Principles of Analog Electronics

499

Amplifiers: Cascading Stages

A very weak signal cannot be sufciently amplied by a single stage amplier, like

those we have dealt with so far—ampliers made only of a single biased transistor

forming a complete network. Single-stage transistor ampliers can be inadequate

for meeting most design requirements for any of the four amplier types (voltage,

current, transresistance, and transconductance). If a microphone produces about

20 mV under normal conditions, you will need a single stage of amplication, but

if it produces only 1 mV, you will need two or more stages of amplication, that is,

a multistage amplier, to provide greater amplication than a single stage could

provide by itself.

In the previous chapters, we dealt with congurations with two transistors,

such as the Darlington pair (Section 13.1.1 in Chapter 13), current source

(Section 13.1.2), cascode ampliers (Section 13.3), and differential ampliers

(Section 13.4), which can be considered “similar-multistage,” since you cannot

“isolate” a single transistor, with its own bias network, forming a complete isolated

amplier. Real multistage can be considered more properly as a cascade of single

stages.

We treated single stages made of a single BJT congured as C.E. or C.C. or C.B.,

and single stages made of a single MOSFET congured as C.S. or C.D. or C.G.

When considering a multistage amplier, the choice of the number of stages to

be cascaded, and the choice of the type of transistors to be used and their relative

conguration, strictly depends on the application, so no “golden rule” exists. But

we can take into account the requested overall gain and the matching conditions

that we need to satisfy. Useful rules of thumb can come from the general informa-

tion furnished in Tables 11.6 and 11.7 of Chapter 11, and from the observation

14

K18911_Book.indb 499 27/12/13 6:32 PM

Principles of Analog Electronics

500

that, in general, the MOSFET has a higher input resistance than a BJT, which, in

turn, can provide lower output resistances.

Since a BJT type transistor can be congured as a C.E., C.B., or C.C. topology,

it follows that two stages can be arranged according to nine different combina-

tions: C.E.–C.E., C.E.–C.B. (see the cascode), C.E.–C.C., C.B.–C.E., C.B.–C.B.,

C.B.–C.C., C.C.–C.E., C.C.–C.B. (see the differential amplier), and C.C.–C.C. (see

theDarlington pair). But, of course, some combinations are more useful than oth-

ers. This is because we have to ensure the required overall gain and, thus, the right

resistance (or impedance) matching. So, the n-stage’s output resistance

R

on,

and the

n 1

)

(

+−

stage’s input resistance

R

in

,1+

must be in the proper ratio:

,1 ,

RR

in on

→∞

+

or

0

,1 ,

RR

in on

→

+

, depending on the type of requested gain.

Regarding the rst three combinations with C.E. as the rst stage, we have that

C.E.–C.E. is useful to obtain a high voltage gain, C.E.–C.C. has a good voltage gain

and a low output resistance, and C.E.–C.B. (cascode) has a good voltage gain and

a large bandwidth. But a C.E. amplier cannot drive a low resistance (or imped-

ance) load directly, otherwise it will be overloaded, because of its typical of AC

output resistance value. So C.C. as a second buffer stage between the C.E. ampli-

er and the load prevents the overloading effect.

A useful general reference table is shown in Table 14.1.

The designer sometimes prefers to realize double-stage ampliers using npn-

and pnp-type BJTs. This is because the temperature sensitivity of the amplier

can be greatly reduced. The reason relies on the subtraction of the voltage drifts

with temperature variations of the two BJTs, since each transistor behaves in

opposition, due to the opposite nature of npn versus pnp.

14.1 COUPLING

Each stage can be treated as a two-port network, so the stages can be hooked

up according to different types of interconnections: series, parallel, and cascade

(Section5.5 in Chapter 5). Series and parallel types generally involve non-trivial

mathematical calculations, so the cascade type is usually preferred.

The way the connection between two following stages is realized is called the

coupling method.

The coupling can be realized directly, so that the output of the previous stage

is connected to the input of the following stage without any additional circuitry, or

can be realized by means of ad-hoc circuitry.

Coupling capacitors, inductors, and transformers causes low frequency gain

loss in multistage ampliers. This is for series coupling capacitors, because of their

TABLE 14.1

Characteristics of Some Double-Stage Configurations

Stages Input resistance Output resistance Voltage gain

C.E.–C.E. Medium Medium Very high

C.E.–C.B. Medium High High

C.E.–C.C. Medium Low Medium

C.C.–C.E. High Medium Medium

C.C.–C.C. Very high Very low < 1

K18911_Book.indb 500 27/12/13 6:32 PM