1

Operational amplifier is one of the most common electronic building blocks used by
engineers. It has two input terminals: V+ and V-, and one output terminal Y. It
amplifies X to give Y, i.e. Y = AX.

An ideal op-amp has infinite input impedance Ri, zero output impedance Ro and an
infinite gain A. For an ideal op-amp, for all possible output Y, X is assumed to be
zero because A is infinite.

We can model an op amp as shown here:

X is the voltage difference = (V+ - V-).

2
The output Y is AX, where A is the gain. Since A is very large, X is generally very
small. For example, if the output voltage Y is +5V, A is 105, X is only 50µV. The
scope you have been using at home cannot measure down to this voltage level.

An op-amp is actually very complex inside. However, as a user of op amp, and if you
use it properly, you can simply ignore its internal complexity and treat it more or
less like a perfect amplifier (i..e amplifying the difference voltage, and it is called a
differential amplifier).

Op-amps belong to a type of electronic components known as integrated circuits

(ICs). The packaging is as shown here where pin 1 is always on the left of the notch
in the package and/or indicated with a dot.

3
Before we consider how to build an amplifier with an op-amp, let us consider the
concept of negative feedback. Simply put, when an event causes something to
change, the change itself will counteract the original event.

Shown here is the op-amp using negative feedback. The output Y is connect to the
V- input of the op amp. If output Y falls (the event), it will cause V- to fall (the
change). However, V- falling will increase the difference voltage (V+ - V-). This
causes Y to rise, thus counteracting the initial fall in Y.

The calculation shown here demonstrate that provided A is large, Y = X. Connecting

Y to the V- input ensures that this is always true with this circuit.

Note the golden rule: by feeding the output back to the NEGATIVE input of the op-
amp which has very large gain, the circuit makes sure that (V+ - V-) approaches zero!

4
Op-amp requires power supply for it to work. For the sake of simplicity, we assume
that the op-amp uses dual voltage supply with a +ve and a –ve voltage source as
shown here. We will relax this assumption later.

To conduct analysis on op-amp circuits, we have to make some assumptions as

shown here.

5
Shown here is one of the most commonly used op-amp circuit to provide non-
inverting amplification.

We check through the steps considered in the previous slide. This circuit has
negative feedback. Input voltage X is connected to V+. Therefore the voltage at V- is
also X.

Input current is zero. Therefore X is the voltage divider of Y as shown here.

For such non-inverting op-amp circuit, Gain is always given by:

Y R
Gain = = (1+ 2 )
X R1

For the circuit shown here, the gain is x 4.

6
There is a special case for the non-inverting op-amp circuit. If you make R2 = 0, and
remove R1, then the gain is 1. In fact even if R2 is larger than zero (say 1k), the ratio
R2/R1 is still zero since R1 is infinite.

This circuit is known as voltage follower, or voltage buffer – output Y always follow
input X.
This circuit may first appear pointless – Y is the same as X, why not just use X in the
first place?

The reason why this circuit is useful is because it ISOLATES the output from the
input by presenting a high impedance to the source (Ri is high) and low impedance
to the load, hence behaving like an ideal amplifier, but with gain of 1.

Let us consider the example here. The source voltage U has a resistance of 99kΩ. If
we connect this directly to a 1kΩ load, the output Y = 0.01U (voltage divider
principle). The drop of voltage is due to the loading effect of the 1k resistor on the
source.

When you put the voltage follow between U and the 1k load resistor, the source U
sees the very high input impedance of the op-amp (>10MW), therefore the input X
is effective U. The output resistance of the op-amp is low. The negative feedback
also helps. If the loading effect of the 1k resistor causes Y to drop, this will cause V-
input to drop, and raising Y, thus correcting the loading effect.

7
This is the circuit for an inverting amplifier using an op-amp. Applying KCL at the V-
node gives the following equation:

Y R
Gain = =− ( 2)
X R1

Since the V- node is at the same voltage potential of the V+ node, which is ground (or
earthed), we call V- node in this circuit the virtual earth or virtual ground.

Unlike the non-inverting amplifier case, which MUST have a gain ≥1, inverting
amplifier like this can have any gain, larger or smaller than 1. However, the gain is
ALWAYS negative.

8
With inverting amplifier, since the V- node is virtually zero, you can connect multiple
sources to this node via a resistor and produce a summing amplifier.

Apply KCL at V- node and you see the summing effect immediately. Each voltage is
weighted by the ratio of the feedback resistor RF and the feeding resistor Ri (i.e. the
weighting is - RF/Ri ).

The fact that the V- node is held at a fixed voltage (in this case 0v, but it could have
been a different voltage, as we will see later), the effect of input sources X1, X2 etc
is isolated from each other.

9
We can combine the structure of the inverting AND non-inverting amplifier together
to form this differential amplifier. The best way to understand this is to apply the
principle of superposition.
Consider the effect of X on output Z with Y set to zero. The output due to X alone is:
R2
ZX = − X
R1
Consider the effect of Y on output Z with X set to zero. The output due to Y alone is
more complicated. Firstly, Y is reduced by the voltage divider before reaching the
V+ input.
R4
V+ = Y
R3 + R4

This is amplified according to the non-inverting amplifier gain:

R R4
ZY = (1+ 2 ) × ( )Y
R1 R3 + R4
Now assume that R2 = R4 = 3k, and R1 = R3 = 1k.
R R2 R
ZY = (1+ 2 ) × ( )Y = 2 Y
R1 R1 + R2 R1
Therefore
R2
Z = Z X + ZY = ( ) × (Y − X) = 3× (Y − X)
R1

10
So far, we assume that the op-amp behaves like an ideal amplifier with infinite gain
A, infinite input resistance Ri, and zero output impedance RO.

These assumptions only hold if the resistors we use to construct the circuit are
sensible. So how do we choose these resistor values?

If we use too low a resistor value, the output current required is too much – no
good.

If we use too high a resistor value, the input current into the op-amp is no longer
negligible. Then the gain equation we derived using KCL at the input node is no
longer valid.

In general, we use resistor values in 1kΩ to 100kΩ region.

In the next slide, we will consider the assumption of infinite gain A.

11
For a practical op-amp, the gain is only very high at low frequency. Shown here is
the Gain vs Frequency characteristic of our MCP6002 op-amp.

The gain is over 100,000 at frequency below 10Hz. However, the characteristic is
similar to the of an RC low pass filter. The corner frequency is only 10Hz. It then
falls off at -20dB (or x 0.1) / decade. The gain at 1MHz becomes around 1 (i.e. it
stop behaving as an amplifier).

12
The detail analysis of the impact of this falling gain on bandwidth when we use
negative feedback is beyond the scope of this module. However, it is worth stating
the following observation:

Due to negative feedback, the produce of the overall circuit gain and the effective
bandwidth (i.e maximum frequency for this to behave like an amplifier) is a
constant. This product:

Gain x Bandwidth = gain-bandwidth product = constant.

For MCP6002, the product is around 1MHz, or 106.

The result of this is shown in the plot here.

13
14