Study Unit

Oscillators
By

Joseph A. Risse, P.E.

and

J. A. Sam Wilson, M.S., CET

The oscillator plays a wide-ranging role in contemporary elec-
tronics. Depending on basic circuit arrangements, it can
handle demands for low-frequency, intermediate-frequency, or
high-frequency signals. Also, depending on its configuration,
it can produce one of a variety of output waveforms. This flex-

Preview
ibility makes the oscillator very useful in a host of products,
ranging from radios and television sets to computers.

When you complete this study unit, you’ll be

able to
• Explain the principal differences between several types
of oscillator circuits
• Describe the flywheel effect and how it’s produced
• Calculate the resonant frequency of a basic oscillator
circuit
• Explain the operation of complex RLC tuned circuits
• Explain the operation of oscillators that have LC feed-
back circuits
• Explain the operation of oscillators that have RC feed-
back circuits
• Discuss the basic applications of oscillator circuits
• Describe how a frequency synthesizer works

Remember to regularly check “My Courses” on your student homepage.

Your instructor may post additional resources that you can access to
enhance your learning experience.

iii
 Contents
WHAT IS AN OSCILLATOR? 1

OSCILLATORS WITH LC FEEDBACK CIRCUITS 2

The Operation of an LC Oscillator
Starting the Oscillating Current in the Circuit
A Mechanical Analogy
The Flywheel Effect
Calculating the Resonant Frequency in an LC Circuit
Series-Tuned and Parallel-Tuned LC Circuits
Complex Parallel-Tuned LCR Circuits

LC OSCILLATOR APPLICATIONS 21
Armstrong Oscillators
Hartley Oscillators
Colpitts Oscillators
Crystal-Controlled Oscillators
TCTB Oscillators
Frequency Multipliers

OSCILLATORS WITH RC FEEDBACK CIRCUITS 38

RC Phase-Shift Oscillators
Wien Bridge Oscillators
Twin-T Oscillators
Unijunction Transistor Oscillators
Frequency Synthesizers
Phase-Locked Loops

POWER CHECK ANSWERS 59

v
Oscillators

WHAT IS AN OSCILLATOR?
In simple terms, an oscillator is a circuit that produces a
repetitive waveform as its output, with only a DC supply volt-
age as the input. Thus, an oscillator converts DC electrical
energy to AC electrical energy. A basic oscillator circuit con-
tains an amplifier (usually either a transistor or an op-amp)
and a positive feedback circuit that produces phase shift. The
feedback circuit may be either an RC circuit or an LC circuit.
An oscillator circuit’s output voltage can be either sinusoidal
(that is, in the form of a sine wave) or nonsinusoidal in shape
(such as a pulse, square wave, or sawtooth wave). The design
of a particular circuit and its components combine to deter-
mine the shape of the output wave. A typical sine wave in the
time domain is shown in Figure 1A. This waveform is some-
times called a pure sine wave, and it has no harmonic
content. The same waveform is shown in the frequency
domain in Figure 1B.

FIGURE 1—(A) shows a

representation of a pure
sine wave in the time
domain, and (B) shows a
pure sine wave in the
frequency domain. The
frequency domain shows
the edge view of the sine t f
wave.
(A) (B)

Oscillator circuits are used to produce the many types of

signals that are required for the operation of electronic equip-
ment. For example, oscillator circuits are used in the tuning

1
 stages of radio and television receivers. In transmitters, oscil-
lators generate signals that can travel over many miles to
various receivers. Some tape recorders use oscillator circuits
to erase tapes, and some electronic musical instruments use
them to generate musical tones. Finally, oscillators are used
widely in radar applications, in computers, and in many
other electronic devices.
In this study unit, we’ll first take a look at LC oscillators—
that is, oscillators that use LC feedback circuits. These
circuits contain inductors and capacitors. Then, we’ll look at
RC oscillators—devices that use resistors and capacitors.

OSCILLATORS WITH LC
FEEDBACK CIRCUITS
An LC oscillator has four essential elements: a resonant circuit,
an amplifier, a feedback system, and a power supply. The reso-
nant circuit generates the oscillations (that is, the input signal)
and determines the signal’s frequency. The amplifier increases
the input signal and delivers to the resonant circuit the exact
amount of power needed to sustain continuous oscillation.
The feedback system takes a part of the output signal from
the amplifier and returns it to the input (the resonant circuit).
The power supply is the actual source of the power that’s
used to replace the power that’s lost in the resonant circuit
during each cycle of oscillation. The power is lost in the form
of heat. This heat is produced when current flows through the
resistance of wires, and also through leakage resistance in the
capacitor dielectric. Figure 2 illustrates this cycle.
The feedback system ensures that the returned signal is in
phase with—and therefore, able to reinforce—the input signal
in the resonant circuit. This is called positive feedback or
regenerative feedback, and it’s provided by a special type of
coupling network within the feedback system. In a conven-
tional amplifier, the output signal is 180 degrees out of phase
with the input signal. That’s why two amplifiers are needed
in the circuit shown in Figure 2. Each amplifier shifts the
signal 180 degrees, for a total of 360 degrees. A phase shift
of 360 degrees is the same as 0 degrees, so the feedback sig-
nal is in phase with the input signal.

2 Oscillators
 REGENERATIVE FEEDBACK PATH

FEEDBACK TUNED
SIGNAL CIRCUIT

AMPLIFIED
INPUT SIGNAL
SIGNAL

TWO
AMPLIFIERS

STARTUP FEEDBACK
BEFORE OSCILLATION

POWER
SUPPLY

FIGURE 2—The loop shown here—from resonant circuit to amplifiers to

feedback and back to resonant circuit—is repeated over and over as the
oscillator performs its job. The output of the two series-connected ampli-
fiers is in phase with the input signal. The feedback signal is in phase
with the input signal, so it reinforces the input signal. The feedback
signal replaces the power lost in the system. At startup, a few cycles of
feedback signal starts the system into oscillation.

The Operation of an LC Oscillator

Now, let’s take a closer look at how an LC oscillator works.
The best place to start is with the resonant circuit. This cir-
cuit, shown in Figure 3, consists of a capacitor and an
inductor arranged in parallel. Its function is to determine
which frequency in a group will be selected for amplification
in preference to the others. The resonant circuit also deter-
mines how this selected frequency is regulated.
When the circuit shown in Figure 3 is started, a back-and-
forth current flows between the inductor and the capacitor.
This back-and-forth current flow is oscillation. In our example,
the LC circuit oscillates at the sinewave frequency. In fact, it’s
the only part of the oscillator circuit where oscillation occurs.

Oscillators 3
 FIGURE 3—In this reso-
nant circuit, a capacitor
and an inductor are
arranged in parallel. The
+
circuit constants, C1 and
L1, determine the fre-
quency of oscillation in L1 C1
the circuit.

Let’s assume that an external power source is connected

across the resonant circuit. In one moment, the capacitor C1
is charged and energy is stored in its dielectric in the form of
an electric field. Then, when the external power source is
removed, the capacitor begins to discharge (Figure 4).
Current then flows through inductor L1 in the direction indi-
cated by the arrows in the figure, developing a magnetic field
around the turns of L1. The current stops flowing when the
charges on the plates of C1 equalize. Then, the magnetic field
around L1 starts to collapse and induces a voltage across the
inductor, thus sustaining the current in the circuit. As
shown in Figure 5, L1 becomes a voltage source, taking on
the polarity indicated. The collapsing magnetic lines cut

FIGURE 4—In this illustra-

tion, capacitor C1, is
discharging.

+
+
L1 DISCHARGE C1
CURRENT

FIGURE 5—In this illustra-

tion, the collapsing
magnetic field produces
current, and L1 becomes
a voltage source in the
circuit.
L1 COIL C1
CURRENT
+ +

4 Oscillators
across the inductor turns to produce a current that builds an
excess of electrons on the upper plate of C1. (Note that C1 is
now charged in the opposite direction.)
After the initial surge of current (as C1 discharges in the
opposite direction), a magnetic field develops to maximum
level across L1. This is shown by the arrows in Figure 6. As
discharge current starts to decrease, the magnetic field
begins to collapse. This develops an inductor current in the
same direction as the capacitor current. When the capacitor
has discharged, the collapse of the magnetic field continues—
maintaining current and at the same time charging the
opposite plate of the capacitor. This alternate charging and
discharging of C1 will continue for several cycles. When cur-
rent flows back and forth in the circuit as a result of these
voltages, it’s sometimes called the flywheel current.
FIGURE 6—After the
charges on the plates of
C1 equalize, the capacitor
discharges through the
inductor.
L1 DISCHARGE C1
CURRENT
+ +

In a way, this circuit operates by exchanging energy back and

forth between the inductor and capacitor. In each case, the
stored energy is converted to electrical energy during the fol-
lowing cycle. During each cycle, some of the electrical energy is
converted to heat energy as current flows through the resist-
ance in the circuit. This heat is dissipated in the surrounding
area. The heat energy can’t be recovered in the system, so the
total energy in the circuit decreases with each cycle.
Eventually, all the electrical energy is converted to heat
energy and the back-and-forth current stops.
Keep in mind the fact that energy is never completely “lost.”
Energy can’t be created or destroyed—it can only be changed
from one form to another. Thus, the electrical energy in this
resonant circuit isn’t lost—it’s simply converted to heat energy.
The rate at which the electrical energy is converted to heat
energy is a measure of the amount of power lost in the circuit.

Oscillators 5
 If you could somehow gather all of the dissipated heat energy
and convert it back to electric energy, the back-and-forth cur-
rent would continue forever. However, the power that’s lost in
the form of heat can never be recovered. Therefore, it must be
replaced by power from the power supply. The oscillator cir-
cuit replaces the heat energy that’s dissipated during each
cycle by applying a small amount of power to the tuned circuit
once each half-cycle. The replacement power that’s needed to
maintain the oscillations comes from the power supply.
Note that the amplifiers in an oscillator circuit are transis-
tors. However, you can think of them as acting somewhat like
variable resistors. When the input sine wave is on the posi-
tive half-cycle, the maximum current flows through the
amplifiers. In other words, the amplifiers have less resistance
and offer less opposition to the flow of current at that time.
When the input sine wave is on the negative half-cycle, less
current flows through the amplifiers. The amplifiers have
more resistance and offer more opposition to current flow at
this time. In this way of thinking of the transistor amplifier
action, the input signal controls the amplifier’s resistance to
the flow of current from the power supply.
This is important because it helps you to understand that the
amplifier in an oscillator circuit controls the flow of current
from the power supply. This current from the power supply is
used to replace the energy lost by the resistance of the reso-
nant circuit. Usually, when we talk about bipolar transistors,
JFETs, and MOSFETs, we talk about them being used as am-
plifiers. This is because the output signal strength is usually
higher in amplitude than the input signal. However, as we’ve
shown here, these devices can also be used to control the
amount of power that’s delivered from a power supply to a
load resistance or resonant circuit.
In an LC oscillator circuit, the frequency of the oscillation is
determined almost completely by the inductance of L and the
capacitance of C. (In making that statement, we’re disregard-
ing the effects of distributed capacitance and inductance, the
internal capacitance of a transistor, and the effect of coupling
capacitors.) The transistor serves only to control the power
supply output to the LC circuit. It’s not an important factor in
determining the resonant frequency of the oscillator circuit.

6 Oscillators
Starting the Oscillating Current in
the Circuit
Now, let’s look at two ways in which oscillation can be started.
One way of starting oscillations (that is, an oscillating cur-
rent) is to move a permanent magnet through an inductor.
Figure 7A shows a magnet at rest in the center of an inductor.
The inductor is connected to a capacitor. When the magnet is
pulled through the inductor as shown in Figure 7B, a voltage
is induced across the inductor. That voltage charges the
capacitor and starts an oscillating current in the circuit. The
charged capacitor becomes the source of voltage for the next
part of the oscillation. The fully charged capacitor produces a
current through the inductor. The magnetic field around the
inductor serves the same purpose as the field created when
the magnet was withdrawn—it represents stored energy.

FIGURE 7—One way of

starting an oscillating
N current is to move a per-
manent magnet through
an inductor. (A) shows a
magnet at rest in the cen-
HOLLOW ter of an inductor. The
CARDBOARD inductor is connected to a
COIL FORM S
capacitor. When the mag-
net is pulled through the
N + inductor as shown in (B),
a voltage is induced
+ across the inductor. That
L C L C voltage charges the
capacitor and starts an
oscillating current in the
S circuit.

(A) (B)

Figure 8 shows another way to start resonant-circuit oscilla-

tion. In Figure 8A, a voltage source is momentarily connected
across a capacitor while switch SW is open. The voltage
source charges the capacitor. When the switch is closed, the
capacitor discharges and produces current through the
inductor, starting oscillation (Figure 8B).
Now, look at the graphs shown in Figure 9. These graphs
trace the capacitor voltage in the oscillator circuit over time.
In Figure 9A, the capacitor is assumed to be charged (time 0).

Oscillators 7
 FIGURE 8—In (A), a volt-
age source charges the SW
capacitor when switch SW
is open. In (B), the charg-
ing voltage is removed +
and the switch is closed. L V
The capacitor discharge
current flows through the
inductor, and oscillation is
started. (A)

SW

(B)

FIGURE 9—These graphs

trace the capacitor voltage + +
in the oscillator circuit over
time.
~
~

~
~
*
VC

VC
0 0

*
0 1 2 3 4
*
0 1 2 3 4
t t

(A) (B)

+ * + *
~
~

~
~

* * *
VC
VC

0 0

0* 1 2 3 4 *
0 1 2 3 4
t t

(C) (D)

+ *
~
~

* *
VC

0* 1 2 3 *4
t

(E)

8 Oscillators
As shown in the illustration, the maximum negative voltage
is at the bottom of the capacitor. In Figure 9B, the capacitor
discharges through the inductor between time 0 and 1, and
a magnetic field is established around the inductor.
At this point, we’re at the instant of time marked 1 in
Figure 9B. The capacitor is completely discharged, and the
current from the capacitor tries to stop flowing through the
inductor. However, a self-induced voltage in the inductor
(caused by the collapsing magnetic field around the induc-
tor) keeps the current flowing in the counterclockwise
direction from time 1 to time 2. In Figure 9C, the capacitor
is completely charged at time 2. Note that the polarity of the
charged capacitor is now the opposite of what it was in
Figure 9A.
In Figure 9D, the self-induced voltage in the inductor is
depleted and the capacitor is fully charged, so the capacitor
discharges through the inductor between time 2 and time 3.
The counterclockwise discharge current flows through the
inductor, and a magnetic field again surrounds the inductor.
However, the polarity of the magnetic field around the induc-
tor is now the opposite of what it was in Figure 9B. This is
because the capacitor discharge current is in the opposite
direction.
At time 3 in Figure 9D, the capacitor is discharged and the
counterclockwise discharge current tries to stop. The collaps-
ing magnetic field around the inductor causes a self-induced
voltage that keeps the current flowing between time 3 and
time 4. This period is shown in Figure 9E. Observe that the
situation at time 4 in Figure 9E is the same as the situation
at time 0 in Figure 9A. From this point on, the cycle repeats.

A Mechanical Analogy
A mechanical analogy may help you to better understand the
operation of resonant circuits. Figure 10 shows a weight
that’s suspended on a spring. The other end of the spring is
attached to a stationary platform. The weight stretches the
spring, which in turn causes the spring to exert a force on
the weight. This force is proportional to the extension of the
spring. The spring is eventually extended to a degree where it

Oscillators 9
 STATIONARY
PLATFORM
PAPER MOVEMENT

SPRING

PEN

WEIGHT
REGION REGION
A B

FIGURE 10—This illustration shows the oscillating movement of the

weight and spring. By attaching a pen and moving a strip of paper past
the weight, you can make a record of the oscillations of the weight. If we
assume that there are no losses from friction in the system, the oscilla-
tions would continue indefinitely and the tracing on your paper would
look like region A. In reality, friction in the system would cause the
actual trace to look more like the damped waveform shown in region B.

exerts a force equal to—and opposite of—the force of gravity

exerted on the weight. At that point, the system is motionless
and is said to be in equilibrium.
If you now pull the weight down farther, the force exerted by
the spring increases (because the extension of the spring has
been increased). The spring now exerts a force greater than
that of gravity. Thus, when the weight is released, it moves
upward. When the spring reaches the equilibrium point, the
net force on the weight is again zero. The weight, however,
doesn’t stop; because of inertia, the weight continues to move
upwards. At the same time, some of the potential energy pre-
viously stored in the spring is transferred to the weight as
kinetic energy (energy of motion). This reduces the tension of
the spring, and therefore its upward tension. The spring ten-
sion is now exceeded by the downward gravitational force of
the weight, and the weight eventually comes to rest and then
starts to descend. It passes through the lowest equilibrium
position, eventually reaching its lowest point. The whole cycle
of operations is then repeated.

10 Oscillators
The movement of the weight from the lowest point to the
highest point and back again to the lowest position consti-
tutes one complete cycle. The movement of the weight can be
described as simple harmonic motion. (The term “harmonic”
implies sinusoidal oscillations.) The time required to com-
plete this one cycle is called the periodic time (or simply
period ) of the oscillating system.
You could make a record of the oscillations of the weight by
attaching a writing pen to the weight, as shown in Figure 10.
Then, if you moved a strip of paper past the weight at a uni-
form speed in a horizontal direction, the oscillations would be
traced on the paper. If we assume that there are no losses
from friction in the system, the oscillations would continue
indefinitely and the tracing on your paper would look like
region A in the figure. Note that the trace has the same wave-
form as a sine wave graph. In reality, friction in the system
would cause the actual trace on the paper to look more like
region B in the figure.

The Flywheel Effect

The concept that you’ve seen illustrated in this lesson is
called the flywheel effect. The flywheel effect is a circulating
current that continually reverses direction and that produces
a voltage with a sinusoidal waveform across the capacitor. A
sinusoidal voltage is also produced across the inductor. The
flywheel effect takes its name from the mechanical flywheel.
The flywheel and another well-known mechanical device, the
pendulum, are commonly used to explain the electronic prin-
ciples of resonant circuits.
The momentum of a flywheel causes it to coast or “free-
wheel” through several cycles of revolution. Similarly, the
momentum (the potential energy built up by completed alter-
nation of swing) enables the pendulum to return through the
next alternation to the other end of its swing. How does this
compare with the action of a resonant circuit? To understand
the comparison, consider the potential energy of the magnetic
field that’s created by the capacitor discharge current as it
passes through the inductor. The collapsing magnetic field
around the inductor causes current to build up a charge on
the capacitor in the opposite direction.

Oscillators 11
 Incidentally, the resonant circuit, the flywheel, and the pen-
dulum all have something in common—the presence of the
damping effect. Damping refers to the reduction of energy in
a mechanical or electrical system. The flywheel and the pen-
dulum lose energy through absorption. In the pendulum,
this absorption is caused by friction in the bearings; in the
flywheel, the absorption is caused by air friction. In the
same way, resistance in the resonant circuit (mostly in the
inductor coil) causes power to be lost in the form of heat.
Gradually, because of the power lost, the current amplitude
decreases and the oscillation eventually stops. If you were to
plot the voltage across the capacitor in an oscillating circuit,
you’d get a waveform with the decreasing amplitude shown
in Figure 11.

FIGURE 11—Some resist-

ance is always present in
the inductor coil of a res-
+
onant circuit. As a result,
the amplitude of the gen-
erated waveform
gradually declines as
energy is consumed by
this resistance.

TIME

Calculating the Resonant Frequency in

an LC Circuit
In a series-tuned LC circuit, at resonance, the inductive reac-
tance (XL ) equals the capacitive reactance (XC ). Also, the
voltage across the inductive reactance equals the voltage
across the capacitive reactance. That should be no surprise,
since the same current flows through the two components—
that is, IXL equals IXC.
Since the condition for resonance in a series-tuned LC circuit
is that XL must equal XC, we can use that condition to derive
an equation for the resonant frequency ( fr ) of the series-

12 Oscillators
tuned circuit. The following is the derived equation that can
be used to calculate the resonant frequency of an LC circuit:

1
fr 
2 LC

In this equation fr stands for the resonant frequency of the

LC circuit in hertz, L stands for the circuit inductance in
henrys, and C stands for the circuit capacitance in farads,
and  is a constant with a value of 3.14.
The equation for resonant frequency is very important in
electronics. Let’s look at an example problem to see how the
equation is used.
Example: Determine the resonant frequency of a series-
tuned LC circuit that contains a 1 mH inductor
and a 10 F capacitor.
Solution: First, convert millihenries to henrys and micro-
farads to farads.
1 mH  0.001 henry
10 F  0.00001 farad
Substitute these values into the resonant-frequency equation
and solve.

1
fr 
2 LC

1
fr 
2 (0.001 H)(0.00001 F )

1
fr 
2 0.00000001

1
fr 
2(0.0001)

1
fr 
2(3.14)(0.0001)

1
fr 
2(0.000314)

Oscillators 13
 1
fr 
0.000628

fr  1592.36 Hz
Example: Suppose that you have a 10 F capacitor, and
you want to make a series-tuned circuit that
has a resonant frequency of 1,592 Hz. What
value of inductance would you need in the
circuit?
Solution: To calculate the value of inductance (L), we can
use the following variation on the resonant fre-
quency formula:

1
L
4 f r 2C
2

Substitute the known values into the formula and solve.

1
L
4 (1, 592 Hz )2 (0.00001 F )
2

1
L
4 (2, 534, 464)(0.00001 F )
2

1
L
4 (25.34)
2

1
L
4(3.14)2 (25.34)

1
L
4(9.8596)(25.34)

1
L
4(249.84)

1
L
999.36

L  0.001 H
Note that this answer corresponds to the value of L that was
given in the previous example problem.

14 Oscillators
Series-Tuned and Parallel-Tuned
LC Circuits
Figure 12A shows a series-tuned LC circuit, and Figure 12B
shows a parallel-tuned LC circuit. You can identify these cir-
cuits by the position of the generator. In the series-tuned LC
circuit, the generator is connected across both components,
and in the parallel-tuned LC circuit, the generator is con-
nected across each component.

L
GENERATOR
GENERATOR
C L

(A) (B)

FIGURE 12—One way to determine if an LC circuit is series-tuned or

parallel-tuned is to look for the position of the generator. In the series-
tuned LC circuit in (A), the generator is connected across both
components, and the same generator current flows through both compo-
nents. In the parallel-tuned LC circuit shown in (B), the generator is
connected across each component, and the current divides between the
L branch and the C branch of the circuit.

Another series-tuned LC circuit is shown in Figure 13. A

series-tuned LC circuit will pass signals that have a narrow
range of frequencies at and near the resonant frequency,
and reject the frequencies that are outside that range. The
impedance of a series-tuned LC circuit is minimum at the
resonant frequency. The resonant frequency of a series-
tuned LC circuit can be varied by varying the value of
inductance or capacitance. The effect of series resistance on
the resonant frequency of a series-tuned LC circuit can be
disregarded. When the series resistance in a series-tuned
circuit increases, the Q of the circuit decreases and causes
it to tune over a wider range of frequencies. Decreasing the
resistance increases the Q.

Oscillators 15
 FIGURE 13—This figure
demonstrates the effect
of resistance on reso-
nance. (A) shows a I
series-tuned LCR circuit.
The characteristic curve L LOW R
in (B) shows the effect of
resistance on the circuit.
You can see that a higher C
HIGH R
resistance causes the cir-
cuit to tune more broadly. R
This means that the Q of
the tuned circuit I
decreases as the resist- FREQUENCY
ance increases. fr

(A) (B)

A parallel-tuned LC circuit is shown in Figure 14. A parallel-

tuned circuit will reject signals that have a narrow range of
frequencies at and near the resonant frequency, and pass the
frequencies that are outside that range. The impedance of
parallel-tuned LC circuit is maximum at the resonant fre-
quency. The resonant frequency of a parallel-tuned LC circuit
can be varied by varying the value of inductance or capaci-
tance. The effect of parallel resistance on the resonant
frequency of a parallel-tuned LC tuned circuit can be disre-
garded. When the parallel resistance in a parallel-tuned

HIGH R

L C R
LOW R

I FREQUENCY
fr

(A) (B)

FIGURE 14—In the parallel-tuned LCR circuit shown in (A), the resistor is
connected across both L and C. The characteristic curve in (B) shows that
the high resistance causes the circuit to tune sharply, and the low resist-
ance causes it to tune more broadly.

16 Oscillators
circuit decreases, the Q of the circuit decreases and causes it
to tune over a wider range of frequencies. Increasing the par-
allel resistance increases the Q.

Complex Parallel-Tuned LCR Circuits

Now, let’s look at the complex parallel-tuned LCR circuit
shown in Figure 15. The following equation can be used to
calculate the resonant frequency of this circuit:

1 R L 2C  L
fr 
2 LC RC 2C  L

FIGURE 15—In this com-

plex parallel-tuned LCR
circuit, the values of
resistance affect the
resonant frequency. A
complex parallel-tuned
LCR circuit may contain
one or both resistors, and
L C
either or both of these
resistors may be variable,
to adjust the resonant
frequency.

RL RC

Observe that in this equation, whenever the value of RL2C is

equal to the value of L, the value of RL2C  L will equal 0.
Thus, the numerator of the fraction under the large radical
would equal zero, which would mean that there was no reso-
nant frequency for that particular circuit. In contrast, there’s
a resonant frequency for every positive value combination of
the values of R, L, and C in the circuits shown in Figure 13
and Figure 14.
Also, if the numerator of the fraction under the large radical
equals the denominator, the value under the radical would be
1. Since the square root of 1 is 1, the resonant frequency of

Oscillators 17
 that particular circuit would be the same as the resonant
frequencies for the circuits in Figure 13 and Figure 14.
Thus, you can see that the resonant-frequency of the com-
plex circuit in Figure 15 depends on the values of RL and/or
RC. Therefore, it follows that the resonant frequency of that
circuit can be adjusted by making either or both of those
resistors variable.
In an early television receiver model, a manufacturer used
the complex parallel-tuned circuit of Figure 15 as a resonant-
frequency trap. In the alignment procedure, the resonant
frequency of the trap was adjusted by changing the setting of
RL. Technicians who weren’t familiar with the resonant-fre-
quency equation given above were thoroughly confused by
the use of a variable resistor to adjust a frequency. Thus, you
should take some time to study the resonant-frequency for-
mulas thoroughly so that you understand them well.
In discussions of oscillator circuits, you’ll sometimes run
across the term “induced current.” Some electronics techni-
cians would argue that there’s no such thing as an induced
current—there can be only an induced voltage. However, for
the purposes of your studies, we’ll define an induced current
as a current that flows as a result of an induced voltage. This
makes “induced current” a legitimate term.
Remember that an oscillator is sometimes defined as a circuit
that converts DC to AC. This is a good way to think about the
oscillator circuit’s job.
Now, take a few moments to review what you’ve learned by
completing Power Check 1.

18 Oscillators
 Power Check 1
At the end of each section of Oscillators, you’ll be asked to pause and check your under-
standing of what you’ve just read by completing a “Check Your Learning” exercise. Writing
the answers to these questions will help you to review what you’ve studied so far. Please
complete Power Check 1 now.

Questions 1–6: Indicate whether each statement is True or False.

______ 1. A parallel-tuned LC circuit will pass signals that have a narrow range of frequencies at
and near the resonant frequency, and reject the frequencies that are outside that
range.

______ 2. Sinewave oscillations can be described as simple harmonic motion.

______ 3. The impedance of a series-tuned LC circuit is maximum at the resonant frequency.

______ 4. Oscillators can be defined as circuits that convert AC to DC.

______ 5. In a conventional amplifier, the output signal is 180 degrees out of phase with the
input signal.

______ 6. In an LC oscillator circuit, the frequency of the oscillation is determined almost com-
pletely by the inductance of L and the capacitance of C.

(Continued)

Oscillators 19
 Power Check 1
Questions 7–14: Fill in the blanks in each statement.

7. The rate at which the electrical energy is converted to heat energy is a measure of the
amount of lost in the circuit.

8. The components that are used as amplifiers in an oscillator circuit are .

9. The of a series-tuned LC circuit can be varied by varying the value of inductance or

capacitance.

10. The major components in an LC oscillator circuit are and .

11. In an oscillator circuit, resistance in the resonant circuit causes power to be lost in the form of
.

12. In an oscillator circuit, the feedback system ensures that the returned signal is in phase with
the input signal in the resonant circuit. This is called or feedback.

13. The four essential elements of an LC oscillator are the , the , the ,
and the .

14. When the series resistance in a series-tuned circuit increases, the of the circuit
decreases and causes it to tune over a wider range of frequencies.

Check your answers with those on page 59.

20 Oscillators
LC OSCILLATOR APPLICATIONS

Armstrong Oscillators
Now that we’ve discussed resonance in LC and LCR circuits,
we can turn our attention to oscillator circuits. Let’s look at
how the components in an oscillator circuit work together to
produce oscillation. First, you should be aware that in order
to get an oscillator circuit into full operation, a startup period
is needed. Because the capacitor opposes any change in volt-
age across its terminals and the inductor opposes any
change in current through it, it may take one or more cycles
to get an oscillator into final operation. After the startup
cycle, the voltage across the capacitor and the current
through the inductor will continue at full amplitude during
each half-cycle.
You’ve already seen how a resonant circuit can function as a
frequency-control device. Add the other basic requirements
for oscillation—amplification, feedback, and a power supply—
to such a circuit, and you’ll have a circuit that will oscillate.
The simplest example of how these elements work together is
the Armstrong oscillator circuit shown in Figure 16.

The Operation of the Armstrong Oscillator

In this Armstrong oscillator circuit, when switch SW is
closed, a positive voltage is delivered to the transistor’s collec-
tor through LF, and a positive bias voltage is delivered to the
transistor’s base through RB. (This is an example of basic-
bias operation.) As a result, transistor Q starts to conduct.
When the transistor conducts, collector current flows through
LF and causes the magnetic field around LF to expand. The
expanding magnetic field around LF cuts across the turns of
wire in L and induces a voltage across that inductor. This is
how oscillation in the LC circuit starts.
In the LC tuned circuit, regenerative feedback occurs when
the transistor amplifier conducts through LF. The power
that’s lost due to resistance in the circuit (in the form of heat)
is replaced when the transistor conducts. (Note that the two

Oscillators 21
 LF
RB

VCC
C1

CC
+ + SW
Q
OUTPUT +
L C
SIGNAL

RE
CE

(A)

SINE WAVE VOLTAGE

AT THE TOP OF C AND
ON THE TRANSISTOR BASE DC BASE
BIAS VOLTAGE

0V

(B)

FIGURE 16—An Armstrong oscillator circuit is shown here.

inductors, LF and L, form a transformer, with LF as the primary

and L the secondary. This “transformer” provides the regener-
ative feedback path for the oscillator.)
The induced voltage across L starts the flywheel current in
the LC circuit. As the current increases during startup, the
feedback from LF to L produces a positive voltage at the top
of L. Therefore, there’s also a positive voltage at the top of C.
This positive voltage is delivered to the base of the transistor
through coupling capacitor CC, and the transistor collector
current increases further. The arrows in the figure show the
paths of the startup current.

22 Oscillators
As the collector current increases, the positive voltage at
the top of L (and C ) increases, causing a further increase in
the positive base voltage. As a result, the transistor is
driven toward saturation. At the same time, capacitor C is
charged and the magnetic field around LF reaches its maxi-
mum value. When the transistor is saturated, there can be
no further increase in the collector current, and LF can’t
induce any more voltage in L. The current in L tries to stop
flowing. However, the collapsing magnetic field around L
produces a self-induced voltage that causes the current in
L to continue flowing. The self-induced voltage continues
to charge C.
The voltage at the top of C causes the transistor to go into
saturation when the top of capacitor C is at its maximum
positive value. When the top of C is at its most negative
value, the transistor is driven toward the cutoff point, but
it doesn’t reach cutoff. The reason the transistor doesn’t
reach the cutoff point is that there’s always a positive
voltage on its base. That positive voltage is obtained
through RB .
From your knowledge of LC oscillating currents, you know
that the voltage at the top of C alternates between positive
and negative values. However, you must understand that
the negative voltage that’s delivered to the base of the tran-
sistor through CC does not cause the base voltage on the
transistor to become negative. Remember that a positive
base voltage exists by way of RB. The sinewave voltage,
arriving through CC, combines with the base voltage as
shown in Figure 16B. As you can see, the base voltage is
never negative.
Note: If the base of an NPN transistor is allowed to become
negative with respect to the emitter, the transistor will proba-
bly be destroyed.
Transistor amplification provides a feedback signal just large
enough for the oscillator to continue to oscillate. Again, the
purpose of feedback in an oscillator circuit is only to compen-
sate for losses in the tuned circuit. In practice, only a very
small feedback signal is required. The output of the tuned
circuit—and thus the transistor—will be at the resonant fre-
quency of the tuned circuit.

Oscillators 23
 Armstrong Oscillator Components
Now, let’s take a closer look at the Armstrong oscillator circuit’s
components. Let’s start with the transistor. There has been
much discussion about a transistor being a current-operated
device. In other words, an increase in base current causes an
increase in collector current; and a decrease in base current
causes a decrease in collector current. This is true; however, an
increase in base current is usually caused by an increase in the
voltage delivered to the base. Likewise, a decrease in base cur-
rent is usually caused by a decrease in base voltage. Therefore,
transistor action can be described by the action of the base
voltage AND the collector current. You can think of it this way—
the action of the bipolar transistor is based on currents inside
the transistor, but circuit action is often described by voltages
outside the transistor. The operation of the transistor in this
lesson is described on the basis of base voltages and collector
currents. (Note that you’ll troubleshoot transistor circuits by
making measurements of voltage—not current.)
You should be aware that a transistor circuit that uses basic
bias (like the circuit in Figure 16A) can produce a thermal
runaway that could quickly destroy the transistor. To prevent
thermal runaway, the emitter combination of RE and CE is
used in the circuit. This RC circuit is sometimes called a
temperature-compensating circuit. To be very precise, it’s the
resistor that prevents thermal runaway. If the transistor cur-
rent starts to increase rapidly (which is a characteristic of
thermal runaway) the voltage across RE rises rapidly to
decrease the difference between the base voltage and the
emitter voltage. This prevents a further rise in the transistor
current. The capacitor (CE ) helps to hold the emitter voltage
at a steady DC value. This in turn prevents a loss of amplifi-
cation in the amplifier that would be caused by the presence
of a signal voltage across RE. (Note that you shouldn’t call the
temperature-compensating circuit a bias circuit. The bias for
the transistor is provided through RB.)
The portion of the Armstrong oscillator circuit that contains
L and C is sometimes called the tank circuit. This is the part
of the circuit where oscillation actually takes place. The fre-
quency of oscillation is dependent upon the resonant
frequency of that tank circuit.

24 Oscillators
Now, let’s look at the coupling capacitor (CC ) in greater
detail. If the coupling capacitor wasn’t present in the circuit,
the base bias on the transistor would be grounded through
the wire windings on inductor L. The coupling capacitor
couples the signal from the LC tank circuit to the base of
the transistor amplifier. It’s important to note that the
coupling capacitor doesn’t have anything to do with oscilla-
tion in the circuit. It simply couples the signal voltage from
the tank circuit to the transistor base, and at the same
time, it prevents the DC base voltage from being grounded
through L.
Another very important thing to remember about coupling
capacitors is that they don’t reverse the phases of signal
voltages. If the voltage on the left side of the capacitor is
positive-going, then the voltage on the right side of that
capacitor is also positive-going. Likewise, when there’s a
negative-going voltage on the left side of the capacitor,
there’s also a negative-going voltage on the right side.
The feedback inductor in this circuit (LF ) is sometimes called
the tickler coil. It’s shown in Figure 16A as a separate wind-
ing, but LF is actually usually wound on top of L. This
arrangement provides better magnetic-field coupling for the
two windings.
In an Armstrong circuit, either L or C can be made variable
to adjust the frequency of oscillation. In this particular cir-
cuit, the inductor L is a variable component. Variable
inductance is sometimes used instead of variable capaci-
tance in oscillator circuits to adjust the resonant frequency
of a tuned circuit. This is particularly true for portable elec-
tronic equipment such as car radios and portable radios. A
variable inductance is less affected by vibration. Figure 17
shows how an adjustment screw and a slug are used to
adjust the inductor to obtain the inductance that’s needed to
produce the desired resonant frequency. The slug may be
made of powdered iron or a ferrite material.

Regenerative Amplifiers
An Armstrong oscillator circuit can be modified to make a
regenerative amplifier circuit (Figure 18). If the feedback
control is adjusted so that the circuit doesn’t go into oscil-

Oscillators 25
 FIGURE 17—This illustra-
tion shows how an
adjustment screw and a
ADJUSTMENT
slug are used to adjust
SCREW
the inductor in the
Armstrong oscillator
SLUG
circuit to obtain the
inductance that’s needed
to produce the desired
resonant frequency.

(A) (B)

MAXIMUM MINIMUM
INDUCTANCE INDUCTANCE

RF
FEEDBACK
CONTROL

LF

RB
VCC
C1

CC
SW
Q

INPUT L C
SIGNAL

OUPUT
SIGNAL

FIGURE 18—As shown here, an Armstrong oscillator circuit can be modi-

fied to make a regenerative amplifier. Similar modifications are sometimes
made on other types of oscillator circuits. A regenerative amplifier will
produce a very high voltage gain.

lation, the regenerative feedback will be limited, and the

feedback signal will reinforce the input signal. This will pro-
duce a very high voltage gain for the circuit. Regenerative
amplifier circuits have a very high voltage gain, but unless

26 Oscillators
measures are taken to prevent oscillation, it will occur in
the circuit. Oscillation will make the circuit useless as
an amplifier.

Troubleshooting an Armstrong Oscillator Circuit

Today, it’s possible to purchase a reasonably priced frequency
meter with a digital display. This type of meter is used to
measure the frequency of oscillation. These meters are par-
ticularly useful for measuring the frequency of an oscillator
while L or C is adjusted for the desired frequency. Using this
method, the frequency can be very accurately adjusted. How-
ever, an Armstrong oscillator isn’t highly stable when there’s
a change in ambient (surrounding) temperature, so the fre-
quency of the oscillator is likely to drift away from the
adjusted value.
To troubleshoot an Armstrong oscillator circuit like the one
shown in Figure 16, follow these steps.

Step 1: Start with a careful visual inspection. Look for bad

solder connections, broken wires, burned areas, and
other problems that can disrupt the operation of the
oscillator. Look carefully at the connections for LF
and L. If the circuit is new and has never worked,
the transformer made with the two inductors may
be wired incorrectly and may be producing degener-
ative feedback. Compare the connections of the
numbered terminals on the inductors with the cir-
cuit schematic to make sure that they’re correctly
connected.
Step 2: After a thorough visual inspection, check the power
supply voltage. If the supply voltage doesn’t match
the circuit specifications, correct the problem. This
is particularly likely to be a problem in battery-
operated circuits. You can waste a lot of time
looking for a source of trouble elsewhere if the
power supply voltage isn’t correct.
Step 3: Check the emitter, base, and collector voltages at
the transistor. Even if the supply voltage is correct,
the voltages at the transistor leads may wrong. For
example, if LF is open, there won’t be a collector volt-
age. The emitter voltage should be positive because
of the voltage across RE. Remember that the positive-
base bias voltage must be supplied through RB.

Oscillators 27
 Troubleshooting an oscillator circuit, like troubleshooting any
electronic circuit, should be a logical procedure. One of your
most important troubleshooting tools is your knowledge of
what the components do and how they work.

Hartley Oscillators
As you just saw, an Armstrong oscillator circuit uses a primary
winding that’s sometimes called a tickler coil (LF ), as well as a
secondary winding (L ). However, this transformer arrangement
isn’t always needed for oscillation. The Hartley oscillator shown
in Figure 19 uses a tuned circuit that consists of a variable
capacitor (C ) and a tapped inductor (L ). This tapped inductor
is sometimes called an autotransformer. The lower part of the
tapped inductor is referred to as LF, and it serves as the trans-
former primary. Keep in mind that the tapped inductor
consists of one continuous winding. It’s not two inductors
connected in series, as the illustration might suggest.

RB2 RB1

CC
B
RFC
L
OUTPUT C
+
CE SW
RE
+
LF
C1

FIGURE 19—This is the circuit for a Hartley oscillator. It uses an auto-

transformer for feedback. Observe that the base bias voltage is obtained
with a voltage divider made of RB1 and RB2.

Now, let’s trace the current paths in the Hartley oscillator

circuit. When switch SW is closed, the current flows out of
the battery into the voltage divider (RB1 and RB2). Then, the
current flows out of common and back through the closed
switch to the positive battery terminal. The voltage divider
provides the DC bias at point B for the transistor base bias.

28 Oscillators
Next, current flows out of the negative battery terminal,
through the PNP transistor (from collector to emitter), and
back to the battery through SW. Since the transistor is prop-
erly supplied with base- and collector-operating voltages, it
can be used as an amplifier.
Part of the startup current flows from RFC through C1,
through LF (a radio-frequency choke inductor), and back to
the positive battery terminal. Capacitor C1 is charged during
this time. Since the startup current flows through the auto-
transformer’s primary (LF ), it induces a voltage across L. The
voltage across L starts a flywheel current between L and C
(C is located in the tank circuit). The resonant frequency of
the oscillator is determined by L and C.
During the half-cycle we just described, C becomes fully
charged. It then discharges through L to continue the fly-
wheel effect. The sine-wave voltage at the top of the tank
circuit is coupled through CC to the base of the transistor.
During the positive half-cycle of input, the transistor con-
ducts and produces a voltage across RFC. This voltage is fed
back to the bottom of LF, and it produces a current in LF.
The expanding magnetic field of the current through LF
induces a voltage across L that replaces power lost in the
tank circuit. This power is lost in the form of heat when cur-
rent flows through the resistance of the inductor wires (most
of the loss) and through the dielectric of capacitor C (only a
small amount of the loss).
During the part of the cycle when the capacitor is fully dis-
charged, part of the energy of the discharge current is stored
in the inductor’s magnetic field. When the capacitor dis-
charge current drops to zero, the inductor’s magnetic field
begins to collapse. This sets up a current in the tank circuit
that charges the capacitor with an opposite polarity. When
the energy of the magnetic field is completely used up, the
capacitor is fully charged again, and the capacitor discharge
cycle starts over.
The amplifier, delivering power through the LF portion of the
inductor, is sufficient to replace any losses in the tuned cir-
cuit once it has started to oscillate. This transistor current
will be in phase with the charge-discharge current of the
tank circuit at all times. (In other words, the feedback will be

Oscillators 29
 positive.) The RFC inductor presents a high impedance to the
oscillator current. On the other hand, the RFC inductor offers
little resistance to the DC flow from the power source.
You’ll see Hartley oscillators drawn in circuit diagrams in
different ways. To identify a particular circuit as a Hartley
oscillator, look for the split inductors. It should be pointed
out that in the circuit shown in Figure 19, an air-core trans-
former will usually be used to couple the Hartley oscillator to
the following stage (or circuit). Air-core transformer coupling
is common at high frequencies, such as radio frequencies
between 200 kHz and several hundred megahertz. Iron-core
or ferrite-core transformers can be used at lower frequencies.

Colpitts Oscillators
As you’ve just seen, the Hartley oscillator uses a tapped
inductor to provide the positive feedback necessary to main-
tain oscillation. Another way to provide positive feedback to
sustain oscillations is to tap a split capacitor as shown in
Figure 20. The split capacitor is actually two capacitors
(sometimes mounted on a common shaft so that both capaci-
tors can be adjusted at the same time). An oscillator that
uses a tapped capacitor (or split capacitors) is called a

RB2 RB1

CC
Q RFC

+
C
+
OUTPUT L CE
RE ECC
+
CF +
+

C1

FIGURE 20—In this Colpitts oscillator the split capacitors replace the split
inductor in the Hartley circuit. Except for that difference, the circuit is
very similar to the Hartley oscillator in Figure 19.

30 Oscillators
Colpitts oscillator. Often, C and CF are variable capacitors,
and L is a fixed inductance. However, in the Colpitts oscilla-
tor shown in Figure 20, L is the tunable element, and the
tapped capacitors are fixed.
Oscillations are started in this Colpitts oscillator by the first
pulsation of current through the transistor (Q). This causes
the feedback capacitor (CF ) to be charged, as indicated by the
right-hand set of polarity signs. After CF is fully charged by
the first current pulse, current stops flowing momentarily.
The feedback capacitor then discharges downward through L.
This current is aided by the magnetic field of L, even after the
charge on CF is drained off. Thus, current continues to flow
until the magnetic field of L drops to zero. At this time, the
two capacitors will be charged with the polarity indicated in
the figure. Current will then flow in the opposite direction
until the charges on the capacitor are reversed. As in the
Hartley oscillator, this charge-discharge cycle continues for
as long as the transistor is operating.

Crystal-Controlled Oscillators
A typical crystal-controlled oscillator is shown in Figure 21.
The feedback signal from the collector is coupled through the

RB1

CC
TO
NEXT
STAGE
Q

C L LF
CS
RB2
CE
RE
L
CP
CRYSTAL
ECC C1
+
RS

FIGURE 21—The crystal is the equivalent of the components that are

shown within the broken line. For a given size and thickness, a piezoelec-
tric crystal has a certain mechanical resonant frequency. Generally, the
smaller the dimensions of the crystal, the higher its resonant frequency.

Oscillators 31
 tickler coil (LF ) and the series-connected crystal to the base
of the transistor (Q). When this feedback voltage is applied
across the plates that hold the crystal, the voltage causes the
crystal to vibrate. The crystal vibrates at a frequency that’s
determined by its physical dimensions.
The collector load of transistor Q is a parallel resonant circuit
that contains the variable capacitor C and the inductor L.
The signal that’s developed across this tuned circuit is coupled
to the next stage through capacitor CC. (The values of L and
C are selected so that their resonant frequency matches the
resonant frequency of the crystal.) Often, the variable com-
ponent in the tuned circuit (in this case, capacitor C ) is
adjustable only over a very narrow range, and is adjusted by
means of a very precise mechanical linkage. It’s then locked
in place so that the frequency of the tuned circuit is slightly
above the crystal frequency. This is accomplished by decreas-
ing the capacitance of C. Thus, the resonant circuit (or tank)
presents an inductive reactance to the crystal base circuit,
guaranteeing that the feedback voltage will be of the proper
phase to sustain oscillations. These oscillations occur at the
resonant frequency of the crystal.
At one time, crystal-controlled oscillators weren’t practical for
low-frequency applications, such as timing clocks. The large
size of the required crystal and the cost were the limiting fac-
tors. In addition, only a single frequency was obtainable
because of the fixed-frequency characteristics of the crystal.
However, these are no longer serious disadvantages today.
Off-the-shelf digital circuits (called “count-down” or “divide-
by” circuits) can be used to reduce the oscillator frequency to
a desired lower frequency. For example, a 3 MHz or 4 MHz
crystal oscillator circuit frequency can be reduced to 60 Hz
in order to operate a clock. The advantage, of course, is the
high accuracy that can be obtained. The single-frequency
output isn’t a disadvantage in this case.

TCTB Oscillators
Now, let’s look at the TCTB oscillator. The TCTB oscillator is
an LC oscillator that combines frequency variability and fre-
quency stability. (The letters TCTB stand for tuned-collector
tuned-base.) A typical TCTB circuit is shown in Figure 22.

32 Oscillators
 B+

R1

C2

L2

C1 L1 C5 R2 C3 OUTPUT

C4

L3

B+

FIGURE 22—The TCTB oscillator is similar to the crystal-controlled oscilla-

tor; however, both its base resonant circuit and collector resonant circuit
can be varied. So, the frequency of the signal output can be changed by
readjusting the base and collector capacitors.

The tuned input and output circuits are coupled by capacitor

C4, which places both circuits at ground potential. This
establishes the bottom of the collector resonant circuit (C3
and L2), and maintains the RF path between the collector and
the emitter. An isolation RF choke (L3) is placed in series with
the resonant circuit and the power supply.
In order for the TCTB oscillator to work, one of the resonant
circuits must be tuned slightly off frequency. To understand
the reason for this, assume that the base resonant circuit (C1
and L1) is tuned to resonance. The inductive reactance of L1
would have a value equal to the capacitive reactance of C1.
The two reactances, which have opposite phase character-
istics, would then in effect cancel each other out. Thus, the
base circuit becomes primarily resistive (though with some
capacitive reactances created by the interelements and by
external capacitor C2). Nevertheless, C2 has a very low reac-
tance for the frequency of the oscillator.
Now, assume that we tune the collector tank to resonance:
it too will become resistive, because the effective reactance
has been canceled again. Only resistance and stray inter-

Oscillators 33
 element capacities remain. Therefore, oscillations can’t occur.
Remember, there can be no flywheel effect if there’s no inter-
change of energy between the inductor and the capacitor.
To create oscillations, the resonant circuit in the collector can
be tuned slightly above the frequency that the oscillator is to
produce. (When tuned above the oscillation frequency, the
collector resonant circuit will operate below its resonant fre-
quency, since the oscillation frequency is lower than the
resonant frequency of the resonant circuit in the collector.)
This increases the capacitive reactance of C3, makes the
collector-tuned circuit less inductive, and decreases the
impedance that formerly existed in the parallel circuit. The
signal energy, finding less opposition in the inductor, flows
through it. Finally, the circuit, now that its inductive opposi-
tion has been decreased, offers a lower opposition to
signal-energy flow.

Frequency Multipliers
Before we leave our study of LC (inductive-capacitive) reso-
nant circuits and go on to RC (resistive-capacitive) oscillators,
let’s take a brief look at frequency multipliers.
In many transmitters, the output frequency of the oscillator
stage differs from the desired carrier frequency. When this
happens, the output frequency is first changed to the desired
carrier frequency, and the output is then raised to the required
power level by RF amplifiers. Thus, a frequency multiplier may
be described as an RF amplifier whose output frequency is
n times its input frequency (where n is some whole number
larger than 1). For n  2, the multiplier stage is called a
doubler; for n  3, a tripler; and for n  4, a quadrupler. The
obtainable output from a multiplier stage decreases rapidly,
however, as the order of multiplication increases. Accordingly,
values of n that are greater than 4 are rarely used.
One important thing to remember about a frequency multi-
plier is that only exact multiples of the input are obtainable.
Thus, frequency multiplication to 20 MHz, 30 MHz, or 40 MHz
is possible from a 10 Mhz input, since these values are exact
multiples of 10. Frequency multiplication of 18 MHz or 32 MHz
wouldn’t be possible from a 10 MHz input, since these values
aren’t exact multiples of 10.

34 Oscillators
Another important fact to keep in mind about a frequency
multiplier is that it can produce many output frequencies at
the same time. However, the tuned output circuit of the mul-
tiplier selects only the desired frequency. The undesired (or
spurious) frequencies are suppressed by the tuned output
circuit. Thus, in the doubler-tuned circuit of Figure 23, the
output circuit would be tuned to 2f1. The other frequencies
(3f1, 4f1, and so on) would be suppressed and limited to a
very small fraction of the desired output. A figure of 60 dB of
spurious suppression is desirable and isn’t uncommon.

OUTPUT
INPUT
2f1
f1

CIRCUIT FREQUENCY DOUBLER-TUNED CIRCUIT

TUNED MULTIPLIER TUNED TO 2f1
TO f1

FIGURE 23—Frequency multipliers are used in circuits to change the fre-

quency of an RF signal. When used in the doubler-tuned circuit shown
here, all frequencies other than 2f1 will be suppressed.

A bel is a unit that’s used to express the ratio of two values

of power. A decibel (dB) is equal to one-tenth of a bel. The
decibel is used to compare output powers. So, for example,
for 60 dB of spurious suppression, a 100 W transmitter
would be limited to a radiation of no more than 100 W at
any spurious frequency.
Now, take a few moments to review what you’ve learned by
completing Power Check 2.

Oscillators 35
 Power Check 2
Questions 1–8: Fill in the blanks in each statement.

1. An RF amplifier whose output frequency is n times its input frequency is called a .

2. In a transistor circuit that uses basic bias, a temperature-compensating circuit may be used
to prevent .

3. The portion of an Armstrong oscillator circuit that contains L and C is sometimes called the
.

4. A Hartley oscillator uses a inductor to provide the positive feedback necessary to

maintain oscillation.

5. In a crystal-controlled oscillator, digital circuits called or circuits can be

used to reduce the oscillator frequency to a desired lower frequency.

6. In an Armstrong oscillator circuit, if the feedback control is adjusted so that the circuit
doesn’t go into oscillation, the circuit will be modified into a circuit.

7. The feedback inductor (LF) in an Armstrong oscillator circuit is sometimes called a

.

8. When you’re troubleshooting an Armstrong oscillator circuit, after you perform a careful
visual inspection, you should check the voltage.

(Continued)

36 Oscillators
 Power Check 2
Questions 9–14: Indicate whether each statement is True or False.

______ 9. A frequency multiplier can produce many output frequencies at the same time. The
desired frequency can be selected by a tuned circuit.

______ 10. A tapped inductor is also sometimes called an autotransformer.

______ 11. The first step that should be taken when troubleshooting an Armstrong oscillator
circuit is to check the emitter, base, and collector voltages at the transistor.

______ 12. It may take one or more cycles to get an oscillator into final operation.

______ 13. A frequency meter with a digital display is usually used to measure the frequency of
oscillation in an Armstrong oscillator circuit.

______ 14. A Colpitts oscillator uses a tapped inductor to provide positive feedback to sustain
oscillation.

Check your answers with those on page 59.

Oscillators 37
 OSCILLATORS WITH RC
FEEDBACK CIRCUITS
So far, all of the oscillators we’ve considered have used LC
combinations somewhere in the circuit. This is fine for generat-
ing high-frequency signals, and even for high audio-frequency
signals. At low frequencies, however, either the inductor or the
capacitor must be quite large. For a given frequency, if the
capacitance is low, the inductance is high, and vice versa.
Large-value inductors and capacitors tend to be bulky and
expensive, and large variable capacitors and inductors aren’t
always available. Therefore, in order to produce variable,
low-frequency oscillators, circuit designers use RC rather
than LC combinations.

RC Phase-Shift Oscillators
A special RC oscillator called an RC phase-shift oscillator
is shown in Figure 24. Each of the three capacitors in the
circuit (C1, C2, and C3) will produce a phase shift of

VCC +

2
OUTPUT
RB
RL
C1 3 C2 4 C3

Q
5
1 R3
R1 R2
RE
CE

VCC

FIGURE 24—Although a capacitor produces a phase shift, a resistor

doesn’t. Therefore, the phase shift produced by an RC combination will
always be less than 90 degrees. However, by connecting several RC com-
binations in series, the individual phase shifts combine to produce the
180-degree phase shift needed for oscillator action.

38 Oscillators
approximately 60 degrees. As a result, the signal that’s
coupled back from the collector to the base will be in phase
with the original signal on the base, and in phase with the
collector signal.
The original signal in Figure 24 would be the first cycle of
feedback from the collector (waveform 1). You’ll note that
waveform 2 is 180 degrees out-of-phase with this signal, and
that each succeeding waveform (waveform 3, waveform 4, and
waveform 5) is shifted forward by 60 degrees. Thus, waveform
5 is once again in phase with waveform 1. Consequently, it
will add to waveform 1 to maintain an input high enough to
keep the system oscillating.
The requirement that each RC section have the same phase
shift must satisfy the following relationship:
R 1C 1  R 2C 2  R 3C 3
In some circuits, this condition is satisfied by making R1
equal to R2 and R3, and C1 equal to C2 and C3. The RC
phase-shift network is a loss network, however, so an ampli-
fier gain is needed to sustain oscillation. If more RC sections
are used, a lower phase shift is needed in each section, and
the losses are reduced accordingly. For this reason, it’s
fairly common to find more than three sections in the phase-
shifting network.
In the phase-shift oscillator, the amplifier output voltage is
applied across the series combination of C1R1 to produce the
phase relationship shown in Figure 24. The voltage across
the resistor R1 leads the applied voltage by some angle that
depends on the ratio of capacitive reactance XC1 to R1. Since
XC1 varies with frequency, there’s one (and only one) fre-
quency for which the phase angle will be 60 degrees. This
60-degree out-of-phase voltage across R1 in the oscillator is
applied to the combination of C2R2, where another 60-degree
phase shift occurs across R2. This voltage is finally applied
across C3R3, which produces a third 60-degree phase shift,
thus reaching the total of 180 degrees necessary for positive
feedback. The voltage across R3 is the input signal voltage to
the transistor amplifier. If the gain of the transistor is high
enough to make up for the loss in the RC network, the cir-
cuit will oscillate.

Oscillators 39
 If in the phase-shift oscillator the resistors R1 R2, R3, and RL
are equal or nearly equal, and the capacitors C1, C2, and C3
are also equal or nearly equal, the frequency is approximately

50
f ⬇
RC

In this formula f stands for the frequency in hertz, R stands

for the resistance in ohms, and C stands for capacitance in
farads.
Let’s look at an example problem. Assume that in Figure 24,
the value of R is equal to the value of RL, which is 5 k.
Assume that R1 is also set at 5 k and that C1, C2, and C3
all have values of 50 F. To find the frequency, substitute
these values into the formula and solve as follows:

50
f 
RC

50
f 
5 k  50 F

50
f 
5000   0.00005 F

50
f 
0.25

f  200 Hz
The RC phase-shift is normally fixed in frequency, but it can
be made variable by including ganged variable capacitors or
resistors in the phase-shift network. An increase in the value
of either R or C will produce a decrease in the output fre-
quency. Conversely, a decrease in the value of either R or C
will produce an increase in the output frequency.
The chief advantages of the phase-shift oscillator are its
economy (it uses fairly inexpensive components) and the fact
that it produces output waveforms of fairly good quality over
the range of frequencies for which it’s designed. However, at
the extreme ends of its oscillation range, the waveforms tend
to become distorted to the point that they’re no longer pure
sine waves.

40 Oscillators
There are better low-frequency RC sinewave generators than
the phase-shift oscillator, such as the Wien bridge, the twin-
T, and the UJT oscillator. The UJT oscillator works very well
at low frequencies, but it doesn’t have a sinewave output.
Let’s take a closer look at these other circuits.

Wien Bridge Oscillators

If the frequency of an oscillator voltage must be variable, the
simultaneous adjustment of all the resistors or capacitors in
a phase-shift oscillator network isn’t convenient. In addition,
three-ganged variable resistors or capacitors of suitable value
aren’t readily available. Consequently, it’s more convenient to
use an oscillator of the Wien bridge variety. With it, capacitor
values are selected by two-ganged switches, and fine fre-
quency variations are achieved using readily available
two-ganged variable resistors.
The Wien bridge oscillator shown in Figure 25 is an RC oscil-
lator whose feedback network forms two arms of a Wien
bridge. An even number of stages is required in the amplifier
so that there’s no phase shift between the input and output
voltages. The RC network forms an AC voltage divider; thus,
a fraction of the output is fed back to the input of the am-
plifier. Because of the even number of stages, the output
voltage is in phase with the input voltage. Connecting the
bridge to the input of the amplifier produces positive feed-
back at one frequency. Oscillations at that frequency are
sustained when the gain exceeds three.

FIGURE 25—Voltages A
and B are in phase at only AMPLIFIER
one frequency: f ⴝ 1⁄2␲CR.
Connecting the RC arms
to an amplifier with an
even number of stages
forms the Wien bridge
R
oscillator.
C
A

C R B

Oscillators 41
 A practical Wien bridge oscillator circuit with a variable fre-
quency is shown in Figure 26. You may recognize the first
stage as a Darlington pair. Its current gain and input imped-
ance ensure that the lower half of the Wien bridge isn’t made
unstable by the load. Negative feedback keeps the voltage
amplification independent of frequency. A high initial gain
makes up for the fall that occurs when the negative feedback
is applied by R1 and R2. The low output impedance obtained
from the use of negative feedback minimizes the loading
effect of the Wien bridge. A fraction of the output voltage is
fed back in series with the emitter of the first stage.

+ 28 V

680

3.3 k 680

50

1 .1 .01 1000 100

pF pF
6.8 k Q3 R1
Q1
t°
10 k
Q2 50 k
10 k

OUTPUT
680 R2 1.2 k
1 .1 .01 1000 100 220 50
pF pF

FIGURE 26—For the capacitor values given in this circuit, the frequency ranges covered are from 15 Hz to
200 Hz, from 150 Hz to 2 kHz, from 1.5 kHz to 20 kHz, from 15 kHz to 200 kHz, and from 150 kHz to 2
MHz. The output voltage is about 1 volt RMS (root-mean-square). For a change in supply voltage of 4 V,
the change in output voltage is less than 1 percent, and the change in frequency is less than 2 percent.

In the circuit, a thermistor (R1) provides amplitude control.

(Note that the thermistor symbol is distinguished by a tem-
perature symbol—t°—under the standard resistor symbol.
It’s also common to see only the letter T used with the resis-
tor symbol.) Any rise in output voltage causes an increase in
the current through R1 and R2. The inevitable rise in tem-

42 Oscillators
perature decreases the resistance of the thermistor. The
feedback fraction thus increases, automatically reducing the
gain. This reduces the output of the amplifier to almost its
former value.
Decreases in voltage output, on the other hand, cause a rise
in the resistance of the thermistor. They also cause a
decrease in the fraction of the voltage fed back to the emitter.
A rise in gain results, which returns the amplitude of the
output voltage almost to its former value.

Twin-T Oscillators
The oscillator circuit shown in Figure 27 is called a twin-T
circuit (the “T” stands for “transistor”). However, the circuit
actually has only one transistor amplifier (Q1) located in the
middle of the circuit. The second transistor (Q2) is used only
as a buffer.

R1 A

R2
C1
R5
Q2 C3 C6
Q1
C2
C4
R6 C5
R3
R4

FIGURE 27—A twin-T oscillator circuit is shown here.

In the circuit, transistor Q1 uses a form of basic bias sup-

plied by R2. The load resistor for the amplifier is R1. Observe
that the bias resistor is on the transistor side of the load
resistor. That causes some degenerative feedback for the
amplifier, to protect it from a thermal runaway. If the collec-
tor current started to rise (as with a thermal runaway), the
increased voltage across the load resistor would make the
voltage at point A less positive. This means that the base of

Oscillators 43
 the transistor would be less positive and the rise in the col-
lector would decrease, preventing a thermal runaway. Despite
the degenerative feedback, the amplifier has a relatively high
gain because its emitter is directly connected to common.
In Figure 27, point A is the summing point at which the volt-
age is 180 degrees out of phase with the input voltage. When
this voltage is fed back to the base of the common-emitter
(Q1), the collector voltage will shift another 180 degrees.
Therefore, it will be the same as the input voltage. This is the
way proper phase connections are accomplished in the twin-T
oscillator.
Transistor Q2 serves as a buffer to prevent oscillator loading.
Oscillator loading occurs when its output circuit drags the
oscillator output off frequency. It occurs when the output cir-
cuit has an inductance and/or a capacitance that affects the
output frequency. The buffer circuit acts as a shield to pre-
vent the effect of loading. Transistor Q2 is called an
emitter-follower. Its output signal is delivered to Q1 through
coupling capacitor C2.
Now, let’s examine the frequency-determining network at the
output of amplifier Q1. The output circuit of Q1 is made with
two back-to-back T-filters, which is where it gets the name
“twin-T” oscillator. Figure 28A shows the T-filter configura-
tion, and Figure 28B shows a -filter configuration. The
T-filter and the -filter are both popular configurations for
filter circuits.
The two T-filters that form the output circuit of Q1 are shown
in Figure 28C and Figure 28D. The T-filter circuit shown in
Figure 28C is called a high-pass filter, because low frequen-
cies have a difficult time passing through the two capacitors,
but high frequencies can easily pass through C3 and C4. The
T-filter circuit shown in Figure 28D is called a low-pass
filter, because the high frequencies are shorted to common
through C5, but the low frequencies aren’t passed to com-
mon by that path. The characteristic curve for the high-pass
T-filter is shown in Figure 28E, and the characteristic curve
for the low-pass T-filter is shown in Figure 28F. When the
two T-filters are connected back-to-back to form the output
circuit of Q1, only one frequency (or a very narrow range of
frequencies) can pass through the filter.

44 Oscillators
 FIGURE 28—(A) shows a
X X X T-filter configuration, and
X X X (B) shows a ␲-filter con-
figuration. (C) shows a
high-pass T-filter circuit,
(A) (B)
and (D) shows a low-pass
T-filter circuit. Together,
these two T-filter circuits
C3 C4 R5 R6 form the output circuit of
Q1, previously shown in
C5 Figure 27. (E) shows the
R4 characteristic curve for
the high-pass T-filter cir-
cuit, and (F) shows the
(C) (D) characteristic curve for
the low-pass T-filter cir-
cuit. (G) shows the
combined characteristic
curves, with the passed
frequency marked with a
broken line.

(E) (F)

g f

FREQUENCY
(G)

Figure 28G shows the combined characteristic curves of

Figures 28E and 28F, with the passed frequency marked
with a broken line. The single frequency that passes goes to
the input of the buffer transistor (Q2), and then to the ampli-
fier. The overall result is regenerative feedback at only one
frequency.
To prevent one phase-shift leg of the twin-T from loading the
other, it’s best to follow this practice: Make the resistance of
R5 at least 10 times larger than the resistance of R6. Also,
make the capacitance of C3 at least 10 times larger than the
capacitance of C4. If these conditions are met, the frequency
of oscillation can be calculated using the following equation:

1 1
fO 
2 R1R 2C1C 2

Oscillators 45
 In this equation, fO stands for the oscillation frequency in hertz,
C stands for the capacitance in farads, R stands for the resist-
ance in ohms, and  is a constant that’s equal to about 3.14.
Since the best results are obtained when R1 equals 10R2 and
C1 equals 10C2, the equation can be expressed as

1
fO 
20R 2C 2

Now, let’s look at some example problems that demonstrate

the use of the twin-T oscillation equations.
Example: In Figure 28, assume that the larger resistance
in the low-pass circuit is 15 k and the larger
capacitance in the low-pass circuit is 0.1 F.
Also, assume that the resistance in the high-
pass circuit is 1.5 k and the capacitance in
the high-pass circuit is 0.01 F. Determine the
frequency of oscillation.
Solution: To solve this problem, first convert the compo-
nent values to ohms and farads.
R1  15 k or 15,000 
C1  0.1 F, or 0.0000001 F
R2  1.5 k or 1,500 
C2  0.01 F, or 0.00000001 F
Then, substitute the values into the formula and solve.

1
fO 
20R 2C 2

1
fO 
20(3.14)(1, 500 )(0.00000001 F )

1
fO 
20  3.14  1, 500   0.00000001 F

1
fO 
62.8  1, 500  0.00000001

1
fO 
94, 200  0.00000001

46 Oscillators
 1
fO 
0.000942

fO  1,062 Hz
Example: In a circuit similar to the one shown in Figure 27,
you’re looking for 1,000 Hz frequency of oscilla-
tion. If the capacitance of C2 is 0.01 F,
approximately what resistance is needed for
resistor R2?
Solution: Since the capacitance is given in microfarads,
first convert this value to farads.
0.01 F  0.00000001 F
You can make this problem easier to solve by transposing the
values of fO and R2 in the formula. This will give you the fol-
lowing variation of the formula:

1
R2 
20FOC 2

Then, substitute the values into the formula and solve.

1
R2 
20FOC 2

1
R2 
20(3.14)(1, 000 Hz )(0.00000001 F )

1
R2 
20  3.14  1, 000 Hz  0.00000001 F

1
R2 
62.8  1, 000  0.00000001

1
R2 
62, 000  0.00000001

1
R2 
0.000628

R2  1592.4  or 1.6 k

Note: Your answers may vary slightly from ours, due to

rounding off and variations in calculator manipulation.

Oscillators 47
 Unijunction Transistor Oscillators
A typical circuit of a unijunction transistor (UJT) oscillator is
shown in Figure 29. The waveforms that are generated by
this oscillator somewhat resemble the teeth of a saw; conse-
quently, it’s often referred to as a sawtooth oscillator.

FIGURE 29—A UJT oscilla-

tor may be regarded as a
S
voltage-controlled oscilla-
tor, since its current is
controlled by the junction
RB2 R
voltage. +
B2 E
VBB

B1 OUTPUT
RB1 C

CAPACITOR
DISCHARGE PATH

A UJT oscillator is a type of relaxation oscillator. Relaxation

oscillators produce repetitive, nonsinusoidal waveforms (such
as square, triangular, or sawtooth waves) by gradually charg-
ing and quickly discharging a capacitor or an inductor through
a resistor. The term “relaxation” is used because during the
generation of the waveform, there’s a sharp transition from one
state to another, followed by relative quiet. This cycle—a sharp
transition followed by relative quiet—is then repeated.
You can compare a relaxation oscillator to the human heart.
As the heart beats, there is a period when the blood is
pumped through the heart’s chambers and out into the
arteries. This is then followed by a brief interval in which the
heart muscles relax momentarily before the next pump. This
alternating pumping and relaxing activity forms the pulse.

Calculating a UJT Time Constant

The frequency-determining components in Figure 29 are the
capacitor (C) and the resistor (R). In general, the smaller the
values of C and R, the higher the frequency of the output
waveform. The RC time constant graph shown in Figure 30

48 Oscillators
 100
90
80
FULL CHARGE
70
PERCENT OF
63 60
50
40
30
20
10
0
1 2 3 4 5
NUMBER OF TIME CONSTANTS

FIGURE 30—As the curve shows, it takes a capacitor one time constant to
charge to approximately 63 percent of full charge.

demonstrates the manner in which the voltage builds up

across a capacitor in series with a resistor. You can calcu-
late the capacitor-charge time constant using the following
equation:
RC
In this equation,  stands for time constant in seconds, R
stands for resistance in ohms, and C stands for capacitance
in farads.
As an example of the use of the time-constant equation, let’s
look at the circuit shown in Figure 29. Assume that resistor
R has a value of 1.5 M, and that capacitor C has a value of
0.2 F. You can then determine the value of the time con-
stant by substituting these values into the time constant
formula.
Note that the megohm and microfarad values must first be
converted to ohms and farads as follows:
1.5 M  1,500,000 
0.2 F  0.0000002 F
Then, substitute these values into the formula:
RC
  1,500,000   0.0000002 F
  0.3 second

Oscillators 49
 Practically speaking, it takes about five time constants for a
capacitor to become fully charged. So, in our example, it
would take 5 times 0.3 second, or 1.5 seconds, for the RC
combination to become fully charged.
In actual practice, the values selected for R and C are deter-
mined by UJT characteristics. Typically, R will range in value
from about 2 k to 2 M. To ensure frequency stability, the
value of C must be greater than 0.05 F. The values of R and
C are also selected based on the supply voltage and the UJT
to be used, in order to yield a desired frequency of oscillation.
Usually, the time constant of the RC combination will be
much shorter than the time period of oscillation.
The time period is the reciprocal of the frequency and can be
approximated by using the following formula:

1
T
f

In this equation, T stands for the time period of oscillation in

seconds, and f stands for the frequency of oscillation in hertz.
Now, let’s look at an example where an output frequency of
1,000 Hz is required. The time period of oscillation is calcu-
lated by substituting the appropriate values into the formula
as follows:

1
T
f

1
T
1, 000 Hz

T  0.001 second, or 0.1 millisecond

T

5

0.001 second

5

  0.0002 second, or 0.2 millisecond

To achieve this time period, the RC combination must be
such that its time constant is no greater than 0.2 millisec-
ond. The time constant calculated using the values of the

50 Oscillators
previous example was much too high; therefore, either one or
both values must be decreased.

Negative Resistance
There’s one important point that you must understand in
order to fully understand the UJT oscillator. This is the
concept of negative resistance—that tendency of certain
materials and circuits to offer less resistance to current as
the applied voltages decrease. That is, current increases as
applied voltage decreases, and vice versa. You may have
noticed that this concept is contrary to Ohm’s law. Ohm’s
law, you’ll recall, states that current is directly proportional
to the applied voltage and inversely proportional to the resist-
ance of a device or circuit.
A typical UJT emitter characteristic curve that displays this
negative resistance is shown in Figure 31. This curve is the
emitter VI (voltage-current) characteristic. It shows the rela-
tionship between emitter current (IE ) and emitter voltage (VE ).
The two important points on this curve are the peak point
and the valley point. The peak point is the transition from
cutoff to conduction; the valley point is the change from neg-
ative resistance to saturation conduction. The peak-point
emitter current (IP ) is the minimum emitter current that’s

VE
PEAK
POINT
VP

NEGATIVE
RESISTANCE
NEGATIVE
LEAKAGE
THROUGH SATURATION
EMITTER REGION
DIODE VV

VALLEY
POINT

IP IV IE

FIGURE 31—The negative-resistance characteristic of the UJT makes it

well-suited for use as an oscillator.

Oscillators 51
 required to turn on the UJT (that is, cause it to conduct).
Then, the current IE increases as VE decreases, thus exhibit-
ing the characteristic of negative resistance.
The end of the negative-resistance region shown in Figure 31
is the valley point. The current and voltage at that point are
referred to as the valley current (IV) and the valley voltage
(VV ). Both of these values (as well as IP and VP ) can be con-
trolled over a small range by changing VBB (the interbase
voltage), or by adding resistance in series with B1 and B2, as
shown in Figure 29. The advantage of low valley current and
power is that the oscillator will have low power dissipation
when conducting or in the ON condition.

Operation of the UJT Oscillator

When the switch in the circuit shown in Figure 29 is closed,
capacitor C begins to charge through resistor R to the level
of the supply voltage (VBB ). The voltage appearing across the
capacitor is emitter voltage VE as shown in Figure 32. Until
the value of VE equals the value of VP (the UJT peak-point
voltage), the UJT is cut off, or reverse-biased. (You can see
this even more graphically in Figure 31.) When the value of
VE equals the value of VP, the PN junction becomes forward-
biased. The resistivity between the emitter and base B1
decreases rapidly, providing a low-resistance discharge path
for the capacitor through RB1 and base B1. This action consti-
tutes positive feedback. With the increase in conductivity, VE
decreases as IE increases—and this is negative resistance.

FIGURE 32—Note that the

trailing edge of the curve
decays at a much faster
rate than the leading
VBB
edge rises. The result is C CHARGE C DISCHARGE
the sawtooth waveform VP
that’s typically produced VE
at the emitter of a UJT
oscillator. VV

TIME

The capacitor then attempts to discharge to zero. However,

when the value of VE is approximately equal to the value of
VV (the UJT valley-point voltage), the number of holes

52 Oscillators
injected into the base decreases while the voltage drop
between the emitter and base B1 increases. This voltage drop
causes base B1 to become more positive with respect to the
emitter. Thus, the PN junction that’s nearest to base B1
becomes reverse-biased. As more of this junction is cut off,
the injection of holes declines rapidly until the junction is cut
off completely. The UJT is restored to its original condition,
thus completing one cycle of operation. The capacitor then
starts to charge once more, and a second cycle is initiated.
Both the leading and trailing edges of the VE waveform shown
in Figure 32 are exponential curves. The time constant of the
capacitor-discharge path through B1 (RB1 times C) shown in
Figure 29 is normally much shorter than the time constant of
the capacitor-charge path RC.

Frequency Synthesizers
A basic Pierce crystal oscillator is shown in Figure 33.
Generally equivalent to the Colpitts oscillator, the Pierce
crystal oscillator operates in a parallel-resonant mode and
requires no tuning. The crystal, which is connected between
the base and collector of the transistor, acts as the tuned
circuit.

FIGURE 33—The Pierce

+V crystal oscillator may be
modified to use several
crystals to provide a
R2 multifrequency output.
C3
R1
OUTPUT
C1
CRYSTAL
C2

It’s possible to synthesize or combine stable frequencies by

using two or more crystal oscillators. As an example of fre-
quency synthesis, consider a system that consists of two
Pierce oscillators, a mixer circuit (diode or transistor), and two
bandpass filters. One oscillator is a three-crystal, 50 kHz-per-
step circuit with frequencies of from 3.0 MHz to 3.10 Mhz.

Oscillators 53
 The other oscillator is a five-crystal, 10-kHz-per-step compo-
nent circuit with frequencies of from 7.0 MHz to 7.04 MHz.
Both oscillators feed into the mixer circuit. The output of the
mixer will be four frequencies: the original two, the sum of
the two, and the difference between the two frequencies. The
bandpass filter frequencies are from 3 MHz to 5 MHz and
from 9 MHz to 11 MHz, respectively.
If the 10-kHz-per-step crystals are switched into their Pierce
circuit along with the 3.0 MHz crystal in the second oscilla-
tor, and if the signals from the oscillators are passed through
the mixer, then the output frequency from the 9 MHz to 11
MHz bandpass filter will be from 10.000 MHz to 10.040 MHz
(in 10 kHz steps). By switching in the 3.05 MHz crystal
instead of the 3.0 MHz crystal, the output frequency will
change to 10.050 MHz to 10.090 MHz; with the 3.10 MHz
crystal, the output signals will range from 10.100 MHz to
10.140 MHz. Thus, with six crystals (five crystals in the
10-kHz-per-step oscillator plus the 3.0 MHz crystal in the
50-kHz-per-step oscillator) five output signals can be synthe-
sized (10.000 MHz to 10.040 MHz). In this example, eight
crystals are available, so the seventh crystal (the 3.05 MHz)
will allow synthesis of ten frequencies (in 10 kHz steps) from
10.000 MHz to 10.090 MHz. The eighth crystal will allow
synthesis of fifteen frequencies, from 10.000 MHz to
10.140 MHz.
If the output from the mixer is switched to the low-bandpass
filter (3 MHz to 5 MHz), the difference of the two signals will
be synthesized. The fifteen output frequencies will then range
from 4.000 MHz to 4.140 MHz.

Phase-Locked Loops
Another important oscillator circuit is the phase-locked loop,
commonly referred to by its abbreviation PLL. This type of cir-
cuit is used to synthesize highly stable RF alternating current
for applications in such devices as amplitude modulation (AM)
detectors, horizontal-sweep automatic-frequency controls
(AFCs), and frequency modulation (FM) modulators.
The basic PLL system, shown in Figure 34, consists of a
voltage-controlled oscillator (VCO), a crystal reference oscilla-
tor, a phase detector or comparator circuit, a low-pass filter,

54 Oscillators
 VCO
1 MHz
OUTPUT
DC ERROR SIGNAL

REFERENCE PHASE
OSCILLATOR DETECTOR

DC LOW-PASS DC
AMPLIFIER FILTER SIGNAL

FIGURE 34—A basic PLL circuit is shown here. The phase detector com-
pares the VCO and reference signals and then generates an error signal.
Phase-locked loops like this one are available as integrated circuits.

and a DC amplifier. The frequency of the VCO can be varied

by varying the DC voltage applied, but the crystal oscillator
has a fixed frequency.
Operation of the basic PLL consists of feeding the AC signals
from the VCO and the reference oscillator to the phase detec-
tor. When the frequency and phase of the signals from the
VCO and the reference oscillator are equal, the DC output
voltage from the phase detector will have a particular value.
This DC voltage is fed through the low-pass filter to the DC
amplifier, and then to the voltage input circuit of the VCO. If
the VCO then tries to shift off frequency or phase, the phase
detector generates a DC “error” voltage that corrects the VCO
output frequency, returning it to the exact value of the refer-
ence frequency. Thus, the VCO frequency is “locked in” not
only with the frequency of the reference oscillator alternating
current, but also to the same phase.
A typical PLL synthesizer circuit consists of the basic PLL cir-
cuit with the addition of two transistor-transistor logic (TTL)
divider circuits, a mixer, and a second low-pass filter. One
TTL circuit divides a fixed number of times; the other is a
variable TTL. The second low-pass filter feeds AC signals from
the mixer to the variable TTL. The mixer adds or subtracts
AC signals from the VCO and the reference oscillator.

Oscillators 55
 In PLL synthesizer operation, the number of times that the
variable TTL divides the AC frequency is determined by the
amount of DC voltage applied to the control terminals of the
variable TTL. (The amount of voltage supplied to the control
terminals is determined, in turn, by a switching circuit.)
Output signals from the variable TTL and the fixed TTL are
fed to the phase detector, then through the low-pass filter
and DC amplifier to the VCO. Finally, the output signal from
the VCO is passed through the mixer and the low-pass filter
to the variable TTL, thus locking in the desired frequency
and phase.
Now, take a few moments to review what you’ve learned by
completing Power Check 3.

56 Oscillators
 Power Check 3
Questions 1–8: Indicate whether each statement is True or False.

______ 1. One of the main advantages of the phase-shift oscillator is its economy.

______ 2. A UJT oscillator works very well at low frequencies and has a sinewave output.

______ 3. A high-pass T-filter circuit rejects low frequencies and passes high frequencies.

______ 4. In a Wien bridge circuit, amplitude control can be provided by a transformer.

______ 5. In an RC phase-shift oscillator, a decrease in the value of either R or C will produce a

decrease in the output frequency.

______ 6. The leading and trailing edges of a UJT sawtooth waveform are exponential curves.

______ 7. A PLL can be used as a frequency synthesizer.

______ 8. When a high-pass T-filter circuit and a low-pass T-filter circuit are connected back-to-
back to form the output circuit for a transistor, only one frequency (or a very narrow
range of frequencies) will be able to pass through the filter.

(Continued)

Oscillators 57
 Power Check 3
Questions 9–18: Fill in the blanks in each statement.

9. Approximately time constants are required for a capacitor to become fully charged.

10. In a UJT emitter characteristic curve, the point is the transition from cutoff to con-
duction, and the point is the change from negative resistance to saturation
conduction.

11. A type of oscillator that produces repetitive, nonsinusoidal waveforms by gradually charging
and quickly discharging a capacitor or an inductor through a resistor is called a
oscillator.

12. A basic PLL system consists of a oscillator, a crystal reference oscillator, a phase
detector or comparator circuit, a low-pass filter, and a DC amplifier.

13. If a capacitor has a time constant of 0.2 second and requires five time constants to become
fully charged, it will require a total time of to charge completely.

14. In a oscillator, a crystal that’s connected between the base and collector of the tran-
sistor acts as the tuned circuit. The oscillator circuit operates in a parallel-resonant mode and
requires no tuning.

15. The tendency of certain circuits to offer less resistance to current as the applied voltages
decrease is called .

16. A technique that’s used to combine two stable frequencies to get a different output is called
frequency .

17. In order to produce variable, low-frequency oscillators, circuit designers prefer to use
component combinations.

18. A oscillator circuit is used to synthesize highly stable RF alternating current for
applications in AM detectors, AFCs, and FM modulators.

Check your answers with those on page 60.

58 Oscillators
 Power Check Answers 1
1. False
2. True

Answers
3. False
4. False
5. True
6. True
7. power
8. transistors
9. resonant frequency
10. inductors, capacitors
11. heat
12. positive, regenerative
13. resonant circuit, amplifier, feedback system, power supply
14. Q

Power Check Answers 2

1. frequency multiplier
2. thermal runaway
3. tank circuit
4. tapped
5. count-down, divide-by
6. regenerative amplifier
7. tickler coil
8. power supply
9. True
10. True
11. False
12. True
13. True
14. False

59
 Power Check Answers 3
1. True
2. False
3. True
4. False
5. False
6. True
7. True
8. True
9. five
10. peak, valley
11. relaxation
12. voltage-controlled
13. 1 second
14. Pierce crystal
15. negative resistance
16. synthesizing
17. RC
18. phase-locked loop

60 Power Check Answers