Ix Transistor Circuits

IX.
TRANSISTOR CIRCUITS
Prof. H. J. Zimmermann
Prof. R. B. Adler
Prof. S. J. Mason
A.
VOLTAGE-REGULATED
J. Blair
J. B. Cruz
J. Gross
C. R. Hurtig
POWER SUPPLIES
Design studies of transistorized voltage-regulated

circuits were continued.
C. T. Kerk
R. B. Martindale
R. F. Meyer
power supplies for transistor
A circuit which requires no external voltage supplies was
developed and is shown in Fig. IX-1.
The regulator employs a pnp power transistor
(type 2N57) in series with the rectifier and the load.

fier is used in the feedback loop.
A two-stage, direct-coupled ampli-
The reference voltage is obtained from the reverse-
biased breakdown characteristic of silicon-alloy junction diodes.
In the supply shown,

two silicon diodes (D1 and DZ) are used in series to produce a reference voltage of 15. 6
volts.
The output voltage may be varied by means of the potentiometer
from 20-25 volts
with only small changes in the regulation characteristics of the supply.

current bias for the power transistor is
The quiescent
supplied by the silicon transistor used in
the output stage of the direct-coupled amplifier.
The bias for the silicon transistor
is obtained by use of the breakdown voltage of silicon diode D3 and the 5. 6K resistor
connected to the base lead of the silicon transistor.
The capacitor from the collector
to the emitter of the 1859 transistor suppresses high-frequency oscillations.
The fre-
quency response of the voltage regulator is essentially flat up to 10, 000 cps.
The voltage-regulation characteristic
Table IX-1.
as a function of load current is
given in
The variation in the output voltage as a function of line voltage variation
is negligible.
At a load current of 100 ma the power transistor dissipation is 0. 83 watt

and the efficiency of the regulator (excluding rectifier losses) is 66 per cent. At 140 ma
the dissipation is 0. 43 watt and the efficiency is 84 per cent.
This high efficiency is
the result of the ability of the power transistor to operate close to zero collector voltage.
The lower limit upon the input voltage to the regulator (point A) is determined by the
desired output voltage, the peak voltage of the ac ripple, which increases with load
current, and the static characteristics of the transistor at maximum current.
IN9
..
43
.
43
9K
li5V
60-'
LOAD
Fig. IX-1.
Voltage-regulated power supply
(IX.
TRANSISTOR CIRCUITS)
Voltage Regulation Characteristics
Table IX-1.
Load Current
(ma)
Direct Voltage
(v)
Ripple Voltage (v)

(peak-to-peak)
B
(change from nominal
value of 20 v)
40
0.30
<0.001
50
34
-0.02
0.55
<0.001
100
28
-0.05
0.70
<0.001
140
23
-0.08
0.80
0.001
Voltage regulators of this same general configuration having nominal output voltages
from 3-25 volts were designed and tested.
The lower limit of the output voltage is
determined by the availability of silicon diodes with low values of reverse breakdown
The upper limit of the output voltage is determined by the maximum voltage
voltage.
ratings of the transistors.
In the circuit shown the maximum load current is
limited
by the maximum value of the reverse current of the diode D 3 , and the gain of the series
transistor.
The current in the diode is maximum at no-load and its value is determined
by the desired change in load current and the large signal gain of the series transistor
over this range.
Since the breakdown voltage of the reference diodes has some temperature dependence (the ones used varied by about 0. 08 per cent/C), the output voltage will have at
least an equal temperature dependence.
pensation by means of thermistors,
employed.
To reduce this temperature dependence, com-
or by auxiliary forward-biased diodes,
may be
Alternatively of course, the diodes may be replaced by batteries having low-
temperature coefficients.
C. T. Kerk
B.
A MICROSCOPIC INTERPRETATION
OF POWER GAIN IN JUNCTION
TRANSISTORS
In amplifiers of the vacuum-tube variety, it is usually possible to understand the
process of power gain by tracing particles and discussing their energy changes as they
move.
While this procedure is not usually the best way to achieve a quantitative des-
cription of the device, it often serves as the root of our qualitative understanding of the
gain mechanism.
In some cases, it may indeed have been the keystone of the invention
of the original device.
Peculiarly enough, however, this viewpoint has certainly never
been stressed for the transistor, and to the best of our knowledge it has not even been
discussed.
The following compressed outline presents the major features of an attempt
(IX.
to treat the junction transistor through a discussion of the particle energies involved (1).
We recognize first that "class A" power gain shows up inside a resistive amplifying
device as a change in the internal power dissipation.
tion process is therefore required.
A microscopic view of the dissipa-
We achieve this view by writing an energy balance
relation for clouds of holes and free electrons in a semiconductor which is slightly out
of thermal equilibrium.
In notation which is now standard (2), statistical mechanics leads to the concept of
quasi-Fermi levels for holes and free electrons out of equilibrium
=
+ (kT/q) In (p/n.)
(1)
- (kT/q) In (n/n
S=
i)
The hole and free electron current densities are then given by
P = -q Lp p V p
(2)
In = -q
n V n
leading to the notion of generalized nonequilibrium forces F
p, n
Fp = -q V4p = -q Vi - k V[T In (p/pi)]

(3)
Fn = +q Vqn =
+q
VY - k V[T In (n/ni)]
which are conservative.
The potential energies of the carriers at a point s in the non-
equilibrium force field are defined in volts as
I (s) =
q
p
d9 =
(s) -
pmin
min
(4)
s
I1n (s) = q
Fn
di = -[(s)n
- nn min
in]
min
where the zero point is taken at the position where
of particle.
4 is minimum for each type
The organized drift velocities are
-p
Vdp
qp
p
(5)
Vdn
-I
qn
n Vn
(IX.
The assumption of small departures from equilibrium implies that organized transport
kinetic energy terms 1/2 m vd are negligible. Therefore the work done by nonequilibrium forces goes into the lattice as "conductive dissipation"
-F
cp
dp
Pq
pp
pl
(6)
In
Pnp
Vdn
Fn
qn7
n
2
inn
If we consider the hole and free electron clouds as open systems, the energy of configuration plus thermal energy of each particle can be shown to be
E
w
g
2q
/2 kT
q
(7)
w
where E
= 3/2
2q
is the energy gap, and hole and free electron masses are taken to be equal.
Thus the entire energy per particle is

E
g + 11
Zq
p
Eg
kT
pp = 3/2 q
kT +
n 3/
=3/2
--
+ 1
(8)
We now note that the energy of a unit volume of a cloud may change by four processes:
1.
recombination of particles, 3.
convection of particles, 2.
with the lattice (conductive dissipation), and 4.
collision of particles
change of 0 with time at constant par-
The expression of the total energy balance of the clouds per unit
ticle concentration.
volume then becomes

q at (p
p)
-V.
p)-
qr(np-
n p)p
+ I
Vp
+ qp
p
0)n
n) - qr(np - n
(In
q a (n
Vn
n + qn
(9)
at
at
If Eqs. 9 are expanded fully, they become nothing but the continuity equations for the
flow of holes and free electrons.
--
p at
p p
Expanding only the left side, however,
qr(
qr(np-
np
yields
PVp
0
(10)
=v. (
n) - [qr(np - noPo)
+ n
V n]
(IX.
TRANSISTOR
CIRCUITS)
The term in brackets, as we have seen, is the total dissipation per unit volume, and
the divergence term represents convection of energy into the volume. Thus the left
sides are interpreted as the rate of increase of stored (recoverable) energy per unit
volume.
If we treat the temperature T and the gap E
nonequilibrium potentials HI,
on, we may write Eqs.
p
n
=
=
' +I
' +
as variables independent of the

g
since the latter can be altered by light, voltage, and so
8 as
(11)
1n
Because the energy balance relations (Eq. 10) are linear in
separated into two sets, one in
p and
' and one in II, for each particle type.
they can then be
The equations
involving IH have been checked for dissipation in an idealized transistor under quiescent
conditions, and as a low-frequency amplifier. In both cases the dissipation checks with
computations from the equivalent circuit.

More work remains to be done on the fully time-varying case,
and in connection
with the equations involving only }'.

R. B. Adler for J.
Blair
References
1.
See J. Blair, Theory of power gain in junction transistors, S. M. Thesis, Department of Electrical Engineering, M. I. T., Sept. 1, 1955, for further details.
2.
W. Shockley, Electrons and Holes in Semiconductors (D. Van Nostrand Company,

Inc., New York, 1950).

Ix Transistor Circuits

Uploaded by

Copyright:

Available Formats

Ix Transistor Circuits

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ix Transistor Circuits

Uploaded by

Copyright:

Available Formats

IX.

Design studies of transistorized voltage-regulated

power supplies for transistor

A circuit which requires no external voltage supplies was

developed and is shown in Fig. IX-1.

The regulator employs a pnp power transistor

(type 2N57) in series with the rectifier and the load.

A two-stage, direct-coupled ampli-

The reference voltage is obtained from the reverse-

biased breakdown characteristic of silicon-alloy junction diodes.

In the supply shown,

from 20-25 volts

with only small changes in the regulation characteristics of the supply.

supplied by the silicon transistor used in

the output stage of the direct-coupled amplifier.

The bias for the silicon transistor

as a function of load current is

The variation in the output voltage as a function of line voltage variation

At a load current of 100 ma the power transistor dissipation is 0. 83 watt

Voltage-regulated power supply

Voltage Regulation Characteristics

Ripple Voltage (v)

The lower limit of the output voltage is

ratings of the transistors.

In the circuit shown the maximum load current is

To reduce this temperature dependence, com-

or by auxiliary forward-biased diodes,

Alternatively of course, the diodes may be replaced by batteries having low-

OF POWER GAIN IN JUNCTION

of the original device.

Peculiarly enough, however, this viewpoint has certainly never

The following compressed outline presents the major features of an attempt

A microscopic view of the dissipa-

We achieve this view by writing an energy balance

leading to the notion of generalized nonequilibrium forces F

Fp = -q V4p = -q Vi - k V[T In (p/pi)]

which are conservative.

The potential energies of the carriers at a point s in the non-

equilibrium force field are defined in volts as

4 is minimum for each type

The organized drift velocities are

Thus the entire energy per particle is

with the lattice (conductive dissipation), and 4.

change of 0 with time at constant par-

volume then becomes

Expanding only the left side, however,

as variables independent of the

Because the energy balance relations (Eq. 10) are linear in

separated into two sets, one in

' and one in II, for each particle type.

they can then be

computations from the equivalent circuit.

with the equations involving only }'.

W. Shockley, Electrons and Holes in Semiconductors (D. Van Nostrand Company,

You might also like