1

3. Waveform Generators and

Nonlinear Circuits
In this section, we consider a class of circuits that involves nonlinear operation of the
operational amplifier in order to produce various non-sinusoidal waveforms. These circuits are
useful in a wide range of applications, particularly in instrumentation and measurement.

3.1 Zero-Crossing Detectors

Zero-crossing detectors give an indication of the time a signal passes through zero. The
high open loop gain of the operational amplifier makes this circuit element very suitable for this
implementation.

Non-inverting Zero-Crossing Detector

The op amp in Fig. 1 (a) operates as a comparator. Its non-inverting input compares
voltages Vi with a reference voltage of 0V. (Vref = 0V). When Vi > Vref, Vo equals the positive
saturation voltage +Vsat of the op amp. This is because the voltage at the (+) input is more
positive than the voltage at the inverting (-) input and hence the differential voltage VD = Vi - Vref
is positive. Consequently Vo is positive. The converse occurs when Vi < Vref. Thus, the polarity
of Vo indicates if Vi is above or below Vref. The transition of Vo tells when Vi crosses the
reference and in what direction. For example, when Vo makes a positive transition from -Vsat to
Vsat it indicates that Vi just crossed 0 in the positive direction.

Inverting Zero-Crossing Detector

In the op amp of Fig. 1(b) the inverting input compares Vi with a reference voltage of
0V. (Vref = 0V). This circuit is an inverting zero-crossing detector. Thus, if Vi > Vref, Vo = - Vsat.
When Vi crosses the reference-going positive, Vo makes a negative-going transition from +Vsat to
-Vsat.
 2

Figure 1. Zero-Crossing Detectors

Positive and Negative Voltage-level Detectors

In Fig. 2, a positive reference voltage Vref is applied to one of the op amp's inputs. This
means that the op amp is set up as a comparator to detect a positive voltage. If the voltage to be
sensed, Vi, is applied to the op amp's non-inverting input, the result is a non-inverting positive-
level detector. When Vi is above Vref, VO equals +Vsat. When Vi is below Vref, VO equals -Vsat. If
Vi is applied to the inverting input, the circuit is an inverting positive-level detector. For this
circuit when Vi is above Vref, VO equals -Vsat. If a negative reference voltage -Vref is used
instead, then negative voltage detectors result as shown in Fig. 3.
 3

Figure 2. Positive Voltage Level Detectors

 4

Figure 3. Negative Voltage Level Detectors

Precision Rectifiers
The major limitation of ordinary silicon diodes is that they cannot rectify voltages below
0.7V. A circuit that acts like an ideal diode can be designed using an op amp and two ordinary
diodes. The result is a powerful circuit capable of rectifying input signals of only a few
millivolts. These circuits can be grouped loosely into the following classifications: Linear half-
wave rectifiers and precision full-wave rectifiers.
The Linear half-ware rectifier circuit delivers an output that depends on the magnitude
and polarity of the input voltage. The output is inverted with respect to the input. The circuit is
also called a precision half-wave rectifier and acts as an ideal diode. The precision full-wave
rectifier circuit delivers an output proportional to the magnitude but not the polarity of the input.
This circuit is also called an absolute value circuit.
 5

Linear Half-Wave Rectifier

Linear half-wave rectifier circuits transmit only one-half cycle of a signal and eliminate
the other by holding the output at zero volts. The input half-cycle that is transmitted can be either
inverted or non-inverted. The inverting amplifier is converted into an ideal (linear precision)
half-wave rectifier by adding two diodes as shown in Fig.4. When Vi is positive, (positive half-
cycle), diode Dl conducts causing the op amp's output voltage, VOA , to go negative by one diode

Rf=R

Vi R

D1
I D2 Vo
VoA
RL

Figure 4. Inverting Linear Half-Wave Rectifier: Positive Output

drop (=0.7V). This forces D2 to be reverse-biased. The circuit's output voltage Vo equals zero
because input current I flowing into R given by 1 = Vi/R flows through D1. The load RL must
always be resistive. If it is a capacitor, inductor, voltage, or current source, then Vo will not be
zero. When Vi goes negative, this forces the op amp output VOA to go positive and D2 conducts.
The circuit then acts like an inverter since Rf = Ri and Vo = -(Vi) = Vi. Since the (-) input is at
ground potential, diode Di is reverse biased. Input current is Vi/Ri and gain is Rf/Ri. The diode
turn-on voltage of D2 is virtually eliminated by placing it in the op amp feedback loop. VOA goes
to whatever voltage is required to turn on the diode. Thus millivolts of input voltage can be
rectified.
If both diodes are reversed, the circuit operation is essentially the same except that only
positive input signals are transmitted and inverted. VO equals zero for all negative inputs.

Signal Polarity Separator

 6

R Vo2

Vi R D2

D1
Vo1

Figure 5. Signal Polarity Separator

This circuit (Fig. 5) is an expansion of the previous one. When Vi is positive, diode DI conducts
and an output is obtained only at V0l, V02 = 0. When Vi is negative, D2 conducts, V02 = Vi and VO
= 0.

Precision Rectifiers : The Absolute Value Circuit

The precision full-wave rectifier transmits one polarity of the input signal and inverts the
other. Thus both half-cycles of an alternating voltage are transmitted but are converted to a single
polarity at the circuits output. The precision full-wave rectifier can rectify input voltages with
millivolt ampltudes.

R R R

Vi R Dp
A Vo
B
Dn
R

Figure 6 Absolute Value Circuit

The first type of precision full-wave rectifier or absolute-value circuit is shown in Fig.6. This
circuit, consists of two op amps and equal resistors and has an input resistance equal to R. For Vi
 7

positive, diode Dp conducts so that both op amps A and B act as inverters and Vo = +Vi. For
2
negative input voltages, diode DN conducts and the input current I is made up of I from op
3
1
amp A and I from op-amp B which acts as an inverter. Thus output voltage Vo is positive for
3
either polarity of input Vi and Vo is equal to the absolute value of Vi. To obtain negative outputs,
for either polarity of Vi, simply reverse the diodes.

High-impedance precision full-wave rectifier

R2=R R3=R R4=2R

R1=R Dp Dn
Vo
A B
Vi

Figure 7. High-Impedance Full-Wave Rectifier

This second type of precision rectifier is connected to the non-inverting op amp inputs to obtain
high input impedance. For Vi positive, Dp conducts and the inverting inputs of both op amps are
at Vi. Hence no current flows through R2,R3 and R4. DN is off. Therefore, VO = Vi for all
positive input voltages. When Vi goes negative, Vi and Ri set the current through both R 1 and R2
at -Vi/R. Dp is off and DN is on. The potential at a is then -2Vi. Since that at b is Vi the current
through R3 into DN is Vi/R. The current through R4 is therefore Vi/R and the voltage drop across
R4 is Vi/R.2R = 2Vi. This gives Vo = +Vi. This VO is always positive despite the polarity of Vi.

AC to DC Converter
This type of-circuit computes the average value of a rectified AC voltage. It utilises a full-wave
rectifier (absolute-value circuit) and averaging circuit and is sometimes referred to as a mean-
absolute-value circuit.
 8

R C

Vi R
R R/2 R
Dn Dp
Vo
A B

Figure 8. AC to DC Converter

To construct an AC to DC Converter, we begin with a precision rectifier. For positive

inputs, op amp A inverts Vi. Op amp B sums the output of A and Vi to give VO = Vi. For
negative inputs op amp B inverts -Vi and the current output Vo is +Vi. Note that the potential at
a is 0 Volts and therefore no current flows in R2. Thus the circuit output VO is positive and
equal to the rectified or absolute value of the input. If a large-value low leakage capacitor C is
added across the feedback resistor in op amp B, the resulting circuit produces the mean absolute
value of the rectified output of op amp B. It takes about 60 cycles of input voltage before the
capacitor voltage settles down to its final reading. If the sinusoidal, triangular and square
waveforms are applied to this circuit, its output will be the mean absolute value of each input
signal.
 9

Waveform Generators
Astable Multivibrator

Rf

Vo
C
Vc R1

R2

Figure 9 Astable Multivibrator

A free-running or astable multivibrator is a square wave generator. Resistors Rl and R2 form a
voltage divider to feedback a fraction of the output to the non-inverting input. When V0 is +Vsat,
the feedback voltage is called the upper-threshold voltage VuT given by

VUT 
R2
 Vsat  (1)
R1  R2
Resistor Rf provides a feedback path to the inverting input. When VO is at +Vsat, current I+ flows
through Rf to change capacitor C toward VUT. As long as Vc< VUT, the output voltage remains at
+Vsat. When Vc changes to a value slightly greater than VUT, the inverting input goes positive
with respect to the non-inverting input and the output switches from +Vsat to -Vsat. The (+) input
is now held negative with respect to ground because the feedback voltage is negative and given
by

VLT 
R2
 Vsat  (2)
R1  R2
Just after V0 switches to -Vsat, the capacitor has an initial voltage equal to VUT. Now current I-
discharges C to 0V and recharges C to VLT. When Vc becomes slightly more negative than VLT,
V0 switches back to +Vsat. This now discharges the capacitor from VLT to 0V and then recharges
 10

it to VUT and the process repeats itself. The output voltage V0 is a square wave whose frequency
is determined by the time to change and discharge C.

Figure 11 Output Waveform of Astable Multivibrator

Resistor R2 is chosen equal to 0.86 R1, to simplify the calculation. Thus,

T = 2RtC for R2 = 0.86 RI (3)

f = 1/T = 1/2RfC (4)

Example
Show why T = 2RfC when R2 = 0.86R1
Solution
The time required for a capacitor C to charge through a resistor Rf from some starting voltage
towards some charging voltage to a stop voltage is expressed generally as
 ch arg e  start 
t  R f C ln  
 ch arg e  stop 
  Vsat   VLT 
 R f C ln  
  Vsat   VUT 
If the magnitudes of +Vsat and -Vsat are equal, then
 11

 Vsat 
R2
 Vsat 
R1  R2  R  2 R2 
t  R f C ln  R f C ln  1 
 Vsat 
R2
 Vsat   R1 
R1  R2
Since ln 2.718  1, the ln term can be reduced to 1 if (R1 + 2R2)/R1 = 2.718 or R2 = 0.86R1
Now tl = RfC and t2 = RfC if R2 = 0.86 R1. Therefore T = tl+t2 = 2RfC.
(DUTY CYCLE = %time V0 high).

Monostable Multivibrator
A Monostable or one-shot Multivibrator generates a single output pulse in response to an
input signal. The length of the output pulse depends only on external components connected
tothe op amp.

D1 D3 RD=10k Vi

C=0.1F Rf=100k t
-Vi
D2 Vo
Vi
Ci=0.01F R1=10k

Ri=10k
R2=1k8

Figure 12. Monostable Multivibrator

In the stable state, V0 is at +Vsat. Voltage divider Rl and R2 feeds back VLTT to the (+) input. Dl
clamps the (-) input at approximately +0.7V. VUT therefore holds V0 at +Vsat. If V I goes negative
to -Vip, a value approximately equal to or greater than VUT, then the voltage at the"(+) input will
be pulled below the voltage of the (-) input. Once the (+) input becomes negative with respect to
 12

the (-) input, V0 switches to -Vsat. With this change, the one-shot is now in its timing state. For
best results, the time constant RiCi should be 1/10 or less than the time constant RfC.
The timing state is an unstable state since the one-shot only remains therefore for a short
period. Resistors R1 and R2 feedback a negative voltage (VLT = -2.2V) to the (+) input. The
diode Dl is now reversed - biased by -Vsat and is essentially an open circuit. Capacitor C
discharges to 0 and then recharges to an opposite polarity. As C recharges, the (-) input becomes
more and more negative with respect to ground. When the capacitor voltage is slightly more
negative than VLT, VO switches to +Vsat. The one-shot has now completed its output pulse and is
back to the stable state. If R2 is made about one-fifth of R1, then the duration of the output pulse
is given by
T = RfC/5 if R2 = R1/5 (5)
The exact equation is
  Vsat   0.7V 
T  R f C ln   (6)
  Vsat   VLT 
For the circuit given
T = (100k x 0.1uF)/5 = 2ms (7)
For test purposes, Vi can be obtained from a square wave or pulse generator. Diode D2 prevents
the one-shot from coming out of the timing state on positive transistors of Vi. If the diodes are
reversed, the one-shot will give a positive output pulse for a positive input signal.
After the timing state is completed, the output returns to +Vsat. However, the circuit is not
ready to be retriggered reliably until C returns to its initial state of 0.7V because it takes time for
C to be discharged from VLT = -2.3V to 0.7V. This time interval is called recovery time. It can
be improved by adding a discharge resistor RD in parallel with RF (through a diode DD).
Typically, if RD = 0.1Rf, recovery time is reduced by one-tenth. Diode DD prevents RD from
affecting the timing cycle T.

Bistable Multivibrator
The circuit shown in Fig. 13 is a simple bistable Multivibrator. It has two stable states
with the amplifier output at positive or negative saturation. It is held in one or other of those state
by the positive feedback applied via R1 and R2. A triggering pulse of suitable polarity applied to
the inverting input causes the circuit to regeneratively switch states.
 13

Ci
Vi Vo

R1
Ri

R2

Figure 13 Monostable Multivibrator

Ramp Generator
An inverting op amp with a capacitor connected in the feedback loop is classified as an
INTEGRATOR.

C=1F
+ -
I Vc
Vo
Vi Ri=1M
1V

Figure 14 Integrator

In the circuit shown, the current I = Vi/R flows into the capacitor C thereby charging it. For a
capacitor Ic = C.dV/dt and therefore
dVc I c Vi / R
  (8)
dt C C
Vi V
 Vc   dt  i t (Vc (0)  0)
RC RC
For the circuit in Fig. 14,
Vo  Vc  (Vi / RC )t (9)
which is a ramp. Note that the load current is furnished from the op amp’s output terminal and not
 14

the capacitor and that Vo goes negative only to –Vsat. Even when Vi=0, Vo ramps due to bias
currents. To prevent this, short circuit C until the circuit is operated.

Triangular Wave Generator

C=1F
nR=20k
Ri =1M
R=10k
Vcomp
Vramp

Figure 15 Triangular Wave Generator

The circuit in Fig. 15 combines a symmetrical comparator with an integrator. The output of the
compactor is a square wave and this provides the input to the integrator (Fig 16).

Figure 16 Output Waveform

When Vramp crosses VUT, the comparator output snaps positive to +Vsat since then the voltage at
the (+) input of the comparator will be just positive. This causes the ramp generators' output to
ramp down until Vramp drops just below VLT at time C. Then Vramp snaps negative, causing the
ramp generator to ramp back up to VUT. This completes one cycle of the triangular wave. The
design equations are
 15

VLT = -(+Vsat /n) (10)

VUT = -(-Vsat /n) (11)
If the magnitudes of +Vsat and -Vsat are equal, the frequency of oscillation is
f = n/4RiC (12)
For the circuit, n=2
VLT = -15/2 = -7.5V
VUT = -(-15/2) = + 7.5V
f = 2/(4 x 106 x 1 x 10-6) = 0.5Hz
Note that if n is made equal to 4, then f = 1/RiC.