given in the table below. 108 . At some time in the future devices other than valves or transistors may coine into general use. Manufacturer's published data quotes typical values for the h parameter as hte 1· 3 kO. SMALL SIGNAL EQUIVALENT CIRCUITS To compare these values we shall consider a common general purpose transistor. hre 90. hre 5 x I0. 1 = 90 90 rc = 125 5 X X I06 = 720 kQ I0. and further examples appear in the remaining chapters of the book. If they are fundamentally fourterminal networks it may be convenient to measure the h parameters but it is quite possible that the z. Vg 2 V or (b) RL 60 kO. These will be obtained in the same way by reference to the characteristics. y or g parameters will be more convenient.125 X 10_6 (1 + 90) = 936n ~ Other transistorsmall signal equivalent circuits are used.. 1t is to be operated with an H. The applications of equivalent circuits to complete circuit arrangements will be discussed in Chapters 4 and 5 covering valve and transistor work respectively. One such device is the field effect transistor for which manufacturers are quoting the y parameters. but only for high frequency applications and these will be considered in a later chapter.4 re = 125 X I06 = 4 Q Tb= 5 X 104 1 300. the general methods outlined in this and subsequent chapters will be applicable. In all such cases.LOW FREQUENCY. For each case. Using the equations above: (/.umhos. Where such devices are used under small signal conditions equivalent circuits will be used to represent them. and hoe 125 . EXAMPLES Example 3.T.4 .1. of 450 V and (a) RL 30 kO. the OC75. Vgk 12 V.
In each case. (With respect to hre these characteristics are not typica1 as they have been exaggerated to simplify graphical measurements. The d. VAK h (mA) for VGK 0 2 4 6 I 0 50 1100 150 1200 250 1300 350 1400 450 0 5·0 0 13·7 3·0 9·5 0 2·2 0 8 10 17·0 6·5 14·0 1·5 4·5 11·8 0 12 14 1·3 0 4·0 10·0 1·2 3·6 0 16 11 0 8·3 3·1 1·4 7·2 3·5 0·1 1·3 Ans. operating point is given by (a) VcE 4·5 V. 110).c. 36. determine the small signal h parameters. 1·5 mA.) 109 . Example 3.29·5. 24. Ic 8·8 mA. ra 31 kQ. (a) 153 V.15·8.2. Av. A transistor connected in the common emitter configuration has the input and output characteristics given in the tab1e (p. Av. ra 6·6 kf!. (b) 361 V. /B 40 A. 9·8 mA. or (b) VBE 130 mV.EXAMPLES determine the operating point. the components of the small signal equivalent circuit and hence the voltage amplification.
450 . 97 . 35. 45.umho. iB (pA) for . 5 X I03. (b) 1 200 0. (a) 470 0.1 0 o 90 18 14 22 4 2 0 8 lO 100 50 250 235 214 200 190 200 345 135 117 105 98 150 200 60 50 43 38 Ans.VcE 0 1·5 3·0 4·5 6·0 VBE(mV) 360 331 313 200 300 1 lc (mA) for lB (J1A) VcE (V) 0 40 80 120 160 200 240 I 0 0 0·2 1·2 24 3·6 4·8 6·0 76 0·2 0·2 1·4 2·8 4·1 56 6·9 84 0·4 6·0 0·2 0·2 1·85 1·5 39 29 5·7 4·3 5·85 7·6 7·25 9·45 8·65 11 ·5 1·0 .umho. 6·7 X I03 .
and oc0·97. The published data for a transistor connected in the common emitter configuration includes the following h parameters at the desired operating point. htz 5 x 104 . h22 80 . (a) 6·25 0. 0·993. rb 500 0. hrc + 1. (a) htb 30 0. 120. 244 0. rc 1 MO. hrb 5 X 104 . Ans. hu 1 000 0. hzt 120. hrb 0·97. 1·5 MO. hrc 33·3. 244 0.umho.3.umho. (b) htc 550 0.umho Determine the components of (a) the common base and (b) the common emitter equivalent T circuits. Determine the h parameters for the transistor connected in (a) the common base configuration and (b) the common collector configuration. Example 3. 12·5 kO. hob 1 . 111 . The T parameters for a transistor are given as re 15 0. hoc 33 . Ans. (b) 6·25 0.EXAMPLES Example 3.4.
Triode valve and small signal equivalent circuits electrode potentials and are hence known as small signal a. Any change in operating point will modify the values of the constants ra.c.0. The circuits developed were suitable for small changes of ~k 9m a Vgk . using this method. g g ~ fl Vgk k (c) (a) Figure 4. gm.0b 112 . we shall consider an exact method for solving any valve circuit operating within the limits of the equivalent circuits. In this chapter. This procedure is not only time consuming. simple valve circuits were solved by graphical means. The problern is further complicated by the presence of reactive components which may affect the performance at certain operating frequencies. 1 r. we shall analyse a range of amplifier circuits commonly found in electronic systems. the steps being as follows: (1) For each valve in the complete circuit. and p. operating point. . draw the appropriate equivalent circuit.c. Chapter 3 showed that valves could be represented by equivalent circuits. but becomes exceedingly complex when more than one valve is used in the circuit. at the particular d. These circuits represent the valve only. The procedure for using the equivalent circuits is always the same. equivalent circuits.4 USE OF THE VALVE EQUIVALENT CIRCUIT In Chapter 1. Either the constant voltage form Figure 4. Then.
but in some instances other current definitions lead to simpler equations.s algebraically.der should include all the details shown until he is familiar with the method. If a numerical solution only is required. circuits having most elements in series are best solved by using the constant voltage form. but the different technique involved will be discussed later in the chapter. the sense of the unknown node voltages must be chosen. but the shortest path usually Ieads to the quiekest solution.f. Vgk should now be expressed in terms of any applied voltages or signal sources together with any potentials existing across components placed between cathode and grid. In general.0c may be used and experience will show which will Iead to the simplest solution. the frequency or phase response. the reader should insert values in the equations before solution. potential at the grid. source of e. while those having more parallel components are more suited to the use of the constant current equivalent. resistance of a battery or power pack is negligible and may be regarded as a short circuit. The technique is to indicate the sense or direction of allsuch potential differences and applied e.s and e.s and proceeding from cathode to grid. In the nodal form such potentials will simply be Va or Va . add such p. (2) Taking each electrode in turn.USE OF THE VALVE EQUIVALENT CIRCUIT or the constant current form Figure 4. Ifthe constant current circuit is to be used. The d. positive H.c. indicate the sense of the unknown currents.d.i2)Z1.Vb. The re<1. Examples showing both forms are given later in this and other chapters. These will normally be Maxwell's circulating currents in the clockwise direction.m.f. connect it to earth through any component.c.T. but in the mesh form they will be in terms of the unknown currents as in iaRk or (h . Remernher the potential across an impedance is positive if measured in the opposite sense to the direction of current flow. the output voltage or power. and the input impedance. the equivalent circuit is again used. (3) If the constant voltage circuit is being used. measured with respect to cathode.c.f. These results will normally lead to the solution of the problern for information concerning the voltage amplification. or current source shown in the complete circuit. Any path between cathode and grid willlead to the correct result.m. If the amplifier output impedance is required. since the a. line is taken as earth. This 113 .m. (5) Write the mesh or nodal equations for the circuit and solve for the unknown currents or voltages using Substitution or determinant methods where necessary. (4) In either type of circuit the generator is given in terms of Vgk· This means the a.
7) = and the valtage gain. The full circuit is given in Figure 4. c. f shows the unknown current ia and the positive sense of the potentials iara. 114 . The valve has amplification factor fl and anode resistance ra.1) (Note. c. =''=. Step 4 in the procedure is to obtain Vgk· From the diagram.5) lf the output valtage is required.. ra + RL + Rk(1 + fl) (4..3) Collecting terms in ia on the right hand side of the equation and fles = ia[ra + RL + Rk(l + ft)] (4. The derivation of standard formulae may involve particular steps or even approximations which have to be memorized. d.4) (4. Example 4. this would have led to Vgk = iaRk +es and Vo = iaRL.:.. and Vo flesRI. Av = The minus sign indicates that the output valtage will be 180° out of phase with the input signal es.) Continuing with step 5: flVgk = iara + iaRL + iaRk (4.C.. Vo = iaRL. Consider a single stage anode loaded amplifier with cathode bias and no decoupling capacitor..la and the steps in drawing the equivalent circuit in b. iaRk. and iaRL due to a positive ia. if ia had been chosen to circulate in an anticlockwise direction. Note that the output valtage v0 is measured at the anode with respect to earth and that this is given by +iaRL. d and e show the connection of grid..1. cathode and anode respectively to earth..2) Substituting for Vgk ft(iaRk +es)= iara + iaRL + iaRk (4. and e.6) Vo = ftRL es ra + RL + Rk(1 + fl) (4. proceeding from cathode to grid we obtain: (4.USE OF THE VALVE EQUIVALENT CIRCUIT will usually result in a simpler solution.
(a) Triode amplifier with undecoupled cathode resistor. (b) to (/) Steps in the drawing of the equivalent circuit for (a) using the constant voltage form. (d) (e) ( f) (g) Figure 4. (g) The constant current form of equivalent circuit for (a) 115 .1.__.USE OF THE VALVE EQUIVALENT CIRCUIT HT+ (c) (b) (a) a.
ra rearrangement will lead to the result obtained using the constant voltage generator. Mesh analysis therefore resulted in the simplest solution. A triode valve having f1 of 20 and ra 30 kO is to be used as a simple cathode follower with a cathode Ioad of 5 kO. (b) (a) Figure 4. These are the simple cathode foliower or grounded anode amplifier and the grounded grid amplifier. The simple cathode follower.8) Then by putting Ya 1 = . Figure 4.Vz +es The reader should solve this for practice and obtain Vo = vl = Ya( yk + gmesYk (4. Two other series forms will be given as examples.USE OF THE VALVE EQUIVALENT CIRCUIT This circuit is essentially a series circuit.m.2.2a shows the full circuit and Figure 4. the equivalent circuit obtained is that shown in Figure 4. If the constant current form is used.1 g.2.2b the required equivalent circuit. There are now two unknowns and Vgk is given by: Vgk = .s. (a) Full circuit and (b) the small signal equivalent circuit Calculate the voltage gain and the load power if the input signal is l volt r. 116 . THE CATHODE FOLLOWER Example 4.
flVgk  Vgk + es (4. the voltage gain = Vo = 0·74 es and the power dissipated in RL is found from 20 ia2RL= ( . this is the best to use. volts.9 is simpler. In addition there will be d. and kQ 20es = ia[30 + 5(1 + 20)] Now.11) Both forms are correct. Writing the circuit equation (4.THE CA THODE FOLLOWER Using Figure 4.iara 1 + fl (4.9) or and Vgk Vgk(l =  + fl) = es  giving iara.c. Since this may be of the order of for example 5 mA resu1ting in 125 mW dissipation the load resistor should be rated at! W to allow a margin of safety. power in the load. but since result 4. Vo = iaRL since +ia.10) iara = es.c.135 X I03 )2 X 5 X 1Q3W = 0·11 mW This result gives only the a.2b (4. Vo = + 20 135 Ses = 0·74e8 V Av.RL is measured with respect to cathode and v0 is measured with respect to earth.13) Inserting values and working in mA.12) (4. 117 . power due to the direct anode current.
g Vo vr (a) ( b) Figure 4. Derive expressions for the valtage gain. the terminal input impedance. (a) Full circuit and (b) the small signal equivalent circuit Figure 4.15) ß)] (4.3b shows the complete equivalent circuit for the grounded grid amplifier shown in Figure 4.17) . but a general method of calculation of output impedance will also be introduced.3.3b Vgk = (4. and internal resistance Rs. The anode is loaded with resistance RL and the valve has amplification factor and anode resistance offland ra respectively. and the output impedance of a grounded grid amplifier driven by a source of open circuit valtage es.es)= ia(RL es(l + fl) .14) +iaRs.USE OF THE VALVE EQUIVALENT CIRCUIT THE GROUNDED GRID AMPLIFIER Example 4.16) (4. This may be used directly to determine the valtage amplfication v0 fes or the terminal valtage amplification v0 /v' and the terminal input impedance v' /ia.3.es writing mesh equation rearranging es. la = ia[ra + = ra + RL + es(l + fl) RL + Rs(l 118 + Rs R s(l + + fl) + ra) (4.fl( +iaRs. The grounded grid amplifier. The output impedance may be deduced from the first of these results. Considering Figure 4.3a.
.4). connected to a Ioad RL (Figure 4.19) Ia and since z _ ::: _ ln  ia v' =es.18) giving the overall voltage amplification Vo (1 + p. Circuit for discussion on output impedance generator of v volts and internal \mpedance Z..CALCULA TJON OF OUTPUT IMPEDANCE and (4..:::::(4.:. Consider a simple Figure 4.) + RL 119 . The output voltage is given by vRL/(Z + RL). V This results in a value for the terminal voltage amplification Avt = (1 + p.20) If it is required the terminal voltage gain Avt may be obtained by calculating v' in terms of e8 using v' = iaZtn.22) ra + Rs(1 + p.)RL =~~~ es ra + RL + Rs(l + p.( 1 + p.) _ R s + RL + p...) (4.iaRs R _ ra + RL + Rs(I {1 s  ra Ztn = .21) CALCULATl ON OF OUTPUT IMPEDANCE The output impedance of an amplifier is the effective internal impedance of the equivalent voltage generator.)RL Vo = ::':~.20) (4.) + p.:.) The terminal input impedance v' Ztn =: (4. es(l + p.. and putting Vo Avt = . Now consider the expression for the output voltage of the grounded grid amplifier.4...)RL ra + RL (4.
and to add it in parallel to the calculated result.USE OF THE VALVE EQUIVALENT CIRCUIT which would be the result for a generator of e8 (1 internal impedance ra + fl) volts and + Rs(l + fl) (4.23) The output impedance of the grounded grid amplifier is thus ra + Rs(l + fl)D.m.5.e.25) (4. (See Figure 4.) I I I l ~Re g l I I I I Figure 4. replaced by their internal impedances).. sources suppressed (i. This logical approach is not always so convenient and an alternative method may be adopted.5. Equivalent circuit for determining the output impedance of the grounded grid amplifier (3) Calculate the current I due to this generator and find E/1 = Zout· In practice it is often convenient to omit RL in the calculation. (2) Connect a generator of E volts and zero internal impedance to the output terminals. The procedure for determining the output impedance is then as follows: (1) Redraw the equivalent circuit with any external generators or e.26) ./Rs) rearranging and E = /[ra Zout = /(ra + Rs) + Rs(l + fl)] E =I= ra + Rs(l + fl) 120 (4. Following this procedure for the grounded grid amplifier: Since the external generator has been suppressed Vgk = IRs and writing the mesh equation E + fl(.24) (4.f.
Calculate the input impedance and the output voltage if an r.4 bypass capacitor may be assumed to have negligible reactance at all signal frequencies. The valve has .THE MODIFIED CATHODE FOLLOWER which is the same result (4.6a isthat of a cathode follower. all 121 . Full and equivalent circuits of the modified cathode foliower circuit in Example 4. signal of two volts is applied between grid and earth. + Rs(l + . and a two valve circuit known as a long tailed pair. R5 (1 . The direction of the unknown mesh currents are chosen to make the output voltage v0 = +(ia + i 8)Rk. The equivalent circuit for the problern is shown in Figure 4.27) This method of calculating the output impedance of a circuit is completely generat and will be used in subsequent chapters.23) as was obtained using the logical approach. THE MODIFIED CATHODE FOLLOWER Example 4. The 1 kQ resistor provides the correct d.u 50 and ra 10 kQ.s.6b.m. bias and the (a) (b) Figure 4. Taking all resistors in kQ. designed to give a high resistive input impedance.u) (4. Examples involving the solution of two mesh currents will now be considered.4.6.u)] RL + ra.c. The circuit shown in Figure 4. The overall output impedance including the effect of RL will therefore be Zout = + + RL[ra. These are a form of cathode follower.
31 .m.31. = .31) 25 OOOis + 15ia (in equation 4.15 . Vo = 1·886 volts THE LONG TAILED PAIR Example 4.ls = .s in volts. VI = 505is + 5ia (4.28) (4.5. The long tailed pair or cathode coupled amplifier shown in Figure 4.30 VI Zin Also • 18 = 500is VI = :.32) substituting in equation 4.29) Rearranging and collecting terms.30 VI therefore • Ia = 5VI ( 505 = 8 834 Vt + 5ia ) 1 . we have: Vgk = 500is [or 5(ia + is) + vt] writing the mesh equations: + 5(ia + is) 500ttis = 10ia + 5(ia + is) = 500is VI (4.f. Show that the voltage between the two anodes is proportional to the difference between the two input signals e 1 and e2 • Assurne that the two valves and their loads are identical. 25 000 . therefore Vo = 0·189vi X 5 = 0·943vl and since v1 is 2 volts.8505 834 = 0·189vi but ia ~ is.) Using substitution methods. +5is has been neglected.3 .USE OF THE VALVE EQUIVALENT CIRCUIT currents in mA. 5 000 . and all potential differences and e. 122 .7a is tobe used as a difference amplifier.ls Ia (4.30) 0 = (4.= /s + 5 ( 5 ~ i + i") = 8 834iR 8 8·834 MQ VI = 8 834 mA Substitute in equation 4. from equation 4.
40) These two expressions have the same denominator.11e1 .Rk2(1 + /1)2 i 2 = 11e2[ra + RL + Rk(l + ß)] + ßeiRk(l + f1) [ra + RL + Rk(l + /1)] 2 .34) Vgk2 = (h + i2)Rk + e2 a1 (a) (b) Figure 4.3 9) (4.7.38) = hRk(l + ß) + i2[ra + RL + Rk(l + ß)] The solution requires an expression for the voltage between the two anodes.37) (4.35) (4.f1Rk(h .Rk2(1 + /1)2 (4 . Full and equivalent circuits for the long tailed pair amplifier discussed in Example 4.11e2 .7b. we find (4.5 The mesh equations may then be written: . If substitution methods were used. Solution is best achieved by the use of determinants. a term +Rk2(1 + 11)2 will appear.THE LONG TAILED PAIR Taking h and i2 in the directions shown in Figure 4. By inspection this is given by Vo = i1RL  i2RL = RL(h  i2) Thus the equations above must be solved for h and i2.33) Vgkl = (h + i2)Rk + e1 and (4. and if the first term of this is expanded.f1Rk(h + i2) = h(ra + RL + Rk) + i2Rk + i2) = hRk + i2(ra + RL + Rk) (4. the equivalent circuit. On 123 .36) rearranging: 11e1 11e2 = h[ra + RL + Rk(l + ß)] + i2Rk(l + ß) (4. h and = 11el[ra + RL + Rk(l + ß)] + 11e2Rk(l + ß) [ra + RL + Rk(l + f1)] 2 . very unwieldy expressions would appear.
ei)(l + . The cathode bias circuit consists of a I kO resistor 124 . at low frequencies. depends on the magnitude of the particular reactance relative to its series or parallel resistive component. .USE OF THE VALVE EQUIVALENT CIRCUIT collecting terms. A triode valve having . This problern would have been much less cumbersome if it had been solved numerically. CIRCUITS CONTAINING REACTANCE All the circuits discussed so far in this chapter have been nonfrequencyconscious. Similarly.u) 11 .u 39 and ra 10 kO is loaded with RL 20 kO. [ra + RL + Rk(l + . high or medium. Usually the effect of these can be neglected over certain frequency ranges. but over other ranges both gain and phase shift will be modified.u)] .12 = .u(e2 . and stray reactances due to wiring etc.41) From this result. The student should try this for hirnself using typical values for the components and valve constants.12 = . those for minimizing negative feedback. Effect of the Bias Decoupling Capacitor Example 4.u)](e2. and the second term in the denominator will cancel. series inductors may be neglected.u(e1. The question of whether a frequency is low. this.U (ra + RL)2 + 2(ra + RL)Rk(l + .u)] + RL + 2Rk(l + . In practice most circuits contain those reactive components necessary for interstage coupling. and at high frequencies shunt inductors are ignored.ei)[ra (ra RL)[ra Vo = (h  + (4. Two examples involving reactive components will be given.e2)RL i2)RL = + RL ra . then at low frequencies shunt or parallel capacitors can be ignored.6.e1) + . The first demonstrates the effect ofthe bias decoupling capacitor. Initially the full equivalent circuit should be drawn.uRk(e2. it can be seen that this circuit behaves as a single stage amplifier with an input of (e1 . .u) rearranging numerator and denominator Thus + RL + 2Rk(l + . and a Ioad RL. Detailed working will show how the decoupling capacitor may be regarded as a shortcircuit at all frequencies above a certain Ievel. Therefore .e2) volts. 11 . The second example will give a detailed analysis of resistance capacity interstage coupling. and at high frequencies series capacitors can be ignored.
42) 1 = wC = 1 kil So. Av = ra At 159 Hz. flRL + RL + Zk(1 + p) Xe (4.j 2 2 Therefore from equation 4. working in kil: zk = j x 1 = j(1 + j) = 1. connection between grid and earth.8a shows the complete circuit.CIRCUITS CONTAINING REACTANCE in parallel with a 1 flF capacitor.1.j 1. The equivalent circuit in (b) (a) Figure 4.1 except Rk has been replaced by Zk.e.6 Figure 4. 1 590Hz and 15·9 kHz.j) = 5 . Circuits for Example 4. i. Figure 4. This circuit is identica1 to that obtained in Example 4.7. equation 4.c. Sketch the gain and phase response over the frequency range 100 Hz to 20 kHz.42 39 X 20 39 X 2 78(5 + j2) Av = 10 + 20 + ~/(1 .j2 = 25 + 4 = 2·69(5 + j2) = 13·45 .8.8b shows the bias circuit as Zk. Calcu1ate the voltage gain at 159Hz. We can therefore use the result obtained in Example 4. Rg is necessary to provide a d. where Zk is the parallel combination of 1 flF and 1 kil.j5·38 = y(l3·452 + 5·382)/tanlfJ:s in the third quadrant Av = 14·53 L 202° At I 590Hz Xe= 1000 125 .
9·25 + 0·16 = 8·3(3·04 + j0·4) = and = = 25·2 . To sketch the required frequency response curves. the voltage gain tends rapidly towards a value where Zk(l + p.) ~ ra + RL.7. In coupling circuits this does not necessarily apply and in practice a capacitor may be neglected when Xe > 5R for parallel combinations (Z === R). Gain and phase responses for circuits in Figure 4.78(3·04 + j0·4) . the magnitude of Av and the phase shift are best displayed when plotted to a base of log frequency as shown in Figure 4.j0·4 .. Derive generat expressions for the voltage gain of an audio frequency amplifier using resistance capacity coupling to 126 . THE RESISTANCE CAPACITANCE COUPLED AMPLIFIER Example 4. In generat the effect of the bias circuit may be neglected if Xe < Rk/10.USE OF THE VALVE EQUIVALENT CIRCUIT Therefore j0·1 j0·1(1 +j0·1) . Zk=1j0·1= 1·01 =::=:0·01JO·l 39 X 20 Av = 10 + 20 + 40(0·01 + j0·1) 78 3·04 .9.8 In this problem. ~ 160 100 1 00010 000 100 1000 10 000 Hz Hz (a) ( b) Figure 4. and to Xe < R/5 for series combinations (Z === R).:. 30 .:.9a and b. low frequencies would be less than say 2 kHz and high frequencies those above 2 kHz. and 39 X 20 0 A V = 20 + 10 = 26 L 180 Thus for this circuit.j3·3 = 25·4 L 187° 36' It can be seen that as frequency increases. the reactive component is effectively multiplied by (1 + fl) due to the valve action.
The complete circuit is shown in Figure 4.10a. Rg =50 kO. connection to earth. but capacitance between connecting wires and earth is included in this component. If the factor of 10 used in Example 4. ra = 10 kO. The second valve will normally have cathode bias and the grid must have a d. If Rk X.c. certain components may be neglected as either opencircuit or shortcircuit. The effects of interelectrode capacitance will be considered in Chapter 9. The first step then. the high positive voltage at the anode of the first valve must be blocked or isolated. This is provided by Rg across which H. If the output voltage v0 is to be applied to the grid of a second valve. Ce= 0·1 fkF.10b could be solved by mesh analysis. is to simplify this circuit by examining the component values over certain frequency ranges.5. Sketch the gain and phase responses and calculate the 3 db frequencies. The full equivalent circuit shown in Figure 4. =  10 10 then wC = 50000 1 106 w = 500 C = 50 000 = 20 rads/sec e andf = 3·18 Hz.10. but the expressions obtained would be cumbersome and difficult to use. Ce the coupling capacitor is included for this purpose. given that: !' = 20. Cs represents the stray capacitance which is principally caused by the interelectrode capacitance of the second valve. 127 . (1) The cathode bias circuit. is applied.THE RESISTANCE CAPACITANCE COUPLED AMPLIFIER a second stage. (b) (a) Figure 4. output voltage will be developed.T. and Cs = 100 pF.c. RL = 20 kO. Full and equivalent circuits for valve amplifier with RC coupling network the a.
The cathode bias circuit is an effective short circuit to a. low frequencies. If 1 CB = lORg W 1012 1 w== lORgCs And J= 5 X X 105 100 = 2 x 104 rad/sec 3180Hz But capacitive reactance increases as frequency is reduced. both capacitors and the bias circuit can be neglected. C8 can therefore be regarded as open circuit for all frequencies less than 3 180 Hz.USE OF THE VALVE EQUIVALENT CIRCUIT As shown in Example 4. The coupling circuit is in parallel with the valve which suggests the parallel or constant current form of equivalent circuit. so the coupling capacitor can be neglected at all frequencies above 318 Hz. Rg wCe = lO 107 10 w== RgCe Therefore J= 50 X 1(}3 X 0.5. If these results are examined. amplification. 128 .c. (3) The stray capacitance Cs is effectively in parallel with Rg. at all frequencies above 3·18 Hz and since this is an audio frequency amplifier this applies to the whole range. This range will be known as the medium frequency range and the voltage gain in this range as Avm· At frequencies greater than 3 180Hz only Cs need be included. it may be regarded as open circuit. and high frequencies. If the reactance is greater than lORg. Therefore if Xce is less than one tenth of Rg it can be regarded as a shortcircuit. (2) The coupling capacitor Ce is effectively in series with Rg the grid resistor. We can now draw equivalent circuits for medium frequencies. 1 = 2000radfsec 318Hz Now Xe is inversely proportional to frequency. This will be the high frequency range with a voltage gain of Avh· The range 3·18 Hz to 318Hz is the low frequency range with gain Avl· At these low frequencies only Ce need be considered. for all frequencies greater than this the cathode bias circuit has negligible effect on the a.c. it can be seen that for frequencies in the range318Hz to 3 180Hz.
11.44) gmRe Inserting numerical values in equations 4.= . and Vo = . If the parallel combination Figure 4. High frequency equivalent circuit for RC coupled amplifier 129 . The three resistors in parallel may be combined to form one resistor Re.gmesRe Rg Figure 4.= 2 mA/V ra 10 Avm = 2 X 5·89 = 11·78 In Figure 4.43 and 4.43) =++Re ra RL Rg Vgk = es.gmVgkRe = .44: 1 Re giving Therefore = 1 10 1 1 + 20 + 50 = 10+5+2 100 (4. Equivalent circuit for RC coupled amplifier at medium frequencies V0 • k Therefore medium frequency gain Avm = (4. the high frequency equivalent circuit is shown with the three resistors combined as R e.45) Re= 5·89 kQ ft 20 gm = . 1 1 I 1 (4.THE RESISTANCE CAPACITANCE COUPLED AMPLIFIER The required medium frequency equivalent circuit is shown in Figure 4.12.11.12.
the high frequency gain given by: Avh = Avh is (4. thus Re I = Xcs = wCs and I w== ReCs 1012 IOO X 1()3 X 5·89 100 (equatton .jXcs Dividing numerator and denominator by jXc8 . then Avm Avh = I +j = Avm v 2 L 45 o = 11·78 L_180° v 2 L 45 o = 8·33 L I35 0 This is known as the upper 3 db point since a voltage reduction by a factor of I/v2 is a reduction of 3 db in the Iogarithmic scale. 2. Iet Re/Xcs = I.) The upper 3 db frequency occurs then when Re/Xcs = I. (See Appendix 1. 4) But 5·89 was arnve 5. and 5. (4. 0·5.47) lnserting values for Xcs at various frequencies would Iead to the gain and phase shift at these frequencies. The frequencies at which these values occur can then be found and the required frequency responses plotted.USE OF THE VALVE EQUIVALENT CIRCUIT of Re and Cs is denoted as Ze. For example. This approach is somewhat laborious and it is much easier to select values for Re/ Xcs such as I. From these values I + jRe/Xcs can be readily calculated giving the modified gain and phase shift. d at f rom 17 . 4. Therefore the frequency f is given by: f I012 X J03 X I7 I 700 = IOO X I03 X IOO X 211 = ~ kHz f= 27I kHz 130 . 0·2.46) gmZe _ gm(jXcsRe) Re.
and tan1 5 from 180°. tan1 2. Cs becomes open circuit but Ce must be included. k Figure 4. y'(12 + 0·52). Table 4.c. This is shown in Figure 4. further values of Re/ Xcs may be taken. leading to the result shown in Table 4.THE RESISTANCE CAPACITANCE COUPLED AMPLIFIER This may be sufficient to complete the response curve. anode voltage Va. To obtain the low frequency equivalent circuit. The voltage gains are obtained by dividing Avm by y'(1 2 + 0·22). where = JXcc wCc 131 Xcc (4.1.48) . and y'(12 + 52). but if more accuracy is required. The frequencies are obtained by using the value for the 3 db frequency calculated above and then multiplying or dividing by 2 and 5.13. the value of Re/ Xcs was changed. The new phaseangle is obtained by subtracting tan1 0·2.1 Re f(kHz) Avh X es 0·2 11·53 L169° 0·5 10·5 L 153° 30' 1·0 8·33L135° 2·0 5·25 L 116° 30' 5·0 2·3 L 101 o 20' 54·2 135·5 271 542 1 355 To obtain these points. y'(12 + 22). Vo = 1 RgVa Rg  .13. Low frequency equivalent circuit for RC coupled amplifier First an expression for v0 is obtained in terms of the a. tan1 0·5.
132 . +RgTaRL R + R R 1s t e para e com matwn Ta g L g of Ta. is given by the product of the current .gmVgk and the total impedance presented by Ta. + RL = Rg + Ta + RL Let and since . + RL) (4. + RL) TaRL Ta.52) = Avm.44 Avl = Avm :~ I _ jXcc R' (4.Rg + RgRL we obtain A But vl = I _ gmRe jXcc(Ta. X ( TaRL ) + R _ g Ta +RL ·x. and Rg. it is Re.USE OF THE VALVE EQUIVALENT CIRCUIT Now Va. RL.jXcc) Rg Vo = Rg.jXcc) cancels and Vgk =es.53) This resu1t can now be used in a simi1ar manner to that app1icab1e to the high frequency range. h 11 1 b' . So if we divide the numerator and the denominator by Ta.RL Ta. RL.51) . Values for Xcc/R' are chosen for simplicity of calcu1ation and the appropriate frequencies subsequently determined.e.RL + Ta..RL Ta+ RL = R' (4.49) The term (Rg.gmVgk ( Ta+ RL (Rg. J cc (4. from equation 4. T he expresswn Ta (4.jXcc(Ta. i. and Vo Avl== es TaRL ) gmRg ( Ta+ RL ~~~~~ ( TaRL ) Ta+ RL +Rg jXcc Mu1tip1ying equation 4. Therefore TaRL ) . . and Rg.Rg RLRg + + TaRL + TaRg + RLRg Ta.gmRe Rg + Ta.50) R L + RL) gmRgTaRL Ta.Rg + RLRg.50 by (Ta A vl = TaRL + .jXcc .
first consider the point where Xcc R' =1 I 1 R =Xcc=wCc I w = CeR' rad/sec Therefore and giving 1 106 J= 2TTCcR' = 27T X 0·1 X 56·7 x }()3Hz = 28·1 Hz When the value for Xcc/R' is multiplied by a constant. The phase shift in this case must be greater than 180° since the imaginary term in the denominator is negative. now .1wCc = 2R' 1 and w = .2RgCc 133 . and R' = = ra raRL + RL + Rg 200 30 +50= 56·7kQ Xcc R' Avl /(Hz) 0·2 0·5 1 2 5 11·53L191° 10·5L206° 30' 8·33L225° 5·25 L243° 30' 2·3 L258° 40' 140·5 56·2 28·1 14 5·6 The values for Xcc/ R' are the same as those chosen for Re/ Xcs in the high frequency case. the frequency must be divided by the same constant. tan. This new phase is obtained by adding 180° to tan1 0·2. For example: put Xcc/(R') = 2.THE RESISTANCE CAPACITANCE COUPLED AMPLIFIER As before Avm = 11·78 L 180°. The corresponding values for voltage gain will therefore be the same.1 0·5 etc. To calculate the frequencies.
Example 4. An alternating voltage of 50 mV r. EXAMPLES Example 4. A triode valve has 1140 and ra 20 kQ. By definition. 33·3.9. is applied between grid and cathode of a triode valve whose parameters are 134 .8.m. The gain and phase responses are given in Figure 4. 20. the reader will find that quite complicated circuits may quickly be reduced to a set of equations. Gain and phase response curves for the RC coupled amplifier discussed in Example 4. 13·3. Once the technique has been mastered. Examples on more involved circuits will be found in later chapters.Hz Figure 4.USE OF THE VALVE EQUIVALENT CIRCUIT Thus the frequency must be divided by 2. 41 225" 180" 135° 2 10 f. (b) 20 kQ and (c) 100 kQ. Calculate the valtage gain when it is connected as a simple common cathode amplifier and the anode load is (a) 10 kQ.s. From the tables: Upper 3 db frequency =271kHz Lower 3 db frequency = 28·1 Hz These give a measure of the useful frequency range or bandwidth ofthe amplifier. 7 This chapter has shown how the valve equivalent circuits may be used to solve a variety of simple amplifier circuits. Ans.14.14 plotted in each case against frequency on a logarithmic scale. the 3 db frequencies are those at which the voltage gain falls to l/v2 of the minimum value.
6·72 V. Ans. Example 4. Example 4. If the anode load is 20 kQ and the valve parameters Jl and ra are 85 and 12 kQ respectively. 51·0. lf the input signal is 0·8 V. Take Jl and ra as 100 and 8 kQ respectively. calculate the voltage gain and the output impedance of the circuit. 323 0. Example 4.EXAMPLES gm 3 mA/V.m. 135 . 34·3 kQ.13. Ans. Determine the value of grid leak resistor which will result in an input impedance of 5 MO. A modified cathode foliower as shown in Figure 4. Example 4. ra. 17·1 kO. The capacitive reactance may be neglected and a signal of one volt is applied between the valve grid and earth. If Jl and ra are 25 and 10 kQ respectively.6a has a cathode load of 2 kO. 0·3 V having internal impedance 600 Q.11. Ans.15. capacity coupled to the load. 17·5 kO. Calculate the output impedance of the circuit described in Example 4. 0·95. Hence find the voltage across a 500 Q resistor. 97 0. Example 4. h 2 mA. Ans. 0·75 V. If the anode load is 27 kQ.3a) is driven by a source of e. and VaK 3 V. determine the constants Jl and ra of a suitable valve. calculate the voltage gain and the overall output impedance of the amplifier. Calculate the anodecathode voltage when the anode is loaded with (a) 10 kQ and (b) 50 kQ. Ans.f. A triode valve employs cathode bias without decoupling to provide the desired operating point of VAK 110 V. Ans. Find also the voltage gain and output impedance assuming the source impedance tobe negligible. 1·25 V. 455 kQ. A simple cathode follower (Figure 4. 0·49 V. 10 kQ.2a) operates with a cathode load of 2 kQ. find the equivalent Thevenin generator 'seen' between anode and earth.10. Ans. 0·808.14. A single stage amplifier is required to produce 20 V signal in a Ioad of 33 kQ or 15 V signal in a Ioad of20 kO. 80 0. gm 6 mA/V and ra 10 kO.14 when the source impedance is 100 kQ. 10·5. Example 4. A grounded grid amplifier (Figure 4.12.
Take fl and ra as 60 and 10 kQ respectively. e1 = 0·5 sin wt and Ans. Determine the voltage gain using this Vo Figure 4. Ans.15 isthat of a long tailed pair connected as a difference amplifier.16 e1 e2 = I sin wt and e2 = 0·9 sin wt. 136 . The d. 4·3 sin (wt + 41 o 48'). 28·3.17. for each valve.c. amplifier circuit shown in Figure 4. Circuit for Example 4. calculate v0 (a) if Figure 4.17 approach and check the result by solving the complete equivalent circuit.USE OF THE VALVE EQUIVALENT CIRCUIT Example 4. The circuit shown in Figure 4. Example 4. Assuming that the valves are identical with ß40 and ra 10 kQ. or (b) if = 0·4 sin (wt + TT/4). 1·01 sin wt.15.16. Circuit for Example 4.16 can be considered as a cathode follower and a grounded grid amplifier connected in cascade.16.
calculate the voltage gain and input impedance of the amplifier shown in Figure 4. Circuit for Example 4. If f1. 35·6.Cl Figure 4. Ans. 9·62. Estimate the frequency range over which the voltage gain is frequency conscious and calculate the gain above and below this range. 4·88 kQ. Ans.18.19.+ Vo 1kfi 1 Figure 4. A grounded cathode amplifier employs decoupled cathode bias using 2 kQ in parallel with 0·1 ttF for this purpose. 30·4. 50 Hz to 10kHz.18. The direct coupled amplifier shown in Figure 4. Draw the complete equivalent circuit and calculate the small signal voltage gain v0 fe 8 • Ans. 13.17.Cl 50k.19 Example 4. H:T. Circuit for Example 4.18 Example 4. and ra are 50 and 10 kQ respectively. 500.18.IlXAMPLES Example 4.20.17 employs identical valves with tt 36 and ra 12 kQ. 137 . The valve has tt 40 and ra 15 kQ and is loaded with 47 kQ in series with the anode.
19. Calculate the gain and phase shift of the amplifier shown in Figure 4. The valve parameters are f.lF.23.22. 39·15. 0·37 mA/V.10a has the following components.USE OF THE VALVE EQUIVALENT CIRCUIT Example 4. An AF amplifier having three identical stages is required to have an overall gain of 1 000 and a 3 db bandwidth of 100kHz. Take f. 138 . 0·01 f. Cs 330 pF.l90 and ra 30 kQ and it may be assumed that Rk is effectively decoupled at all signal frequencies..l and ra as 50 and 8 kQ respectively. Ans. Draw complete gain and phase frequency response curves and hence find the maximum gain and 3 db frequencies. calculate the minimum gm for the valves and the common value of RL. The reactance of the coupling capacitors can be neglected and the grid leak resistors will be much greater than the required RL. 5·5 Hz.lF. Rg 100 kQ.. H. 37kHz. 31·9. RL 30 kQ.. An RC coupled amplifier is required to have a useful frequency range from 60 Hz to 500 kHz. Ans. Example 4. 9·6 kQ. Ans.21 Example 4. An RC coupled amplifier having the circuit shown in Figure 4.24. 27 kQ. 12·1 L 241° 42'.+ Figure 4. Ce 0·25 f. Ans. Taking gm as 5 mA/V and ra as 20 kQ calculate suitable values for RL and Ce and the resulting maximum gain. Circuit for Example 4. The total shunt capacitance is 50 pF and the grid resistor for the next stage is 330 kQ.21..19 if the signal frequency is 5kHz. Ifpentodes having ra 1 MQ are used and the shunt capacitance per stage is 30 pF.T. Example 4.
Ce 0·1 p. Cs 100 pF and Rg 120 kO. The coupling components are RL 50 kQ.25.EXAMPLES Example 4. An RC coupled amplifier having two identical stages employs valves with ra 40 kQ. 139 . Ans.F. Calculate the frequency range over which the overall phase shift is less than ±100. 128Hz to 7·5 kHz.
base and collector connections. if correctly applied. if the transistor is to be used in the common base configuration. This will usually depend upon the available information. If however. or even. Experience will show whether conversion is worthwhile. Both the hybrid or h parameter equivalent circuit and the equivalent Twere shown tobe suitable for representing the transistor. hrb. The equivalent circuit method to be outlined. for example. operating point and that there may be considerable variation in these values for transistors of nominally the same type.c. particularly the sense or direction of all voltages and currents. Show all details. GENERAL METHODS (1) Select the form of equivalent circuit to be used. applied or generated. When the correct h parameters are available.5 USE OF TRANSISTOR EQUIVALENT CIRCUITS In Chapter 3 small signal equivalent circuits were developed for valves and transistors. produce the correct solution in all cases. the components of the comrnon collector equivalent T. lt is appreciated that some readers will be principally interested in transistor circuits so the important ideas and techniques used for valve circuits in Chapter 4 will be repeated in this chapter where they apply. the general solutions discussed in Chapter 2 may often be used with considerable time saving. will. lt must be stressed that the component values for such circuits will only apply at a particular d. the best hybrid parameters to use are h1b. rb and cx is suitable for this circuit. draw the chosen equivalent. and hob· Similarly the simple T equivalent involving only re. (2) Foreachtransistor in the full circuit. but the solution will be found more quickly if the appropriate circuit is used. and mark clearly the points representing the ernitter. rc. i. the available information includes the common emitter hybrid parameters.e. there is no need to convert to the required form. hrb. 140 . Any form of equivalent circuit rnay be used for any problem.
When the h parameters are given. the transistor having h parameters hte 900 n.c. We shall now consider a simple example showing the complete application of the generat method.c. R2 and Ra provide the stabilized d. or as iZ or v Y products together with external generators appearing in the complete circuit. connect it to earth through any components or generators shown in the complete circuit. Assurne the reactance of the capacitor shown tobe zero at the signal frequency. Remernher that batteries or other d. Find the terminal voltage and current gain and the output voltage and current. hre 5 X 104. power supplies are short circuit to a. Figures 5. The circuit shown in Figure 5. (hre ib. With hybrid parameters these unknowns will normally be h and V2. Note also that the hre generator is expressed in terms of the voltage at the collector 141 .1.GENERAL METHODS (3) Taking each electrode in turn. (5) Express any equivalent generator currents or voltages directly in terms of the unknown currents and voltages. the generat solution derived in Chapter 2 may be used. and that reactive components such as capacitors can sometimes be neglected at the applicable signal frequency. The output voltage is obtained across RL and the amplifier is driven by a source es of internal resistance 500 . These methods will normally Iead directly to the calculation of voltage and current gain. to earth.lc. To draw the correct equivalent circuit we must apply steps 1 to 3 in the general method. operating point as discussed in Chapter 1.Q. Figure 5. collector and base respectively.) the unknown mesh current or node voltage should correspond to the direction indicated on the equivalent circuit. is shown as a short circuit. Note that the negative supply line is at earth potential to a. Components R1.c. and the coupling capacitor. d and e show the connections of the emitter. (4) Indicate the sense of unknown mesh currents or node voltages. hre 90 and hoe 125 fliDho.la shows a single stage amplifier. and of the input impedance. using substitution methods or determinants. having zero reactance. In any case. Example 5. hre Vce. Note that all details must be included if errors are to be avoided. but with T equivalents. the two mesh currents are usually more convenient. If the output impedance is required the method outlined in Example 4.3 must be used. if an equivalent circuit generator is expressed in terms of a circuit variable. (6) Write the circuit mesh or nodal equations and solve. etc.c.lb shows the h parameter equivalent circuit.
Vcc c RL 4 k!l. e (b) (a) c b c b (f) (e) Figure 5.la 142 . (a) Circuit for Example 5. (b) to (f) Stages in the equivalent circuit solution for the circuit in Figure 5. 1'" RE 1k.O.USE OF TRANSISTOR EQUIVALENT CIRCUITS ..1.1.
The hre generator is expressed in terms of ib flowing into the base. we must examine the equivalent circuit generators and express them in terms of our unknown currents or voltages. lf proceeding against the current arrow.m.lf shows the complete equivalent circuit including this simplification. we find the current in the emitter load is (ib + ic) and the current in the hoe branch is (ic . For mesh analysis the hoe branch is expressed as an impedance from I z =y = I25 1 X J06 = 8 kQ For paragraph 5 of the general method. 143 . These may be found by proceeding from the emitter to the collector around the circuit by any route. Before continuing the solution. The hre generator is expressed in terms of Vce: this may be expressed in terms of ib and ic in one of three alternative ways. Putting R1 and Rz in parallel we obtain: 20 X 5 + 5 = 4 kO RB = 20 Now applying Thevenin's theorem. Applying Kirchhoff's current law. The unknown currents chosen for mesh analysis are ib and ic.f. The hre current generator is already expressed in terms of ib and the rule is satisfied. = 0·445 kO and e' = esRB RB+ Rs = I X 4·5 4V = 0·89 V Figure 5. it is convenient to simplify the input section of the circuit.) and the signal source. take the product as negaitive. This is achieved by the application ofThevenin's theorem at points XX looking away from the transistor.GENERAL METHODS with respect to emitter.90ib).s. and by adding any iZ products and generated e. not with respect to earth. This includes the bias components R1 and Rz (in parallel to a.c.
1) 0 = 4 OOOic + 8 OOO(ic . e' io = .3) 1 OOOib (5. lb .04 = 1·74 X 105e' A The output current is normally expressedas fl. collecting terms and inserting the value of R'.9 X 104 lc .5 X 104 X 8 OOO(ic. _ 13 X 103 . In this case.90ib) V.) 144 .4 Substituting in 5. or (b) Vce = [ 1 OOO(ic + ib) .71 .360ib + 1 345ib + 1 OOOib + 1 OOOic = 1 985ib + 1 004ic 0 = 4 OOOic + 8 OOOic .90ib) + 1 OOO(ic + ib) (5.90ib) = (900 + R')ib + 1 OOO(ib + ic) (5.USE OF TRANSISTOR EQillVALENT CIRCillTS and if proceeding with the current arrow.e' .4) _2 • 10 lc X + 1 004ic e' ic = 1040 A and ib = 1·81 x I05e' 1. Each of these results wou1d lead to the correct solution. The three alternatives are: (a) Vce = 8 OOO(ic .4 OOOic] V. _ .71·9 X 104ib e' and .4 OOOic] V.3) (5. e' = 19·85 X 1·8lic (5.ic = 1 040 A . or (c) Vce = [5 X I04vce + 900ib + R'ib .72 x 104ib + 1 OOOic + = 13 X 103ic.owing towards earth.1 81 From 5. Inserting the va1ue found for e' and expressing the answer in !LA ic = = 0·89 X 106 1 040 fLA 856fLA (Note the minus. indicating 180° phase shift. take the product as positive. but the shortest path is usually the most convenient.3. Taking (a) above and writing the mesh equations in the normal manner: e'. = 4ic .2) Rearranging these equations.
If however the common emitter parameters are given and the transistors are connected 145 ..1040 X 1·74 = .1·74 X 105 X 445 = 0·89 .GENERAL METHODS The terminal input current is ib. since the input voltage must not only drive ib into the transistor. will always Iead to the correct solution. Any circuit involving transistors with known h parameters may be solved in this way.1f App1ying Kirchhoff's voltage law: V'= e'.::::=: At = 1·74 X 105e' ib 105 At=. correctly applied.55 "2 The output voltage v0 is now expressed with respect to earth.ibR' = 0·89 . This analysis may seem very long. and substituting for terminal current gain e' 1 040 ic = :=::. but must also produceieRE volts across the emitter resistor. Vo = icRL = 856 X lQ5 = 3·44 L180° V X 4 X 1()3 V To obtain the terminal voltage gain we require the terminal input voltage V' shown in Figure 5. V = 3·87 L180° Note that we could also obtain the terminal input impedance from: V' Ztn =:lb lQ3 = 1·74 X lQ5kQ = 57·5 k!l This high input impedance is to be expected. Thus with an amplifier having an unbypassed emitter resistor we can say the input impedance will be of the order of hreRE (since ie ~ hreib). On the other hand this technique..0·00775 ~ 0·89 V Av = Vo. but techniques to be introduced in Chapter 7 will reduce the solution of this problern to a few lines.
2b shows the complete equivalent circuit for the amplifier. 146 . hre 2 X I0. Circuit for the two stage amplifier in Example 5.2a. Find also the output impedance assuming a source impedance of 500 n. (3) The required input and output voltages and currents have been indicated Vtn.2 The transistors are identical and have the following common emitter parameters.c. Assurne all capacitive reactances to be zero at the signal frequency. (1) Since capacitive reactances are zero..4. Determine the overall voltage gain. The following points should be noted. Figure 5. hte 1·3 kO. and hoe 105 . as have the Ioad resistors. current gain and input impedance for the amplifier shown in Figure 5. This also eliminates the emitter resistors. itn and io.umho.2. The next example showing a two stage amplifier is particularly suitable for this approach.USE OF TRANSISTOR EQUIYALENT CIRCUITS so that the emitters are earthed to a. (a) 1(b) Figure 5. the h parameter general solutions may be used.2. Vo. these components have been shown as short circuits. (2) The bias resistors have been combined and expressed as admittances. Example 5. hre 110.
Applying this rule. and io = Vo YL2 find the current gain i 0 Ji1n.umhos. and a nodal equation at v0 • v' = 1 300ib2 + 2 110ib X 104 v0 = V0 (105 + 303)106 Substituting for v0 in 5·5. (3) Find the input admittance to Trl and hence the overall input impedance Z1n· (4) Using the effective Ioad found in 2 find the voltage gain v'/v1n for Trl.4·08 = 1 246 Q lb2 ~ 147 (5. I V = 1 Z1n2 = 300 . lb2 + 2 X 104( 110ib2) 408 X lQb 220 v' = 1 300 . Alternatively taking current division between parallel admittances and using the input adrnittances found in 1 and 3 write a single expression for the current gain.umho h2 1 ()()() 1 000 21. (5) Deterrnine the voltage gain for Tr2 and hence the overall voltage gain vo/VIn· (6) From i1n = V1n/Z1n.+ 4·7. = YLl YBl = Yn2 = 1000 = 3T = 303 . This circuit could be solved by converting the hre voltage generators to current generators by Norton's theorem and using nodal analysis.= 47·6 + 212·7 ~ 260 . The reverse conversion is: To obtain the impedance of a component in kilohms divide 1 000 by the number of .5) (5.umho Now writing a mesh equation for v'.umhos. To obtain the adrnittance of a component in .GENERAL METHODS (4) For convenience the terminal input voltage to Tr2 has been shown as v'. divide l 000 by the resistance in kilohms. Before applying these steps to the problern in hand.6) . a more convenient method for multistage amplifiers is as follows: (1) Find the input admittance to Tr2. (2) Calculate the resulting effective 1oad to Tr 1. A simple rule for this conversion is as follows. we must find the admittance values for the bias and Ioad components shown in our equivalent circuit.
using the general solution found in Chapter 2.6 u 2 Ytn llO Y 2 + Voltage gain Avl = ~ Vtn = X 10_4 V 2 100 1 300(1 471)106 .1·3 = X 1·471.0·022 58·2 or 58·2 L 180° 148 .( 1 366 Ytnl = 1000 1.umho The effective Ioad Yd for Trl is the parallel combination of YLl. and 110 X Ztnl = 1 300 . for step 3.umho Now from the general solution.hoe + YL 110 X 2 X I04 = I 300.285 2 X 104 + 105) 10_6 = 1 285 0 = 778 . Ytn2 = 1000 I· 246 = 803 .100 X 2 X I04 110 . = llOibl = Vtn + 2 X 104 V2 V2(I05 + 1 366)106 1 300ibl (5.8) Substituting for ibl in 5.USE OF TRANSISTOR EQUIVALENT CIRCUITS Note that this result could have been obtained directly.umho Including YBl for overall input impedance 1000 Ztn = 778 260 = 0·964 kO + For step 4. we require a mesh equation for Vtn and a nodal equation at v2.(303 + 105) 10_6 = I 246 0 Proceeding to step 2. Ztn2 hrehte = hte.7 _ I 300(1471)10. YB2 and Ytn2· YL·1' = 303 + 260 + 803 = I 366 .7) (5.
58 ' 2 Using this generat solution for Av2 110 Av 2 = I 300(105 + 303)10. io Vo YL2 At=:== AvYL2Ztn ltn Vtn Ztn At = 12 600 X 303 X 106 X 0·964 X 1()3 = 12·6 X 30·3 X 9·64 = 3 680 L 0° The alternative procedure.110 X 2 X 104 = .hrehre 110 = 1 300(1 471)106 . 260 + 303 + 105 = I·471 lbl = 803 + 110ib2 X 303 110 X 3·03 = . when voltage gain is not required is to find the input admittances as before and applying the current splitting rule write: itn Ytnl iin778 7·78 . hre Av = . lbl = Ytnl + YBl = 778 + 260 = 10·38 ltn .I 10 IIO X 2 X 104 = 2I6·5 or 2I6·5 L180° The overall voltage gain ~= Vtn Avr X Av2 = 2I6·5 X 58·2 = I2 600 L0° For the current gain.6 . refer to step 6.~chte(hoe + YL) . IR 2 io IlOibl X 803 11 X 8·03.GENERAL METHODS This result could also have been öbtained from the appropriate generat solution. .3:::0=3+:1=05~ 4·08 ib 2 io At = itn = = 7·78 10·38 X Ilx8·03 1·47I 3 680 L0° 149 X 1Ix30·3 4·08 .
Output tmpedance = 400 = 2·5 kQ Overall output admittance I 50 = .umho Including YL2.4 = + 260) .umho = 97·2 .umho I 000 . hrehre Yo = hoe .hte + Zs In this case we proceed from the input end taking the following steps: (1) Find the equivalent Zst including the bias resistors.28 . (2) Find Yot and hence the equivalent Zs2· (3) Find Yo2 and including h2.umho X 10 = 92·4 .umho Zs2 Now 110 X 2 X 104 I 300 + 442 .umho Ys2 = (92·4 + 303 = 655·4 .42 .umho 1·53 kQ _6 Yo2 = 105 X 10 =   110 X 2 X I04 I 300 + I 530 .umho 1000 Zst = 2 260 = 0·442 kQ For Step 2: Y01 _6 105 = 220 105 .USE OF TRANSISTOR EQUIVALENT CIRCUITS For the output impedance calculations weshall rely simply on the general solution. 97·2 + 303 . 3 .umho ~400 .umho 220 I05 .I 7 . the overall output admittance and impedance.umho = = 1 000 655 . For Step I: Yst = (260 + 2 000) .
3. may be neglected or at least rounded off. Finally. further components are added in parallel. Av= hre hreZL . Thus for a multistage amplifier we can neglect hoe for all transistors except the last. First note the effect upon the general solutions if hre is assumed to be zero.2 is accurate but not really practical. The equivalent generator solution for Example 5. having an impedance much greater than h1e. is that the effect of hre on the results is very small.3 where es is the source Figure 5. . the open circuit output voltage is the overall gain multiplied by Vtn· But esZtn es X 964 Vtn = Z1n + Zs = 500 X 964 V = 0·658es E = 0·648 X 12 600e8 = 8 300e 8 This solution for Example 5..2 voltage. the load and bias resistors will probably have a 20 per cent tolerance on the stated value. the neglecting of hre might cause 10 per cent error. The first thing to notice. In the calculation of input impedance or output admittance for a stage. we shall reconsider Example 5. and hoe is neglected. E.GENERAL METHODS Thus the complete amplifier can be reduced to the Thevenin equivalent generator shown in Figure 5. so the error in the effective load admittance or source impedance is very much less. In any case.2 as far as the calculation of voltage gain. .. . To show the effect of these approximations. the bias components. A further valid approximation is to note that the effect of hoe on Avt and Z1n1 is small and is in any case smaller than the effect of component tolerances.__ hJe(hoe + YL) 151 . In each case however.
152 .1. 600 12 600 X 100 per cent = 4·75 per cent This is much less than the probable error when component and parameter tolerances are allowed for. Low frequency analysis could be achieved by using either common emitter parameters and normal equivalent circuit methods or by use of the general solution in terms of the common base parameters h1b. hrb. Avz Z1n2 = llO 1·300(105 + 303)106 = 208 = 1 300 = 770 flmho YL1 err = (770 + 303 + 260) flmho = Y1n2 and so and ZLl = 0·75 kO Avt = 750 90 X .2.1 300 Overall voltage gain = = 1 333 flmho 63·5 63·5 x 208 = 13 200 This result must be compared with the accurate result of 12 600 and may be expressed as a percentage error. a summary of the performance would be useful for comparison with the other configurations. hrb and hob· The resulting properties are listed in Table 5. Having considered the common emitter amplifier.USE OF TRANSISTOR EQUIVALENT CIRCUITS Now applying these results to the problem. Properfies of Common Emitter Amplifiers Input impedance Output impedance Voltage gain Current gain Phase change Medium Medium High High 180° 50002 kfl 5 kfl20 kQ up to 500 up to 250 Common base amplifiers are only used in practice at very high frequencies. Table 5.
Figure 5. hoe 250 . Analysis may be achieved either by use of the common emitter parameters. Investigate the performance ofthe common collector amplifier shown in Figure 5. firstly.4.4b shows the complete equivalent circuit using common emitter parameters. In the next example.umho and hre 65. by determining the common collector parameters and using the general solutions.3. or by determining the common collector parameters and using the general solutions.2. these alternative methods will be compared. and secondly. Properfies of Common Base Amplijiers Low Input impedance Output impedance Voltage gain Current gain Phase change 5300 0·52Mfl up to 500 up to 0·99 High High Low oo The common collector or emitter foliower is a very important circuit and it is used in many applications.3 emitter parameters. Example 5.4a. Circuits for the common collector amplifier in Example 5. by direct use of common 250 t~ V' (a) (b) Figure 5.GENERAL METHODS Table 5. Take hte 800 n. Proceeding with the general method we must first find Vce in terms of ic and ib our unknown currents. Since common collector parameters are not normally quoted we shall have to consider the best approach to solving common collector circuit. hre 103 . Since the collector is connected to earth we can see that Vce 6 = Vo = 2 OOO(ib 153 + ic) .
258. + ic)2 + 2(ib + ic) V' = 0·8ib .umho Now for the overall current gain and input impedance we must include the effect of the bias components.4a. VI = 2·8ib + 2 V' Z1n = .USE OF TRANSISTOR EQUIVALENT CIRCUITS Now writing the mesh equations and working in mA and kQ. = 6'b lc Now Output current = ib = + ic ib ( 1 + 2~ 8 ) = 44ib .umho = 60 . = 2·8ib + 2ic 0 = 258ib + 6ic V' (neglecting the 2 x I03 terms) From the second equation.umho . 16·7 kQ Overallinput admittance = YBI + and the overall input impedance = 14 kQ. From Figure 5. :ie Current gam lb = 44 To find the input impedance we need to know ib in terms of V'. YBl is given by the two bias resistors in parallel. and 1 000 1 000 YBI = 50"" + 25 .I03(ib 0 = 4(ic  65ib) + 2(ib + ic) Collecting terms. so substituting for ic in the first equation. 154 Y1n = 71·25 .= lb Y1n = 2·8 X 6258 ib + 86 = 88·8 kQ 11·25 . .umho ZBl = 60 = 1000 .
(a) I Figure 5. and apply a generator of I amps to the output terminals. Then by calculating the resultant output voltage V we find I Yo = V This equivalent circuit is more easily understood if the branches are shown in parallel between the emitter terminal and earth.3 155 . suppressing any external generators. Av = 6·95 X 2 X 1()3 X 71·25 X I06 = 0·99 L0° To find the output adrnittance we must redraw the equivalent circuit. as shown in Figure 5.5a.5.GENERAL METHODS By current splitting. Equivalent circuit solution for the output admittance of the common collector amplifier in Example 5. the overall current gain is given by 44 X 11·25 Al= 11·25 But Input voltage and Output voltage Av + 60 = 6·95 LO o = hnZin = ioZL ioZL = :Z = A1ZL Y1n lln ln Voltage gain.
USE OF TRANSISTOR EQUIVALENT CIRCUITS This equivalent circuit (Figure 5.3. Table 5.65 X 384 X I06V +I= V(500 + 250 + 384)10. Properfies of Common Collector Amplifiers Current gain Valtage gain Input impedance Output impedance Phase shift I Medium 520 (limited by bias components) Low 0·80·99 High Bias components Low 20100 il 156 .1 and 5. The result of this conversion is shown in Figure 5.) Now writing the nodal equation. care must be taken not to lose ib since this is required forthehre generator.5a) is of mixed form and is best converted for nodal analysis by applying Norton's Theorem at XX. First Vce = .6 Neglecting the first term.5b.2.384 X I06V =:::= . I= V(500 + 250 + 384 + 24 950)106 The output admittance Yo I =V= 26·1 mmho and the output impedance Zo = 38·4Q These results may now be summarized for comparison with Tab/es 5.384 X I06 V and I y =( (This is effectively neglecting hre. so the sense of the hre generator can be reversed and the value changed to I03 V. Applying Kirchhoff's current law: ib =I'. 0·384 X 106.VY' = 0·384 X I06V.V. IOav Now ]' = ( 16·7 X 2) = 0·384 X l06V 0"8 + 16·7 + 2 1000 h 16·7 X 2) flill 0 0 "8 + 16·7 + 2 = 384 ßmhos At this stage.
The conversion from he parameters to hc parameters can be simply achieved by redrawing the common emitter equivalent circuit with the collector as the common terminal as shown in Figure 5.GENERAL METHODS The alternative procedure is to find the common collector h parameters and to use the general solutions. This leaves hoc = hoe· Also since Vec = Vce. Thus Vce is also zero and the hre generator disappears. ib and ie as shown hlc I = . All the current from the hre generator. Thus and + hreib) hrc = (hre + 1) and h1c ie = (ib = h1e Now for ib zero.. the hre generator becomes hreVec. There is no volt drop across h1e so: Vbc = Vec  hreVec and 157 hrc = 1 . Vec.I = V~c hrc lb Vec=O ie lb Vec=O First if Vec is zero we have a short circuit across the emitter collector terminals. the base terminal is open circuit and the hre generator disappears. Circuit for determining the common collector h parameters from the common emitter h parameters W orking with the required voltages and currents.6.6. The normal h parameter definitions may then be applied to find the hc parameters. hre Vce Figure 5.hre . together with ib flow into the short circuit. Vbc.
umho = 38·2 0 This method is obviously very much quicker and is therefore preferable even if the necessary conversion factors have tobe found. ALTERNATIVE COUPLING METHODS Two further examples of the application of equivalent circuit techniques to practical configurations will now be given.umho Now 1000 560 kQ Z8 = Y0 = 250 + 1 788 + 800 . including Ioad.umho = 1·788 kQ 66 X 106 = 250 + 25 500 .umho For overall Y0 . The properties of such a combination will become apparent from the example.umho 1 .103 ::= 1 Applying the generat solutions Av = 800(250 + ~:)I06 + 66 = 0·99 Ztn = 800 + (250 + 500)I0.USE OF TRANSISTOR EQUIVALENT CIRCUITS Now applying these to Example 5. we require the effective Zs. The second circuit isonein which two transistors are interconnected and used as one. but Ys is given by Ys = 500 + 40 + 20 .c. With slight modifications it can be used as a difference amplifier. amplifier.6 = 66 LOo 88·8 kQ For the output admittance.3: htc = 8000 htc = 66 hrc = hoc = 250 . The first circuit is known as the emitter coupled or long tailed pair amplifier. or super <X pair. Y0 and Zo = 25 750 + 500 . 158 . This combination is sometimes known as a Darlington connected. a phase splitter or a d.
umho and hre 90. The transistors have the following hybrid parameters at the d. so we shall draw the equivalent circuit neglecting hre and solve using nodal analysis. the general solutions can be applied in the normal manner. hoe 125 . Circuits for Example 5. 7a. by the capacitor shown (assuming Xe~ 4 kO).c. supply.7. Determine the voltage gain and input impedance of the emitter coupled amplifier shown in Figure 5.c. Thus if the hc and hb parameters are known. One possible approach is to take Tr2 as a common base amplifier and Tr1 as a common collector amplifier. The output of the first stage is taken from the (b) Figure 5.4. hre 5 X 104.4 emitter and the collector is earthed through the d. hte 1 300 0.7a could be treated in a number of ways. In this example only the he parameters are available.ALTERNATIVE COUPLING METHODS Example 5. 159 . The amplifier circuit shown in Figure 5.c. The base of Tr2 is also earthed to a. operating points determined by the circuit.
which changes the units to volts. kil and mA: is = l·08v1 .A. Now following the general method for writing nodal equations as stated in Chapter 2 on page 59. and va. Note the components representing h1e have been expressed as admittances.0·125va 75vz + 0·625va 160 . equating currents entering the node from generators to currents leaving the node as voltage admittance products: is = (VI  vz)830 + VI 250 The admittances are expressed in p. vz and va are measured in volts the currents will be given in p.0·83vz 0 = 0= 75·5vl + 153vz . we may write and ibl = (v1  ibz = vz)830 pA vz 830 p. as have the combined bias components on Trl and the loads on both stages. therefore if VI. The ratio of va to VI will give the required terminal voltage gain and that of v1 to is the input impedance.A Now for node I.vz)830 + 90( 830vz) = vl830 + vz(830 + 125 + 2 000 + 125 + 830)  125va and for node 3: 90( 830vz) = 125vz + 625va Now collecting terms and dividing by 1 000. For node 2: 90(vl .7b. vz. Note also that ibl and ibz are clearly indicated as are the unknown node voltages v1. The input is shown as being supplied with a current i 8 • This is necessary.USE OF TRANSISTOR EQUIVALENT CIRCUITS The required equivalent circuit is shown in Figure 5. since for nodal analysis the only permissible generators are current generators. For ib1. The next step is to express any equivalent generators in terms of the unknown voltages. circuit admittances and external generators.mho. vz and va in terms of is. The solutions to the nodal equations will express v1.
ALTERNATIVE COUPLING METHODS Using determinants.0·125 X 75) + 0·83 X 0 + 0 1·08(153 X 0·625.0·125 X 75) + 0·83(75·5 X 0·625) + 0 ~~~~~~~~~~==~~~~~~~~~ 87 = 54·8 = 1·59 kQ This answer might seem low for a common collector stage. For the voltage gain we require the ratio va/vl and in determinant form this is given by va Av=Vl = 1·08 75·5 0 0·83 +153 75 is 0 0 is 0 0 0·83 +153 75 0 0·125 +0·625 The denominator has already been found in the numerator of the expression for Ztn· Av 1·08(0) + 0·83(0) + is(75·5 X X75.= ls 1 :ls is 0 0 0·83 +153 75 0 0·125 +0·625 1·08 75·5 0 0·83 +153 75 0 0·125 +0·625 1 Ztn =:. but the effective load on the stage is the input impedance to a common base stage which is also very low. since each stage of a 161 .0) 87is = = 75·5 X 75 87 = 65 This expression for voltage gain is positive and there is therefore no phase reversal.X ls X is(153 X 0·625. the input impedance Ztn is given by Ztn Vl = . This is to be expected.
hoe 150 pmho and negl\gible hre· Tr2 v.8. Assurne the transistors to have h1e 2 0000. An electronic circuit shows two transistors connected as a Darlington pair.w 'h. and hence find the conditions leading to high current gain and high input impedance.~· it. hre 120. the equivalent circuit is shown in Figure 5.USE OF TRANSISTOR EQillVALENT CIRCUITS cascaded common base. (d) (c) Figure 5. Determine the h parameters for the composite unit used in the common emitter configuration. Example 5. hoe'.8a. Compound connected transistors and circuits for determining the overall h parameters The required Darlington connection is shown in Figure 5. (b) (a) !2 ho. hoe" and hre ".' irQ' "1 v2 ho•w h .. hre' and Tr2 to have parameters h1e ". common collector amplifier has no phase reversal... conditions and for calculation it may be included in the h1e for Tr2.5.8b. Assuming Trl to have parameters h1e'. 162 . A resistor R' is normally included to provide the required d.c.
_h_"_ 1 ~ h. oe II le (5.d. Here Vz ib"=11 1 hte h. across hte'. for h21 and hn.. hzz are + V2h12 hh21 + V2h22 hhn V1 = h = We can see that the required conditions for finding the parameters.:. Figure 5. Under these conditions h and ib are zero.8c we can see that hoe and hte" form a potential divider across v2.. 1 Now h12 = vll v2 It=o From Figure 5. Thus the hre1 ib 1 generator is eliminated. are open circuit input. but first we must find h22. this can also be deleted.ALTERNATIVE COUPLING METHODS Rewriting the h parameter equations. + oe Now.. + 1 h Oe + 1 + hte "hoe 1 hte II hre"hoe 1 + h. h12. ib" in terms of the node voltage. and short circuit output. where hn.8c shows the modified equivalent circuit with the input open circuit. the parameters for the overall circuit. where Y is the combination of hte ".9) we can write a nodal equation.1 + 1 + hoe 1hle " le 163 (5. hz1. hoe1 and /z and h = h re "V2 h le II + h oe _ fz _ 22  = V2  h " oe 1 + v2 ( hoe " h + le oe 1 h ") + .:. oe 1 hre "hoe 1 hoe 1hte II 1+ + h 1 oe hoe ".. for h12 and h22. and since there is no p. so h12 may be written: 1 h12 To find = hte"hoe 1 hte "hoe 1 + 1 hte" h 1 + h.10) .
The redrawn equivalent circuit is shown in Figure 5. V' + hoe'hte 11 (5. the input impedance II V' +ib' V! h h I .. we shall see that the last term of this result is the predominant part.. For h21....12) Once again. The hre ib" generator can have no effect on h since the entire generator current flows in the short circuit...USE OF TRANSISTOR EQUIVALENT CIRCUITS When practical values are inserted.= 11 = le h Writing a nodal equation for V'....I6) . (5. By inspection. Proceeding to hu and h21 we require the short circuit output condition. numerical values will prove the last term to predominate.I3) h I• I 1 + re lb hte 11 12 due to hre'ib' alone = _____. In this case. • I + h I• I lb re lb V' ( hte' 1 = V' I + hre' I h. h = ib'· Now applying the Superposition theorem.8d.I h . le h _ h .14) = hre "ib" = ib'(l 1 and +hre' ib' + hre')hte 11 + hte V' • II lb =" hte 164 hoe' 11 (5.I5) (5. I . 11  + hoe ') le + hoe (I (5.11) I + hre')hte 11 + I+ hte"hoe' (5.II. oe + h le " 12 due to hre"ib" alone But from equation 5.. we must find the value of the current in the short circuit output. h in terms of h. 12 is given by the sum of the currents flowing in the short circuit due to the two current generators and ib' taken separately.
practica1 va1ues will approximate hz1 to hre"(1 + hre')o Taking the va1ues for the parameters given in the question: h 12 = I 2 000 X 150 X 106 X !50 X 106 + 2 000 _6 800 1 Oo 23 I 120 X 150 X 106 I 50 X 106 mho + 150)10 + 2 000 + I + 2 000 X h 22 = ( 150 = Oo 3 = J.ALTERNATIVE COUPLING METIIODS Substituting from equations 5oi5 and 5oi6. 5o14.3 = x 106 OoOI8 + """13 mho = Oo8 + 138 mmho = 14°6 mmho Note. hu = 2000 + 121 2000 1°3 = X 188kQ From equation 5018.h22 + YL 11 200 X 0o231 _ (1 406 + 10) X 10_3 Q .0o3 + 120 X 121 = = YL 1°3 o o Thus from the genera1 so1ut10ns. this result is equiva1ent to an output impedance of 68o5 Qo From equation 5oi2. smce A1 11 200 h21 YL + h 22 a ho1gh current gain will be obtained if YL is of the same order as. h _ h ib'  _ hre'. or greater than h22o In this case a suitable value of Ioad would be 100 Q or lesso Using this value. and 5017.(I + hte "hoe')hte" hre "(I + hre')ib' I + hte"hoe' (5oi7) Now adding equations 5013.83 kQ 165 . the input impedance is given by: Ztn = _ Z1n  3 _ 188 X 10 hi2h21 hu . h "0 "_ hre"(I + hre')hie"ib' re lb .hoe'hte" + hre"(l 21 I + hte"hoe' + hre') (5018) In this case. 1121 = 120.
Example 5. as the large h21 permits a very low d. The effect of capacitors used for decoupling emitter resistors is exactly the same as those used in the cathode circuit of valve amplifiers. but the effect of an overall shunt capacitance Cs will be included in the next example. Capacitors. the coupling capacitor.9a is to be used at frequencies from 100Hz upwards. except in phase. (b) the high frequency above which the gain is more than 3 db below the maximum value. The conversion of parameters is obtained by using the results found in Example 5. base current for the first transistor. determine. where shown.c. The output section of the second transistor is not required since the absence of reactive components in the Ioad make the gain. The total shunt capacitance is 1 200 pF.6 X 10_3 = 538 H2 All the examples so far considered. An even higher input impedance can be achieved. hre 110. independent of frequency. Tr1 bias components have no effect on the voltage gain of the stage and only modify 166 .Q h12c = 1 h21c = 11200 h22c = 14·6 mmho 0·231 = 0·769 The current gain is unchanged. have neglected the effect of reactances. and that due to transistor properties.9b shows those parts of the equivalent circuit essential to the general analysis. have been assumed to have negligible reactance at all signal frequencies. by using the composite transistor in the common collector configuration. have not been shown.USE OF TRANSISTOR EQUIVALENT CIRCUITS This of course will be modified by the bias components in parallel with the input. The audio voltage amplifier shown in Figure 5. Investigate methods by which the gain and phase responses can most easily be recorded.3. Detailed analysis of high frequency performance will be considered in a later chapter. These may have far higher values than are necessary for a single transistor. since at low frequencies their reactance is very much greater than shunt resistive components. Stray capacitance due to wiring. This analysis is similar to that used for the RC coupled valve amplifier in Chapter 4. and hre negligible. and the input impedance becomes 188 X 11 200 X 0·769 1()3 + 24 . huc = 188 k. At this point then.6. without loss of current gain. By means of a general analysis. (a) a suitable value for the coupling capacitor Ce. we shall investigate the effects of Cs and Ce.umho. Av2 . hoe 100 . The transistor he parameters are h1e 1 000 0. Figure 5.
20) hoe and let where and Jet + YL + R'= h 1 oe + YL Ytn2 + Ztn2 (5. Jet 1 Re = :=c=(5. By definition.6 For simpJicity.19) Vcc 2kil Tr2 t t V. 0 I Xc=O (a) v' (b) Figure 5.22) . Y1n2 is the total input admittance to the second stage and will include the bias components Ym and YB2· Since hre is zero. certain components may be combined for the anaJysis.ALTERNATIVE COUPLING METHODS input impedance.9. The RC coupled amplifier for Example 5. the terminal input impedance to Tr2 becomes hte· Ym Ytn2 = 1 + YB2+hle (5.21) 1 Ztn2=y tn2 I RL 1 = hoe + 167 YL (5.
Av1 and Avh respectively.   VcZ!n2 Ztn2 jXcc may be found in exactly the same way as v' was found at medium and high frequencies. X ce and Ztn2 act as a potential divider circuit V . as low frequencies. The corresponding valtage gain for these frequency ranges will be referred to as Avm. jXcsRe v = hrelb x R ·x e . First. Ve shown in Figure 5.USE OF TRANSISTOR EQUIVALENT CIRCUITS Also. .9b at medium frequencies: 1 Xce = C ~Z1n2 CO and c Xcs = 1 C '.jXce) Ztn2 v = . Avh hre = hRe le = ( 1R 1 _ _e_ jXc Avm (5. .9b is the same as v'. At low frequencies.23) At high frequencies. X es is in parallel with Re. ·x X (Z ·x ) Vc RL + Ztn2 168 J Ce ln2  J Ce .J es J Dividing numerator and denominator by jXcs.>Re v' = hreib X Re Thus CO s but v' hre Avm ==Re VI hte (5. frequencies at which both Ce and Cs may be neglected. those at which Cs only may be neglected.24) "R 1+~ X es At medium and high frequencies. considering Figure 5. h . . will be referred to as medium frequencies. RL'(Ztn2 . and those at which Ce only may be neglected as high frequencies.telb X .
25) l _ jXcc R' Equations 5.23. the value of X es or X cc could be calculated and used to determine the complex gainin the polar form AL8.24 and 5. Avm A vl = ':'::: (5.ALTERNATIVE COUPLING METHODS The terms in brackets cancel and ib is the same as that for medium frequencies. let Re/Xcs = 1. Avm = 110 X 1 ()()() 540 Now considering equation 5. 106 Re = 100 + 250 + 1 000 + 500 Q = 540 Q and from equation 5.) Thus from equation 5.25 provide a convenient means for plotting graphs of the variation of gain and phase shift of the amplifier as the frequency is changed. Forageneral investigation.26) 169 .22.21.24.20. and 5. from equation 5. it is more convenient to choose a simple numerical value for Xcc/R' or Re/Xcs and then to determine the complex gain and the frequency at which it applies. But (Equations 5. At any particular frequency.20.23. 5. 5. This results in Avh = 59·3 1+j Wh = _ 59 _3 be the frequency at which 59·3 Ll80a = V 2 L 45 a = 42 Nowif L135 a (5.23. In this numerical example.19. 5.
etc. At low frequencies. this corresponds to the frequency at which X ce/R' = 1.4 on page 171. and 5wh respectively.F Thus in practice a 1 p. 3wh.USE OF TRANSISTOR EQUIVALENT CIRCUITS 1 1012 Then Wh = CsRe = 1 200 X 540 = 1·54 X 106 rad/sec Wh /h = 277 = 245 kHz Other values for the complex gain at particular frequencies may be obtained by putting Re/ X es as !. 106 I06 + 100 + 200 + 50 + I 000 Q I 670 + 800 = 2 470 Q R' = 500 = Ce= 27T X 106 100 X 2 470 p.2I. 2..24. tan. 2wh.I9 and 5. These results are shown in Table 5. At these frequencies.. 3..F = 0"63 p. y'5. the problern requires that the amplifier should be suitable for use down to 100Hz. Inspection of equation 5. I Avl I= antilog10 0·15 = y2 Avm IAvl I = IAvml V2 With reference to equation 5.F capacitor would be suitable. _1_=R' 27T/Ce I Ce = 277JR' Farad and From equations 5.1 2. In practice. tan1 3 and tan1 5 respectively..26 shows that the corresponding angular frequencies will be given by wh/2. y'IO and y'26 respectively with phase shifts changed from 180° by tan1 0·25. 5. this implies that the gain shall not fall by more than 3 db (see Appendix 1). the gain may be obtained by dividing Avm by y'l·25. I I 3 = 20 log10 Avm Avl Thus at 100Hz. 170 .. .
1 2 63° 30' = 18.1 5 78° 30' 10 IAvml = 5·93 tan110 84° 18' 20 IAvml = 2.ALTERNATIVE COUPLING METHODS Table 5.1 3 71° 30' 3wh 735kHz 5·0 IAvml = 11 .5 3·0 IAvml = 18 . Xcc Avt 7i.6 tan1 5 78° 30' Swh 1·015 MHz 10 IAvml = 5 _93 tan11o 84° 18' IOwh 2·45 MHz 20 IAvml = 2.' Avm = 59·3 Phase shift = 180° 0 wherc tJis + 1·0 IAvml = 42·0 tan11 45° o5 IAvml = 53·1 tan. 8 tan.1 3 71 ° 30' IAvml = ll6 tan.1 20 87° 10' 20wh 4·9MHz so v' 2501 tan.1 0·5 IAvml = 57·4 2·0 IAvml 3·0 IAvml 5·0 Ll80° !=~ where w is I I = R'Cc lOOHz 26° 30' 2w1 200Hz tan10·25 14° 4w1 400Hz = 26·5 tan.8 tan.0 tan1 45° 0·5 IAvml = 53 .1so 88° 49' 50wh 12·2MHz 0·25 v'I·0625 y'S v'10 y'26 y'IOI v'401 IAvml = 1' 19 171 .4.10·5 26° 30' T tan10·25 14° 4 61kHz tan12 63° 30' 2wh 490kHz y'2 v'l·25 IAvml 1 ReCs Wh= Wh Wh 245kHz 122kHz 2·0 = 57 "4 IAvml = 26 .1 tan.96 tan120 87° 10' IAvml = 11 9 tan1 50 88° 49' 0·25 50 v'2 y'0·25 v'I ·0625 v'5 v'IO y'26 v'IOI v' 401 y'2501 Avh w1 W1 2 W1 T W1 5 W1 iO W1 2o Wt 5o 50Hz 33Hz 20Hz lOHz 5Hz 2Hz where 0 is 1·0 IAvml = 42.96 tan.
2.=::::: hle(hoe 110 = 1 000(100 + 500)106 = 184 L 180o Thus the overall gain at any frequency is given by the result shown on Table 5. Av2 = hre YL) .4 multiplied by 184 and the phase shift may be found by adding 180° to the angle given in the same table. Considering the single stage first. These results arealso shown in Table 5. Xcc _ _1_ _ 1 h Also. since we are puttmg R ' . etc. t e corresponding frequencies will be found from 2w1. for the single 172 .2 .10b shows both gain and phase shift plotted in this manner. while the gain scale would be multiplied by 184. i.e. we could plot graphs of !Avl and phase shift against frequency. 2. An improved display is obtained by using logw of frequency as a base. . ~\ ~1 respectively where w1 = c:R. 3. of the two stage amplifier. but since most of the change in gain occurs between 1 and 400 Hz and between 100 kHz and 10 MHz. the additional phase shift will be added to 180° since the j term in the denominator is negative instead of positive..hrehre + ::::. Figure 5.10a where we can see that all detail is lost at frequencies below 1 MHz. This may be found by application of the generat solution in terms of the h parameters. The scale for the phase shift graph would range from 90° (at the top) to +90°. values of !. This is shown in Figure 5. The gain and phase variations may be shown simultaneously by considering !Avl L if> as a vector and drawing the appropriate vectors for a number of frequencies. We must now consider the question of displaying the information obtained from the above analysis. The resulting gains will have the same values as those found at high frequencies but in this case. To calculate the overall gain v0fv~. may be assigned to Xcc/R' and the corresponding complex gain and frequencies determined.wCcR' . a linear frequency scale hides much of the information at low frequencies.4. must be multiplied by Av2. at each frequency. This is shown.USE OF TRANSISTOR EQUIVALENT CIRCUITS Further points on the frequency response graphs may be obtained in the same way as that used for the high frequency range. The same graphs are correct for the two stage amplifier under consideration if the vertical scales are changed. the gain of the second stage. etc. 3. the gain of the first stage.
plotted in the complex plane. 30 180° e 20 150° 10 120° 10 1·0 100 2·0 1000 3·0 10~ 4·0 105 5·0 f Hz log f Figure 5.40 > 3. 4 6 f.10. the locus is a circle. in Figure 5.6. This locus is shown in Figure 5. we obtain the locus of the gain vector. The locus for the two stage amplifier will be obtained by multiplying each vector by 184 and adding 180° to the phase shift. If the ends of all these vectors are joined. with the origin on the circumference.llb.Jla.ALTERNATIVE COUPLING METHODS (a) 60 .MHz 8 10 12 (b) 50 240° 40 210° > 3. In this case. Both methods of presenting the information are useful and examples of their application will appear in later chapters. Graphs of the variation of gain and phase shift with frequency for Example 5. 173 . (a) Frequency on a linear scale and (b) frequency on a log scale stage.
Gain vectors for the amplifier in Example 5. An alternative form of bias circuit for a common emitter amplifier is shown in Figure 5 . Examples of the use of the y parameter circuit and the hybrid n circuit will appear in Chapter 9.7. R (a) L Vo (b) Figute 5. but to show that the general method is applicable to all equivalent circuits.O.12.11.7 using the transistor T equivalent circuit 174 . Circuits for Example 5.USE OF TRANSISTOR EQUIVALENT CIRCUITS 245kHz 0 (b) ~"'"''"' 100Hz 0 5kHz ) increasing 245kHz 59·3x184 100Hz Figure 5.6 ALTERNATIVE EQUIVALE NT CIRCUITS The examples so far considered in this chapter have made use of the h parameter equivalent circuit. In Chapter 3 other possible circuits were mentioned. Example 5. If the transistor employed 4k.12a. we shall complete this chapter with a solution using the common emitter T equivalent circuit.
and ir. Figure 5.ALTERNATIVE EQUIVALENT CIRCUITS has re 30 !1.12b shows the amp1ifier equivalent circuit drawn using the common emitter T equiva1ent for the transistor.ie) 15ir.ie + rx'ib Writing the three mesh equations.29) Rearranging and collecting terms. ie.lo R L 0 = lere V! (X Putting rc/rx' = 10 kil. The unknown currents have been se1ected as ib. . rc +. In terms of these three. the other circuit currents shown are given by: io i' = ir + ib  ie = ib . = 0·03ie + 0·7ib 0 = 14ie + 814ib + 4ir 0 = 10ie .463 + 175 = 24·2vl mA 175 .10(8lib..ie = ibrb + iere . + l. determine the amp1ifier input impedance and voltage gain.27) (5.810ib + 15ir So1ving by determinants for ie.28) (5. V! 0 0 = 0·7ib + 0·3ie = 0·03ie + 10(81ib  = ie) + 4(ir + ib . substituting for i 0 and i'. rb 800 kil and rx' 80. inserting values and working in kil and mA. V! V! ie = i e = 0 0 0 +0·7 +814 +4 810 +15 0·03 +0·7 0 14 +814 +4 +10 810 +15 mA Vt(814 X 15 + 810 X 4) mA 0·03(814x 15+810x4)0·7(14x 154x 10)+0 15 450vl .ie). = ib(1 + rx') .0·7ib (5.
however. In the majority of the examples.0·03Vt 0.ie)4 V = 22·5 X 4Vt For the input impedance. This should be allowed for and if a tight specification is required.203·5 x 0·392v1 = 1·3v.28. component tolerances and spread of transistor parameters willlead to a possibly large degree of error. ib = Vi. 176 . In fact. the input current is ib + ir.n Vt Vt = :lln = 0 ' 392Vt + 1' 3Vt = 590 Q But the overall input impedance must include the 15 kO bias resistor.7 X 24·2 = 0·392v1 mA Substituting in equation 5. circuits using devices other than transistors may be analysed in the same way provided the appropriate equivalent circuit parameters are available. feedback methods should be used. In the design of a practical amplifier. Vo Vo Av = . mA = ioRL = (ir. we have investigated the use of small signa1 equivalent circuits for the solution of a wide range of transistor amplifiers. The methods used. . the typical values quoted in the manufacturers' published data should be used. ir = 14ie . It should be stressed that the parameters quoted in each example are those for a particular transistor at a particular operating point. These methods will be discussed in the next two chapters. are equally applicable to any other form of equivalent circuit. However.59 kQ = 568 Q SUMMARY In this chapter.USE OF TRANSISTOR EQUIVALENT CIRCUITS Substitutingin equation 5.814ib 4 = 3·5 X 24·2v. it is therefore given by: Ztn = 15 X 0·59 15 .= 90 Vi Z.ib.27. the h parameters have been used since these are the parameters that are usually available.
. 560.8.8 using RLO. 0.10. 65·6. The common emitter amplifier shown in Figure 5.umho. Example 5.9. 6·72. and At against RL. 177 . Av. and oo. Repeat the calculations performed in Example 5. The common collector amplifier shown in Figure 5. 960 Q. 6·83 kQ. 25·7 Q. v. 65·6. Ans. 258 n.13. hoe 90 . 960 Q. Repeat Example 5. 100 n. Ans. Circuit for Example 5. 10 kQ. 8·83 mV. hre 75. 0·68. Calculate the output voltage and the output impedance.11. 1·97.EXAMPLES EXAMPLES Example 5. 862 Q.8. Example 5. 130. 0. Ans.8 i/p If RL is 2 kQ calculate the amplifier input impedance and voltage and current gain.14. 645 n. Hence sketch graphs showing the variation of Ztn. hre I03 .14 emp1oys a transistor having the same parameters as that in Example 5. Figure 5. Ans. Figure 5.13 employs a transistor having hte 1 100 n.8 taking Xe as being infinite. 56·5. 3 130. Assurne Xe = 0. Circuit for Example 5. 36·2.11 Example 5.
2·22 kQ. 1100 Q. A two stage RC coupled amplifier has the following components. 1 528 Q. Collector Ioads. 264 Q. shunt bias resistors 8 kQ 178 . Ans.15.16. The corresponding h parameters are hte 1·8 kQ. Repeat Example 5. A three stage common emitter amplifier has identical transistors for each stage. Ans. If the transistor parameters are h1e 1 kQ.umho.14. Example 5. 656 Q. hre 0.USE OF TRANSISTOR EQUIVALENT CfRCUITS Example 5.13 Calculate the current gain and the amplifier input and output impedances.7a.15. 76. hre 90. hre 5 X 104. The transistor shown in the amplifier circuit in Figure 5. Example 5. 11·35. 8·83 mV. 1. 147. calculate the voltage gain and the input impedance. Assurne that the output is taken between Tr2 collector and earth and that Xe = 0.12.11 by finding the common collector h parameters and using the general h parameter solutions.13. 3·3 kQ. 650 X 1()3. 496 X 1()3.umho and negligible hre· Each stage has a collector Ioad of 2 kQ and the bias components effectively shunt the input of each stage with 10 kQ. hoe 80 . hoe 120 . hoe 80 . hre 150.umho. Calculate the overall voltage and current gain and the input impedance. Circuit for Example 5. hre 110. The Ioad on Tr2 is 3 kQ. Figure 5. Ans.umho. A long tailed pair amplifier has the circuit shown in Figure 5.15 has h1e 1 300 0. the emitter resistor is I kQ and the shunt bias components 47 kQ and 68 kQ. Example 5. 90 . 25·7 Q. Example 5. Ans.
hre 103 . Find the h parameters of the equivalent transistor (inside the broken 1ine) and hence determine Av and Z 1n with C considered as either short circuit or open circuit. 5 260. hre 5 X 104. If the transistors have h1e 1·2 kQ. Circuit for Example 5. 1 130 L 3° 30'. Example 5. Ye Example 5.18. 6 600. 63 kn.17. 162 L 168°. 105 Hz. hoe 130 ftmho. 444 kHz. 7 370. and IX 0·992. 89. lt is connected as a COffimon emitter amplifier and loaded with 1·5 kQ. Example 5.19. Calculate the voltage gain and input impedance at a frequency of 4kHz. and hoe 130 ftmho. 2·10 kil. hre 90. The equivalent T parameters of a transistor are 20 Q. calculate the voltage and current gain and the input impedance. The bias components place 15 kQ in parallel with the input and the transistor parameters are h1e 1·3 kil. Yb 350 Q. Ans.19 and hoe 125 ftmho. effective interstage shunt capacitance 500 pF. and hre 0. lf the shunt bias components total 20 kil. The amplifier shown in Figure 5. coupling capacitor 0·5 ftF. hre 125.16. 119 ftmho. hre 80. Ans. A singlestage common emitter amplifier is loaded with a 2 kQ resistor in parallel with a 0·01 ftF capacitor. Ans. 1 123 Q. 179 . Ans. 63·3.16 employs a matched pair of transistors having h1e 1 300 Ü. per stage. 20·4 x tos. Figure 5. Yc 750 kQ. 5. calculate Avm and tbe 3 db frequencies. 1 122 n.EXAMPLES effective.
Q. 218.USE OF TRANSISTOR EQUIVALENT CIRCUITS Example 5. 88 . 180 . Ans.Q. 14·6 k.Q. (b) 0·973. the current gain and the input and output impedance if (a) a load of 400 .umho. h1e 1· 5 k. take the combined shunt bias components to be 100 k. 1 270.umho.90 k. hre 4 X 104.Q. hoe 110 .8a. In each case. hre 130 and Tr2 (power) h1e 200 .Q. Assurne Zs to be 10 k. Calculate the voltage gain. The transistor parameters are Trl (small signal).Q is connected in the combined collector or (b) if the same load is connected in the emitter load of Tr2.Q. hte 70.20.Q. 4·0 . (a) 41·0. hre 103 . hoe 500 .Q. A Darlington pair of transistors is connected as shown in Figure 5.
The input impedance is the ratio of input voltage to input current. manufacturing tolerances may result in a considerable change in exact value. The passive components may vary with temperature and time. the probability of obtaining identical parameters is most unlikely. If the source internal impedance is much greater than the amplifier input impedance.c. the output impedance determines the suitability of a particular load. and of the device parameters. or that the alternating current in the Ioad will be greater in magnitude than the alternating current flowing into the input terminals.c. operating conditions. and in the event of a replacement. In addition the parameters may be very sensitive to d. current gain or both. In either case the gain will be a function of the passive amplifier components. Before defining and investigating feedback it will be useful to review these properties. In this context we mean that the alternating voltage across the load will be greater in magnitude than the alternating voltage at the input terminals. supply voltage may result in a considerable change in gain.6 THE THEORY OF FEEDBACK AMPLIFIERS In the preceding chapters. a nurober of properties of electronic amplifiers have become apparent. Impedance Amplifiers also have both input and output impedance. THE PROPERTIES OF AMPLIFIERS Amplification An amplifier will have voltage gain. The input impedance determines the suitability of the amplifier for use with a particular source. The device parameters may also vary with time and temperature. The output impedance is best compared with the internal resistance of a signal generator. Similarly. In this 181 . and a change of d. and if one is replaced by another of nominally the same value. the terminal input voltage may be so small that the amplifier output is less than the original open circuit source voltage.
These impedances will be sensitive to changes in both passive components and device parameters in the same way as the gain.1. some degree of distortion will always be introduced. Definition of Feedback In this chapter we shall see how all these amplifier properties may be modified by the use of feedback. These will occur at frequencies which are multiples of the desired signal frequency. In general a feedback amplifier or system is one in which the terminal input signal is the sum of an external signal and a feedback signal proportional to the output signal. Distortion and Noise Since all active devices are basically nonlinear. The large box represents the amplifier having a gain A. The impedance will usually increase at very low frequencies and be reduced at very high frequencies. 182 . At this stage it is convenient to work in general quantities or signals rather than in voltages or currents~ The basic feedback amplifier is shown in Figure 6. supply or due to electrical noise occurring in either active or passive circuit components.c. Phase Shift Amplifiers introduce a phase shift between input and output current or voltage. where in general A = lAI L 0. Other unwanted signals may also be introduced within the amplifier due to residual mains variation or hum in the d. In general the gain will be smaller at very low and very high frequencies. At medium frequencies this will be either 180° or 360° depending upon the number of stages in the amplifier. This takes the form of additional alternating voltages known as harmonics in the output. These variations are principally due to changes in the reactance of capacitors in the amplifier circuits. FrequencyResponse All the properties discussed above may vary with frequency.THE THEORY OF FEEDBACK AMPLIFIERS context the Ioad may weH be the input impedance of another amplifier or electronic circuit. and their configuration. The phase shift will rise with lower frequencies and be reduced at higher frequencies. particularly at high frequencies. but device parameters may also be frequency sensitive.
1. S is the input signal applied from an outside source.ßA AS So= 1.ßA Finally and the overall gain with feedback So Ar=s = A 1. Usually. lßl is less than one. the complete expression for gain with feedback Ar= lAI L O 1lßAI LO +c/> 183 (6. S' is the terminal input signal.2) .ßA) = S S' and I so S = s 1.1) So= AS' = S + ßAS' S'(l. where cp. Thus Iß~o I= lßl and ßSo leads So by a phase angle ß. The small box represents a feedback network having an attenuation in general ß = lßl L cp. but not essentially.Thus I THE PROPERTIES OF AMPLIFIERS I~~ = lAI and So leads S' by a phase angle 0.ßA Remernhering that both ß and A will be complex. and So is the output signal. s 8> s' So ß So I I ß I I F/gure 6. Block diagram for an amplifier with feedback From the diagram: S' But = S + ßSo (6.
the connection of which makes (J + cp = 180°.3 suppose ß is suchthat \ßAI ~ 1. With reference to equation 6. Simple Negative Feedback When the term is used. Amplijication with Feedback Application of simple negative feedback may be shown to improve the gain stability of an amplifier.THE THEORY OF FEEDBACK AMPLIFIERS + The modulus of (1 . and in particular negative feedback. Now \Ar\ ~ A \ßA! = 1 Tßi But ß is normally a fraction resulting from a simple resistive network and will not be subject to changes resulting from device ageing or replacement. \Arl \Ar\ > < \A\. so application of feedback can either increase or reduce the overall gain. i. the gain becomes less susceptible to the changes of parameters etc. Feedback positive \A\. These would normally be the 'medium frequencies' at which the amplifier phase shift is 180° or 360°. We shall now consider the effects of feedback.1. or temperature changes. supply changes. discussed above.2 becomes !Ar\ lAI = l + \Aß\ (6. 184 . Feedback negative It is important to realise that the same feedback system may result in negative feedback at some frequencies and positive feedback at other frequencies.) Example 6.3) This condition will be referred to as simple negative feedback. Thus equation 6. In this respect the feedback is said to be positive if the overall gain is increased and negative if it is reduced. so in general it is not correct to refer to a 'negative feedback amplifier'. d. on the properties of amplifiers discussed above. A two stage transistor amplifier is constructed using transistors with a nominal hre of 125 resulting in an overall current gain of 3 000. (Since all resistors will change in approximately the same proportion. the feedback is negative. i.e.c.\ßA\ L (J cp) may be either greater than 1 or less than 1.e. Simple negative feedback is applied using aß of 1/600. Under these conditions the ß factor is usually provided by a simple resistive network. the implication is that over the normal operating frequency ranges.
SERIES AND PARALLEL FEEDBACK
If a change in the direct supply voltage reduces hre to 80, determine
the percentage change in overall gain with and without feedback.
Weshallfirst find the change in gain without feedback. Since gain
per stage is proportional to hre, we can say that the overall gain will
be proportional to hre2
3 000
=
K(l25)2
where K is a constant.
K
=
3 000
= 0·191
1252
the new gain A2 = 0·191(80)2 = 1 220.
Percentage change in terms of the original gain
30001220
3 000
X
Percentage change
100 per cent
= 59· 3 per cent
Now applying simple negative feedback to the original amplifier
Au
and
=
A 2r =
3 000
3 000 = 500
1 + 600
1 200
I 220 = 402
1 + 600
. gam
. wlt
. h fieedback = 500  402
Percentage change m
500
X
100 per cent
= 19·5 per cent
Thus the application of feedback has reduced the percentage change
in gain from approximately 60 per cent to approximately 20 per cent,
which is a considerable improvement.
Before we can consider in detail the values of ß and A and the
effects of feedback on gain, input and output impedance, we must
lookback at the definition of a feedback amplifier.
SERIESAND PARALLEL CONNECTED FEEDBACK
From the definition of a feedback system, the terminal input
signal is the sum of an input signal and a feedback signal. With
electronic amplifiers, the feedback signal may be added in series
7
185
THE THEORY OF FEEDBACK AMPLIFIERS
or in parallel with the input signal. lf two electrical quantities are
to be added in series, the two quantities must be voltages.
Alternatively, if they are to be added in parallel, the quantities
must be currents. In general amplifiers will amplify both current
and voltage, but depending on the method of application of the
feedback signal they must be treated as either voltage amplifiers or
current amplifiers.
Thus if the feedback signal is added in series, the circuit will be
treated as a voltage amplifier and the terminal current gain will be
unchanged. The overall current gain however will be modified by
changes in input and output impedance due to the voltage feedback.
Similarly current feedback will not change the terminal voltage gain.
VOLTAGE AMPLIFIERS WITH FEEDBACK
To investigate these effects, it is convenient to introduce a general
equivalent circuit for firstly, a voltage amplifier. This is shown in
Figure 6.2, the equivalent for the amplifier being that part of the
diagram within the box.
Figure 6.2. Equivalent circuit for a voltage amplifier with voltage
derived feedback
Our equivalent voltage amplifier has three properties, an input
impedance Ztn, an open circuit output impedance Z 0 ' and an open
circuit voltage gain Av'· If this were to represent a multistage
transistor amplifier, these properties would be given approximately
by hte in parallel with the bias components for Ztn, hoe1 for Z 0 ',
and the terminal voltage gain with infinite finalload for Av'·
The feedback voltage may be obtained from the output in one
of two ways. Figure 6.2 shows the first of these; the feedback circuit,
R1, R2, is in parallel with the output and the feedback voltage is
said to be derived from the output voltage. The alternative will be
186
VOLTAGE AMPLIFIERS WITH FEEDBACK
referred ta as current derived feedback and is abtained when the
feedback circuit is in series with the amplifier laad. This is shawn in
Figure 6.4.
Voltage Derived, Series Applied Feedback
Referring again ta Figure 6.2 we shall investigate the effects af
valtage derived series applied feedback upan the averaU valtage
gain Avr and input impedances Ztnr. ß is defined as that fraction
af the output valtage that is fed back and added ta the input signal
voltage. Inspectian af the circuit shaws:
=
Rt
ß=
Rt
ßVo
R2
+ R 2 Vo
R2
+ R2
(6 .4)
Naw the gain withaut feedback Avo is given by
(6.5)
where
(6.5a)
>
since in practice Rt + R2
RL. (This is not always the case and
shauld be checked.)
Naw summing valtages at the input,
V'= es+ ßVo =es+ AvoßV'
V'=
but
Vo
=
(6.6)
es
1 Avoß
,
Avoes
AvoV = 1Avo ß
Gain with feedback
=
Vo
es
=
1  Ap~ vo
(6.7)
Fram this result, since ß is positive (equatian 6.4) simple negative
feedbackwill accur if Avo is negative. If Avo is positive, an alternative
cannectian may be used ta make ß negative.
Input Impedance With Feedback
From equation 6.6
es = V'(l  ßAvo)
187
THE THEORY OF FEEDBACK AMPLIFIERS
If this equation is divided by i the input current,
es
V'
:l = .
(1
l
 ßAvo)
But inspection of Figure 6.2 shows that V' fi is the amplifier input
impedance Ztn· Also e8 Ji must be the input impedance with feedback
Ztnr. The previous equation therefore becomes:
Ztnf
= Ztn{l
 ßAvo)
(6.8)
Thus with simple negative feedback the amplifier input impedance
is increased.
Output Impedance with Feedback
Todetermine the output impedance, we must redraw the equivalent
circuit with any external generators suppressed, and apply a generator
of E volts to the output terminals. By calculating the resulting current
I the output impedance is given by E/I. The redrawn equivalent
circuit is shown in Figure 6.3.
Figure 6.3. Circuit for determining the output impedance of a
voltage amplifier with voltage derived feedback
We shall first determine the terminal output impedance with
feedbackZur as given by Efl'. The overall resulting output impedance
can then be found by adding RL and R1 + R2 in parallel with Zor.
If Zs + Ztn ~ R2 and if Zs ~ Ztn we can say
V' = ßVo = ßE
If not, R2' should be used in determining
R2
1
R2(Zs
+ Ztn)
(6.9)
ß'. Where
= ==='R2 + Zs + Ztn
188
VOLTAGE AMPLIFIERS WITH FEEDBACK
V'= ß'E
and
Zs
Ztn
Ztn
+
Formost practical purposes equation 6.9 is satisfactory.
Writing a mesh equation:
E  Av' V' = I'Zo'
From equation 6.9
E(l  ßAv') = I'Zo'
E
Zo'
Zor = ji = l  ßAv'
(6.10)
The overall output impedance can thus be found from
1
1
Z = Zor
1
1
+ RL + R1 + Rz
Alternatively the loaded output impedance without feedback, and
the loaded voltage gain Avo may be used in equation 6.10 and the
same result will be achieved. In practice however where simple
negative feedback has been used Zor ~ RL < R1 + Rz and the
shunting effect of RL and R1 + R2 is negligible.
Current Derived, Series Applied Feedback
We must now consider the effects of current derived feedback.
The required circuit is shown in Figure 6.4.
I
.;!, 1
I··
RL
l
Vo
I
I
c0 t
E
I
I
I
v,r.,       J
Figure 6.4. Equivalent circuit for a voltage amplifier with current
derived feedback
In this case the output current ftows in both the load and a resistor
RF. The resulting voltage ioRF is applied in series with the signal as
ßVo.
189
THE THEORY OF FEEDBACK AMPLIFIERS
Thus if RF ~ Zs
+ Ztn (as is usual)
ßVo
Vo
ioRF
ioRL
=
ß=
RF
RL
(6.11)
In addition, Avo is modified from equation 6.5 to,
Av'RL
Avo
= RL + RF + Zo'
(6.12)
With these new values for Avo and ßthe results for gain (equation 6.7)
and input impedance (equation 6.8) are unchanged.
Output lmpedance with Feedback
For the output impedance calculation the generator E (shown in
Figure 6.4) is applied and the source es suppressed. As with the
voltage derived case RL can be included afterwards if required.
Now since I is in the opposite direction to io,
V'
=
IRF (if Ztn
+ Zs ~ RF and Zs ~ Ztn)
and writing the mesh equation
E Av'( IRF)
= l(Zo' + RF)
arranging and collecting terms,
E
or
Zor
= IZo' + /RF(l
 Av')
= ~ = Zo' + RF(l Av')
(6.13)
With simple negative feedback this can result in a considerable
increase in terminal output impedance. However the shunting effect
of RL on the overall output impedance normally makes this effect
unimportant.
Example 6.2. An amplifier having input impedance 1 kQ and open
circuit output impedance 10 kO has a voltage gain of 800 when
loaded with a 4 kO resistor.
A feedback voltage is derived from (a) a potential divider chain
in parallel with the Ioad having Rt 9·9 kO and R2 100 n, or (b) a
100 n resistor in series with the Ioad. In each case the series connection at the input ensures simple negative feedback.
190
m. For the amplifier without feedback. so we must find RL' and Avo from equations 6. 623 Avr = 1 + 623 X 0·01 = 86 ' 2 and Zrn = 1(1 + 623 X 0·01) kQ = 7·23 kQ 191 . the voltage derived feedback.86 = 623 from equation 6. Since Rl + R2 = Avo Since R2 ~ (Zrn +Z 8).5 and 6. and the change in output current if the amplifier is driven from a source of 1 mV e. for both methods.5 Av'4 800 = 4 + 10 Av I = 800 4X 14 = 2 800 Now consider case (a).4.Applying equation 6. the overall terminal voltage gain. The effect of R1 + R2 cannot be neglected. = 2 800 10 X 2·86 + 2. from equations 6. 10 X 4 RL = ~ = 2·86 I 10 kQ.f. 100 ß= . and internal impedance 1 kO. the modified input and output impedance. we must first find the open circuit voltage gain Av'· .14) To investigate the effects of the feedback circuits.A and V'= i1nZ1n = 0·5 p.VOLTAGE AMPLIFIERS WITH FEEDBACK Determine.5a.A X 1 kQ = 0·5 mV Since .7 and 6.= 10000 0·01 As the feedback is negative. the input current i 1n is given by: 1 mV i1n = 1 kQ + 1 kQ = 0·5 p.8. Voltage gain = 800 Output voltage v0 = 800 O~tput X 0·5 mV = 0·4 V 0·4 current io = 4 000 A = 100 pA (6.
1 795 = 192 .23 311 mV The output current is given by vo/RL'. it would appear that the current gain has increased! This is in fact so.THE THEORY OF FEEDBACK AMPLIFIERS Now and New input current = Overallinput voltage Vo New output current 8 ~2~:n = = 0·122 0·122 f1.7 and 6. 6. and the reason is simply that the Ioad has been changed from an RL of 4 kQ to the RL' of 2·86 kQ. Zor =I lOkQ + 6. and using equations 6.24 In part (b) of the example we have current derived feedback.12. ß= Avo = RF 100 n RL = 4 kQ = 0·025 2 800 X 4 14. If the original input current 0·5 flA was applied.A Note that although this current is considerably less than that found for the amplifier without feedback (equation 6.A x 7·23 mV = 0·88 mV = 0·88 X 86·2 mV = 75·8 mV = 75·8 mV 4 kO = 19 fJ. 6. the current gain is unchanged.A With reference to equation 6.A = AvrV!n 1 kQ{l X = 0·5 ftA X 1 kQ{l = + 6·23) 623 + 6·23) X l + 6. To return to the problem.8 we find.11.14.23 = 1·38 kQ The overall output impedance including the Ioad is found by shunting Zor with RL' 1·38 X 2·86 Overall Zo = kQ = 930 Q 4 . the overall input voltage is given by: V!n and Vo = 0·5 fJ.14). io 311 mV = 2·68 kQ = 109 fJ.10. we must find the modified output impedance using equation 6.
Finally for the output impedance with current derived feedback. Ztn 1 000 0.13.1. From which Zor = 10 kO + 100 0(1 + 2 800) = 290k0 . Av 38·1 Ztn 20·8 kO Zo 3 9500 193 .VOLTAGE AMPLIFIERS WITH FEEDBACK 795 = 1 + 795 X Ztnr = 1(1 + 795 Avr and = 38 ' 1 0·025) kO = 20·8 kO 0·025 X For the output current calculation and 1mV .8 mV .1. lo  As before. This is given by 4 X 14 10 = 2·86k0 Summarizing these results : No feedback: Av 800. we refer to equation 6. as has the input current. _ ~ _ 38·1 X 20·8 A _ 9. 1 A RL 4 X 21·8 f. f. Av 86·2 Ztn 7·23 kO Zo930 0 (b) Current derived feedback. 290 X 4 Overall output tmpedance = 294 kO = 3·95 kO This resu1t shou1d be compared with the loaded output impedance of the original amplifier without feedback. as a result of the increase in input impedance. ltn = 21·8 kO Vtn = v0 itnZtnr = 38·1 X 20·8 = 21·8 mV 20·8 21 . Z 0 2 860 0 (a) Vo1tage derived feedback. the current gain has not been reduced.
Fora multistage transistor amplifier.THE THEORY OF FEEDBACK AMPLIFIERS Thus in each case negative feedback reduces the voltage gain and increases the input impedance. /li0 io YL Y. The equivalent circuit properties are an input admittance Y1n. but with current derived feedback the output impedance is increased.n  v' G2 Figure 6. Yo' =::= hoe for the final transistor. The equivalent circuit for a current amplifier is shown in Figure 6.n=:=h le in parallel with the bias components. CURRENT AMPLIFIERS WITH FEEDBACK If the feedback signal is applied in parallel with the input signal.'. a terminal output admittance (excluding final Ioad) of Y0 '. A.' === hre of the last stage multiplied by the current gain of all preceding stages. and the amplifier must be treated as a current amplifier. Circuit for a current amplifier with current derived feedback As with the equivalent voltage amplifier shown in Figure 6. With voltage derived feedback.2 the amplifier is represented by that part of the diagram enclosed in the box. 194 . the output impedance is reduced.5. I' is l•. and a short circuit current gain A. the two quantities to be added must be currents.5. these would be given by: I Y.
17) Ato = Yo' + YL' YL(G1 + G2) where (6. The feedback current G" ßio = Gl G2 io + ß= Gl G2' + G2' (6. Adding the currents at the input we obtain and But I'= is + ßio = is + ßAtol' is = /'(1 . and G1 and G2 form the current derived feedback network.ßAto) is .= Ytn and .ßAto . the valtage across G1 Bic = (v' .ßAto Current gain with feedback Au = 1 :ßAto (6.= Ytnr Vt Vt the input admittance with feedback Ytnr = Ytn(l .19 by the terminal input valtage Vt.CURRENT AMPLIFIERS WITH FEEDBACK Current Derived Current Feedback In Figure 6.= . is I' . YL is the finalload.20) To find the input admittance with feedback we divide equation 6.21) 195 .v1)G2 = v'G2 (6.19) Atois 1.18) YL' = h + G1 + G2 which in practice may often be approximated to YL.ßAto) Vt Vt I' is But .16) Next we require Ato the current gain without feedback.(1 . lo = (6. I = 1. This is given by: At'YL' (6.ßAto) (6.5.15) Note: The input admittance Ytn is negligible in the determination of ß since Vt is very much less than v'.
with simple negative feedback.e. I' Y. This circuit arrangement is shown in Figure 6. First note that if Ys ~ Y1n where Ys is the source admittance Gz' if not use = Gz + Y1n + Ys G2(Y1n + Ys) (6. and the input admittance is increased (i.20 and 6. The reader may have noticed by this stage. The effect of YL. Calculation of the resulting output voltage V Ieads to the output admittance Yor being calculated from Yor = 1/ V. G1 and Gz may be included afterwards to give the overall output admittance if required. Output Admittance with Feedback For the output admittance with feedback. Circuit for determining the output admittance of a current amplifier with current derived feedback First we shall calculate the terminal output admittance given by 1"/V. the current source must be replaced by its internal admittance. the current gain is reduced.6.22) ß" where ß" = ß'Yln Ys + Yln 196 . that this derivation is proceeding along identical lines to that for the voltage amplifier with voltage derived feedback.5 can be regarded as the dual of Figure 6. Since admittances and currents are being used Figure 6.2. and a generator of I amps connected to the output terminals.24.THE THEORY OF FEEDBACK AMPLIFIERS Note from equations 6. the input impedance is reduced).n G2 Figure 6.6.
197 ß and . Y. Yor( G1 + Gz') Overalloutput admlttance = Yor + Gt + G2 ' + YL (6.ß"At' (6. is shown in Figure 6.ß"A'I" = = VYo' VYo' and the output admittance with feedback Yor is given by I" Yor = V = Yo' 1 .'RENT AMPLIFIERS WITH FEEDBACK Now writing a nodal equation for V: I"+ At'I' I".CUR.7. Voltage Derived Current Feedback The alternative connection for current amplifiers. we must first define A 10 • Examination of the circuit shows that io and since ßio Vt = VoYL = VoGF ~ Vo.n . .24 will approximate to YL. resulting in voltage derived feedback. I i~ YL iV I I I Figure 6.7. is I' I ~Ys io y.24) With typical values. YL will be the predominent term and equation 6.23) The overall output admittance can then be obtained by adding G1 + Gz' in series with Yor and YL in parallel with the result. Circuit for a current amplifier with voltage derived feedback As with the current derived feedback.
2 is to have feedback connected in shunt with the i/p by one of two alternative methods. I + GF'(l. Example 6.25) Also. the overall current gain. The amplifier described in Example 6. and calculating the resulting V. Ytn Ys + GF( Ytn + Ys) GF + Ytn + Ys and putting A ~ " = A. In each case the connection of the feedback network ensures that simple negative feedback is applied.3. (6. Under these conditions I' = VGF' X where GF' = Ytn. suppressing the source generator.THE THEORY OF FEEDBACK AMPLIFIERS Thus (6. input admittance.At") (6.26) Using these values of ß and A10. Output Admittance The output admittance is calculated in the usual manner by connecting the generator of I amps to the output terminals.21 respectively.1 Ytn Ytn + Ys Overalloutput admlttance =V= Yo' + h Writing a nodal equation and . the current gain and input admittance with feedback may be determined from equations 6.27) Once again the last term will usually predominate and which with negative feedback makes the output impedance very low. 198 . and output admittance. Assuming Ys ~ Ytn determine in each case. The two networks are (a) a 200 Q resistor (Gt) connected in series with the load and a 50 kQ resistor ( G2) connected in parallel with Gt to the input termirrals and (b) a 100 kQ resistor (GF) connected directly from the output terminal to the input terminal.20 and 6.
umh o 2 800v' 10 000 isc = ltn :.umho = + 251 (representing an input impedance of 560 Q).umho ZL For part (a) app1ying equations 6. 6.umho = 238 + 100 = 197 Now app1ying equations 6.20 and 6. 6.umho 1 251 = 238 .100 .18 Gz 1 = SO kQ = 20 .16. 197 A~t = 197 = 110 1 + 251 and Ytnr = 1 000 ( 1 197) . Also 1 YL = .umho 20 ß = 20 + 5 000 = and YL . and itn is the current flowing into the input terminals without feedback.umho ln Yo At . .= . 199 1 786 . Ato = 250 (5 020) 250 + 5 020 280 X 238 G1 and 1 = 200 Q = 5 000 .= 250 .umho . remembering that ßA 10 will be negative.CURRENT AMPLIFIERS WITH FEEDBACK We must first find the parameters of our equiva1ent current amplifier.17 and 6.21. . 1 = Zo' = .15.= 280 V I 000 where isc is the current that wou1d flow in a short circuit connected across the output terrnina1s. 1 Ytn = 2 = 1 000 .
Au= 194 194 = 22·1 1 +25 Y1nr = 194) 1 000 ( 1 + 25 = 8 780 . This should be compared with the original overall output admittance of 100 + 250 = 350 .22 and 6.24: Yo = 56 X 5 020 5 020 + 56 + 250 = 305 . ßand A10 are obtained from equations 6. For the voltage derived feedback in case (b).' and GF' 100 + 250 + 9·9(1 (representing an output impedance of 320 Q).umho = 56 . applying equations 6.uV (representing an input impedance of 114 Q).21.umho (representing an output impedance of 3·28 kQ).umho 1 25 280 X 250 250 + 100 + 10 = 194 from equations 6.27.umho 1 + 251 and the overall output admittance from equation 6. GF' = 1001kQ 10 ß = 250 = and Alo = = 10 .20 and 6. = Yor = At" = = 9·9 .26. 200 .umho 101\ 0 A.umho + 280) = 3 130 . The output admittance with feedback can now be determined from equation 6.25 and 6.THE THEORY OF FEEDBACK AMPLIFIERS For the terminal output admittance.23 19·6 ß'=5000+ 19·6::!:=251 100 Yor = 197 .umho.
a reduction (say) in terminal output voltage causes a reduction in ßv0 and hence in V'. when the feedback network is connected in parallel with the output (voltage derived). and the parallel current by ßAtv Y 1n leading to the increase in input admittance shown by equation 6.8. A parallel addition would similarly be expected to reduce the input impedance. the series voltage will be ßAviZin leading to the result given in equation 6. If however the feedback signal is added in parallel. We shall now review these results and see how the effects upon input and output impedance can be explained. (b) Voltage derived feedback At22·1. Y0 350 . and since the input voltage V' is given by iZtn. the change in feedback current is amplified 201 . Ytn 1 786 . but in this case the reduction is amplified by the active circuit. since a larger current change has been produced by the given reduction in terminal voltage.umho. This makes it appear to have a lower impedance. the output impedance is reduced. Whenever the feedback signal is applied in series with the input.umho.umho.umho. A parallel circuit would normally reduce an impedance. In this case the input current I' is given by v Y1n. and the input impedance is increased. Y1n 1 000 . the current gain and input impedance are reduced.EFFECl'S OF FEEDBACK ON AMPLIFIE!t IMPEl>ANCES Summarizing these results: No feedback 280 X 250 At = 250 100 = 200. + (a) Current derived feedback At 110.umho.21. Yo 3 130 . With a current amplifier. With negative feedback provided by a positive ßand negative Av. Y1n 8 780 .umho. Yo 305 . SUMMARY OF EFFECTS OF FEEDBACK ON AMPLIFIER IMPEDANCES In the previous section we have seen by mathematical analysis how negative feedbackmodifies the gain and input and output impedances of amplifiers. A series addition would be expected to increase the input impedance. Now moving to the output terminals. the voltage gain is reduced. Phase inversion in the amplifier leads to an increase in Av'V' driving more current through Zo'.
naturally increases the output impedance. since the gain is 5 L 0. These effects are shown in equations 6. FREQUENCY RESPONSE OF FEEDBACK AMPLIFIERS We now come to the question of frequency response of amplifiers with feedback. At three different frequencies. The procedure in each case is to draw respective1y vectors representing V'. determine for each case whether the feedback can be said to be positive or negative. The effects of ßA > I with positive feedback will be discussed in the next section.ßvo and by vector subtraction the es vector may be determined. In this case the volt drop across the feedback network is effectively amplified in a similar way to the modification of the input impedance. and that is that ßA must be less than one.4. Positive Feedback All the effects discussed above have been the result of negative feedback. 202 . ßvo is one tenth of Vo and the angle cp is expressed with v0 as the reference vector. oo. Thus for both current and voltage amplifiers current derived negative feedback increases the output impedance.27. First we must be clear about the information supplied. the values of 0 and cp are respectively (a) +135°. If however the feedback is positive all the effects are reversed. 135°. (c) +45°.10 and 6. An amplifier having a voltage gain of 5 L 0 employs a feedback circuit having aß of 1/10 L cp.ßA) shou1d be a vector calcu1ation. Then since V'= es+ ßvo es= V'. 90°. v0 = A L 0 V'. the output vo1tage is five times the terminal input voltage V'. and in Chapter 8. and the feedback signa1 ß L cp Vo. Remernher first that both ßand A are vector quantities. (b) 90°. Example 6. By means of sketched vector diagrams. A feedback network connected in series with the output (current derived). and the angle 0 is expressed with V' as the reference vector. and that the calculation of Af(l . There is a limit to the amount of simple positive feedback that can be applied. This may be illustrated by an example.THE fHEORY OF FEEDBACK AMPLIFIERS increasing the current in Yo' in the same way with a similar result.
11 for Example 5. An alternative procedure is to consider the locus of the gain vector as shown in Figure 5. IJvo ~ ~vo t's (b) f5ro es Yo Yo ftvo Figure 6. ß is usually constant having an angle cp of 0° or 180°. and by measurement of vectors or by calculation. Thus the procedure outlined in Example 6. Vector diagrams for the feedback amplifiers in Example 6.4 could be repeated for a large number of frequencies.8. 203 . the gain and phase shift with feedback determined.8 shows the constructed vector diagrams from which the results are (a) negative. (b) negative.FREQUENCY RESPONSE OF FEEDBACK AMPLlFIERS The gain without feedback is given by feedback by lv0 fesl.4 To investigate the frequency response of a multistage RC coupled amplifier. and (c) positive. Forasinglestage amplifier. if or if I::1 < I~~~ I::1 > I ~~~ lvofV'I and that with the feedback is negative and IV'l < lesl the feedback is positive and IV'l > lesl Figure 6. the gain and phase shift at each frequency could be determined by repeated use of the general solutions found in Chapters 4 or 5.6.
then the required addition will lead to yet another circle with its origin at X. having its origin on the circumference. From this we shall deduce the locus of the gain vector for amplifiers with two or more stages. and /h respectively. fm . 02Q.THE THEÖRY OF FEEDBACk AMPLIFIERS this is a circle with the origin on the circumference.(l.ßA) vector.ßA) vectors corresponding to f1. the scale of Figure 6. Now if Am were 100 L 180° and ßio L ()0. Since this is the locus of (. and 02R are the (1 .ßA) vector would remain at 01.ßA).9b iscorrect. The origin for this locus is 01 and note that the reversal results in /1 appearing at the top of the diagram instead of /h as in Figure 6.atfm.9a ft (a) Figure 6.ßA).9a. 02P. Next the locus of ßA is required.9b. for each frequency.ßA) and divide it into the A vector for the same frequency. The origin of the (1 . origin 02 is the origin for the (1 . we must add the vector (1) to each vector ( ßA). each vector must be reversed leading to yet another circle as shown in Figure 6.ßA)is3 L 0°. with the vectors corresponding to three frequencies shown. we must find.9a. Locus of gain and feedback vectors for single stage amplifier with feedback also represents the locus of ßA provided of course that the scale is changed. and we require (1 . But first we will use this representation to investigate the frequency response of a single stage amplifier having simple negative feedback at medium frequencies. the locus of the vector ßA will be another circle.9. and the same result may be achieved by shifting the origin by one unit to the left. The resulting diagram becomes somewhat confusing.9b. Assuming ß to be a simple fraction having zero phase shift.andatjiandfhitis2·28 L +26·5° 204 . Thus Figure 6. This is shown in Figure 6. Assuming that the values of ß and A are such that this unit vector is given by the line 01X. In Figure 6. the vector (I . For this. To obtain the response ofthe amplifier with feedback.
If Figure 6.9b shows that 11 . the connection of the ß network ensuring simple negative feedback at medium frequencies. A table may now be drawn up.10a and b.5. The length of these vectors may then be squared and the angles doub1ed to construct the cardioid which is the 1ocus of the gain vector for the two stage amplifier. it is applied over two or more stages.4. At our medium frequency (5kHz) point ßA = (3 520 X 8·5 X I04) = +3 Since this is represented by 4 cm. Example 6. draw the overall gain and phase response. This is shown in Figure 6. the locus of the gain vector for a two stage amplifier must be determined.= 880. Before we can apply the methods outlined above. This may be constructed by drawing a circle of any convenient diameter.ßA) Ar 100 L 180° 3 L oo 33·3 L 180° /h /I 70·7 L 135° 2·28 L 26·5° 31 L 161·5° 70·7 L 225° 2·28 L 26·5° 31 L 198·5° Thus changes in both gain and phase shift have been reduced by the application of feedback. The origin must be shifted by one unit or by t cm to 02 shown on the diagram.FREQUENCY RESPONSE OF FEEDBACK AMPLIFIERS and 2·28 L 26·5° respectively.ßA I can never be less than one. the scale must be one centimetre represents ! of one unit. the resultant vectors will be IAI2 L 20.ßA) by changing the scale and shifting the origin. In practice if feedback is used to improve frequency response. the gain must be squared.10b is to apply to the problern in hand. having transistors and components identical to stage one of the amplifier analysed in Example 5. The scale must therefore be 1 cm = . Examination of Figure 6 . employs negative feedback to improve the frequency response.1 the frequencies corresponding to points on the cardioid can be inserted.6. A two stage amplifier. Since at each frequency. This will now be illustrated by an example.ßA) and (1 . If the ß factor is 8·5 x I04 . say 2 cm and drawing the gain vectors for a single stage at I oo intervals. and hence find the 3 db bandwidth of the system. the medium frequency gain vector must represent 3 520 (59·3)2 or 3 520. 205 . showing the gain and phase shift with and without feedback /m A (1 . With reference to Table 6. the feedback is therefore negative for all frequencies. The same cardioid may be used to represent (.
ßA).1 f(kHz) IArl Or 0·033 0·05 342 869 129° 87° 0·01 972 32° 0·02 870 15° 0·03 879 70 206 5 61 122 245 490 735 880 879 870 972 869 342 oo 70 15° 32° 87° 129° .THE THEORY OF FEEDBACK AMPLIFIERS For any frequency the complex values of A and (1 .5 Therefore at 122kHz gain with feedback 2 820 L 53o o L 38 o = 870 L 15 Ar= 3. Table 6.24 Repeating this procedure for the remaining spot frequencies shown lead to the results shown in Table 6.ßAI = 4·18cm x! = 3·24 . Construction vectors (a) A scale= 880 200Hz L___j (1.ßA) may be found and the resulting gain and phase shift determined.10. cp' = 38°. For examp1e consider the 122kHz point: The angle of (1 lAI = 3·22 cm x 880 = 2 820 () = 53° 11.6 A) scale = 0·75 (b) L___j Figure 6. Locus of gain and feedback vectors for Example 6.1.
11 shows graphs of gain. lAI at any particular frequency is divided by lAI for the medium frequency. :.11. it is convenient to plot graphs of gain. Lower 3 db frequency Upper 3 db frequency No FB FB 130Hz 38 Hz 150kHz 600 kHz Bandwidth ~150kHz ~600kHz From these results we can see that from a practical point of view the bandwidth of an RC coupled amplifier is given by the upper 3 db frequency.. ~co (/) co· fU~ <llOl o. Graphs of gain.I:.lfl ~ +250° 10 +150° 0·8 0·707 (3db) 0·6 +50° 00 50° 0·4 1500 250° Figure 6... 207 . The increase in bandwidth is most evident from the comparison of relative gains with and without feedback..FREQUENCY RESPONSE OF FEEDBACK AMPLIFIERS To compare these results with the gain without feedback. relative gain and phase <ll <ll > ..r:.. Figure 6.. From these curves the following results may be obtained. phase shift. relative gain and phase shift for two stage amplifier with and without feedback shift plotted against log frequency with and without feedback. To obtain the relative gain in each case. and relative gain against log frequency.:c .
Oz.11. Remernhering the definitions of negative and positive feedback we can see that in this region the feedback changes from negative to positive. ßA and hence ß must be 208 .ßA) for three and four stage RC coupled amplifiers employing overall feedback. Examination of Figure 6.ßA) vector is nearly tangential to its locus. In each case the distance to 01 must be unity.10b shows that this will occur when 11 . are possible origins for the (1 . and a study of multistage feedback amplifiers will provide a useful introduction to the phenomena of oscillation.ßA) scale.10b shows that a change in frequency reduces the gain Ao. This can only occur if 02 is moved to 01 representing infinite feedback which is impossible. In general this is true for any two stage RC coupled amplifier with negative feedback at medium frequencies.ßA) locus depending upon the amount of feedback applied. This gives rise to the two 'humps' shown in Figure 6. Figure 6. 03.12b the amplifier must have four stages. Such a circle should be drawn about the 02 origin using the (l .ßAol is also reduced.THE THEORY OF FEEDBACK AMPLIFIERS It is of interest to consider 'how' the application of feedback has increased the bandwidth. allowing the gain with feedback to be maintained. If the origin for (1 .ßAI is less than one which obtains when the locus ofthe (l . then 11 . the components in the coupling networks vary from stage to stage. and since it is not symmetrical. the 'amount' of negative feedback 11 . was cut by the locus. 04 etc. but it cannot rise to infinity. the origin for ßA. This is the point corresponding to infinite and zero frequency where in any·case the gain is zero. Consideration of Figure 6. For the locus shown in Figure 6. Figure 6.ßA). Over certain sections the reduction in negative feedback is morerapid than the reduction in gain since the (l . This is shown on the graphs of actual gain where Ar becomes greater than Ao. In each case the medium frequency point occurs in the right hand plane since the feedback network ensures negative feedback at this frequency.ßA) vector falls within a circle of unit radius. 02. The origin 01 isthat applicable to the ßA locus.12 shows the form of the locus of ßA and (1 .12a is the locus for an amplifier having three identical stages and coupling networks. we can expect the gain to be greater than that without feedback. In this region the locus only touches the real axis at 01. Thus at the highest and lowest frequency ranges. Iostability of Feedback Amplifiers If feedback is applied over more than two stages the above is not necessarily true. and if this distance is smaller.ßA I would be zero and the gain would become infinite.
for those frequencies less than /1 and those greater than /2 the gain will be increased.ßA I is about .ßAI is at 03.ßAI is zero and Ar A A =. but it can be less than one. Locus of (1 .ßA) for three and four stage feedback amplifiers For this particular case the minimum value of 11 0·75.FREQUENCY RESPONSE OF FEEDBACK AMPLIFIERS larger. With the origin at 02 li . (a} fm (b} f reducing Figure 6. moving the origin from 02 to 03 with the three stage case. 209 . since the gain is infinite. The unit circle drawn on 02 shows the frequency range over which this applies. oo Thus at frequencies /3 and / 4. In other words the amplifier has become a generator of alternating voltaJ!eS at two frequencies simultaneously. Now Iet us consider what happens as the feedback is increased.= 0 = 1ßA .ßAI cannot equal zero.e. i. 03 is the origin for (1 ..ßA) with more feedback than 02. giving positive feedback. no input is required for an output to be present. i.e. If however ß is increased so that the origin 11 then the minimum value of li .12.
This discussion will also loosely apply to the introduction of electrical noise and 'mains hum' within the amplifier. Such distortion is usually expressed as a 210 . three times and higher multiples of the signal frequency. that is troublesome. or second harmonic. lf the feedback is increased still further moving the origin to 04. the amplifier will be unstable if the locus encloses or cuts the origin. From the above discussion we can see that feedback may be safely applied over two stages of amplification. With most devices it is only the double frequency component. REDUCTION OF DISTORTION BY FEEDBACK The last property of amplifiers to be considered is the introduction of harmonic distortion by the amplifier. the amplifier will be unstable. These remarks are all applicable to the four stage case as well and a summary of the results referring to Figure 6. An alternative form known as Nyquist's criterion of stability is given as: If the locus of ßA is drawn in the complex plane and it encloses or cuts the point 1 + jO. This aspect will be considered in more detail in Chapter 8. positive FB at high and low frequencies.12b is given below.THE THEORY OF FEEDBACK AMPLIFIERS Since these oscillations are present at all times. but also voltages at twice. we might expect a stable condition again. the amplifier can no Ionger be used for amplification and is said to be unstable. Unstable.ßA) is drawn in the complex plane. but in practice the amplifier gain adjusts itself until the oscillating condition is maintained. In general the nonIinearity of the device characteristics result in harmonic distortion. bandwidth increased. Criterion of Stability From the situation discussed above we can deduce the condition under which a feedback system is stable as follows: lf the locus of (1 . but if it is applied over three or more stages. oscillation at /1 and /2 simultaneously. care must be taken to ensure the stability of the amplifier. oscillation at /1 only. This means that the output voltage will contain not only a voltage at the signal frequency. but in some instances the third harmonic must also be allowed for. U nstable. Origin Position 0203 Oa04 040s and beyond Result Negative FB over certain ranges.
Consider the system shown in Figure 6.ßA) and (6.30 into equation 6. for example an amplifier having an input of IO mV at I kHz might have outputs of 5 V at 1kHz.29) S'(l .13. This would be expressed as 10 per cent second harmonic distortion and 2 per cent third harmonic distortion.ßA Substituting for S' from equation 6.30) 1. and O·I V at 3 kHz. 0·5 V at 2kHz. Si s' Distorting amplifier S 0 :AS'+ D ~So Figure 6. and the input signal amplitude must be increased to restore the output signal to the original Ievel. So = AS' +D (6.28) Now writing the usual equation for the input junction: S' =Si+ ßSo = St + ßAS' = St + ßD S' = S.13.REDUCTION OF DISTORTION BY FEEDBACK percentage.28 So = AS' 1 . Under these conditions the distortion D can be greatly reduced by feedback.ßA AßD ßA + 1211 +D . So must contain a component AS'. The overall distortion D is then given by: D = y(D22 + Da2) = v(100 + 4) = 10·2 per cent In Chapter I we found that appreciable distortion only occurred with large signals and that if the amplification at the output is reduced the distortion is also reduced. In addition there will be a distortion signal D whose amplitude is proportional to the amplitude of So. With a negative feedback system the gain is reduced. Block diagram for demonstration of the effect of feedback on nonlinear distortion Since we have an amplifier. + ßD + ßD (6.
This will also Iead to an explanation of how negative feedback reduces harmonic distortion. using the methods described in Chapter 1.ßA With simple negative feedback equation 6.AßD I . 2 I• Feedback . It is connected in series with a Ioad RL and a cathode bias resistor RK having values 38 kQ and 2 kQ respectively.ßA D (6.. Figure 6. Weshall now verify these feedback formulae by solving a problern graphically. But from equation 6. Example 6.14..6 212 .THE THEORY OF FEEDBACK AMPLIFIERS Putting the last two terms over a common denominator ASt So = I .6.32) lf St is increased to restore the output to the originallevel then D will also be restored to the originallevel. A triode amplifier has the anode characteristics shown in Figure 6.31 becomes: ASi So D = 1 + ßA + 1 + ßA (6.31) + 1.14.32 the output distortion is now only 1 : ßA' and has therefore been reduced by the use of feedback.ßA ASt = I . Characteristics and graphical solution for Example 6.ßA AßD + +D .
since the value of D depends only upon the ratio of ß to ß'. It may be shown that if the two half cycles are denoted ß and ß' as shown in Figure 6. In Figure 6. Severe distortion :. In Chapter 1 it was shown that the distortion introduced amplified one half cycle more than the Ä = Ä' A»tf :. Input and output waveforms for distorting amplifier other. these lengths may be measured directly along the Ioad line. supply is 200 V and the signal es is sinusoidal. and es is 2·5 sin wtV. Compare the results with those obtained by feedback theory. and (b) when negative feedback is applied by the removal of the decoupling capacitor and e8 is increased to give the same output voltage.14 the following steps have been taken: 213 .15a shows the applied distortionless signal varying about zero. Now toreturn to the problem. No distortion (b) (a) Figure 6. on Figure 6. Calculate the voltage gain and harmonic distortion (a) when RK is decoupled by a suitable capacitor.T.REDUCTION OF DISTORTION BY FEEDBACK The H. Before this problern can be tackled two facts concerning second harmonic distortion must be stated. Figure 6. and ß and ß' are the two peak values when the signal is present.15 then the fundamental or required signal output is given by ß+ ß' 2 and the second harmonic distortion by ßß' 2(6.15. + ß') x 100 per cent Further.15b Vao is the quiescent anode voltage in the absence of any signal.
78·5 2 = 57·75 V which is approximately the same as in the first case.THE THEORY OF FEEDBACK AMPLIFIERS (a) The d.c. the operating point moves between VBK 0·5 V to VaK 5·5 V. ß= ZF ZL = 2 38 = 1 19 (c) For the graphical solution we must plot two further lines of VaK = hRK t es and hRK es Since the resultant v0 must be unchanged. Now the VAK excursion is from 78·5 V to 194 V. 17 x 100 per cent ~ = 5·7 per cent Now when the capacitor is removed.76 2 = = 57·5 V and therefore 57·5 Valtage gain Avo = 2 . This form of circuit will be discussed in the next chapter. The resulting construction lines are shown on the graph. The change from 40 kQ to 38 kQ is small and has been neglected. The resulting d.11. Peak v0 = 194.D. Trial values show that if ·es is increased to 5·5 V a satisfactory solution is obtained.c.c. has been drawn. but ß may be obtained from equation 6. operating point is VAK 139 V. VaK 3 V. load line for 40 kQ.5 = 23 Also along the load line = 2·32cm ~~ = 1·85 cm 047 S. load line. bias line for 2 kQ. Peak Vo 191 . current derived negative feedback is applied. = 2 x 4 .c. From the graph the extremes of the VAK excursions are 76 V and 191 V. But now Avt = 57·75 5:s = 214 10·4 .H. as has the d. (b) Without feedback. along a 38 kQ a. h 1·5 mA. es must be increased.
= 2 x 4 . Figure 6. since the output voltage is also slightly larger in the second case. In practice this merely means the second harmonic component is 180° out of phase with that introduced by the valve itself. When this signal is amplified and inverted it tends to cancel out the distortion produced by the amplifier. with and without feedback and for ßvo and V' in the feedback case.16 shows the actual waveforms for es and v0 . v0 with feedback is slightly distorted. In this chapter we have discussed the effect of feedback upon amplifiers. ß and ß' being 2·4 V and 2·8 V respectively.D. From this we can determine how feedback reduces distortion.REDUCTION OF DISTORTION BY FEEDBACK If feedback formulae are used Avt 23 = . The terminal input signal V' given by es + ßvo also contains a distortion component. The distortion in Vgk can actually be measured on the graph.H. negative feedback results in an input signal distortion such that the distorting amplifier produces an undistorted output. In general the application of negative feedback appears to have considerable advantages provided the changes in impedance and gain are acceptable.24 x 100 per cent = 2·8 per cent But from feedback formulae Dt = 5·7% n 1 + 19 = 2·6 per cent The measured distortion is slightly larger.19 + 23 1 = 10·4 Also working from the Ioad line. Notice that in this instance ß' is the larger implying a negative distortion. In other words. Caution must be taken if the feedback is applied over more than two stages or instability may result. ßv0 is also slightly distorted. ß = 2·24cm ß' = 2·0cm 0·24 S. In the 215 .
80.7. and how the properties of such amplifiers may be determined.6 EXAMPLES Example 6. A two stage amplifier is designed to use a matched pair of transistors having hre in the range 50 to 150. Ans. In a prototype.v0 feedback V0 175 no feedback 150 125 100 75 Figure 6. Waveforms for Example 6. V +5 ~Vo / (d istorted) es no feedback (undistorted) es feedback (undistorted) ~k=es+pvo 5 200 (distorted) . 0·012. transistors having hre 100 produce an overall current gain of 2 500.THE THEORY OF FEEDBACK AMPLIFIERS next chapter we shall consider how negative feedback may be applied to practical amplifier circuits. 216 . Feedback is to be used to minimize gain variation with spread of hre such that in no amplifier will the gain be more than I 0 per cent less than that with the highest gain. Calculate the required ß factor and the resulting maximum gain.16.
Ans.EXAMPLES Example 6. This consists of a 200 n resistor placed in series with the equivalent output generator and load. Find also the amplifier outputimpedance when the driving source has an internal impedance of2000. (a) by finding At and Ytn with feedback. 169 0.8. For the circuit shown in Figure 6. 91·2. Voltage derived feedback is provided by a potential divider chain of 19·5 kO and 500 0. 1 013 . determine the current ratio i0 /i8 .umho achieves a current gain of 950 in to a load of 1 kO. Example 6. Ans. The amplifier described in Example 6. Calculate the resulting overall current gain and input impedance. find the new voltage gain and input and output impedance.!l Figure 6.17. 47·5 kO. 24·5. A 5 kO resistor is connected from this 50 n resistor to the input terminal such that simple negative feedback occurs. Determine the overall voltage gain and input impedance if the amplifier is loaded with 5 kO. Ans. Find also the output admittance when the amplifier is used with a source having internal admittance 100 . This 500 Q resistor is connected in series with the amplifier input such that simple negative feedback is applied.umho. 4·96 kO.11. Circuit for Example 6. The amplifier input admittance is 2 mmho and feedback is provided by a 50 n resistor in series with the load. 48·8 0.umho. If the load and source are unchanged. 25·8 kO. The voltage across this component is added to the input to provide negative feedback as before. An amplifier having input impedance 800 Q has an open circuit voltage gain and output impedance of 5 000 and 12 kO respectively. Example 6.9.11 8 217 .8 is modified by replacing the feedback system with a current derived system.10.17. and 400kfi 10' i ' 2k. Example 6. 38·7. A current amplifier having an open circuit output admittance of 130 .
Example 6. calculate (a) the frequency ranges over which feedback is positive if ß is 5 x 10. 218 . Example 6. Measurements on a three stage amplifier give the following figures for gain and phase shift: f(kHz) A/6 0·025 420L 45° 0·5 0·2 0·1 0·05 1 100L65° 2 100 L 90° 3 300 L 130° 4 400 L 155° f(kHz) A/6 10·0 5 000Ll80° 100 4 500 L l40° f(kHz) A L (} I 000 1 000L0° 250 3 450L90° 500 2100 L 45° 2000 420 L 45° If feedback is provided by a purely resistive network such that simple negative feedback occurs at medium frequencies. An amplifier has two identical stages each having a short circuit current gain of 95 and a resistive input impedance of 1 k. ß = 0·059 L 245° By means of vector diagrams.4 . whether the feedback is positive or negative. 161. The total loading per stage is 600 . 2·6 MHz.3 . (c) neither. Example 6.Q.12. Check your answer by calculations. is applied. determine for each case. 1/1 760. 450 kHzoo. Assuming the coupling capacitor to have negligible reactance at all signal frequencies and using graphical methods. (b) the minimum value of ß for oscillation to occur and the frequency of oscillation. Ans. 10. 1 MHz. ß L 200°. Ans. (a) negative. (b) positive.13. 060 Hz.14. phase shifting.Q resistance in parallel with 200 pF capacitance.THE THEORY OF FEEDBACK AMPLIFIERS (b) by finding the equivalent Norton generator at the terminals marked XX. Amplifiers and their associated feedback systems have the following complex gain and feedback factors: (a) A (b) A = 35 L = 50 L (c) A = 24 = t L 20° 30°. I I 50. ß = 61ö L 90° I40°. Ans. calculate the maximum gain and 3 db bandwidth when simple negative feedback. Ans.
Use graphical methods to calculate the output voltage and percentage distortion if the decoupling capacitor is removed and e8 is increased to 3 sin wt. 32 per cent.T. and RK 1·2 kO. 47 V peak to peak. 219 .B.EXAMPLES Example 6. calculate the value required for RF.15. Checkthese results using feedback theory. 3·4 per cent (3 per cent calculated but Iarger output voltage increases distortion). Determine also by how much the open circuit voltage of the driving generator must be increased to restore the required Output level.16. 43 V peak to peak. lf RK is adequately decoupled. The source impedance is 600 n. RL 13 kO. A triode valve having the characteristics given in Example 1. 2·12 MO. If the gain and input impedance of the amplifier without feedback are 3 200 and 800 n respectively.8 is connected in series with 250 V H.. is to be used to reduce this figure to 4 per cent and is to be applied by connecting a resistor RF from the final stage collector to the first stage' base. A three stage amp1ifier having a fina1load of 500 Q introduces 7 per cent harmonic distortion in the output current. Ans. calculate the output voltage and percentage distortion when es = 1 sin wt. Example 6. N.F. 8·9 per cent. Ans.
Various methods of analysis of the resulting complete circuits will also be examined. The equation for terminal input current is given by: i' and since = is + ßio ßio = Aßi' . the amplifier must be treated as a current amplifier.ßA so for simple negative feedback.lb shows the feedback signal ßvo in series with the input indicating a voltage amplifier.1 shows possible forms ofinput circuit for a common emitter stage. The terminal input equation in this instance is given by v' =es + ßvo (I :I turns ratio) leading to v' .I . l is = I . with feedback derived from a later stage.. Figure 7.la shows a feedback path in parallel with the input. Figure 7. INPUT CIRCUITS FOR FEEDBACK AMPLIFIERS The possible forms of input circuits are determined by the original feedback definitions. Figure 7. In this chapter. if it is added in series it must be a voltage signal. In this situation.7 PRACTICAL FEEDBACK AMPLIFIERS In Chapter 6 the application of feedback to amplifiers was found to be advantageous if it was in the form of simple negative feedback at medium frequencies. we shall investigate the circuits and connections necessary to produce this form of feedback for various amplifier configurations.ßA 220 . if the feedback signal is added in parallel it must be a current signal. either A or ß must be negative.
The arraw natatian shaws that the twa quantities es and ßvo may (b) (a) Figure 7. In this case hawever ßvo is shawn as being measured at earth with respect ta the emitter. (b) and (c) voltage amplifiers (c) be added in series ta give v' as befare. Connections for feedback amplifiers. (a) Current amplifier.1. In practice the use af transfarmers is incanvenient. A number af cases with examples will naw be cansidered.INPUT CIRCUITS FOR FEEDBACK AMPLIFIERS and ance again either ß ar A must be negative far simple negative feedback. 221 . If the feedback valtage is shawn as being measured at the emitter with respect ta earth then ß wauld be negative. and an alternative input circuit far valtage feedback is given in Figure 7.1 c. The way in which the feedback signal is derived fram the autput depends upan the number af stages in the amplifier and whether it is ta be cansidered as a current ar valtage amplifier.
2..1 Since the feedback path is taken directly from the output terminal this is an example of voltage derived feedback. Figure 7.PRACTICAL FEEDBACK AMPLIFIERS Single Stage Current Feedback Amplifier First let us consider a single stage current amplifier. SOkO = 12·5 .umho and negligible hre· RL is 2 kQ. First the equivalent amplifier components are required. Circuit for Example 7.fi h . hre 120.2. Neglecting the effect of the bias components and Ys determine the current gain and the input and output impedances of the amplifier.2.1. Y1n Yo' 1 = hle = 1 250 . the transistor parameters are hle 800 n. For negative feedback {J must be positive. hoe 90 . hre YL smg e stage common emltter amp11 er. l . Since hre is zero. With a . Example 7.umho = hoe = 90 . In the circuit shown in Figure 7. .24 to 6. and the feedback resistor RF = 80 kO.25 {J GF 12·5 = h = 500 = 0·025 222 = 120 . The required circuit is shown in Figure 7.27 inclusive.umho and A1' = hre The feedback admittance 1 GF . t e current gam YL + hoe is phase inverting and therefore A is negative.umho and from equation 6. The formulae applicable to this circuit are given by equations 6.
and with the equivalent circuit problems.') = 90 + 500 + 12·5(1 + 120) .o' Z1nr = 231 n Finally from equation 6.11.'YL YL + GF + 120 X 500 90 + 500 + 12·5 = .7.) Single Stage Voltage Feedback Amplifier lf a single stage vo1tage amplifier is required to have negative feedback.21 provides the va1ue of the input admittance. Figure 7.umho The resu1ting output impedance ior = 476 n.ßA.n.13).o) = 1 250(1 = 4 336 . This suggests that the vo1tage feedbackwas current derived (equations 6. Examp1es of this were shown in Chapters I and 6 using graphical methods and in Chapters 4 and 5 using equivalent circuit methods. 6.99 ' 6 The normal feedback equation now gives the current gain: 99·6 1 + 99·6 X 0·025 A1o Atr = I . part or all of the emitter or cathode resistor may be left unbypassed.8.ßA1o Au= 28·6 Equation 6.1 (Note the va1ue of A 1' wou1d have tobe modified for this calculation if Zs were not much greater than Z.SINGLE STAGE CURRENT FEEDBACK AMPLIFIER Applying equation 6.umho + 99·6 X 0·025) . Y1nr = Y.umho = 2100 . In each case the voltage gain was reduced.n(1 .12 and 6. 6.A. the input and output impedances were increased. . 6.3 shows how the 223 .27: Yor = Yo' + YL + GF(l .26.umho The input impedance with feedback is given by the reciproca1 of Y1nr. or by finding 1 ~pA. A!o = Yo' A.
PRACTICAL FEEDBACK AMPLIFIERS correct phase relationship arises.3. Current derived.RE AvV' RL But Avis negative forasinglestage (hie(hoe es = V' ( 1 + and V' +. since V' is less than es. voltage feedback on singlestage ampliiier Now. To verify the use of feedback methods weshall repeat Example 5. using the notation shown. the overall voltage gain is reduced indicating negative feedback.~) _ hrehrJ· ~: IAvl) es = 1 + lßAvl Thus. The feedback voltage in the sense shown is given by: But 1·· Figure 7. 224 . =V'.1 using the results obtained from Chapter 6.
hoe 125 . Circuit for Example 7.4.90 90 X 5 X 104 = 1300 Note.umho is used but Av' is the open circuit voltage gain with YL zero. we must first find the components of the appropriate equivalent amplifier.4 Z1n = 900..2. A single stage amplifier having the circuit shown in Figure 7. Using feedback methods. 6. hre 5 X 104. the output voltage and current.4 employs a transistor with h1e 900 0.SINGLE STAGE VOLTAGE FEEDBACK AMPLIFIER Example 7. and the output impedance.12. Using the standard h parameter formulae: 90 X 5 X 10. find the terminal voltage 4k. the circuit YL of 250 .Cl 1k .2 and current gain.33 = 225 283 . = 13·33 kil Yo Now from equations 6. Yo' = 125 X 106  Zo' 90 X 5 X lQ4 = 75 .025 + 250) 10_6 = 780 0· Av' = 900(125 + 0)106.7 Avo 1300x4 = 4 + 1 + 13 .n Figure 7.umho 900 =!.. As with the last example.umho and hre 90. for Z1n.11 and 6.
_ Terminal input voltage _ 0·88~ A lbZ1nr .umho Now using potential divider methods. the base input current ib is required.56·2 m = 15·7 .13. . = 17·8 + 200 Z1n = 3·74 kQ Y1n + 50 = 267·8 . and Vo = 3·48 V. Working in admittances.uA To calculate the terminal current gain. 1 Termma A1 880 = 15 . To obtain output current: io = VoYL = 3·48 X 250 X lQ6 A = 880 . = 13·33 + 1(1 + 1 300) = 1 314 kQ Huttor the overall output impedance. RL must be included in parallel Zor z 0 = 1 314 X 4 ~ 4 kQ 1 318 Similarly.7 =56 226 . the terminal input voltage is given by 1 X 3·74 = 0·883 V 3·74 + 0·5 The resulting output voltage is therefore 0·883 x 3·95 V. the overall input impedance should include the bias components.8: Z1nr = 0·78 ( 1 + 283) 4 = 56·2 kQ and from equation 6.PRACTICAL FEEDBACK AMPLIFIERS and RE ß= Avr RL =t 283 283 = 3·95 = 1+4 Also from equation 6.uA .
the overall input impedance becomes: Ztn = 90 X 4 ~ = 3·8 kQ . which would slightly modify ß.55 ' 5 Thus we can see that the terminal current gain is not modified by voltage feedback. 1 X 3·8 T ermma mput vo1tage= 3. This is principally because the current in the emitter resistor is the sum of the output current i 0 and the input current ib. In practice the difference is negligible and if it is remernbered that all components and parameters are subject to wide tolerances. Two Stage Current Feedback Amplifiers Feedback is more commonly applied over two or more stages and the next problerntobe considered isthat of a two stage current amplifier. an even simpler solution may be obtained as follows.5 = 0·88 V Vo =4 X 0·88 = 3·52 V which is only about 1 per cent high.8 + 0. but when the bias components are included. The current gain in this case will be positive and direct 227 . Since ßAv~ l Avo~ Ztnr hreZL hte = 90 X 4 000 900 = 400 = hte(l + ßAvo) = 90 kQ This may seem a large error.TWO STAGE CURRENT FEEDBACK AMPLIFIERS Note in generat terms At hreYL = =YL + hoe 90 X 200 = 200 + 125 = . The results for Avr and Z 1nr are not identical to those obtained from the equivalent circuit which were 3·87 and 57·5 kO. Variations in components and parameters will cause a much larger margin of error than will the use of these or similar approximations. 1.
signs. Since all i' Tr2 io Figure 7. Yor.23.5. The resulting equation at the input terminal is: i' = is + (ßA1A2i') which upon rearrangement leads to the correct relationships for negative feedback. it must split between the two parallel paths provided by G1 and G2. although the terminal current gain will not be affected by their presence. the steps in the solution will be: A12. These must be allowed for in the calculation of A 1'. making approximations where valid. The feedback current ir is considered as flowing in the opposite sense to ßA1A2i' hence the sign change as shown. Current splits in the direct ratio of parallel admittances. Av2 and Ztn2· Ztn2r using ß for 1ocal feedback of RE/RL. as shown in the simplified circuit diagram in Figure 7. the input impedance to Tr2 and thus the overall current gain. since.5. will be. So. The 180° phase change per stage is shown by the + and . Au and Ztnr. Current amplifier with current derived feedback the output current flows into the emitter of Tr2. An and Z1n1. The interstage bias components are included in YL 1 . A further problern arises as a result of G1 and G2 in the emitter lead of Tr2. 228 . The required phase relationship can be obtained by connecting the feedback path to the final emitter. A1' and ß1 for overall feedback of G2/(G1 + Gz). Here. the solutions may be found from equations 6. thus ßis given by G2/(G1 + G2). Yo' (approximately hoe for Tr2). Since we have current derived current feedback.15 to 6. A1 and A2 are the magnitudes of the current gains of stages 1 and 2.PRACTICAL FEEDBACK AMPLIFIERS connection between final collector and first base would result in positive feedback.
The transistor h parameters are hte 1· 5 kil.s.umho. Circuit for Example 7.m. of internal impedance 600 n. hoe 150 .3. If the final emitter resistor is unbypassed. The output current in the amplifier.TWO STAGE CURRENT FEEDBACK AMPLIFIERS Example 7. shown in Figure 7.m. so that the distortion may be reduced to 5 per cent.6. determine the value of the feedback resistor to be connected from the final emitter to the first base. Since hre is negligible this current gain may be written directly by current splitting techniques. hre 140.6.m. and negligible hre· We must first determine the original current gain and hence the original source e. must be changed if the 400. A Ytnl 1 = Ytnl + YBl X hre X hoe + YLl Y1n2 + YB2 X + hre Ytn2 X y L2 + hoe where YBI and YB2 are the combined bias components foreachstage and Ytnl = Ytn2 = l/hte· 667 + 25 + At= 100 667 X I 40 667 X 150 + 250 + 100 + 333 + 667 = 0·84 X 140 X 0·46 X 140 229 X X  0·943 140 X = 7 130 2 500 2 500 + 150 . of a driving source. contains 20 per cent harmonic distortion at the required output signallevel of 5 mA r.f. Determine also by how much the e.!1 Gz Figure 7.f.3 required output current is to be maintained.!1 Tr2 10k!l 200.
we require Av2· The effective load is 600 n making YL I 670 fliDho.32.m. and hence the interstage factor in the expression for A 1 above. 140 Av 2o = l 500(150 + I 670)10. D Dr = I .!i:l') kQ =40kQ = 25 pmho Thus the interstage factor becomes I50 + 250 + 25 IOO + 333 + 25 = 0' 029 Substituting this value in the expression above: 0·029 AI = 7 130 X 0·46 = 450 Now applying equation 6.PRACTICAL FEEDBACK AMPLIFIERS The input current is given by: 5x1000 itn = 7 130 flA = 0·7 pA But the overall input impedance Ztn = Y 1 + lnl Y Bl = 1·26 kQ Source e.6 =51 200 ß = 400 But Ztn2f Ytn2 = 1·5(1 + . Since this component provides current derived voltage feedback for Tr2. = itn(Ztn + Rs) = 0·7(I·26 = 1·3 mV + 0·6) mV The unbypassed emitter resistor for Tr2 will effect the current gain without overall feedback by modifying Y1n2.f.ßA 20 per cent 1 + 450 ß 3 450ß = 4 and ß = 450 or 5 per cent = 1+ 230 150 .
THREE STAGE CURRENT FEEDBACK AMPLIFIERS
But
and
G2
I
and G1 = 200 = 5 mmho
G1 + G2
5 + G2 = G2 X I50
5
G2 = I 49 mmho = 33·5 ,umho
ß 
This represents the overall feedback admittance or its equivalent
resistance of 29·8 kO.
RF = 29·8 kO
In practice the nearest preferred value of 27 kO would be used.
With this feedback resistor in circuit, the current gain and input
impedance becomes:
Au
=
Ztnr =
450
450 = 112·5
I+ I50
1·26 kO
450 = 315 0
1 + 150
But the output current in the final Ioad is to be maintained at
5 000
5 mA, requiring an input current itn, of 112.5 ,uA. The required
source e.m.f. now becomes:
5 000
112.5 (0·6 + 0·315) mV
es= 40·6 mV
Thus to reduce the distortion to 5 per cent the source e.m.f. has
to be increased by a factor of approximately 30.
Three Stage Current Feedback Amplifiers
If negative feedback is to be applied over a three stage current
amplifier, the phase relationships are the same as for feedback over
a single stage. The procedure outline in Example 7.1 cou1d again
be applied. An alternative procedure is to regard the amplifier and
the feedback network as two separate fourterminal networks,
connected in parallel. In Chapter 2, we found that with this circuit
arrangement, the combined network y parameters were given by
the sums of the individual network y parameters. Thus if we can
231
PRACTICAL FEEDBACK AMPLIFIERS
find the y parameters for a three stage amplifier and then for a
feedback network, the general solution obtained in Chapter 2 may
be applied directly.
Example 7.4. A three stage current amplifier employs identical
transistors having hte 1 000 !1, hre 90, hoe 100 ,umho, and negligible
hre· Each stage has a Ioad of 1 kil and the effect of the bias components may be neglected. A 100 kil resistor is connected between
r ,
I
100kil.
I
I
I
I
I
I
L 
I
I
I
!
'
I
I
  



  
    __ J
lk.O.
L            Figure 7.7. Interconnected fourterminal network circuit for
Example 7.4
the final stage collector and the base of the first stage. Determine the
overall current gain and the input and output admittances for the
resulting circuit. Investigate the probable effect of the bias components and hre on these results.
Figure 7.7 shows the circuit represented as two fourterminal
networks in parallel.
lt is convenient to show the Ioad of the final transistor as the Ioad
for the combined network as shown. The y parameters for a network
may be redefined.
These parameters are all admittances and in each case a short
circuit is applied to the opposite end of the network for the purpose
of calculation.
232
TIIREE STAGE CURRENT FEEDBACK AMPLIFIEM
Taking the feedback network first, and connecting a short circuit
to make V2 zero then:
h
1
vl = yu = 100 kil =
10 ,umho
Also remembering the direction or sense required for h,
h
Y21
V1
= 100kil
h
= v1 =
10 ,umho
But the network is symmetrical, so
y22
=
y12 = 10 ,umho
10 ,umho,
Now proceeding to the amplifier, since hre is zero
yu
1
= hte =
1 000 ,umho
where hte is a parameter of the first stage transistor, and
y22
= hoe = 100 ,umho
where hoe is a parameter of the final stage transistor.
To find y21 we require the short circuit output current in terms of
the applied input voltage V1.
Using normal current splitting methods:
V1
hte
h = 
X 
h
re
X
Ytn2
====::
hoe + YLI + Ytn2
X
hre
X hoe
+
Ytn3
+ Ytn3
YL2
X
+hre
Note that the last hre term is positive since the normal convention
requires / 8 to flow into the output terminal.
Y21
=
h
V1
=
X
= 729
90
1000
1 000 X 100 + 1 000 + 1 000
1000
90 X 100 + 1 OOO + 1 OOO X +90mho
1
x 2. 1 x 2 . 1
=
233
165·5 mho
PRACTICAL FEEDBACK AMPLIFIERS
Finally since hre is zero a voltage V2 can cause no short circuit
current h.
y12
=0
The combined parameters may now be written,
yu
y21
=
=
1 010 ,umho
165·5 x 106 ,umho
y12
y22
=
=
10 ,umho
110 ,umho
The general solutions can now be applied taking YL as 1 000 ,umho.
From equation 2.47,
Ytn
10 X 165·5 X 106
= 1 OIO110 + 1 000
= I·01 x 10a + I·49 mho
= 1·49 mho
,umho
From equation 2.49 the current gain may be found:
~ = v2 YL =
y21 YL
h
h
yu(y22 + YL)  y21Y12
I65·5 X 106 X I 000
1 010(I10 + I 000)  ( IO X I65·5
1000
= 1..,..0_10__xII.,..,0
106)
~~~~~~~~~
165
=
X
1000
10·007
I06
=
X
+ 10
IOO
To determine the output admittance equation 2.50 must be used.
Y.o
Y21y12
Ys
= y22 yu +
Ys is not specified in the example so taking the limits of zero and
infinity.
10 X I65·5 X I06
Yo = 110 +
= I·64mho
1010
with Y8 zero or Y 0 = 110 ,umho with Ys infinite. Thus the output
admittance is very dependent upon the value of source admittance
but with a practical value of say I 000 ,umho, the output admittance
will be of the order of 800 mmho.
234
TWO STAGE VOLTAGE FEEDBACK AMPLIFIERS
The above results suggest that if feedback formulae are to be used,
the value of ß is not affected by Ys in the calculation of Ytn and At.
but when calculating Y0 , the current division between Ytn and Ys
should be allowed for.
To complete this example, the effect of the bias components,
and hre would be to reduce the current gain without feedback, and
under the same conditions to increase the input admittance (bias
components) and reduce the output admittance (hre). Since Au
is approximately given by 1/ß the current gain would still be 100.
Ytnr is given by Ytn(1 + lßAtl); the first term is increased and the
second reduced so there wou1d be little change in this result. Yot
is.given by Yo' + YF(1 + JAt'J); here both termswill be reduced.
Thus we can say that the stated approximations will have negligible
effect upon the input admittance and current gain but will cause the
output admittance to be a little smaller than that calculated.
Multistage Current Feedback Amplifiers
If a current amplifier having more than three stages is to employ
overall negative feedback, the connections obviously depend upon
the number of stages. In Examples 7.1 and 7.4, the amplifier had an
odd number of stages; the required phaserelationship was obtained
by the use of voltage derived feedback. This would equally app1y
to any higher odd number of stages. In Example 7.3 the amplifier
had two stages with current derived current feedback. Similar
circuitry would be required for four or any even number of stages.
Of course the usual care would be necessary to ensure that positive
feedback did not give rise to instability at the extremes of the frequency
range.
Two Stage Voltage Feedback Amplifiers
Example 7.2 was concerned with a single stage voltage amplifier
and the negative feedback was current derived. From the discussion
above, we should expect a two stage voltage amplifier to have voltage
derived feedback. This is correct, if the feedback is applied to the
emitter of the firststage as in Figure 7.lc. If an output transformer
were used, as in Figure 7.lb, then current derived feedback would
be essential to obtain the correct phase relationship. An example of
the first case will now be discussed.
Example 7.5. An ideal two stage transistor amplifier has Ioads of
1 kO per stage. Feedback is provided by a resistive potential divider
chain of 9·5 kQ and 0·5 kO in parallel with the output. The 0·5 kQ
235
PRACTICAL FEEDBACK AMPLIFIERS
of this arrangement is connected between the emitter of the first
transistor and earth. The transistor parameters may be taken as
h1e 1 000 0, hre 100, hoe 100 ,umho and hre 0. By means of a general
analysis verify a solution based on feedback theory.
The circuit arrangement for this problern is shown in Figure 7.8.
V,
9·5k!l.
1
Figure 7.8. Voltage amplifier with voltage derived feedback for
Example 7.5
To confirm first of all that this arrangement does provide negative
feedback; the output voltage of the first Stage will be Al vl with
180° phase shift. V0 will therefore be A 1A2 V1 with a further 180°
phase shift giving a total zero phase shift between Vo and V1.
Initially neglecting Tri emitter current, Vr will be in phase with V1
and its value will be given by:
ßVo = ßA1A2V1
0·5
ß = 0·5 + 9·5 = 0·05
Vr
Where
=
Now writing the series equation around the input circuit,
Vs
or
=
V1(l
Vr = (1
+ ßA1A2)
Vs
+ ßA1A2)
Since V1 is less than Vs the overall gain is reduced and the feedback
is negative.
In the above discussion, we have neglected the effect of Trl
emitter current upon Vr. In practice this emitter resistor not only
236
overall gain and input impedance with feedback. but strictly the overall ß should be modified by the output impedance measured at the emitter ofTrl.umho 100 Avi = 1 000(100 + 2 000)10. (3) Av1 and Ztni without local feedback.47' 6 Ztnl = 1 kO For Iocal feedback ß= 500 X 2 000 X 1Q6 47·6 Avu = 1 + 47 .umho 100 Avz = 1 000(1 100 = Ytn2 = YLI etr = Ztnz + 100)106  83 "4 1 kO 1 000 .6 = Ztnu = 1 k0(1 = 1 0·98 + 47·6) = 48·6 kO 237 . (5) Overall gain without overall feedback from Avu X Avz. We shall now follow the procedure for the circuit shown in Figure 7. The proposed technique based on feedback theory will involve the following calculations: (1) Avz and Ztn2 using 1 kO in parallel with 10 kO as the Ioad. This will reduce ßfor this calculation but unless the source impedance Z 8 is low the effect will be negligible. (6) U sing ßof0·05.umho 1 000 + 1 000 = 2 000 . The output impedance can be determined in a similar manner. hz =(I 000 + 100) . (2) YL1 err including Ztnz and the 1 kO Ioad for Trl. resulting in a gain with local feedback of less than one.6 = .8.umho = 1100 . in certain cases this ß may be greater than one.TWO STAGE VOLTAGE FEEDBACK AMPLIFIERS provides a part of the overall feedback network. it also causes 'local' negative on the first stage. This does in fact occur and may be verified with measurement on practical circuits. (4) Avu and Ztnr with local feedback taking ß as ZF Z Letr = 0·5 kO X YL1 etr Note.
Ii• Figure 7.2) 0·05) = 248 kO To verify this procedure we shall use the equivalent circuit shown in Figure 7. i1 the base current for Tri.05 = 16·1 (7.5 First.3) 238 . + 100 X lE = lE = 92VlmA V1 ( 1 !:) mA (7.PRACTICAL FEEDBACK AMPLIFIERS Overall gain without overall feedback = 0·98 x 83·4 = 82 Overall gain with overall feedback = 1 + 8 ~ 2x 0. Equivalent circuit for the verification of the feedback method used in Example 7.9 and solve by nodal analysis.1) and Overall input impedance = 48·6 kO(l + 82 X (7. is given by V1/hte· The output of Tri is loaded by h1 err in series with Y F2· Strictly Y Fl provides additionalloading on Tribut the effect is negligible since YF ~ YF2· This loading is given by YL1 h1 err YF2 + Y F2 eu lE = 2 000 2 000 = ~: X 2 000 + 2 000 ttmho = 1 000 ttmho [1 + hre C~o:o hoJJ Inserting values and writing lE in mA.9.
admittance must be expressed in mmho.524 = 241 v1 (7. since currents are expressed in mA and voltage in volts.3 and 7. 2·105 X 4 750Vl + 0·105 X 92V1 2·105 X 1·205.524 v1 = 3 960V1 Vo = (7. Vo = 3 960V1 Vo 3 960 .7) .O·I05Vo 4 750Vl = 0·105Vr + 1·205Vo Solving by determinants.6. V2 = .5. and l2 hreV1 = IOOV1 hte(hoe + h1 eff) I 000(100 + 2 000)106 V V2 IOOV1 = hte2 = 2·I X 1 000 A (7.TWO STAGE VOLTAGE FEEDBACK AMPLIFIERS Now. Thus YF2 = 2mmho YL2 = 1 mmho YF1 = O·I05 mmho hoe2 = 0·1 mmho Now writing nodal equations for the Vr and Vo nodes and using equations 7. Va = VI(24I + I)= 242Vl But from equation 7. Overall gam = Va = 242 = 16·3 239 (7.0·105Vo 100(47·5)Vl ~ 0·105Vr + Vo(I + O·I + 0·105) Rearranging: 92V1 = 2·105Vr. Substituting from equation 7.5) 92V1 X 1·205 + 4 750Vl X 0·105 Vr = · 2·524 609Vl = 2.4 above: 92Vl = Vr(2 + 0·105).6) The signal vo1tage Vs = Vr + V1.4) i2 = 47·5Vl mA To give the correct dimensions.0·105 X 0·105 10 010 = 2.
Forthis situation.2 is series input and series output which is the correct configuration for summation of z parameters. suggesting the use of g parameters. Since voltage derivation led to negative feedback over an even number of stages. d ance = :Vs = 242 V1 k" 0 vera11 mput 1mpe V :. ll = 242 kO 1 (7 . The remaining procedure then follows normal fourterminal network practice.4 are the parallel input and output form permitting the summation of y parameters. Unfortunately this is not valid. The reader may have wondered why a so1ution based upon fourterminal network theory was not used for this example. and the resulting solutions are extremely close to those obtained by feedback methods. and feedback applied to the first stage emitter (or cathode). This suggests that if the h parameters of the network and those of the amplifier were added. and the circuit for Example 7. Apart from this. h for the network.1 and 248 kO) the difference is only of the order of 2 per cent. Three Stage Voltage Feedback Amplifiers The one remaining simple configuration not yet considered is that of a voltage amplifier with an odd number of stages.7 and 7.8 with results 7.PRACTICAL FEEDBACK AMPLIFIERS and · . the overall h parameters would be obtained. Thus only feedback methods or circuit analysis can be used for this very common circuit. the calculation of g21 is made by neglecting the final emitter resistor and is given by Av1 X Av2 with YLZ zero.8) Comparing results 7. while the current flowing out of the common terminal... The circuits for Examples 7.3 has the feedback network in parallel with the input and in series with the output.1 and 7. The current amplifier for Example 7.2 respectively (16. 240 . Other situations however do permit the fourterminal network approach. Since components and parameters are unlike1y to be known to within 20 per cent this error is negligib1e and the method based upon feedback is adequately verified. the A for the calculation is the normal loaded voltage gain. since h for the amplifier is ib.~. an odd number of stages will require current derived feedback. is ib + ic. The feedback network is connected in series with the input. and in parallel with the output of the amplifier. This may be done provided the effect of the final emitter resistor is included in the calculation of input impedance for Tr2 with local feedback.1 and 7. we shall consider a valve amplifier.
There is no coupling network to Ioad the final stage and so the gain may be assumed constant at 30. Figure 7. A three valve amplifier employs pentodes having gm 3 mAJV. 10k. ra 1 MQ. Three stage feedback amplifier for Example 7.3 :.9) . Foreachstage ra = Rg~ RL = lOkQ = gmRe = 3 X lQ.THREE STAGE VOLTAGE FEEDBACK AMPLIFIERS Example 7.ll.6.10 determine the new medium frequency gain and the gain at the 3 db frequency calculated above.10.6 When no feedback is applied. lf negative feedback is applied as shown in the simplified circuit in Figure 7. determine the overall voltage gain and the upper 3 db frequency. If each stage is loaded with 10 kQ and grid leak resistors of 1 MQ. the results found in Chapter 4 may be used. Overallmaximum voltage gain = (30)3 Re~RL Avm = 27 X X 104 = 30 103 Todetermine the 3 db frequency we need only consider the first two stages. At the 3 db frequency IAh I = I= IAm y'2 241 27 X 103 y'2 (7. and input capacitance 5 pF on Ioad.
Stages one and three have local feedback due to the 200 (! cathode resistors. X lQ30) w2=~~~~~~~ 5 = 12·5 ± y50 X lQ15 = 2·07 12·5 X lQ15 (since negative w 2 not permissible) w = and J(2·07) 1·25 w J=27T = X 107 2·04 MHz Thus at 2·04 MHz the overall gain without feedback is 27 X 1()3 v2 9 OO = 1 1 If the feedback circuit is now considered at medium frequencies.PRACTICAL FEEDBACK AMPLIFIERS But allowing for the two interstage couplings: Am2 Ab 27 X 1()3 = ( 1 + j _e R )2 X Am X es from equation 7.1 5 X 1015 ± y(25 X lQ30 12·5 X 1Q3o X 1Q16w2 = 2 =0 + 2S.2 X 25 X 1Ql6w2 X + 625 104 X =5 6·25 X 1Q30w4 +5 X X 1032w4 X or (7.10) + j2wC Re .9 27 X 1()3 y2 or = But + I 2 8 8 CsRe =5 .I012 X 1 . 242 . and there is overall current derived feedback.w2Cs2Re2) 2 + 4w2Cs2Re2 = 2 J1 (1 27 X 1()3 (1 jwCsRe)2 + jwCsRe)2j2 = J(1 and I = (1 + JW · Cs R e)2 lQ8 +4 25 1Q15w2 .w2C 2Re2J2 = 2 .
12) ..gmZL = 25·35/180°  But local ß for Avr = = 32° 36' RF ZL 0·2 8·45 L 32° 36' . 10 580 Overall voltage gam = 1 + 10 580 X 3.11) 18 "75 Overall gain without overall feedback = ( 18·75)2 = 10 580 X 30 Examination of the circuit shows that 200 ßvo = 200 + 104 X 4 X Overall ß = 1... the overall ßis unchanged.~o: 243 (7.92 = X 10_4 2060 At the upper 3 db frequency. At this frequency.THREE STAGE VOLTAGE FEEDBACK AMPLIFIERS For the local feedback..02 vr 200 where vr = 104 Vo 104 108 X = 3·92 X _4 10 .. 2·04 MHz + j27T X 2·04 x 104 mho = 118·5 L 32° 36' flmho h 1 = YL2 = 104 = (1 + j0·64) X 106 X 5 X 10~12 mho ZLl = ZL2 = 8·45 L 32° 36' kQ without local feedback Avl = Av2 = . but the Ioad on the first valve is changed modifying the local ß for that stage..... in each case ß= 200 104 Avu = Avar = 30 30 X 200 = 1+ 104  (7....
Thus for the 3 db frequency I II Am 1 y2 1 . Now the overall gain without overall feedback at 2·04 MHz Av = 15·5 L 141° 24' X 25·35 i.j2·62 7 350 L I14° 48' = 3·I2 L 49° 51' = 2 357 L I64° 39' This is greater than the medium frequency gain with negative feedback.2·88(cos 114° 48' + j sin 1I4° 48') 7 350 L 114° 48' .1 + I·2I.ßAm = = Ah I 1 .ßAh Am (1 _ 1 + jwC Re)2 6 ßAm (1 + jwCsRe)2 = I(l + jwC~)2 244 ßAm I .11.PRACTICAL FEEDBACK AMPLIFIERS Avu 25·35 L I41° 24' = ::2=s. lt is of interest to note this drop in gain due to feedback is identica1 to that shown in resu1t 7 . This result is as would be expected from the general case discussed in Chapter 6.=3=5x::0:·2:=_ _ _ _ _ __ I  L I41° 24' 8·45 25·35 L I41° 24' I + 0 . Feedback over a singie stage does not improve the bandwidth since ß changes with the 1oad. The local feedback effect on the first stage does not vary with frequency so this can be ignored.7 350 x 3·92 X I04 L II4° 48' 7 350 L 114° 48' Avr = I . The 3 db frequency with feedback may be found by following a similar technique._ 147° 24' X 18·75 L 180° = 7 350 L 414° 48' = 7 350 L 114° 48' The overall ß is unchanged Avr = 7 350 L 114° 48' 1 .6  = + 32° 36' o 15·5 L 147 24 . but is considerably 1ess than the gain at this frequency without feedback.
11. ß for the local Figure 7. 625 w2= and 1·6 X X 1Q32w4 . Single stage amplifier with frequency compensation by negative feedback Ce 245 .6·4w2Cs2Re2 .1·6 X 1Q14w2 .10·4w2C 2Re2 + w4C 4Re4 + 4w2C 2Re2 8 8 8 8 8 w4Cs 4Re 4 . In the last example.ßAml But v'2 4·2 + jwCsRe)2 + 4·21 = li + j2wCsRe . Complex Feedback Factors So far in this chapter we have only considered ß factors with real values.w 2Cs2Re2 + 4·21 5·2 = 1(1 X Taking the modulus and squaring: 2 X 27 = (5·2.27 = 0 But CsRe = 5 x 10s.27 1014 ± y'(2·6 X 1Q28 + 1·69 12·5 X I030 =0 X 1028) ~~~~~ 3·66 w2 = .ßAml ßAm = 3·92 X 104 X 10 580 = v'2ll . at high frequency.x 101 6 12·5 w = 5·31 x 107 rad/sec w f=:::: lOMHz 27T Thus since the low frequency 3 db frequency may be ignored.w2C 2Re2)2 + 4w2C 2Re2 = 27. negative feedback has increased the bandwidth from 2 MHz to 10 MHz while reducing the medium frequency gain from 27 000 to 2 060 which is in approximately the same ratio.COMPLEX FEEDBACK FACTORS = 1(1 + jwCsRe)2 .
Av =GE+ ~WCE GL + JWCL This ratio will be constant if wCE wCL GE = GL or if RECE = RLCL This implies a flat frequency response so long as hreZE ~ 1 h1e Let hre be 100. The voltage gain for a single stage grounded emitter amplifier is approximately hreZL/hle and ß for the current derived feedback shown is ZE/ZL.11. The frequency response for a single stage may be improved if the cathode or emitter resistor is shunted by a small capacitance.PRACTICAL FEEDBACK AMPLIFIERS feedback on the first stage was comp1ex (7.12). is to prevent instability. hreZL . 0·2 w = 2·5 X 108 20 I = 1·3 MHz 2·5 Another way in which a capacitor may be used in a feedback path. If overall feedback is applied to. Taking account of the components shown. a four and J= 2rr X 246 . h1e Overa11 gam = h Z Z 1+~2 h1e ZL The ZLs in the denominator cancel. say. let hreZE = 5 or ~ = 5 hte h1e YE 5 = 11 000(103 ::25 X 1010) j102 + jw2·5 X 10Sj = 0·2 Since the real term can have little effect. Todetermine the approximate useful frequency range. and if hreZE/hle ~ 1 the overall gain = ZL/ZE = YE/ YL. If the shunt capacitance CL is 0·01 ttF then CE must be 0·0025 pF. hte 1 kO. This resulted in no improvement in frequency response for that stage. Consider the circuit shown in Figure 7. RL 4 kO and RE 1 kO.
Figure 7. Examples of amplifiers using composite feedback (c) 247 . COMPOSITE FEEDBACK Other forms of feedback circuit may also be found when two types of feedback may be included on the same amplifier. Many such combinations are possible and may be found in practical circuits. This is sometimes referred to as composite feedback.COMPOSITE FEEDBACK stage amplifier. (b) (a) Figure 7.12. Similar modifications for low frequencies may be achieved with suitable capacitors or inductors. and oscillation results at a particular high frequency. This will reduce the amount of positive feedback at these frequencies eliminating the instability and flattening the gain frequency response. a shunt capacitor may be used to reduce ß at the high frequencies in this range.12 shows some examples of composite feedback circuits.
In Figure 7. The amplifier shown in Figure 7. will now be considered.c. The voltage feedback is applied only over the last two stages.13. Avz and A1z accounting for the local feedback present. and the output voltage if the amplifier is driven from a source which may be represented by a current generator 248 . A final example.12b employs both voltage derived and current derived voltage feedback and a similar procedure should be used taking the voltage derived loop first. The procedure would be to determine first. based on this circuit. Determine suitable values for R1 and Rz. Example 7. The three stage amplifier in Figure 7.12a both current derived current feedback and voltage derived voltage feedback are present. the current feedback loop is 'outside' the voltage feedback loop. The circuit shown in Figure 7. Circuit for Example 7. finally to account for the effect of current feedback.7 stabilize the circuit against changes in the d.7. second to find Av1 and An including local feedback . Since the voltage feedback is applied to the first emitter.PRACTICAL FEEDBACK AMPLIFIERS In each case only negative feedback at medium frequencies is employed. supply voltage Vcc· Feedback through Rz is provided to reduce the second harmonic distortion in the output voltage by a factor of 5.13 employs overall feedback through R1 to reduce the input impedance to 10 Q and to Rz Figure 7.12c includes voltage derived voltage and current feedback. third to calculate the overall gain and input impedance with voltage feedback.
COMPOSITE FEEDBACK of I pA in parallel with an impedance of 50 Q. 140 Ava = I 200(120 + 212)10.umho).89 For local feedback ß= ZFYL = 100 1 395 X X ]Q6 = 0·1395 77 = I + 77 X 0·1395 = 6·55 Z1n2r = 1 200(1 + 77 X 0·1395) 0 = Avzr 14·1 kQ Now for the voltage feedback loop A = Ava X = 2300 Avzr But distortion must be reduced by a factor of 5 1 + ßA = ß= ß = R2 + 100 Rz 9 ßA IOO But and 5 and 4 2 300 + 100 = 2 300 X 4 4 = 2 300 100 = Rz = 57·4k0 249 =4 57 500 Q . hre I40. hoe I20 .77 ·5 YL2 Z1n2 1 = .6 = .mho = 834 and = ~ = ß34 pmho + 212 + 46 + 303 = I 395 . The transistor parameters are h1e I 200 0.umho and hre negligible. Assuming that R1 and R2 will be much greater than 4·7 kQ (212 .umho 140 Avz = 1 200(120 + 1 395)106 = ?? 12000 140 X 834 A12 = I 395 + 120 = . All capacitors may be assumed to have negligible reactance at the signal frequency.352 and and A 13 = 140 X 212 212 + 120 Y1na = 1 200 .
then Ytn = 105 . Since the source impedance is 50 Q and the input impedance of the amplifier is 10 n.umho By current division 140 X 14·2 An = 120 + 14·2 + 212 + 12·2 + 45·5 = .umho The feedback resistor 1 Rl=yF=1·35MQ Now all that remains is to find the Ioad current for the complete amplifier.uA 250 .4"9 Also Ytnl = 834 + 100 + 455 .PRACTICAL FEEDBACK AMPLIFIERS This feedback path further modifies the input impedance to Ztn2r(l + ßA) = 14·1 X 5 kQ = 70·5 kQ Ytn2 = 14·2 . 834 Overall At= l 389 = 4·9 X X 77·5 X 89 20 300 But if the input impedance is to be I 0 0.389 = 72 71 ß = 20 300 = 3·5 and X 103 But for voltage derived current feedback ß = YF YL YF = 3·5 x 10a x 212 .umho = 0·74 . then the input current 50 itn =50+ 10.umho = 1 389 .umho = 1 + ßAt or 1 389(1 + ßAt) 102 = 1.umho Now allowing for current splitting at the input to Tri.
264 Q. 9·62. Many other forms are used. Example 7. In this chapter we have considered the practical circuits for providing negative feedback on a number of amplifier configurations. 4·88 kQ. hoe 125 ftiDho. 1·97. Ans. 11·4. Repeat Example 5. A two stage common emitter amplifier is constructed with the following components.EXAMPLES But c urrent gam . in any particular situation.19 using feedback methods.10 using feedback methods. 525. 17·5 kQ. 10·5. 490 !2. combined shunt bias resistors 10 kQ per stage.10. Repeat Example 5. For convenience. the solution would have involved not less than five simultaneous equations.11 using feedback methods.11. 656 Q. 251 . Ans. 6·72. 6·8 kQ. hre 90 and negligible hre· Calculate the amplifier input impedance and current gain.12.5 x 10_3 = 20 300 = 282 72 = 282 X H ftA = 235 ftA Output voltage = ioZL = 0·235 mA X 4·7 kQ = 1·1 V Load current For comparison. For more general analysis.8. Example 7.13 using feedback methods.20x 300 3. wlt. RL2 1 kQ.9. if the problern had been attempted using the complete equivalent circuit. Example 7. Ans. including those with common base or common collector stages. but the methods applied here may be simply extended in most cases. Example 7. RL1 4 kQ. feedback formulae carefully used probably gives the best approach. EXAMPLES Example 7. h fieedback = 1 + 20 300 . Ans. the interconnected fourterminal network is probably better. Ans. Repeat Example 4. The transistors are identical and have hte 1 250 n. Repeat Example 4. Feedback is provided by connecting a 100 Q resistor in the emitter Iead of Tr2 and a 100 kQ resistor between Tr2 emitter and Tri base.
Figure 7.umho. Each stage has a collector Ioad of 4 k.Q. Calculate the terminal current gain and the input impedance. Example 7. 250. A two stage common emitter amplifier is constructed with the following components. combined shunt bias resistors 8 k.Q.Q. Feedback is now provided by connecting a 1 MO resistor between Tr3 collector and TrI base. The transistor parameters are h1e 900 . 43·5.Q resistor connected between them.17 252 .14. hre 60 and hre 0. Repeat Example 7.Q and 100 . and hre 0.Q in parallel with 8 k.Q. The three stage amplifier described in Example 7.m.14.Q resistors respectively and the two emitters have a 10 k. hre 75.14 is converted into a current a mplifier by removing the emitter resistors on Trl and Tr3 and the 10 k. A three stage common emitter amplifier employs identical transistors having h1e I 000 n.Q.Q resistor in the emitter Iead of Trl and a 20 k.Q. RLl 2 k.f.16. 200 k.15 using interconnected fourterminal network theory. Ans. In the emitter Ieads of Tri and Tr3 are 50 .Q.Q and the shunt bias components total I2·5 k. hoe 100 . Ans. 3·98 k. hoe I25 . 0·004 . Find also the output impedance if the source impedance is 50 .Q per stage. Circuit for Example 7.Q.umho. Feedback is provided by a 400 . Ifthe amplifier is driven by a source of e. calculate the output voltage and the amplifier output impedance. Calculate the overall voltage gain and input impedance.Q feedback resistor. 0·224 . RL2 4 k.Q per stage.15.13. 0·1 m V and internal impedance 2 k.Q. Example 7. Example 7.PRACTICAL FEEDBACK AMPLIFIERS Example 7. 0·475 V.Q resistor between Tr2 collector and Trl emitter. Ans.
4·5 kO. 1 820. 71·5 kO in parallel with 1 kO. and neg1igible hre. hoe 80 . Example 7. Find also the resulting output voltage.15.17. 10k.15. 17 kO. The amplifier shown in Figure 7. (c) 77·3. If the transistor parameters are hte 1·5 kO.O.umho. Ans.14 is required to have an output impedance of 1 0. At. 88·5 n. 95 n.18. 1 310. 10k!1 20k!1 (a) (b) 60k0. hre 140. Ans. (b) 18 4oo. 83 mV. (a) 46·1. kO. 287. calcu1ate the required value of Rp. (c) Figure 7. Each stage of a multistage amplifier has Z 1n 1 kO. Calculate Av. Zout 2 2MO.EXAMPLES Example 7.18 Assurne that the collector Ioad for each stage is 1 kO and that the driving source impedance is also l kO. Circuits for Example 7. 253 . and an open circuit valtage gain of 150. 1·65 kO in parallel with 1 kO. 6·23 kO. Ztn and Zout for the configurations shown in Figure 7.
Circuit for Example 7. Ans. 0·99 MHz.19.16 employs transistors with hie 1 000 n.PRACTICAL FEEDBACK AMPLIFIERS Example 7.O. 212.umho and hre 0. 254 . hre 100. The effect ofthe bias components may be neglected and the effective capacitance 20k. Determine the maximum gain and the 3 db bandwidth assuming that the coupling capacitors have negligible reactance at all signal frequencies.19 in parallel with each load is 500 pF. hoe 125 . The amplifier shown in Figure 7. Figure 7.16.
Electronic oscillators can be constructed to work at frequencies as low as one cycle in 10 minutes or Ionger and as high as 200.f.1) If the gain. 255 . voltage phase shift networks giving 0° or 180° and current phase shift networks giving ooor 180°. Consider once again the expression for the gain of an amplifier with feedback: Ar= 1 lAI LO lßAI LO + cp (8.m. the system will oscillate. This resulted in self oscillation. but the basic principles for all frequencies are the same. or an output when no inputwas present. lßAI must equal 1 and (() + cp) must be zero. Ar. F or many electronic systems a source of alternating e. cp must be 180° or 0° at that frequency only. Then if A ?. (1/ß) at that frequency. or current is required and oscillators based on the positive feedback principle are suitable for this purpose. We shall consider four cases. is to be infinite. two conditions must be fulfilled. RC OSCILLATORS A common class of oscillators employs feedback networks consisting of resistors and capacitors only. we found that under certain conditions positive feedback could result in an amplifier having infinite gain at certain frequencies. For frequencies higher than a few 100 MHz special forms of valve and circuitry are required.8 OSCILLATORS In Chapter 6. In generat () will be either 0° or 180° depending upon the number of amplifier stages. These feedback circuits may then be employed with one or two stage voltage or current amplifiers respectively.000 MHz. For oscillation at a particular frequency.
. Using normal potential division methods.jXc2R2 ß= and if (8.1. we can write jXc2R2 V2 = R2 .. having no imaginary component. V2 V1  jXc2R2 ~~~~~~~~~~~~~~ R1R2 .R .e. Wien bridge voltage phase shift network Multiplying by (R2.2) v2 v1 The angle of this ß factor must be either 0° or 180°.X  The frequency at which this is true is given by w2 and = RIR21CIC2 (radfsec)2 1 /= 27ry(R1R2C1 C2) Hz 256 (8.jXc2). This can only be true if the real part of the denominator equals zero.OSCILLATORS Voltage Wien Bridge Oscillator Consider the voltage phase shift network shown in Figure 8.Xc1Xc2 .X 2. R1R2  XCIXc2 =0 1 wC1 1 wC2 R1R2 = .3) (8. the angle of the numerator is 90°.2. X Vl Amplifier output terminals tV 2 Amplifier input terminals Figure 8.1.J C2 .jR1Xc2 . Thus the whole expression must be real. In equation 8. ß will have the required angle.J X c1 .jXC2 jXc2R2 R1 . if the angle of the denominator can be ±90°.4) . i.jR2Xc1 .
jR1Xcz. So. 1 (8. Unfortunately these circuits are not suitable since the very low input impedance in each case must effectively become R2. the amplifier required must have no phase shift and a voltage gain of at least 3. Suitable values of R1._J wC 257 . and to ensure that the ß network has a negligible loading effect upon the final stage. identic:al capacitors and resistors are used: Av mln = 1 + 1 + 1 = +3 Since this result is positive. To find the order of the loading effect we will consider the case when R1 = R2 = R and C1 = C2 = C. cancelling the j throughout and inverting XCI Rl 1 =++ Xc2 R2 putting Xe = (lfwC).RC OSCILLATORS Note if R1 = R2 and C1 = C2. Av mln c2 R1 = Cl + R 2 + 1 (8. The impedance Z presented by the feedback network to the amplifier output terminals is given by: jR j wC Z=R.5) f= 2TTCR Hz Now at this frequency. the feedback factor V2 ß= h = ß is given by: jXc2R2 jR2Xc1. as before. C1 and C2 for the required frequency then make Av mtn more the amplifier can provide when loaded with the feedback network. The practical solution is to use a two stage common emitter or grounded cathode amplifier.jR2Xc2 The minimum amplifier gain Av must be equal to or greater than 1/ß.6) If. wC R. This suggests either a single stage common base transistor amplifier or a grounded grid valve amplifier.
2 1 . The finalload of 1 kQ makes the amplifier output irnpedance much less than the network impedance. jR2 C C jCR RC j] = 3 R (.J) (8. Figure 8. the second stage gain !Avzl = 30 X 1 10 + 1 258 = 2·7 . Thus a suitable amplifier will be one having an output impedance much less than its input impedance with a voltage gain greater than 3.2. Wien bridge oscillators the network impedance 15y2 L 45° kil..7) But in practice R will be the value of the amplifier input impedance.2a shows a valve oscillator designed to operate at 1 kHz.5. and the amplifier input impedance excluding the Rg of 10 kQ which formspart of the ß network is infinite..J) .. Figure 8. Taking typical 1l and ra values of 30 and 10 kQ respectively.2 shows valve and transistor circuits based upon these principles. The feedback network has equal capacitors. and resistors rnaking 0·0161!F 10kQ [b) [a) Figure 8. (1 . jCR Z=RC =R [( .j) 1 .OSCILLATORS Substituting for w from equation 8. and the minimum amplifier voltage gain 3.
The circuit begins ta ascillate. Making R1 10 kO ensures that the netwark impedance Z will be much greater than the final laad af 1 kO. A stable canditian will be reached when ßA L () + cf> is exactly equal to 1 L oo. and the amplitude of the oscillating sinusaidal signal tends ta rise·ta infinity. It must pass thraugh the value which makes ßA L () + cf> equal ta 1 L oowhere the gain with feedback becames infinite.2b shows a transistar circuit designed ta ascillate at 5 kHz. Amplitude Stability This raises the questian of the behaviaur af these circuits if the gain is mare than the minimum required. a finite time will elapse befare the direct currents build up ta their steady value. supplies switched aff.3. When the supplies are switched an. But as this signal amplitude rises the gain will fall.4 gives the camman value af the equal C1 and Cz fram Naw applying equatian 8.6 ta find the minimumvaltage gain: Av mln 10 =l + 1+1= 12 If the transistars have hre 50.RC OSCILLATORS So pravided the gain af the first stage is greater than 1·1 the circuit will ascillate. and hre Av~hZL le Then Av2 ~ 50 giving mare than sufficient valtage gain. The gain will be zera ar very much less than one. R2 in this case is the amplifier input impedance which will ~e appraximately hte.2a. During this time the gain will rise fram zera tawards the final steady value calculated far the circuit. or because the devices run into cut aff and battaming. either because af change in parameters with large signals. say 1 kO. Figure 8. This process is shawn in Figure 8. First consider an amplifier circuit with the d. Applying equatian 8.c. 259 . Suppase the averaU signal gain of the amplifier is 6 withaut feedback and equal resistars and capacitars are used as in Figure 8.
3e the resulting input is 1·6 V peak to peak but if it was multiplied by 6 the output would be 9·6 V which is greater than the d.c.c. supply and that the d. supply voltage. conditions Figure 8. In Figure 8.c. Figure 8.c. Since ß is one third.3b.b.OSCILLATORS Assurne our amplifier is a two stage transistor amplifier working from an eight volt d. operating point for the second transistor is given by VcE = 4 V. This is amplified by 6 to give the output of 2·4 volts peak to peak shown in vo Bottoming Harmonics attenuated by f.3.3d). network 2 ~'tr'\r'trl4 Vo4= Vo3 VoJ f:JvoJ= (f) Vin4 (g) Stable condition (h) Cut off 6 8 Figure 8. the new signal at the input is 0·8 V peak to peak resulting in an output of 4·8 V (Figure 8.3/ The fundamental of the waveform is attenuated by one third but the harmonics are more severely attenuated thus 260 . The final transistor therefore cuts off and bottoms as shown in Figure 8.3a shows the input signal shortly after oscillation has commenced. Limitation of amplitude by d.
Unfortunately the amplitude of the resulting output signal will be extremely sensitive to any change in temperature or supply voltage. Voltage negative feedback is provided by R1 and Rz with Ce to eliminate any d.3g is nearly sinusoidal. R1 Figure 8...4 shows a circuit employing this technique.t.RC OSCILLATORS Vtn4 in Figure 8.. This will again cause bottoming and cut off resulting in the same value of ßvo as before. the input impedance for Trl will be large.c. The gain is now F_u_n_d_a"""'m. Firstly negative feedback can be included in the circuit to reduce the small signal gain to just greater than 1/ß. Wien bridge oscillator employing negative feedback for amplitude stabilization 1/ß before excessive distortion occurs... 261 . change of parameters can be sufficient to reduce the gain to exactly n. Then as the signal amplitude increases.c. The shunt R of the positive feedback network will therefore be given by RBI and RBz in parallel. This fault may in turn be eliminated by making the amount of negative feedback proportional to the signal amplitude. Since negative feedback is employed.4.en_t_al_o'/~p __8__ 3 i/p . Figure 8.2·67 The resulting distortion in the output is undesirable and additional techniques must be introduced to eliminate it.
w2C1C2 + jwC1G2 + jwC2G1 + jwC2G2 (8. Following the same reasoning.. in this case negative (n.y(RtR2C1C2) ( 8. A suitable circuit is shown in Figure 8.5. When the output voltage approaches the desired maximum value the dissipation V2fRt in Rt rises and with it the temperature of Rt. The negative feedback ß is given by R2/(R1 + R2) making the gain greater than 3 for small signals. jwC2G2 hY2 h G2 + jwC2 h= Y1+ v2= .10) .8) This expression has exactly the sameform as equation 8. it is more convenient to work with admittances as shown. R1 is a thermistor which is a resistor having a very high temperature coefficient of resistance.5.).c. Wien bridge current phase shift network Since we are concerned with current division. An alternative form of Wien bridge oscillator is based on a current amplifier. f 1 = 27TCR as before 262 (8.2. at the required frequency. the phase shift network is required to give zero phase shift between input and ou~put currents.OSCILLATORS path.t. A stable condition is now reached without distortion of the output signal. . In this case.9) If Rt = R2 = R and Ct = C2 = C. The resistance of Rt falls increasing ß and thus reducing the gain. C 2 JW 2 Multiplying numerator and denominator by G2 /2 h= + jwC2 jwC2G2 G1G2. Figure 8.. jwC2G2 Gt + JWCI + G + . Current Wien Bridge Oscillator The oscillator circuits discussed so far are known as Wien bridge oscillators. we may deduce that for zero phase shift w j(G1G2) _  C1C2 1 .
the minimum current gain Atmln is given by Atmln 1 h G1 C1 Rz C1 = ß = h. it must have zero phase shift. Since the Ioad is reactive (8.11) which.13) If the amplifier is a two stage transistor amplifier the shortcircuit current gain Atstc will have zero phase shift.1. current feedback. and to reduce the output admittance as required.12) The amplifier requirements can now be considered. If YL ~ Yo. The correct modifications will be achieved by using current derived. and the phase shift will be zero even if YL is complex. A thermistor can be used to Iimit the amplitude in the same way as for the voltage amplifier. The input a.1 263 . therefore negative feedback may be used to reduce the input impedance.12). ifthe short circuit current gain has zero phase shift. Amplifier circuit for Example 8. for equal components = 3.= 1 + Gz + Cz = 1 + Rl + Cz (8. This may be verified from an expression for current gain: (8. Such an amplifier will have a current gain far greater than that required to sustain oscillation. and the input impedance must either be less than R1 or become R1. 7 the reader can show this to be Ytn = jG(1 + j) (8. Yo may be neglected. the same phase shift will be obtained on load when the amplifier has a low output admittance.dmittance may also be determined for equal components and referring to equation 8.RC OSCILLATORS At this frequency.6 is tobe converted into an oscillator by (a) using a voltage phase shift network. and 12V Figure 8. Example 8.6. The amplifier shown in Figure 8.
All the resistors shown are required to provide the correct d. This will make the input impedance at the base of Tr1 very large and R for the network can be taken as the 6·8 kO and 22 kO bias resistors in parallel. But ß will be approximately 1/3 so R1 will be of the order of 2 kO.1 47·6 = 0·313 264 .e. Now to determine Avo: Av2 = Yu = = Avlo =  = 100 1 017 . conditions and stability. 3 and 47·6 = 1 + 47·6ß ß= 47·6 1 500 (9o + 1 0: ) 106 667 + 37 + 100 + 213 . But feedback must reduce this to 3.c. Design suitable positive and negative feedback circuits for each case if the required oscillator frequency is 5 kHz. 750 n. Assurne transistor h parameters of hte 5 kO. hoe 90 . i. series applied feedback will be employed to reduce the gain to the required value of 3 for equal capacitors and resistors in the phase shift network.OSCILLATORS (b) a current phase shift network. the load can be taken as 1 kO in parallel with (2 + 1) kO. the emitter resistor of Tr1 may be unbypassed and used for the R2 of the feedback network. removing any possibility of loading by the network. Thus for the calculation of Avo.umho 100 1 500(1 100)106 = 60"5 ß =REh=== 1 But local 60·5 Avlt = 1 + 60·5 =::= 1 but Ztnr = 1·5(1 + 60·5) kO = 92 kO Overall gain without feedback = 47·6. Consider first the voltage phase shift circuit: voltage derived. The negativefeedbackwill also make the output impedance very much less than the 1 kO finalload.umho (47·6/3) . hre 100 and hre 0. Working on the assumption that the final gain will be approximately 1/ß.umho.
265 .7. = 92(1 + 0·313 = 1·5MQ Z1nr X 47·6) kil Thus R for the phase shift network is given by the bias components alone.'VI t 0006J. f 21r~R (equation 8. The completed arrangement is shown diagrammatically in Figure 8.8 kil = 5·2 kil The required frequency. by making R1 a negative temperature coefficient thermistor of say 2·7 kil cold.313 . 6·8 X 22 R = 22 + 6.U F = 0·006 p.7. and the resulting change in Avz will have a negligible effect upon the calculation. Now.F 5·2k.t . Solution for Example 8.Q .RC OSCILLATOkS But ß = R 1 ~ Rz = 1 : R 1 (since Rz = 1 kO) 1 R1 = 0 . 2·7k.c.1 = 2·2kil which compares favourablywith the estimated 2 kil.I(a) In practice ß would be adjusted so that the small signal gain was greater than 3.Q. Figure 8.1F 0·006 fLF n.5) = 5kHz = C= 106 27T X 5 000 X 5 200 .
j) kO which will cause a negligible change to the gain and phase shift of the feedback amplifier.j) = 7·8(1 . Forthis calculation. Neglecting hoe. 1 kO. the voltage gain is required. hre Avo=hY !e L and (1 .ßvAvo) = 1+ hreZe YL h y !e = 1 + 100 L X 1 000 1 500 :::= 68 The input impedance and admittance are therefore given by. Turning now to the alternative solution with a current phase shift network. Current derived current feedback can be obtained by unbypassing the emitter resistor of Tr2 and connecting an additional resistor from Tr2 emitter to TrI base.OSCILLATORS Finally. taking Zo without feedbacktobe approximately 750 Q 750 Zor = 1 + 0·313 X 47·6 Q = 47 Q But the Ioad provided hy the phase shift network is JR(1 . Z 1n = 1·5(1 + 67) kO and 266 Ytn = 9·7 . h = 1·5 + jl·5 mmho === 2 L 45° mmho This is very much greater than hoe (0·09 mmho). to confirm that the required output'impedance condition is satisfied.umho . the R will be the Ioad of Tr2. the current gain of Tr2 is therefore approximately hre· Tr2 input impedance will be increased by the 1ocal voltage feedback provided by Tr 2 emitter resistor.12) and putting G as 1 mmho. The Ioad admittance is given by fG(l + j) (equation 8.
umho Ttn = 1·16 kO The required feedback resistor is therefore just Iess than 1 kO. including current splitting for both sets of bias components is given by: 667 At = 667 + 45·5 + 14·7 X i/p bias = 0·918 X 100 9·7 X 9·7 + 90 + 212 + 37 + 100 X interstage bias hte 100 X 100 hre 2·05 = 188 But the required gain of 3 must be given by At Au= I . The overall input impedance will now be given by Ztnt = Ztn 1 .ßAt 1·16 kO X 0·33 ~ 18 O = 1 + 188 267 . Now and Ytn = 667 + 45·5 + 147 = 859·5 .33 mmho ~ 0·5 mmho This would be provided by a 2 kO resistor but the input impedance without feedback must be included in this vaiue.0·33) Y2 = 0·33 1 _ 0 .ßAt 188 3 = 1 + I88ß 188 T1 ß= and But ß is given by Y2/( Y1 + 188 = 0·33 Y2) when Y1 is the 1 kO emitter resistor.RC OSCILLATORS The current gain. 0·33 Y1 = Y2(I . In practice a thermistor having a cold resistance of I kO would be satisfactory.
1k. 268 . Solution for Example 8. 1 f= 2TTCR I C = 106 2TT X 5 000 X 1 000 . The attenuation is greater than that found for the Wien bridge networks.OSCILLATO:RS This is very much less than the series R for the phase shift network which must therefore be included externally.uF = 0·03 . The output admittance with feedback is given by Yo hoe 1.ßA = 1 + 188 x 0·33 = 1"5 . 1k Q Figure 8.umho satisfying the required condition that Y 0 ~ YL. All that remains is to determine the value of C for the network.8. at a particular frequency.10.t.uF The complete arrangement is shown diagrammatically in Figure 8. but a single stage amplifier can be constructed to give sufficient gain for the combined circuit to oscillate.8.O. to give 180° phase shift.c.I(b) Single Stage RC Oscillators An alternative RC network may be used. From equation 8. 0 03 f!F n.
For the same conditions the voltage attentuation.9c and d show current phase shifting networks having 180° phase shift at v6 I 2TTCR Hz and 2TTCRy6 Hz respectively. In each case the network input impedance should not Ioad the amplifier. but forthat in Figure 8. (a) (c) (b) (d) Figure 8.9b. or current.9a. With suitable ratios between the C and R values. Figure 8. such networks may be either voltage. With the circuit in Figure 8.9. the final R will be the amplifier input impedance. phase shifting.RC OSCILLATORS As with the Wien bridge circuits. is I/29th. With these circuits.9. Voltage and current phase shifting networks Figures 8. the ampiifier input impedance should be very much lower than that of the final 269 . The current attenuation will once again be I/29th for equal capacitors and resistors.9a and b are both voltage shifting and the frequency for 180° phase shift is given by 1 v6 2TTCRy6 Hz and 2TTCR Hz respectively for equal capacitors and resistors. ß. The four basic configurations are shown in Figure 8. the input impedance should be very much greater than Xe at the oscillatory frequency. the attenuation may be as low as I/8th but the frequency is of course modified.
From first principles. while that of the current networks is more conveniently achieved by nodal analysis.OSCILLATORS network component. 20 X 5 gx Ztn = 20 X 1 5 g+1 k!l = 800 Q Thus the phase shift network has identical components and may be analysed by nodal analysis as follows: = = 0= + jwC) . Only one example will be considered here.10. The analysis of the voltage networks is by mesh analysis.2. Oscillator circuit for Example 8.10. but the other forms may be analysed by similar methods. Example 8.umho respectively and hre can be neglected. and the network input impedance should be much less than the amplifier output impedance.GVa GVz + Va(2G + jwC) Node 1: ltn V1(G Node 2: 0 GV1 Node 3: 270 (8.14) .GVz } + Vz(2G + jwC) . Figure 8. hte and hoe may be taken as 1 k!l and 100 .2 The amplifier input impedance is given by hte in parallel with the bias components. determine the frequency of oscillation and minimum value for hre for the circuit shown in Figure 8.
G2] + G[ G(2G + jwC).19) .0) + jwC)[(2G + jwC)2. 16) If ß is to be real.....2G3 + j4wCG2  w2C2G  jwCG2 (8.5w2C2G Dividing through by G and substituting from equation 8.jwCG2 (8 .15) + j3wCG2  4w2C2G .= f= 271'CR 12·2kHz (8..0] + 0 and ß= lo ltn = (G + = 3G3 G3 2 jwC)(3G + j4wCG ..jwaca .. and or w = Gy'6 c rad/sec v'6. ß= Ga Ga . G G+jwC G 2G 0 ltn + jwC G 0 0 lo = VaG= G X .2Ga .. the imaginary terms in the denominator must be zero ...w2C 2) ..17) (8..18 ß= G2 1 G2 _ 30G2 = 29 271 (8.G G+jwC 0 G 2G + jwC G 0 G 2G + jwC Io = (G 0 + 0 + Gltn(G2.RC OSCILLATORS Now using determinants. equation 8..14 may be solved for Va and hence / 0 may be determined..18) At this frequency..
17 1 R Xe= GCy6 = V 6 = 3270 c Since the total network impedance must be less than this. Minimum hre = 29 X !~~ = 36 In practice both the frequency and the minimum hre would be modified by the Ioad and hoe and this could be allowed for in the first term in equation 8. the current gain may be taken as At = hre X y tn = Ytn + Yi blas 1000 hre X 1 000 + 250 For oscillation At must be 29.4w2C2G .2G3 .jwCG2 Ga w2C2G' + j3wCG2 .2Ga . first note that the reactance of one shunt capacitor is 1/wC.15 G3 = (G' + jwC)(3G2 + j4wCG = 3G2G' + j4wCGG'  w2C2) .14 by replacing (G + jwC) by (G Putting G' lo ltn = G + YL + jwC + YL + hoe) + hoe and rewriting equation 8. and since RL is parallel with hoe is much greater than this.jwCG2 Equating the imaginary term to zero as before: + 2wCG2 = 4GG2 + 2G2 = 4wCGG' w= 272 waca w2C2 y(4GG' + 2G2) C . · from equation 8.OSCILLATORS To determine the approximate minimum hre.jwaca .
At the resonant frequency of the anode circuit. The simplest form employs a parallel tuned circuit as the load of a single stage amplifier and inductive coupling between input and output for the feedback network. Consider first the tuned anode oscillator shown in Figure 8.2·5 . if an a. Thus.11. The useful range of up to 1 MHz is limited by stray capacitance and amplifier input impedance. the single stage amplifier is not as suitable for stabilization by means of negative feedback. Single stage RC oscillators are not often used in practice for two reasons. signal at this frequency is applied as Vgk. second. the current iL 273 . and the feedback circuit can be either capacitive or through mutual inductance.34·5 . frequency adjustment requires the simultaneous switching of three components.2Ga w2 C 2 G' G =w2C2G' 4w2C2 3G'  = 4·65  G2  c. Neglecting the effect of r. first. . With these. the oscillating frequency is approximately the resonant frequency of the LC circuit involved.4w2 C2G .c. the anode Ioad impedance is purely resistive and is given by (Lt/Cr) n. LC OSCILLATORS Wien bridge oscillators are widely used for audio frequency signal generators. y[(4 X 1·938 X 106) + (2 X 1·56 X lQ6)] Hz 27T X 4 X 108 = 131kHz ! = and fo ltn = 3G2G'  Ga . Radio frequency oscillators are usually ofthe LC type.  1·25 10·7 . the anode voltage ( [IZ)/(ra + Z) will be 180° out of phase with this signal.LC OSCILLATORS But G = 1·25 mmho and G' = 1·55 mmho. the combined effect is to change both the frequency and minimum hre by about 10 per cent. h re = 34·5 M tmmum X 2G 1 34·5 1250 = 43 1 000 lt is interesting to note that although the Ioad admittance and hoe are much less than the network admittance.
the system will oscillate.+ Figure 8. The sign here depends only upon the sense of winding of the two coils. A tuned anode oscillator methods. The equivalent circuit is shown in Figure 8. At the required frequency of 159 kHz the Q factor of each coil is 20. Vgk is given by: Vgk nd .11. Now if M and the valve gain are sufficiently large.12. l = ±jwMi = r +VaJWLa . This current may be analysed using normal equivalent circuit H. A tuned anode oscillator is to be constructed using a triode valve having f1 40 and ra 20 kil. 274 . Figure 8.T. If this is such that Vgk is given by jwMiL then the total phase shift is zero. If two coils have mutual inductance between them.OSCILLATORS through L1 willlag the anode voltage by a further 90°.3. This problern is best solved by finding a general solution for the frequency of oscillation and the maintenance condition using symbols. The available coil assembly has two windings of 5 mH and 0·2 mH with a maximum coupling factor K of 0·1 . a current i in one induces a voltage of ±jwMi V in the other. As a result of the mutual inductance M. taking either the constant current or the constant voltage form. Determine from first principles which of the two coils should be used in the anode circuit and the correct value of tuning capacitor C.
+ w2 = r ra = _1_ ( 1 LaCra LaC The frequency of oscillation f = 27Ty~LaC) Imaginary parts. and taking the negative M we obtain: Mf1 =La+ Crra which is known as the maintenance condition. but if the circuit is oscillating Va ~ 0. + ~) ra J( + ia) 1 ±wMgmra = wLa Hz (8.21) + wCrra Putting grnra = f1. the required results can be obtained. by equating the real and imaginary parts of this equation.LC OSCILLATORS Writing a nodal equation: . Real parts. therefore Va may be cancelled.22) . 275 (8.20) (8. ±JwMgm Va . La r One solution of this equation would be given by Va = 0. The equivalent circuit for a tuned anode oscillator lf both sides of the equation are now multiplied by ra(r we obtain ±jwMgmra = (r + jwLa)(l + jwCra) + ra = r + ra  w 2LaCra + jwLa) + jwLa + jwCrra Now. L a = ra + J VawC + r + jOJ . dividing by w. Figure 8.12. Va Va + JW .
for each coil we can find r. La Crra (8. Similarly Iet L2 and r2 be the inductance and resistance of the 0·2 mH coil. we can see that if !' is sufficiently large. 1 C=47T2. and for each coil C and r can be found from equation 8. from equation 8. Thus for each set of values.OSCILLATO:RS In the problern M and !' are known. Let the 5 mH coil be L1 and the associated resistance be r1. (r/ra) ~ 1 for both r1 and rz.23: fJ. equation 8.21. = 5 X I03 10_4 + 200 X 1012 X 250 X 2 X 104 = 60 10 _4 But the available valve has a !' of only 40. so this arrangement would not oscillate. 276 .f2 C1 and = 1012 C2 = 1012 Also X = 1012 1012 5 X 10_3 pF X 1012 2 X 104 = 200 pF = 5 000 pF M = Ky'(L1L2) M = O·ly(1·0) = 0·1 mH If L1 is used in the anode circuit. Now from wL Q r= and rl = 106 X 5 X 103 = 250 Q 20 106 X 2 X 104 20 = 10 n r2 = Referring to equation 8.23) fJ.f2La Substituting values.20.22 can be rearranged to. 1 f =:::= 27Ty(LaC) Hz and rearranging. and putting 4rr2.=+ M M First.
13.13. (b) h parameter equivalent circuit. the same equation yields: I'= 0·2 X 103 104 + 5 000 X 1012 X 104 X 10 X 2 to4 = 12 Thus if the smaller coil is used in the anode circuit. The mechanism of the operation is identical to that of the valve circuit. of 40 will be more than sufficient to ensure oscillation. The circuit and the h parameter equivalent are shown in Figure 8. the p. but the analysis is a little more involved as alternating currents flow in both coils. (a) e (b) Figure 8. A tuned collector oscillator. The Tuned Collector Oscillator The transistor equivalent to the tuned anode oscillator is the tuned collector oscillator. (c) Simplified equivalent circuit 277 .LC OSCILLATORS If L2 is used in the anode circuit. (a) Full circuit.
278 .c.27) If the circuit is to oscillate ib =F 0 and can be cancelled.: since both currents shown enter the coil at the end indicated. C2 has a very low reactance at the oscillatory frequency. R1. The full equivalent circuit is shown in Figure 8.13a shows the complete circuit.26) Substituting for ic in 8. R2.25) Rearranging equation 8.j) + w3M2C 1 1 1 (8.f. Thus multiplying equation 8. (h!e +OJ MjwLb)ib ( rc + JW c OJ jhreib 1 1 1 ~ (8. 1 ) + .24) and 1 ) .27 by jw2 MC Ieads to: wMhre = (h!e + jwLb)(wCrc + jw2Lc C.OSCILLATORS Figure 8.24 .m. operating point. the real and imaginary components may be selected and equated. The dot notation on the coils Lc and Lb indicates the sense of the mutually induced e. L c + :c JW lb = J. Re and Ce provide the correct d. L c + jwC = lc. ( rc + JW jhreib 1 wC 1 . The simplified equivalent circuit shown in Figure 8. zc = (hle 1 + jwLb)ib . effectively earthing one end of coil Lb.28) Without expansion. (h!e 1 = +J + jwLb)ib OJ M (8.26.13c is obtained as follows: hre is neglected and rb is added to h1e to make the new component h1e hoe is combined with rc and Lc modifying them to rc and Lc This operation will be shown in an example. both jwMi terms will be positive. M JW . Thevenin·s theorem is applied to the remainder of the collector circuit resulting in the voltage generator of: 1 1 • = ftlb ft hreib x 1 • 1 ~C JW jhre = wC The impedance in series with the generator is given by 1/jwC and is represented by C. M" . lb · M"lb + JW (8.13b. By mesh analysis: 1 (8.25 from 8.
LC OSCILLATORS Imaginary terms: 0 2 W = w2LbCrc' + w2Lc'Ch1e'. determine the actual frequency of oscillation and the minimum value 279 . A tuned collector oscillator employs a collector coil of inductance 1·6 mH and resistance 100 Q tuned to a nominal frequency of 106f27T Hz by a shunt capacitor. w2 = Lc'C ( 1 + fc~~~J (8..4.+L c ' M+M h1e'rc'C M =~+Lc' (8.w3Lc'LbC Dividing through by w and substituting for w 2 from equation 8.h1e' h1e' = LbCrc' + Lc' Ch1e' Dividing by h1e' and rearranging.30) and the frequency of oscillation (8.29) In practice Lbrc'fLc' h1e' will usually be much less than one.31) Real terms: wMhre = wCh1e'rc' + w3 M 2C + wLb. The base coil has inductance 0·1 mH and 10 Q resistance with coupling factor K ofO·Ol. 1 Lc'C w2=::= (8.30 Ieads to Therefore the minimum hre necessary to maintain oscillation is given by Lb Lb M hre = h1e'rc'C M . Ifthe transistor h1e is 1 000 Q and the effect of hoe is neglected.32) Example 8.
32. Figure 8.umho.OSCILLATORS of hre.14.umho = 625 . Circuit for the solution of Example 8. Consider the circuit shown in Figure 8. hre = = 1 010 1·01 X X 4 100 X 625 4 x I06 62·5 + 2·5 X X 1012 4 X + 1·6 X 10.4 At the nominal frequency of w = 106 radfsec the circuit will have an impedance which may be represented by rc' and Lc' in series.O. and an external shunt Ioad of 2 kil are included. C = w2Lc = 1012 X 1·6 X I03 = 625 pF and M = Ky(LbLc) = 0·01y(1·6 x 0·1) mH = 0·004 mH From 8. First the values of C and M are required.6 I03 I03 Minimum hre = 15·8 If hoe and YL are included we must find the modified values of Lc' and rc'.umho Rerr = 1·6 kil Yerr 280 . 1·6mH 100fi 2k. First YL and hoe may be 1umped together: = (125 + 500) . Find also how these va1ues are modified ifthe hoe of 125 . Since 1 1012 w = 106.14.29 w2 = LcC (1 + 1 0·1 X 100 ) 1·6 X 1 000 ! 1012 = 1 + 6·25 X I03 === 1012 106 =Hz 27T Also from 8.
y5 and 103 f = 27TV 5 kHz = 71 kHz This is an approximation. 10_2 = 184 These results show that if the design frequency is to be maintained. since this va1ue of w shou1d have been used in the calculation of Lc' and r 0 ' . h _ 1 010 tmmum re  = 808 X X 0·1) = 2·74 800 2·74 X 62·5 2·74 X X + X 106 H 625 X 1012 2·74 101 + 7·53 O· 6 3 X X X 106 104. the tuning capacitor will have to be increased by a factor of approximately 2. Also the available hre is unlikely to be as high as 184 suggesting a maximum shunt loading of say 5 kQ.j160) Q + j753 Q rc' = 800Q Lc' = 0·753 mH 1 2 Now w  0·753 X 103 X 625 X 1012 (1 + __Qi_ X 0·753 800 ) 1000 1012 . 10 281 .j1·6) Q 1·72 + 1·62 = 0·294(170 = 800 + 2 560 + j2 720. The new value of M = O·Oly(0·753 M' .LC OSCILLATORS Also at w = 106 rad/sec XL= 106 X 1·6 X 103 = 1 600 Q Overall Z = = (100 + j1 600)1 600 1 700 + j 1 600 Q 1·6(100 + j1 600)(1·7.0·47(1 + 0·106) w = 106 .
OSCILLATORS Hartley and Colpitts Oscillators There are many other forms of LC oscillator two of which are shown in Figure 8. The Colpitts oscillator shown in Figure 8. and h1e together as h1e'.16b shows the complete equivalent circuit for the given information. The feedback capacitor C' will also M c L IXc=O (b) (a) Figure 8. One further example will be considered.16a employs a transistor having hre 40.umho and negligible hre.5. 282 . Analysis in each case may be accomplished using normal equivalent circuit methods. by expressing the current generator in terms of h instead of ib. For both circuits the transistor employs the normal bias circuit with the emitter capacitor having negligible reactance at the oscillatory frequency. YLl and YL2. Find also the maximum loading beyond which oscillation will cease. Determine the unloaded frequency of oscillation.15.15. and RB2. h1e 1 300 0.16c is obtained by combining h0 e. Hartley and Colpitts oscillators have negligible reactance. Example 8. and the maximum loading for which the frequency will be changed by less than 5 per cent. hoe 125 . and the bias resistors may be neglected or included in the transistor hte and hre parameters. and by combining Rm. The simplified version in Figure 8. Figure 8.
1H (c) (a) L c. (b) Figure 8.34) . (c) Simplified equivalent circuit Also by expressing ib in terms of h. (b) Equivalent circuit.LC OSCILLATORS The values of these components are obtained as follows: YL' = 125 + 200 + YL2 = (325 + YL2) .5.16. (a) Colpitts oscillator for Example 8. 1 1 hre = 40 X 1 2 700 1 300 1 1 + 10 100 + 1 300 283 ~ 25 (8.34) L 200f.umho (8.33) and 1 hte' 1 1 1 = 2 700 + 10 000 + 1 300 mho hte' = 800 n (8.
38) ll=1 hte and from equation 8. But (h I (8...w3LC1C2 + wC1 hte wL 1 1 Rearranging and dividing by w.wL wLh 1 . Node 2: v2 1 .39) Substituting for V2 from 8.40) Following the usual technique. hre Vl jV1 (jwL )( 2 htel = wL + htel .36) .w LC1 + 1 h 1 1 1 ) + JWC2 + jwL V1 • (8..36 (8. V1 cannot be zero and may therefore be cancelled.oC1 + JW :L h Node 1: ..37.38 into 8.37) (8.' + (1 1 1) ( w2LCI) wC2 .+ wC2 .. 1 ) o = v1 ( h + y.h + Cl + c2 le 2 w = I 1 YLI L ( C1C2 ) + C1C2hte' c1 + c2 (8. Now equating the imaginary components: 1 0 = wL + 0 1 = wL + wLh .OSCILLATORS Our equivalent circuit is now in the correct form for nodal analysis. LYL1 w2LC1C2 = .hre ii I =  vl jwL + V2 ..:L JW I c 1 ) + JW 2 + jwL • v1 (8.42) .39 and h from 8.41) c1c2 C1 + C2 = Cett Putting 2 1 w = L_C_e_rr (1 LYL + (:C:::1+C:::2:)h:t:el 1 284 ) (8.
::::::::::::. whichinequation8.40.43) But from equation 8. 200 X t06YL' 1100 X 1012 X 800 and YL' = 0"18 = 0·18 X 4 X 1100 . the term 1·036 must change to 1·036 x 1·05 = 1·09.umho = 793 .umho which represents a shunt 1oad of 2·1 kO.074) 107 co = 1.200 X I06 X 91 1 100 X I012 X 800 1014 = 1·82 (1 + 0.41 1 (8. 101 2 ( 200 X I06 X 325 X I06) co 2 1 + ::::::. hte hte coL Minimum hre' = hte' h'(co2LC1) + co2LC2  1 (8.44) 285 .35 x 1·036 rad/sec and the frequency J= 10 X 1·036 27T X 1.325 = 468 .35 MHz = 1·22 MHz If this frequency changes by 5 per cent. + .42becomes 1·092 = 1·18. jcoL ( JCOC2· j ) . hre' = ( I .CO2LCl)h..LC OSCILLATORS From the information supplied... Cerr = 91 pF and taking YL' when YL2 is zero.umho YL2 = 793. For the second part of the problern we must equate the real parts of equation 8.
In practice a Ioad of this magnitude would certainly stop oscillation.6. network 22·2 kQ. 1 kQ. Frequency stability with change of d. but some of the factors involved are as follows: Frequency stability with change of load. EXAMPLES Example 8. and an accurate analysis should include these components. A two stage valve amplifier has loads of 20 kQ and the first stage has an undecoupled cathode resistor of 1 kQ. The resistance of the coil has been ignored in the analysis since the loading effects of YL' and hte' would make this negligible compared with Ioads greater than 1 kQ. A two stage transistor amplifier has a first stage voltage gain of 1 L 180°. The valves have fL 80 and ra 20 kQ and the effects of RG.D. It has a series arm of 5 kQ and 0·001 fLF and the parallel arms are 20 kQ and 0·02 f. P. supplies.OSCILLATORS Once again inserting values: 25 150 800 = 800h' + 10 10 = YL' = 187 mmho This represents a shunt Ioad of approximately 5 Q. A Wien bridge feedback network is connected to convert the circuit into an oscillator.c. Example 8. The choice of oscillatory circuit for any particular application is beyond the scope of this book.F. and an example of this will be discussed in Chapter 9. Ce and Cs may be neglected. The final stage collector Ioad is 500 Q 286 .7. 3·5 kHz. Tuningrange for available variable capacitor. Determine the minimum gain required from the amplifier and hence design a suitable negative feedback network so that the oscillatory output will be undistorted. Ans. This last effect may be utilized in the design of other forms of oscillator circuits. Find also the frequency of oscillation. this is no Ionger true. 21·25. In the extreme case shown above.J. Effects of valve or transistor capacitances.
287 . Example 8. Calculate the value of R' . The transistor oscillator shown in Figure 8. 0·074 .8.1F Figure 8. If the Wien bridge network shown in Figure 8. calculate the values of C1 and Figure 8. If it is required to oscillate at 2kHz./p Tr2 collector 5k . 0·56 0. Assurne that the input impedance of the first stage is much greater than 1 kQ. R' is included to reduce the current gain to the required Ievel.18.7 and the frequency of oscillation.17 is connected between output and input. Circuit for Example 8.17. Ans.uF. Assurne all other capacitors to have negligible reactance at 2kHz.0.EXAMPLES and the transistors have h1e 1 000 Q and hoe 100 . 1·43 kHz. In the absence of a negative feedback network. 1k l"25J. determine the minimum hre for the second transistor 1 t_rr. 65.18 employs transistors with hre 90. hte 1·3 kO. hoe 125 .8 C2. Ans.umho and hre 0.umho. Circuit for Example 8.
Determine the value of the capacitors and the minimum value for RL if the frequency is to be 175 Hz. Example 8. bias voltage of 4 V. Circuit for Example 8.s. output voltage if RL is exactly 3·5 kQ and the pentodemutual characteristic is given by: h (mA) 65 52·3 40 27·5 17·5 9·5 Vax (V) 0 1 2 3 4 5 5·0 6 (Construct a graph of gm against peak to peak Ans.11. lower it. Ans. The feedback network is to be a three stage voltage phase shift network having eqtial resistors and capacitors. Ans. 61·5 V. 35. Figure 8. 288 .9. Ca1cu1ate the minimum value of RL and the frequency of oscillation. hre 120. The oscillator shown in Figure 8. The transistor h parameters are h1e 1·5 kQ.uF.OSCILLATORS Example 8.19.uF.m.umho. 2·5 7 1·0 8 0·3 9 0 10 Vgk. 1·3 kHz. 383 Q.umho. A single stage common emitter amplifier is to be used as a low frequency oscillator. What effect will the capacitive loading have on this H.) Example 8. If the oscillator frequency is to be 3 kHz.10. hoe 150 .T. hre 0 and the shunt bias resistors are 47 kQ and 12 kQ. 0·0066 . 3·23 kQ. calculate the value of C and the minimum va1ue of hre.19 shows a phase shift oscillator employing a pentode having gm 9 mA/V and ra 380 kQ at the d. + R=10k n C=0005JlF Figure 8.20 employs a transistor with h1e 900 Q and hoe 100 .c. 0·286 .9 frequency? Determine the r.
20. together with the phase shift network shown in Figure 8. Figure 8.I3 was replaced by a transistor having hte 500 n. Example 8. If the valve in Example 8. 970 pF.14.12 Example 8. Neglect the effect of hoe. 16.13. 0·785. Ans. If the frequency of oscillation is to be 250 kHz and the valve parameters are p 60.EXAMPLES ________~ ~~1c Figure 8. Circuit for Example 8. How would these results be modified if a Ioad of I 000 Q was capacitively coupled to the collector and the coupling factor K between the coils was reduced to 0·005? Ans. 46. Determine (a) the tuning capacitor. 8·2 kHz. I 160 pF. Circuit for Example 8.21. Ans. find the values of c and hre for the same frequency of oscillation. I! 289 . Determine the required current gain and the frequency of oscillation.12. ra 5 kQ. 0·02. A tuned anode oscillator employs anode and grid coils of Q factor 10 and inductance 0·3 mH and 0·7 mH respectively. 1 380 pF. hre and the bias components. (b) the minimum coupling between the coils.21. A current amplifier.11 Example 8. is tobe used as an oscillator.
16. hre.15. Ans.umho.22 shows a Colpitts oscillator employing a transistor with hte 1 500 Q and hoe 125 . Example 8.OSCILLATORS Example 8. bias components and coil resistance. Figure 8.22. Show that the approximate frequency and maintenance conditions for a Hartley transistor oscillator are given by: w = y[C(Lb Lb hte = Lc 1 + Lc ± lM)] rad/sec M ±Lc Neglect hoe. 840kHz. 290 . Calculate the value of C1 which will result in an oscillatory frequency of 600kHz in the absence of the load. 1 080 pF. Estimate the minimum value of the load Figure 8. Circuit for Example 8. Find also the frequency when this minimum value is connected. 28 Q.15 if the transistor hre of 30 is to maintairi oscillations.
5 k (a) Figure 9. and the transit time for which the electrons or holes are crossing the active region of the device. The electrodes ofa valve are conductors separated by an insulating medium. These are the effects of the various reactances associated with the construction of valves and transistors. In this chapter. for a valve and cx for a transistor will be considerably reduced._ca~._____ } ____. VALVE EQUIVALENT CIRCUIT AT HIGH FREQUENCIES We shall first consider in detail the effect of the interelectrode capacitance ofa triode valve.1.. Triode valve and equivalent circuit for high frequencies 291 RL . p. a .1a... If the transit time is of the same order as a single period of the signal. The resulting capacitances are denoted Cag.___.9 MODIFICATIONS TO EQUIVALENT CIRCUITS FOR HIGH FREQUENCY OPERATION The use of the equivalent circuits discussed in the preceding chapters becomes inaccurate at higher frequencies for a number of reasons... we shall consider how the small signal equivalent circuits must be modified for use at high frequencies and we shall examine the necessary techniques for the solution of various circuit configurations with such modified circuits.. Cgk and Cak and they can be shown diagrammatically as in Figure 9. The reactances are due to capacitance between the various electrodes and to lead inductance.
2. To analyse this situation.3.5. i2 = jwCag(Vgk  Vo) (9. Cak formspart of Cs and may be allowed for in the normal calculation of high frequency gain. 9. we shall assume initially that Cag provides a negligible Ioad on the output. Typical values are Cag 1·5 pF. including ra.1) Yln= Vgk (9.5) But where ZL' is the total effective Ioad.2.') 292 (9. Cgk 2·5 pF and Cak 1 pF. i (9. calculate the input admittance of the amplifier shown. Cak and any externalload present. 9.4 and 9. Equivalent circuit for the derivation of valve input admittance tance further modifying the gain of the preceding stages. This capacitance provides the principal component of C8 for the previous stage. The reactance of Cag is much greater than RL at normal operating frequencies. From equations 9.2. By inspection.MODIFICATIONS TO EQUIVALENT CIRCUITS Figure 9. but together with Cgk it produces a shunt capacitance across the input.4) (9. In addition it may result in a shunt conduc Figure 9. i = jwCgkVgk + jwCagVgk(l + gmZI. and with reference to Figure 9.1b shows the constant current equivalent circuit with the interelectrode capacitances included.6) .3) A pplying the sense of i2 shown.2) and h = jwCgJcVgk and (9.
umho = 100v(3·25) L 33° 42' .3.wCaggmX' (9. Each stage bas a resistive Ioad of 20 kQ and a grid resistor of 100 kQ. Determine the overall voltage gain and input impedance at an angular frequency w of lOB rad/sec.10 becomes Ytn = jwCgk + = jwCgk + jwCag(1 + gmR' + jgmX') jwCag(l + gmR'). Thus since gmZL' may be quite large the input capacitance may be as high as 100 pF. ra 10 kQ. Interstage wiring capacitance totals 10 pF. gmZL' will be a real number and the input circuit appears as the parallel combination of two capacitors. G=wCaggmX' Figure 9. Equation 9.3. Cgk and Cag(1 + gmZL'). Cak 1 pF. A two stage amplifier emp1oys triodes having gm 10 mA/V.VALVE EQUIVALENT CIRCUIT AT HIGH FREQUENCIES Now applying equation 9. and the 20 kQ load.33° 42' 293 . Cgk 2·5 pF.umho = 150 + jlOO ZL' = Av2 = !.1. If ZL' is inductive. This will be composed of the valve ra and Cak. The input circuit obtained is shown in Figure 9..11) If ZL' is capacitive. X' will be negative and the input admittance includes a positive conductance.1.gmZL' we must first find ZL' for the second valve.j3·08 kQ 55·5 L 180 . In general ZL' = R' + jX' where X may be positive or negative. = 5·~5 L YL 33° 42' = 4·6 . Ytn = _!__ = jwCgk + Vgk jwCag(1 + gmZL') (9. Effective input circuit for valve at high frequencies Example 9.10) If ZL' is approximately resistive. Cag 1·5 pF. YL' = 100 +50+ jlOB X IQ12 X 106 . X' is positive and the resulting input conductance is negative. Since the gain of a stage is given by .
j0·91) = 1·1 L 120° Thus the overall gain Av1 X Av2 = 55·5 X 1·1 L 93° 42' F or Y1n1 we again refer to equation 9. this becomes. in the RC coupled common cathode configuration. Cs and Cak1: Yu' = (4·62 + 0·1 + 0·05 + 0·01) + (j108 x 11 x 1012 x 103) = 4·78 + j8·25 mmho But Zu I Av1 1 4·78 . the resulting negative input conductance may be used to neutralize the Iosses in a parallel tuned 294 .11.11 : Y1n1 = [(j108 X 2·5 X + 1012) + (j108 X 1·5 X 1012 X 0·525) (108 X 1·5 X 1012 X 0·91)) X 1()3 mmho = j0·25 + j0·079 + 0·137 mmho Including the Rg for the first valve.umho + j7·3 mmho But this forms a part of Yu'. Rg. If the anode 1oad is inductive.252 kO = 0·0525 = 10Zu' = + j7·15 mmho • J0·091 kO (0·525 . ra. The changes in input admittance discussed above are the result of internal feedback. the remainder being provided by RL. and this may be utilized to design an oscillator circuit.782 + 8. Y1n = 0·147 + j0·33 mmho = 0·361 L 66° mrnho Z1n = 2·77 L 66° kO Thus at the upper figure limits of a triode. Y1n = [(j108 = 4·62 X 2·5 X + 1012) + (j 108 X 1·5 X 1012 X 47) (108 X 1·5 X 1012 X 30·8)) X 106 .j8•25 = Yu' = 4.MODIFICATIONS TO EQUIVALENT CIRCUITS Applying equation 9. the addition of extra stages produces 1itt1e increase in gain and a very low capacitive input impedance.
A suitable circuit is shown in Figure 9.J. This method is preferable since it eliminates the effect of coil self capacitance. 1 w2=:= LC C 1012 I04 pF = 1014 x = lOOpF :.4a and the valve used in Example 9. This anode Ioad is tuned to a frequency above that of the oscillator and is thus inductive at the required oscillator frequency. In practice. The parallel tuned circuit consists of a coil of inductance 0·1 mH. Valve input capacitance = 100.. Oscillator incorporating Miller feedback Example 9.. I I 1 I (b) (a) Figure 9. A lossless tuned circuit can maintain a nondecaying sinusoidal signal at the resonant frequency. 295 . Q 50 in parallel with a capacitor of 50 pF..50= 50 pF.1.4.VALVE EQUIVALENT CIRCUIT AT HIGH FREQUENCIES grid circuit. 1 I I I .1... ..2. determine the values of X' and R' for an oscillatory frequency of 107 rad/sec. Rg . But valve input capacitance = Cgk + Cag(l + gmR'). First we must find the value of the total shunt capacitance to give the required frequency.4b.4a. the parallel tuned circuit is provided by a quartz crystal and the inductive Ioad is a parallel LCR circuit as shown in Figure 9. Taking the simple circuit of Figure 9..
50= 2 R'=3·1k0 from·which At resonance. and the collector base junctions. two further effects must be considered.MODIFICATIONS TO EQUIVALENT CIRCUITS + 1·5(1 + IOR') pF Inserting values.uH) and the electron transit time.uH 107 10X' G~. The resulting parallel resistance may be of the order of a few 1 ooo n. X'= 0·133 kQ from which L' = 133 = 133 .L But R = Go= WoL Q 20 107 X 104 = X 50 I010 10_4 ll = 20 ~>' mho For oscillation. but these are the cathode Iead inductance (0·005 . In generat there will be capacitances associated with the emitter base. (X' in kO) Pentode Valves The effect of Cag is considerably reduced by the use of pentode valves where this parameter may be ofthe order of0·005pF. 20 X I010 10_4 = W Cag X 1·5 X 1012 X Inserting values. and its form depends upon the way in which the transistor was manufactured. the negative input conductance to the valve. and there is the 296 . values being typically 2 000 and 500 kO respectively. must equal Go. At very high frequencies of the order of 50 kHz. the conductance of a parallel tuned circuit is given by G _ CR o. HIGH FREQUENCY TRANSISTOR EQUIVALENT CIRCUITS When working with transistors at high frequencies.u and ra. Detailed treatment is beyond the scope of this book. similar modifications must be made to the equivalent circuits. These properties both have the effect of introducing a shunt conductance in the input circuit. Pentodes also have the advantage of very high . An exact model suitable for all frequencies is exceedingly complex.
manufacturers quote the upper frequency limit in one of three ways: h is the frequency at which IX has fallen by 3 db. i. depending on configuration. In addition the current gain IX or IX'. 297 .3 000 Q.e.40 mA/V gm is related to IX' by the equationgm = IX'/rb·e· The upper frequency to which this circuit is valid would lie in the range 10 kHz to 100 MHz depending on transistor type. We shall first examine a complete analysis at a particular frequency and then see if this suggests any approximation to simplify calculation. In certain cases variation of y parameters with frequency are shown graphically permitting calculation at any desired frequency. cb'C 5 . The hybrid TT is accurate b Figure 9.250 Q. but they are usually quoted for a specified frequency only. rb·e 250 .HIGH FREQUENCY TRANSISTOR EQUIVALENT CIRCUITS resistance of the base material between the active region and the external connection. 'b'C 2 .40 kQ Cb'e gm 20 . is reduced with increase in frequency.30 pF. For any particular transistor type.5. Hybrid 1r equivalent circuit for transistor in the common emitter configuration from low frequencies upwards. /1 is the frequency at which IX' or hre has fallen to unity. but becomes increasingly less accurate as fa. Typical values for the components shown are: rw 50 .5 000 pF rce 20 .5 Mn. 300 . The y parameters are frequently quoted in manufacturers' published data. In generat either the hybrid TT equivalent circuit or a y parameter circuit is used for high frequency work. /T is the calculated frequency at which IX' will fall to unity if the rate of fall at the upper frequency end of the pass band is maintained at 6 db per octave.5 shows the hybrid TT equivalent circuit for common emitter connection. Figure 9. In this section we shall consider the analysis of circuits using both hybrid TT and y parameter equivalent circuits. is approached. to approximately 0·7.
13) (38 000 . rb·e 790 Q.~.j94) + Vce(28·6 + j()4) (9. gb·c gL = 2011mho.MODIFICATIONS TO EQUIVALENT CIRCUITS Example 9.Vce(0·39 + j94) Vb•e(38 000 . and · Vb•e(1 265 + j9 494) .Vb•e(0·39 + j94) + Vce(8·2 + 20 + 0·39 + j94) where gb•e = 1 265 11mho. 1012 For Cb•e. If the collector load is a parallel tuned circuit of dynamic resistance 50 kQ at the resonant frequency of 1 MHz determine the output voltage. Collecting terms: 1 0 = = Vce = (1 265 1 Vce = = 0·39 11mho.Vce(O· 39 + j94) 38 OOOVb•e = . Repeat with suitable approximations where the 1oad is on1y 2 kQ. volts. fb•e.JXbe the input current may be taken as 1 flA.j94) + j9 494)(28·6 + j94) + (38 000  j94)(0·39 104 X 9·55 L 82° 24' X 9·83 L 73° 6' 3 800 L 9' = 24·8 L 155° 39' + 0·39 + j94 = 22·6 + j10·2 + 0·39 + j94 = 22·2 + j104 = 106 L 102° 298 + + j94) 0·39 '94 +J . ree 122 kQ and gm 38 mA/V. A common emitter amplifier is supplied from a high impedance source having short circuit current of 1 11A at a frequency of 1 MHz. Xe= 1 500 X 27T X 106 = 106 Q = Xbe and Bbe = 9·4 mmho 1012 For Cb·c. Fora frequency of 1 MHz the reactances ofthe various capacitors may be determined.12) (9. Cb'e 15 pF. Working in flA. The transistor hybrid 7T parameters are: rbb' 100 Q. We can therefore solve by nodal analysis using nodes Vb•e and Vce only. Xe = 15 X 27T X 106 = 10·6 kQ = Xbe and Bbc = 0·094 mmho Since the input impedance is less than fbb' jXbefb•e . rb•e 2·6 MQ. Cb·e 150 pF. and 11mho: 1 = Vb•e(1265 + 0· 39 + j9 400 + j94) . gce = 8·2 11mho.
122 + 122 =:= 76 + j9·4 + j0·094(1 + 76) mmho = 1·265 + jl6·6 =:= 16·6 L 85° 42' Now i vb'e = Ytn I06 X 103 = 16·6 X I03 L 85° 42' mV = 0·06 L 85° 42' mV The output voltage is given by = . lf the collector Ioad is sufficiently small.14) If the overall voltage gain is required.gmZL'Vb.13. Neglecting rb·c.e Vce = 38 X 2 X 0·06 L 85° 42' mV = 4·5 L +94° mV Vce (9. where Ztnb' = _yl = 60 lnb' 299 L 85° 42' Q . the additional loading due to Cb'c may be neglected and the circuit may be treated in a similar manner to that for the triode input admittance calculation. This approach will be used for the second part of the question.HIGH FREQUENCY TRANSISTOR EQUIVALENT CIRCUITS The a.c. output voltage 1 = Vce 106 L 102 o = 9·4 L o 102 mV This calculation could ha ve been simplified if Yb•c had been neglected and with it the j94 in the first term ofthe right hand side of equation 9. the effect of rw must be included. In this case ZL' is given by rce in parallel with the gmZL' Ytnb' = 1·265 = 38 X 2 2 X RL of 2 kO. the input admittance at Vb'e is given by where ZL' is the effective collector Ioad.
for transistors intended for high frequency applications. 300 . In the published data. the four parameters are all comp1ex. app1ying equations 9. graphs showing how the y parameters vary with frequency and operating point. Neglecting Yb'c is making a similar approximation to that made by neglecting hre when using the h parameter equivalent circuit. conditions. and if rb'c is neglected calculation is extremely simple.14 and 9. In some cases. =1+ 1·66 L +85°42' = 1·25 + j1·66 = 2·075 L +48° 24' (9. arealso supplied.c. Use of y Parameter Equivalent Circuit A1though the hybrid 7T circuit discussed above may be used for calcu1ation up to the high frequency Iimit of any particular transistor.15 A V = 76L 180° 2·075 L +480 24 = 36 3 0 ·7 L + 1 1 36 I The input impedance may be found from V1n Zin = ~· 103 x 4·5 L +94° = 106 x 36·7 L +131° 36' Q = 123 L 37° 36' This examp1e shows how the hybrid 7T equivalent circuit may be used forahigh frequency calculation. most manufacturers do not quote the hybrid 7T parameters. At low and medium frequencies the effects ofthe capacitance become negligible. At these high frequencies.MODIFICATIONS TO EQUIVALENT CIRCUITS V1n Ybb' Vb·e Zlnb' =1+ = 1+ 100 60 L _ 85 o 42.15) Overall vo1tage gain. the y parameters are usually quoted for a specific frequency and d.
HIGH FREQUENCY TRANSISTOR EQUIVALENT CIRCUITS
and the resulting common emitter equivalent circuit is given in
Figure 9.6.
Amplifier calculations in terms of known y parameters may be
made by use of the general solution obtained in Chapter 2. If feedback
is included in the circuit, either feedback theory, or interconnected
fourterminal network theory, will provide the correct solution.
Two examples will now be considered; first a single stage amplifier
b
c
I~
ib
(,~
Yoe
=_!;be+jwCoe
'"1
Figure 9.6. y parameter equivalent circuit for transistor in the
common emitter configuration
with a resistive load, and secondly a multistage amplifier with feedback.
Example 9.4. An r.f. transistor is to be used as a single stage
amplifier with a resistive load of 2 kQ. At the signal frequency of
200 MHz and at the d.c. conditions applicable to the circuit, the
y parameters are specified as:
Yle(0·028
+ jl3w
X 1012) mho
Yre 500 X I0 6 L 250° mho
yre 0·034 L 292° mho
Yoe(220 X 106
+ j2w
X 1012)
mho
If the amplifier is driven by a source of short circuit current 10 flA
and internal admittance (0·01 + j0·005) mho, determine the output
voltage.
Rewriting the general solutions obtained in Chapter 2,
Yln
and
v2
h
y1zy21
= yu yzz +
YL
 yz1
= yu(y22 + YL) y21Y12
(9.16)
(9.17)
We can first find the input admittance, and hence the input
current. Then, using 9.17 we can find the output voltage directly.
301
MODIFICATIONS TO EQUIVALENT CIRCUITS
At 200MHz,
J!e = 0·028 + j200 X 106 X 27T X 13
= 0·028 + j0·016 mho
and Joe = 220 X 106 + j200 X 106 X 27T
= (220 + j2 500) x 106 mho
X
1012 mho
X
2
X
1012 mho
Now working in millimhos:
.
0·5 L 250° X 34 L 292°
Ytn = 28 + jl60·22 + j2·5 + 0·5
= 28
.
+ Jl6
17 L 182°
2·6 L 74o
= 28 + j16 6·5 cos 108° = 28 + jl6 + 2·02 j6·2
= 30 + j10 mmho
j6·5 sin 108°
Note that the term due to Yre is by no means negligible.
The short circuit source current of 10 tJA must divide between the
source admittance and Ytn·
_
10(30 + j10)
A
ltn (30 + jlO) + (10 + j5) fl
_ 10(3 + j)(4  j1·5) A
16 + 2·25
fl
= 0·548(12
+ 1·5 + j4 
j4·5) flA
= 7·4  j0·274 flA
But for the output voltage calculation, only the magnitude of hn
is required.
lltnl = y(7·42 + 0·2742) =::= 7·4 flA
Applying equation 9.17 and working in mA, mmho and volts.
7·4
V2
X
103
X
34 L 292°
= (28 + j16)(0·72 + j2·5) 34 L 292°
=
0·252 L 292°
x 0·5 L 250° V
20·2 40 + j11·5 + j70 17(cos 182°
0·252 L 292°
 19·8 + j81·5 + 17 + j0·6 V
252
IV21= y(36·82 + 81·52) mV
= 2·8mV
302
+ j sin 182°)
V
HIGH FREQUENCY TRANSISTOR EQUIVALENT CIRCUITS
If required, the voltage gain may be obtained from
!Vtnl
=
jltnl
I Ytnl
=
7·4 X I06
v(32 + 12) X 102 V
= 0·232 mV
2·8
!Avl = 0·232 = 12
For our final example weshall consider a two stage amplifier with
overall current feedback.
Example 9.5. An amplifier is required for signals at 450kHz.
The available transistor has a suitable d.c. operating point of
VcE 6 V, Ic 1 mA. The averagehre is quotedas 150, but to allow
for spread and temperature variations a d.c. stability factor of
K = 0·05 is to be used. The required current gain of 100, into a
resistive load of 1 kO is to be obtained using two stages with overall
feedback to improve stability. Using the published data quoted
below, design a suitable amplifier using a 10 V supply.
Ico 2 pA, VBE 300 mV,
hre 150
g1e 0·25 mmho, Cte 70 pF, IYrel 37 mA/V, ~re oo
goe 1 ,umho, Coe 4 pF, IYrel 4 ,umho, ~re 270°
The proposed circuit is shown in Figure 9.7.
RF 2
Figure 9.7. Circuit for Example 9.5
The first consideration must be to obtain the correct d.c. operating
conditions. This was discussed in Chapter 1, and the results obtained
will be used here.
303
MODIFICATIONS TO EQUIVALENT CIRCUITS
In each case the required operating point is VcE 6 V and Ic
1 mA. Consider first Tr2:
VcE = 6 = 10
Taking lE === Ic
RE2
+ /cRL2 + /E(RE2 +
Rpt)
+ RFl + RL2 = {kQ
RE2 + RFI = 3 kQ
(9.18)
The stability factor
+(X')+ RB= 0·05
K = RE(l
where
RaR4
RB= Ra+ R 4 and
.
Insertmg values 0·05
453
Also V'
where
But
and
=
+ RFI) = 3 kQ
(9.19)
3 +RB
= 3( 1 + 150) +RB
+ RB =
RB
and
RE= (RE2
=
60 + 20RB
397
19 = 20·9 kQ
IBRB IERE
+ VBE
V'
(9.20)
(using conventional polarities)
= =V_c_,c_R_4_
(9.21)
Ra+ R4
fc
f co'
 rJ.IIB=
(/.1
fco'
7
===Ico
lB =
1 000
.sü 
2 ttA
= 4·7 ttA
V' = 20·9 X 0·0047  3  0·3
= 3·4 V
(9.22)
Applying equations 9.19 and 9.21
10R4
Ra+ R4
(9.23)
RaR4
20' 9 =Ra+ R4
(9.24)
3·4 =
and
304
HIGH FREQUENCY TRANSISTOR EQUIVALENT CIRCUITS
From equation 9.23
(9.25)
Substituting in equation 9.24,
20 .9
=
RaR4
2·94R4
Ra= 61·5 kQ
Substitutingin equation 9.25
Ra
R4 = 1.94 = 31·7 kQ
In practice the nearest preferred va1ues would be used, i.e.
Ra
= 68 kQ
= 33 kQ
and R4
Ca must have a reactance that is negligible compared with RE2
at the signal frequency.
Let
Xe= 30 Q
= 2n450
1
X
1Q3C
From which a 0·01 J.tF capacitor will be found to be satisfactory.
A similar procedure may now be conducted for Trl. In this case
RL is not given, but since a high current gain is required RLI should
be large.
By comparison with equation 9.18
Let
RLl + RE1 = 4 kQ
RLl = 3 kQ and RE1 = 1 kQ
From stability considerations:
1 +RB
0"05 = 151 +RB
151
+ RB =
and
20
+ 20RB
RB= 6·9kQ
R1R2
RB= R1 + R2
where
(9.26)
Now applying equations 9.21 and 9.22
V'= 0·0047
where
X
6·9 1 0·3
V'= 10R2
R1 + R2
305
= 1·33 V
(9.27)
MODIFICATIONS TO EQUIVALENT CIRCUITS
From 9.27
R1
+ R2 =
R2
10
1.33 = 7·5R2
X
(9.28)
Substitutingin 9.26,
6.9 = R1R2
7·5R2
R1 = 51·7 kQ
and
So, from 9.28,
R2
=
51·7
6.5
= 8 kQ
In this case the preferred values would be R147 kQ and R 2 8·2 kQ.
By comparison with Ca, C1 should be 0·025 JlF.
The first step in determining the current gain is to decide upon a
suitable value for RF1 so that the input impedance to Tr2 with local
voltage feedback may be found. Since RE2 + RF1 must be 3 kQ.
suitable preferred values would be RE2 2·7 kQ and RF1 330 Q.
This will give a ßv of 330/1 000 === t. Next the app1icable y parameters must be found.
Yte
= 0·25 + j27T450 X 103
= 0·25 + j0·05 mmho
X
X
lQ12
X 1Q3
mmho
4
X
1012
X 1Q3
mmho
=
37 L Ommho
= 0·004 L 270° mmho
Yoe = 0·001 + j27T450 X 103
=== 0·001 + j0·001 mmho
yre
Yre
70
X
From the general solutions in terms of the y parameters,
Ytn2
.
37
X
0·004 L 270°
= 0·25 + j0·05 0.001 + j0·00 1 + 1 mmho
Neglecting the yoe term,
Ytn = 0·25
= 0·25
= 0·25
+ j0·05 0·15 L 270° mmho
+ j0·05 0 + j0·15 mmho
+ j0·2 mmho
(9.29)
To find Av, consider the original so1utions, in terms of the y
parameters.
h(yoe + YL)  0
(9.30)
v1 =
ß.
and
v2 =
0 hyre
ß.
306
(9.31)
HIGH FREQUENCY TRANSISTOR EQUIVALENT CfRCUITS dividing 9. may be obtained from equation 9.JreJte Substituting values and neglecting Yoe At 2 = (0·25 + j0·05)1  37 L oo 37 L 0° X 0·004 L 270° Comparison with the input admittance calculation shows that this denominator has the same value as result 9.32) YL) .30 A vSubstituting values v2. YL1.31. Ytnl and Ytn2 must be included. 37 L 180° L 38° 42' 37 A12 and A12 = 0·25 + j0·2 = 0·32 = 115 L 141° 18' (9. YLI err. Multiplying by YL and dividing by /: Ar= !j . current splitting factors due to bias components. which is unaffected by the local feedback.Yoe + YL 37 L oo = 1 + 0·001 + j0·001 =  Av (9. but this will only give the current gain in terms of Trl collector and base currents. + YLI + Yna + Yn4 = 0·019 + j0·015 + 0·333 + 0·015 + 0·03 mmho = 0·5 + j0·015 mmho YLI eff = Yrn2 307 .32. · First we must find the effective load on Trl.33) To calculate Au we can again use result 9.31 by 9.32) 37 Now allowing for the effect of local feedback with aß oft + + Ytn 0·25 j0·2 Ytn2 = Ytnf = 1 _ ßA = 1 12 .3 mmho = 0·019 + j0·015 mmho (9.yre vl .YteYL = V2YL = h h Jte(Joe + ( 9.29. To find the overall current gain.33) The current gain.
At = At' X An X At" X A12 = 0·82 L 12° 18' X 85 L 126° 26' X 0·0464 L. 0·15 L 270° = 0·25 + J0·05 .33. Ytnl = 0·26 + j0·35 = 0·435 L 53° 30' .5 L 1o 42 .36.35) To determine the input current splitting factor we require Ytnl which may be obtained from the genera1 so1utions. 9.34.MODIFICATIONS TO EQUIV ALENT CIRCUITS From equation 9. = 0·25 + j0·05 + 0·009 + j0·3 :. 37 X 0·004 L 270° Ytnl = 0 "25 + J0·0 5 . 36° 36' X 115 L 141° 18' = 374 L 316° 18' (9.35 and 9. = 0·0 64 L 36 36 0 I (9.0 .36) I The overall current·gain may now be obtained from equations 9.34) The interstage current splitting factor = = 0·019 + j0·015 ==:7::::0·5 + j0·015 " 0·0232 L 78° 18' 4 At = 0 .37) 308 . .' At' = Ytnl = _ 0·435 L 53° 30 YBl + YB2 + Ytnl 0·021 + 0·122 + 0·26 + j0·35 0·435 L 53° 30' 0·435 L 53° 30' 0 = 0·4+j0·35 = 0·53L41°12' =O·S 2 L 12 18 (9.32 neg1ecting Yoe. 37(0·5 + j0·015) An= (0·25 + j0·05)(0·5 + j0·15) + j0·15 37(0·5 + j0·015) 37 L 180° X 0·5 L 1° 42' An= 0·125 + j0·18 = 0·218 L 55° 18' Au = 85 L 126° 26' (9.5 L 1o 42 . 9.0·001 + jO·OOl + 0·5 + j0·015 .
374 L 316° 18' 374ß L 316° 18' l I + 374ß L 316° 18'! = 3·74 !I + 270ß. and 100 = I+ ß will be real.7 provides simple negative feedback if ßA has zero phase angle.j258ßl = 3·74 (1 + 270ß) 2 + (258ß)2 = 3·742 X 104ß 2 + 6·7 X 104ß2 = 14 !I l + 540ß + 7·3 Rearranging and 14 X 10 4ß 2 + 540ß.HIGH FREQUENCY TRANSISTOR EQUIV ALENT CIRCUITS Finally we come to the design of the overall feedback circuit: the connection shown in Figure 9. But ß= GF2 GFl + GF2 an d 0·79 X I02(GF2 GF2 RF2 GFI = 3mmh o + 3) = GF2 0·79 X 1()2 X 3 = l _ 0 .38 is obviously correct.13 ß= 540 ± 540 ± = 0 yi(5402 + 52 X 14 X 104) 28 X 104 + V[(29 728) X 104] 28 X 104 540 ± 27·5 X 102 28 X 104  Taking the positive sign. 309 . At Au= 1 + ßAt In this case the required gain is 100.Q Since this is a preferred value no modification to this result is necessary.79 X 10_2 mmho = 2· 37 x 102 mmho = 43 k.38) So remembering that Ar::!:: 1/ß result 9.8 X 1()2 = 0·79 X 102 (9. ß= 2·21 2 .
Figure 9.F capacitor wou1d be satisfactory. They are not applicable to circuits involving large a.8.1 for an angular frequency of 107 radfsec. The completed circuit design is shown in Figure 9. EXAMPLES Example 9. The reader should appreciate that in practice many approximations could be 10 V ::f'"~ 1k. Should other devices such as field effect transistors come into general use the methods will be equally useful provided suitable equivalent circuit parameters are known.5 made in the above calculations.6. 6·86 L 27° kQ. Ans. 0·01 uF 47 k.c.. 587 L 277° 42'.O.O. Full details have been given to demonstrate the various principles that have been discussed in this book. 310 . Repeat Example 9. the equivalent circuit methods discussed in these chapters provide a very useful tool for the solution of a wide range of electronic circuits.8.MODIFICATIONS TO EQUIVALENT CIRCUITS The remaining components to be selected are C2 and C4 which must have negligible reactances compared with the terminal input impedance to Tr2 and to RF2 respectively. signals or switching circuits. In general. In each case a 0·01 p. Circuit for the solution to Example 9.
Example 9. Calculate the tuning capacitor required to tune it to 80 MHz and the voltage gain if the transistor y parameters at this frequency are Yle (0·4 + j20co X 1012) mho. 2 pF.9. Av 155 L 164° 30'. Z1n 102 L 8° n.f. Example 9.3 is used as a wide band amplifier over the frequency range 100 kHz to 2 MHz. Yre 350 X 106 L 300° mho. 420 pF. 311 . 4·5. 37·5 L 70° 48'.uH and Q factor 25. Ans. Z1n 390 L 49° n. If the collector load is purely resistive of value 5 kil.EXAMPLES Example 9. An r. transistor is loaded with a parallel tuned circuit. calculate the voltage gain and input impedance at these frequency extremes. having a coil of inductance 2 . and use valid approximations at 2 MHz. Yoe(150 x 1()6 + jco x 1()12) mho. Ans.8. Av 34 L 100°.7. If Cag is 3 pF and gm 8 mA/V determine the value of tuning capacitor and the Q factor for an anode coil of inductance 0·05 mH. At the required frequency of 1 MHz the crystal requires an additional 30 pF to produce resonance and the loss component is equivalent to 10 kQ in parallel with the terminals. yre0·02 L 320° mho. A transistor having the hybrid 7T parameters given in Example 9. Ans. A valve oscillator employs Miller feedback and a crystal resonator in the grid circuit. Use the exact method at 100 kHz.
The Bel is an inconveniently !arge unit and the Decibel (db) is roore convenient. Pz will be greater than P1 and the result will be a positive nurober. the corresponding logarithmic units roay be siroply added or subtracted. This roay be defined as: Nurober of Decibels = 10 Iog1o ~: (A 1. It is convenient to express such a range of Ievels in logarithroic units.3 will involve the deterroination of a negative logarithro. If P1 is the input power. Also. voltage or current ratios are coromonly used and under 312 . In both cases the units are based upon power ratios. each roultiplying or dividing the signal by a factor.3) lf the circuit in question is an aroplifier. where a nurober of circuits are cascaded.4) In practice. one using log to the base 10 and the other using log to the base e. lf it is an attenuator.2) The Neper is mainly used in transroission line probleros and will not be discussed further in this book. and P2 the output power. It is simpler under these circurostances to write: Number of db of attenuation = 10 log10 ~: (A1.APPENDIX 1 USE OF LOGARITHMIC UNITS In roany electronic systeros. P1 will be greater than Pz and expression Al. then by definition: p2 Nurober of Nepers = Ioge p 1 {ALl) and p2 Nurober of Bels = log10 p 1 {Al. Two sets of logarithroic units are in cororoon use. the signallevel roay vary froro kilowatts to microwatts in different locations.
5 should only be used with equalload and input resistance. 313 . even where the condition is not maintained. are referred to as the 3 db frequencies and are quoted simply as they are most convenient to calculate. In practice.APPENDIX 1 certain conditions the db scale can be applied.3 may be rewritten as: Number of db = h 2R 10 log10 h 2R = 10 log1o = 1o1ogw (*Y = 101og10 /z = 20 log10 h = 20 log10 /R V22 v12/R (~:r Vz h (A1. wh and WL. the frequency response of amplifiers is frequently quoted in db units. Avm Avh = 1 + j IAvl Number of db = ~.5) Strictly the definition in equation Al. Avm = 20 log10 Avmv2 = 20 logwv2 = 10 log102 = 3·010 db Thus these frequencies. Thus at Wh and WL for the RC coupled amplifiers discussed in Chapters 4 and 5. expression Al. If the input resistance is equal to the Ioad resistance.
. D. B. L. 1050. Pitman Technical Books. 1965. London. J. Sevin. Hakim. Le Croissette. Inc. 1966. 315 . 1966. J. Feedback Circuit Analysis. Burford and H.. Semiconductor Junctions and Devices. Edward Arnold Technical Books. De Witt and A. W. Ryder. S.. Transistors. McGrawHill Book Co. D.. 1957. 1962. PrenticeHall. New York. London. McGrawHill Book Co. Rossoff. New Jersey. Parker. Transistor Electronics. Transistors: Theory and Circuitry. G. Electronics. L. Field Effect Transistors. K. London. J. New York. lliffe Books Ltd. Dean.. New York. 3rd edn... New York. D. 1964. McGrawHill Book Co.WORKS FOR FURTHER READING Electronic Fundamentals and Applications. McGrawHill Book Co. P. Verner. Englewood Cliffs. 1964. S.
17 Emitter follower. 166 single stage. 94 Dynamic input characteristics. 25. current feedback. 223 three stage. 222 single stage. 5. bias line. 11. 12 Amplification. 303 Biasing transistor. 94 for high frequency. 1 Ioad lines. 27. 140. 11. 162 D. 19 operating point. 181 RC coupled. 134. 220 frequency response. 312 Determinants. 2. 114 single stage. 10. 118 long tailed pair. 158. 121 Collector feedback. 14. 19 Decibel. 291 small signal. 158 emitter coupled. 122. 6. 21. current feedback. 166 common emitter. 134. 153. 103 diode.INDEX A. transistor. 291 Feedback composite. 173. current feedback. 122. 194 current derived. transistor approximate solutions. 181. 247 current amplifiers. 14. 15. 296 valve. 195. 146 two stage. 241 graphical analysis. 231 three stage. Ioad lines. 194. 99. 24. transistor. 227. 3. 19. 202. 306 definition. 152 common collector. 3. 21. 221 single stage. 96. 159 feedback. 2. 131. 116 modified. 210. 5. 32. 21. series applied. 235 Anode characteristics. 303 two stage. shunt applied. 124 Bias stabilization.C. 126. 30 Complex feedback factors. 100 transistor. 24. 30 triode. 18. valve. 152 difference. 240 transistor multistage. 245 Cut off. 153. 25. 212 grounded grid. 189 current derived. 156 Equivalent circuits determination by network theorems. 158 multistage. 14. voltage feedback. 8 characteristics. 4. 126. 55 Diode characteristics. 101. 94 Thevenin and Norton. 33 Cathode follower. 15. 10. application to network analysis.C. 31. 34 Darlington connection. emitter resistor stabilization. 7 Bottoming. voltage feedback. 141. voltage feedback. 182 317 . 170. nonlinear. current feedback. 34 reactive Ioads. 127. 235 properties. 151 cathode coupled. 212 Bias decoupling. 229 Amplifier. 122 common base.
emitter resistor bias. transistor. 27. 190. 66. 106 Nyquist's criterion of stability. 231. 277 voltage Wien bridge. 58. 273 tuned collector. 268 tuned anode. 203 voltage amplifiers. 197 Feedback amplifiers input circuits. 241 effect on gain. 262 Hartley and Colpitts. 190 gain stability. 75. 101 general solutions. 68 interconnections. 184 Stability factors. 185. 202. 159 common collector. 210 effect on frequency response. 198 effect on output impedance. 45 Oscillators. 295 current Wien bridge. 42 applications. 28. 81. 196. 184. 198 effect on input impedance. 186. 184. 157. 187 voltage derived. 273 Miller. 188. 208 Four terminal networks. 50.163 238 284 Noise. 190. 143. 77 Hybrid 1r equivalent circuit. 212 ' ' Norton's theorem. 42 notation. calculation. 159. 297 Input resistance.) effect on distortion. 255 practical amplifier circuits. transistor biasing. 220 summary of effects on impedances. 195. 256 Output impedance. 248 effect on output admittance. 303 Simple negative feedback. 87 h parameters.INDEX Feedback(Contd. 261 Negative input conductance. 119. 255 single stage RC. 74. 150. 198 effect on input admittance. 255 amplitude stability. 293 Network analysis fundamentals. 295 Negative feedback for amplitude stabilization. 202. 83. 86 common base. 159 common emitter. 210 248 ' Ohm's Iaw. 152. 241 voltage derived. 185 negative. 259 crystal. 195. 248 input connections. 187. 141. 145 Maxwell's circulating currents 54 Mesh analysis. shunt applied. 282 LC. 291 Kirchhoff's laws. 101 equivalent circuits. 43 Nodalanalysis. 220 instability. 99. 60 318 . 155 Potential divider. 187. series applied. 80. 268 RC. 184 positive. 303 Superposition theorem. 245 g parameters. 240 Frequency compensation. 295 phase shift. 201 vector diagrams. 18 Interelectrode capacitances. 52 ' Miller feedback.
general method. 68. 85. 81. 14 input Ioad line. 140 fundamental current relationships. 79. 103 disadvantages of simple biasing. 24 Transistor circuits graphical analysis. 99. 14. 36. 68 Voltage amplification. 24 simple bias circuit. 84 z parameters. 174 Thevenin's theorem. 21 leakage currents. 300 z parameters. 5 Valve. 303 temperature effects. 15 Two port networks. input admittance at high frequencies. 63. 105 Transistor black box representation. 71 319 .INDEX T equivalent circuits. generat solutions. 104. 300 y parameter equivalent circuit. 29. 27. 3. 102 characteristics. 23 equivalent circuit. 15 output Ioad line. 112 y parameters. graphical analysis. general method. 79. 232. 14 Transistor biasing design considerations. 292 Valve equivalent circuit. 21.