N.B.K.

R INSTITUTE OF SCIENCE & TECHNOLOGY

ELECTRONIC CIRCUITS ANALYSIS
UNIT- II FREQUENCY RESPONSE

Syllabus:
➢Low-Frequency Response of the CS and CE Amplifiers.

➢ Internal Capacitive Effects of the MOSFET and the BJT.

➢ The High-Frequency Model of the MOSFET and the BJT.

➢High- Frequency Response of the CE, Emitter follower.

➢High- Frequency Response of the CS, CD.

➢fβ, fT and gain bandwidth product.

1
2.1 Introduction

Fig: 2.1 Sketch of the magnitude of the gain of a discrete-circuit BJT or MOS amplifier versus frequency.

Fig 2.1 represents the variation of gain verses frequency is called frequency response. The frequency
response can be divided into three regions.

1. LOW FREQUENCY:

In low frequency the gain decreases with decreasing frequency due to coupling capacitors and
bypass capacitors.

➢ At low frequency the reactance of coupling capacitor is high, so V0 decreases,

automatically gain also decreases.

➢ At low frequency the reactance of bypass capacitor is high, approximately open circuited.
Then the emitter resistance provides negative feedback for amplifier.

2. MID FREQUENCY :

In mid frequencies the gain is constant or uniform, because the gain also depends on only
resistance values. i.e., in mid frequencies Cc, Ce relate to short circuit. The resistor values does not
change with frequency. So, gain is constant in this frequency range.

3. HIGH FREQUENCY:

In high frequency, the gain decreases with increasing frequency due to internal capacitances in the
BJT and in the MOSFET and stray capacitance. At high frequency, the reactance of stray capacitance is
low, approximately short circuit then V0 decreases, and finally gain also decreases.

2
 Stray capacitance unwanted and unavoidable capacitance exists in every electronic circuit. It is also
called as “inter electrode capacitance”.

Cutoff frequency (also known as corner frequency or break frequency) is defined as a boundary in
a system’s frequency response at which energy flowing through the system begins to be attenuated
(reflected or reduced) rather than passing through. The cutoff frequency of an amplifier is the frequency
at which the amplifier's gain drops by 3 dB from its maximum value. The cutoff frequency of an amplifier
1
is the frequency at which the amplifier's gain is times of maximum gain (Midband gain AM).
√2

Bandwidth = BW = 𝒇𝑯 − 𝒇𝑳

Where fH is Upper cutoff frequency

fL is Lower cutoff frequency

A figure of merit for the amplifier is its gain–bandwidth product, defined as

𝑮𝑩 = |𝑨𝑴 | 𝑩𝑾

2.2 Low-Frequency Response of Common-Source Amplifier

Figure 2.2 shows a discrete-circuit common-source amplifier utilizing the Voltage divider biasing
arrangement. Two coupling capacitors, CC1 and CC2, and a bypass capacitor CS are employed. At midband
frequencies, these large capacitances have negligibly small, and these capacitances can be replaced by
short circuits from fL onwards.

(a)

3
 (b)

Fig: 2.2 (a) Capacitively coupled common-source amplifier. (b) The amplifier equivalent circuit at low
frequencies. Note that the T model is used for the MOSFET and ro is neglected.

At low frequencies, the reactance1 / jωC of each of the three capacitances increases and the amplifier
gain decreases. The amplifier equivalent circuit shown in Fig. 2.2(b). This circuit is obtained by short-
circuiting VDD and replacing the MOSFET with its T model

The gain Vo/Vsig of the amplifier can be obtained from the equivalent circuit in Fig. 2.2(b) by
starting at the source and working our way to the load, as
𝑽𝑶 𝑽𝒈 𝑰𝒅 𝑽𝒐
= x x 2.1
𝑽𝒔𝒊𝒈 𝑽𝒔𝒊𝒈 𝑽𝒈 𝑰𝒅

where Vg is the voltage between gate and ground, and Id is the drain current. the amplifier input
resistance is 𝑹𝑮 = 𝑹𝑮𝟏 ||𝑹𝑮𝟐 Using the voltage-divider rule Vg is
𝑹𝑮
𝑽𝒈 = 𝑽𝒔𝒊𝒈 𝟏 2.2
𝑹𝑮 + +𝑹𝒔𝒊𝒈
𝒔𝑪𝑪𝟏

𝑹𝑮
𝑽𝒈 = 𝑽𝒔𝒊𝒈 𝟏
2.3
(𝑹𝑮 +𝑹𝒔𝒊𝒈 )[𝟏+ ]
𝒔𝑪𝑪𝟏 (𝑹𝑮+𝑹𝒔𝒊𝒈 )

Multiply and divide eq 2.3 with s

𝑹𝑮 𝒔
𝑽𝒈 = 𝑽𝒔𝒊𝒈 𝟏
(𝑹𝑮 +𝑹𝒔𝒊𝒈 ) 𝒔 [𝟏+ ]
𝒔𝑪𝑪𝟏 (𝑹𝑮 +𝑹𝒔𝒊𝒈 )

𝑹𝑮 𝒔
𝑽𝒈 = 𝑽𝒔𝒊𝒈 𝒔
(𝑹𝑮 +𝑹𝒔𝒊𝒈 ) [𝒔+ ]
𝒔𝑪𝑪𝟏 (𝑹𝑮 +𝑹𝒔𝒊𝒈 )

𝑽𝒈 𝑹𝑮 𝒔
= 𝟏 2.4
𝑽𝒔𝒊𝒈 𝑹𝑮 +𝑹𝒔𝒊𝒈 𝒔+
𝑪𝑪𝟏 (𝑹𝑮+𝑹𝒔𝒊𝒈 )

4
the effect of CC1 is to cause the expression for the signal transmission from the signal source to the
amplifier input to acquire a frequency-dependent factor. this factor as the transfer function of a single-
time-constant circuit of the high-pass type, with a pole frequency 𝑤𝑃1.

𝝎𝑷𝟏 = 𝟏⁄𝑪 (𝑹𝒔𝒊𝒈 + 𝑹𝑮 ) 2.5

𝑪𝟏

In addition to the pole, CC1 introduces a zero at s = 0 (dc). CC1 is included in the amplifier circuit because
it blocks dc.

Figure 2.3 shows a sketch of the magnitude of the frequency-dependent factor in the transfer function of
Eq. (2.6) versus frequency ω.

𝒔
Fig: 2.3 Sketch of the magnitude of the high-pass function versus frequency ω.
𝒔+𝒘𝑷𝟏

the drain current Id, which is equal to the source current Is. Is can be found by dividing Vg by the total
impedance in the source lead, (1⁄𝑔𝑚 + 𝑍𝑠 ),

𝑉𝑔 𝑉𝑔 𝑌𝑠
𝐼𝑑 = 𝐼𝑠 = = = 𝑔𝑚 𝑉𝑔
1 1 1 𝑔𝑚 + 𝑌𝑠
𝑔𝑚 + 𝑍𝑠 𝑔𝑚 + 𝑌𝑠

Where

1 1
𝑌𝑠 = = + 𝑠𝐶𝑠
𝑍𝑠 𝑅𝑆

1
𝑅𝑆 + 𝑠𝐶𝑠
𝐼𝑑 = 𝑔𝑚 𝑉𝑔
1
𝑔𝑚 + 𝑅 + 𝑠𝐶𝑠
𝑆

1
𝑠 [𝑅 𝐶 + 𝑠] 𝐶𝑠
𝑆 𝑠
𝐼𝑑 = 𝑔𝑚 𝑉𝑔
1
𝑔𝑚 + 𝑅
𝑆
[𝑠 + 𝐶𝑠 ] 𝐶𝑠

Thus,

5
 𝟏
𝑰𝒅 𝒔[ +𝒔]
𝑹𝑺 𝑪𝒔
=𝒈 𝒎 𝟏 2.6
𝑽𝒈 𝒈𝒎 +
𝑹𝑺
[𝒔+ ]
𝑪𝒔

Thus, the bypass capacitor introduces a pole with frequency 𝜔𝑝2 ,

𝟏
𝒈𝒎 +
𝑹𝑺
𝜔𝑝2 = 2.7
𝑪𝒔

and a transmission Zero on the negative real axis of the s plane at

1
𝑆𝑧 = − 𝐶 𝑅
𝑠 𝑠

And thus has a frequency

𝟏
𝝎𝒁 = 2.8
𝑪𝒔 𝑹𝒔

𝑠+𝑤
Fig: 2.4 Sketch of the magnitude of the function 𝑠+𝑤 𝑧 versus frequency ω.
𝑃2

since gm is usually large, 𝑤𝑃2 ≫ 𝑤𝑧 That is, 𝑤𝑃2 will be closer to the midband, and thus it plays a more
significant role in determining 𝑤𝐿 than does 𝑤𝑧 . Figure 2.4 shows a sketch of the magnitude of the
frequency-dependent factor of the transfer function in Eq. (2.6).

𝑽𝒐 = 𝑰 𝒐 𝑹 𝑳

using the current-divider rule, the fraction of Id that flows through

RL is
𝑅𝐷
𝐼𝑜 = −𝐼𝑑
1
𝑅𝐷 + 𝑠𝐶 + 𝑅𝐿
𝑐2

6
 𝑅𝐷
𝑽𝒐 = −𝐼𝑑 1
𝑹𝑳
𝑅𝐷 + + 𝑅𝐿
𝑠𝐶𝑐2

𝑽𝒐 𝑅 𝐷 𝑹𝑳
=−
𝐼𝑑 1
( 𝑅 𝐷 + 𝑹 𝑳 ) [1 + ]
𝑠𝐶𝑐2 (𝑅𝐷 + 𝑹𝑳 )

𝑽𝒐 𝑅 𝐷 𝑹𝑳 𝟏
=
𝐼𝑑 ( 𝑅 𝐷 + 𝑹𝑳 ) 1
[1 + ]
𝑠𝐶𝑐2 (𝑅𝐷 + 𝑹𝑳 )

Multiply & divide above equation with s

𝑽𝒐 𝑅 𝐷 𝑹𝑳 𝒔
=
𝐼𝑑 ( 𝑅 𝐷 + 𝑹𝑳 ) 1
𝒔 [1 + ]
𝑠𝐶𝑐2 (𝑅𝐷 + 𝑹𝑳 )

𝑽𝒐 𝑅𝐷 𝑹𝑳 𝒔
=
𝐼𝑑 (𝑅𝐷 + 𝑹𝑳 ) [𝑠 + 𝑠
]
𝑠𝐶𝑐2 (𝑅𝐷 + 𝑹𝑳 )

𝑽𝒐 𝑹𝑫 𝑹𝑳 𝒔
= (𝑹 𝟏
2.9
𝑰𝒅 𝑫 +𝑹𝑳 ) [𝒔+ ]
𝑪𝒄𝟐 (𝑹𝑫 +𝑹𝑳 )

from which CC2 introduces a pole with frequency,

𝟏
𝝎𝒑𝟑 = 2.10
𝑪𝒄𝟐 (𝑹𝑫 +𝑹𝑳 )

and a zero at s = 0 (dc). A sketch of the magnitude of the frequency-dependent factor of the

transfer function in Eq. (2.9) is shown in Fig. 2.5.

𝑠
Fig: 2.5 Sketch of the magnitude of the function 𝑠+𝑤 versus frequency ω.
𝑃3

7
The overall low-frequency gain function of the amplifier can be found by combining Eqs.

(2.4), (2.6), and (2.9),

1
𝑠[ +𝑠] 𝑅𝐷 𝑅𝐿 𝑠
𝑉𝑜 𝑅𝐺 𝑠 𝑅𝑆 𝐶 𝑠
= 𝑅𝐺 +𝑅𝑠𝑖𝑔 𝑠+ 1 𝑔𝑚 1 1
𝑉𝑠𝑖𝑔 𝑔𝑚 +
𝑅𝑆
(𝑅𝐷 +𝑅𝐿 ) [𝑠+ ]
𝐶𝐶1 (𝑅𝐺+𝑅𝑠𝑖𝑔) 𝐶𝑐2 (𝑅𝐷 +𝑅𝐿 )
[𝑠+ ]
𝐶𝑠

1
𝑅𝐺 𝑅𝐷 𝑅𝐿 𝑠 𝑠[+ 𝑠] 𝑠
𝑉𝑜 𝑅𝑆 𝐶𝑠
= 𝑔𝑚
𝑉𝑠𝑖𝑔 𝑅𝐺 + 𝑅𝑠𝑖𝑔 (𝑅𝐷 + 𝑅𝐿 ) 1 1 1
𝑠+ 𝑔𝑚 + [𝑠 + ]
𝐶𝐶1 (𝑅𝐺 + 𝑅𝑠𝑖𝑔 ) [𝑠 + 𝑅 𝑆] 𝐶𝑐2 (𝑅𝐷 + 𝑅𝐿 )
𝐶𝑠

𝑉𝑜 𝑅𝐺 𝑠 𝑠 + 𝜔𝑍 𝑠
= 𝑔𝑚 (𝑅𝐷 ||𝑅𝐿 ) ( )( )( )
𝑉𝑠𝑖𝑔 𝑅𝐺 + 𝑅𝑠𝑖𝑔 𝑠 + 𝜔𝑝1 𝑠 + 𝜔𝑝2 𝑠 + 𝜔𝑝3

𝑽𝒐 𝒔 𝒔+𝝎 𝒔
𝑽𝒔𝒊𝒈
= 𝑨𝑴 (
𝒔+𝝎𝒑𝟏
) (𝒔+𝝎 𝒁 ) (𝒔+𝝎 ) 2.11
𝒑𝟐 𝒑𝟑

Where 𝐴𝑀 ,the midband gain, is given by

𝑹𝑮
𝑨𝑴 = 𝒈𝒎 (𝑹𝑫 ||𝑹𝑳 ) 2.12
𝑹𝑮 +𝑹𝒔𝒊𝒈

Fig: 2.6 Sketch of the low-frequency magnitude response of a CS amplifier for which the three pole

frequencies are sufficiently separated for their effects to appear distinct.

𝑓𝐿 is approximately equal to the highest of the pole frequencies,

8
 𝒇𝑳 ≅ 𝒇𝑷𝟐 2.13
𝟏
𝟏
𝒈𝒎 +𝑹
𝑺
𝒇𝑷𝟐 = 𝟐𝝅 𝑪𝒔

If a dominant pole does not exist, the following approximate expression for 𝑓𝐿 can be used

𝒇𝑳 ≅ 𝒇𝑷𝟐 √𝒇𝟐𝑷𝟏 + 𝒇𝟐𝑷𝟐 + 𝒇𝟐𝑷𝟑 − 𝟐 𝒇𝟐𝒛

Example 2.1

Determine values for the coupling capacitors 𝑪𝑪𝟏 and 𝑪𝑪𝟐 and the bypass capacitor 𝑪𝒔 for a CS amplifier for
which 𝑹𝑮 =4.7 MΩ ,𝑹𝑫 =𝑹𝑳 =15 KΩ, 𝑹𝒔𝒊𝒈 =100 KΩ, 𝑹𝒔 =10 KΩ,and 𝒈𝒎 =1 mA/V.it is required to have 𝒇𝑳 at 100
Hz and that the nearest break frequency be at least a decade lower.

Solution

select 𝐶𝑠 so that
1
1
𝑔𝑚 + 𝑅
𝑆
𝑓𝐿 = 𝑓𝑃2 = 2𝜋 𝐶𝑠

Thus,

1.1 × 10−3
𝐶𝑠 = 2𝜋 ×100
=1.75 µF

For 𝑓𝑝1 = 𝑓𝑝3 =10HZ

1
𝑓𝑝1 =
2 𝜋 𝐶𝐶1 (𝑅𝑠𝑖𝑔 + 𝑅𝐺 )

1
10 =
2𝜋𝐶𝐶1 (0.1+4.7)×106

Which yields

𝑪𝑪𝟏 =3.3 nF

1
𝑓𝑝3 =
2 𝜋 𝐶𝑐2 (𝑅𝐷 +𝑅𝐿 )

1
10=
2𝜋𝐶𝐶2 (15+15) × 103

𝑪𝑪𝟐 =0.53 𝝁𝑭

Finally, calculate the frequency of the zero 𝑓𝑧 𝑎𝑠

1 1
𝑓𝑧 = 2𝜋𝐶 𝑅 = 2𝜋 ×1.75 × 10−6 × 10 ×103
𝑠 𝑠

=9.1 HZ

9
2.3 The Method of Short-Circuit Time Constants

There is a simple method for obtaining an estimate for 𝑓𝐿 without the need to determine the frequencies of
the poles. The method is as follows

1. Set the input signal Vsig = 0.

2. Consider the capacitors one at a time. That is, while considering capacitor Ci, set all the other
capacitors to infinite values (i.e., replace them with short circuits—hence the name of the
method).
3. For each capacitor Ci, find the total resistance Ri seen by Ci. This can be determined either by
inspection or by replacing Ci with a voltage source Vx and finding the current Ix drawn from Vx.
𝑉𝑥
𝑅𝑖 =
𝐼𝑥
4. Calculate the 3-dB frequency 𝑓𝐿 using

𝟏 𝟏
𝒇𝑳 ≅ ∑𝒏𝒊=𝟏 ≅ ∑𝒏𝒊=𝟏 2.14
𝑹𝒊 𝑪𝒊 𝝉𝒊

where n is the total number of capacitors.

2.4 Low-Frequency Response of Common-Emitter Amplifier

(a)

10
 (b)

Fig: 2.7 (a) A discrete-circuit common-emitter amplifier. (b) Equivalent circuit of the amplifier in (a).

Figure 2.7(a) shows a discrete-circuit common-emitter amplifier utilizing the voltage divider biasing
arrangement, together with coupling capacitors CC1 and CC2, and bypass capacitor CE. an estimate of the
frequency 𝑓𝐿 at which the gain of this amplifier drops by 3 dB below its value at midband. the low-
frequency gain of the CE amplifier can be analyze using the equivalent circuit shown in Fig. 2.7(b). This
equivalent circuit is obtained by short-circuiting VCC and replacing the BJT with its T model, while
neglecting ro.

11
Applying the Method of Short-Circuit Time Constants

Fig: 2.8 Circuits for determining the short-circuit time constants for the amplifier in Fig. 2.7.

Setting Vsig = 0 in the circuit of Fig. 2.7(b) and considering each capacitor, one at a time, while short-
circuiting the other two results in the three circuits shown in Fig. 2.8. These circuits can be used to
determine the resistance seen by each capacitor and hence its effective time constant. For CC1 we use the
circuit in Fig. 2.8 (a) and note that rπ is the input resistance at the base when CE is short-circuited.
Capacitor CC1 sees a resistance RC1, which can be found by inspection as

𝑅𝐶1 = (𝑅𝐵 |||𝑟𝜋 ) + 𝑅𝑠𝑖𝑔

and the time constant associated with CC1 becomes

𝜏𝐶1 = 𝐶𝐶1 𝑅𝑐1

12
For CE, use the circuit in Fig. 2.8(b). Here, CC1 shorted [refer to Fig. 2.7(b)], the resistance in the base
𝑅𝐵 ||𝑅𝑠𝑖𝑔
becomes(𝑅𝐵 ||𝑅𝑠𝑖𝑔 ), which can be reflected to the emitter side . The total resistance RCE seen by CE can be
𝛽+1
found by inspection from the circuit in Fig. 2.8(b) as

𝑅𝐵 ||𝑅𝑠𝑖𝑔
𝑅𝐶𝐸 = 𝑅𝐸 || [𝑟𝑒 + ]
𝛽+1

and the time constant becomes

𝜏𝐶𝐸 = 𝐶𝐸 𝑅𝐶𝐸

Finally, the resistance seen by CC2 can be determined by inspection of the circuit in Fig. 2.8(c) as

𝑅𝑐2 = 𝑅𝐶 + 𝑅𝐿

and the corresponding time constant 𝜏𝑐2 as

𝜏𝑐2 = 𝐶𝐶2 𝑅𝑐2

With the three-time constants in hand, the 3-dB frequency 𝑓𝐿 can be found from

𝝎𝑳 𝟏 𝟏 𝟏 𝟏
𝒇𝑳 = = [ + + ] 2.15
𝟐𝝅 𝟐𝝅 𝑪𝑪𝟏 𝑹𝒄𝟏 𝑪𝑬 𝑹𝑪𝑬 𝑪𝑪𝟐 𝑹𝑪𝟐

Example 2.2

Determine values for Cc1,Cc2, and CE for the common-emitter amplifier ,which has RB =100kΩ, RC
=8KΩ, RL = 5KΩ, RE = 5kΩ, β =100. gm =40 mA/V, and 𝒓𝝅 = 2.5kΩ it requires to have fL =100HZ.

Solution

First determine the resistance seen by the three capacitors CC1,CE, and CC2 as follows:

𝑅𝐶1 = (𝑅𝐵 |||𝑟𝜋 ) + 𝑅𝑠𝑖𝑔 = (100 || 2.5 ) + 5 = 7.44 kΩ

𝑅𝐵 ||𝑅𝑠𝑖𝑔 100||5
𝑅𝐶𝐸 = 𝑅𝐸 || [𝑟𝑒 + ] = 5 || (0.025 + ) = 0.071kΩ
𝛽+1 101

𝑅𝑐2 = 𝑅𝐶 + 𝑅𝐿 = 8 + 5 = 13 kΩ

Now, selecting CE so that it contributes 80% of the value of 𝝎L gives

1 1
0.8 𝑓𝐿 = [ ]
2𝜋 𝐶𝐸 𝑅𝐶𝐸

1
0.8𝑥 100 𝑥 2𝜋 = [ ]
𝐶𝐸 71

𝑪𝑬 = 𝟐𝟖µ𝐅

13
if CC1 is to contribute 10% of fL

1 1
0.1 𝑓𝐿 = [ ]
2𝜋 𝐶𝐶1 𝑅𝑐1

1
0.1 𝑥 100 𝑥 2𝜋 = [ ]
𝐶𝐶1 7.41 𝑥 103

𝑪𝑪𝟏 = 𝟐. 𝟏 µ𝐅

if CC1 is to contribute 10% of fL

1 1
0.1 𝑓𝐿 = [ ]
2𝜋 𝐶𝐶2 𝑅𝑐2

1
0.1 𝑥 100 𝑥 2𝜋 = [ ]
𝐶𝐶2 13 𝑥 103

𝑪𝑪𝟏 = 𝟏. 𝟐 µ𝐅

1 1
𝑓𝑍 = [ ]
2𝜋 𝐶𝐸 𝑅𝐸

1
𝑓𝑍 = = 1.1 𝐻𝑧
2𝜋 𝑥 28 𝑥10−6 𝑥 5 𝑥 103

which is very far from 𝑓𝐿 and thus has an insignificant effect.

2.5 Internal Capacitive Effects and the High-Frequency Model of the MOSFET

The gain falloff at high frequencies is caused by the capacitive effects internal to the transistors.
Fig. 2.9, which shows the cross section of an n-channel MOSFET operating in the saturation region, as
signified by the tapered n channel that is pinched off at the drain end. As indicated, there are four internal
capacitances: Two of these, Cgs and Cgd, result from the gate-capacitance effect; the other two, Csb and
Cdb, are the depletion capacitances of the pn junctions formed by the source region and the substrate, and
the drain region and the substrate, respectively.

14
Fig 2.9: A cross section of the n-channel MOSFET operating in the saturation region. The four internal
capacitances, Cgs, Cgd, Csb, and Cdb, are indicated.

The gate (or oxide) capacitance per unit gate area is denoted Cox. When the channel is tapered
𝟐
and pinched off, the gate capacitance is given by 𝑾𝑳𝑪𝒐𝒙 . In addition to this capacitance, there are two
𝟑

other small capacitances resulting from the overlap of the gate with the source region (or source
diffusion) and the drain region (or drain diffusion). Each of these overlaps has a length Lov and thus the
resulting overlap capacitances Cov are given by

𝑪𝒐𝒗 = 𝑾𝑳𝑪𝒐𝒙
Typically, Lov = 0.05 to 0.1L. We can now express the gate-to-source capacitance Cgs as
𝟐
𝑪𝒈𝒔 = 𝟑 𝑾𝑳𝑪𝒐𝒙 + 𝑪𝒐𝒗 2.16

For the gate-to-drain capacitance, the channel pinch-off at the drain end causes Cgd to consist entirely of
the overlap component Cov,

𝑪𝒈𝒅 = 𝑪𝒐𝒗 2.17

The depletion-layer capacitances of the two reverse-biased pn junctions formed between each of the
source and the drain diffusions and the p-type substrate (body). The source diffusion, the source-body
capacitance, Csb is
𝑪𝒔𝒃𝟎
𝑪𝒔𝒃 = 𝑽
2.18
√𝟏+ 𝑽𝑺𝑩
𝑶

where Csb0 is the value of Csb at zero body-source bias, VSB is the magnitude of the reverse-bias voltage,
and V0 is the junction built-in voltage (0.6 V to 0.8 V). Similarly, for the drain diffusion, the drain-body
capacitance Cdb,

𝑪𝒅𝒃𝟎
𝑪𝒅𝒃 = 𝑽
2.19
√𝟏+ 𝑽𝑫𝑩
𝑶

where Cdb0 is the value of Cdb at zero body-source bias, VDB is the magnitude of the reverse-bias voltage,
and V0 is the junction built-in voltage (0.6 V to 0.8 V).

The High-Frequency MOSFET Model

Figure 2.10(a) shows the small-signal model of the MOSFET, including the four capacitances Cgs, Cgd,
Csb, and Cdb. This model can be used to predict the high-frequency response of MOSFET amplifiers. the
model simplifies considerably, as shown in Fig. 2.10(b). In this model, Cgd, although small, plays a
significant role in determining the high-frequency response of amplifiers and thus must be kept in the
model. Capacitance Cdb, on the other hand, can usually be neglected, resulting in significant.

15
simplification of manual analysis. The resulting circuit is shown in Fig. 2.10(c). Finally, Fig. 2.10(d)
shows the high-frequency T model in its simplified form.

(a)

(b)

(C) (d)

16
Fig 2.10 (a) High-frequency, equivalent-circuit model for the MOSFET. (b) The equivalent circuit for the
case in which the source is connected to the substrate (body). (c) The equivalent-circuit model of (b) with
Cdb neglected (to simplify analysis). (d) The simplified high-frequency T model.

2.6 The MOSFET Unity-Gain Frequency (𝒇𝑻 )

A figure of merit for the high-frequency operation of the MOSFET as an amplifier is the unity-gain
frequency, 𝒇𝑻 , also known as the transition frequency, which gives rise to the subscript T. This is defined
as the frequency at which the short-circuit current gain of the common-source configuration becomes
unity.

𝐼𝑜
Fig 2.11 Determining the short-circuit current gain 𝐼𝑖

Here 𝑟𝑜 parallel with short circuit, so 𝑟𝑜 can be eliminated.

𝐼𝑜
Fig 2.12 Determining the short-circuit current gain 𝐼𝑖

Figure 2.11 shows the MOSFET hybrid-π model with the source as the common terminal between the
input and output ports. the input is fed with a current-source signal Ii, and the output terminals are short-
circuited for determining the short-circuit current gain. The current in the short circuit is given by

Apply KCL at drain node

17
 𝐼0 + 𝑠𝐶𝑔𝑑 𝑉𝑔𝑠 = 𝑔𝑚 𝑉𝑔𝑠

𝐼0 = 𝑔𝑚 𝑉𝑔𝑠 − 𝑠𝐶𝑔𝑑 𝑉𝑔𝑠

Recalling that Cgd is small, at the frequencies of interest we can neglect the second term in this equation,

𝐼0 ≅ 𝑔𝑚 𝑉𝑔𝑠 2.20

From Fig. 2.12, 𝑉𝑔𝑠 in terms of the input current 𝐼𝑖

𝐼𝑖
𝑉𝑔𝑠 = 2.21
𝑠(𝐶𝑔𝑠 +𝐶𝑔𝑑)

Equations (2.20) and (2.21) can be combined to obtain the short-circuit current gain,

𝐼𝑖
𝐼0 = 𝑔𝑚
𝑠(𝐶𝑔𝑠 + 𝐶𝑔𝑑 )

𝑰𝟎 𝒈𝒎
= 2.22
𝑰𝒊 𝒔(𝑪𝒈𝒔 +𝑪𝒈𝒅 )

For physical frequencies s = jω, the magnitude of the current gain is

𝐼0 𝑔𝑚
| |=
𝐼𝑖 𝑤 (𝐶𝑔𝑠 + 𝐶𝑔𝑑 )

and it becomes unity at the frequency 𝑤 = 𝑤𝑇

𝑔𝑚
1=
𝑤𝑇 (𝐶𝑔𝑠 + 𝐶𝑔𝑑 )
𝑔𝑚
𝑤𝑇 =
(𝐶𝑔𝑠 + 𝐶𝑔𝑑 )
𝑤𝑇
Thus, the unity-gain frequency 𝑓𝑇 = 2𝜋

𝑔𝑚
𝑓𝑇 = 2.23
2𝜋 (𝐶𝑔𝑠 +𝐶𝑔𝑑 )

18
 Table 2.1 The MOSFET High-Frequency Model

2.7 Internal Capacitive Effects and the High-Frequency Model of the BJT

The Base-Charging or Diffusion Capacitance Cde

When the transistor is operating in the active mode, minority-carrier charge is stored in the base region.
For a npn transistor, the stored electron charge in the base, Qn, can be expressed in terms of the collector
current iC as

𝑄𝑛 = 𝜏𝐹 𝑖𝑐

where 𝜏𝐹 is a forward base-transit time and represents the average time a charge carrier (electron)
spends in crossing the base. Typically, 𝜏𝐹 is in the range of 10 ps to 100 ps. the small-signal diffusion
capacitance Cde,

𝑑𝑄𝑛
𝐶𝑑𝑒 =
𝑑𝑣𝐵𝐸

19
 𝑑𝑖𝑐
𝐶𝑑𝑒 = 𝜏𝐹
𝑑𝑣𝐵𝐸
𝑰
𝑪𝒅𝒆 = 𝝉𝑭 𝒈𝒎 = 𝝉𝑭 𝑽𝒄 2.24
𝑻

where IC is the dc collector bias current at which the transistor is operating.

The Base–Emitter Junction Capacitance Cje. A change in VBE changes not only the charge stored in the
base region but also the charge stored in the base–emitter depletion layer. This distinct charge-storage
effect is represented by the EBJ depletion-layer capacitance, Cje. the depletion-layer capacitance is given
approximately by

𝑪𝒋𝒆 ≅ 𝟐𝑪𝒋𝒆𝒐

where 𝐶𝑗𝑒𝑜 is the value of 𝐶𝑗𝑒 at zero EBJ voltage.

the emitter–base capacitance 𝑪𝝅 = 𝑪𝒅𝒆 + 𝑪𝒋𝒆 2.25

Here 𝐶𝑑𝑒 >> 𝐶𝑗𝑒

𝑪𝝅 ≅ 𝑪𝒅𝒆

The Collector–Base Junction Capacitance Cμ

In active-mode operation, the CBJ is reverse biased, and its junction or depletion capacitance, usually
denoted Cμ, can be found from
𝑪𝝁𝟎
𝒄𝝁 = 𝑽 𝒎 2.26
[𝟏+ 𝑪𝑩 ]
𝑽𝒐𝒄

where Cμ0 is the value of Cμ at zero voltage; VCB is the magnitude of the CBJ reverse-bias voltage, Voc is
the CBJ built-in voltage (typically, 0.75 V), and m is its grading coefficient (typically, 0.2–0.5).

The High-Frequency Models

Figure 2.13 shows the hybrid-π and T models of the BJT, including capacitive effects. Specifically, there
are two capacitances: the emitter–base capacitance Cπ = Cde+ Cje and the collector–base capacitance Cμ.
Typically, Cπ is in the range of a few picofarads to a few tens of picofarads, and Cμ is in the range of a
fraction of a picofarad to a few picofarads. The resistance of the silicon material of the base region
between the base terminal B and virtual base terminal B’ is called as Base spreading Resistance (𝒓𝒙 ).
Typically, 𝒓𝒙 is a few tens of ohms.

20
 (a)

(b)

Fig 2.13 The high-frequency models of the BJT: (a) hybrid-π model and (b) T model.

2.8 The BJT Unity-Gain Frequency

𝐼
Fig 2.14 Circuit for deriving an expression for 𝐴𝐼 (𝑠𝑐) = ℎ𝑓𝑒 (𝑠) = 𝐼𝑐
𝑏

Here 𝑟𝑜 parallel with short circuit, so 𝑟𝑜 can be eliminated

21
 𝐼
Fig 2.15 Circuit for deriving an expression for 𝐴𝐼 (𝑠𝑐) = ℎ𝑓𝑒 (𝑠) = 𝐼𝑐
𝑏

The CE short-circuit current gain, as a function of frequency in terms of the hybrid-π components. Apply
KCL at node C

𝐼𝑐 + 𝑠𝐶𝜇 𝑉𝜋 = 𝑔𝑚 𝑉𝜋

𝐼𝑐 = (𝑔𝑚 − 𝑠𝐶𝜇 )𝑉𝜋 2.28

A relationship between 𝑉𝜋 and Ib can be established by multiplying Ib by the impedance seen

between B’ and E:
𝐼𝑏
𝑉𝜋 = 𝐼𝑏 (𝑟𝜋 ||𝐶𝜋 ||𝐶µ ) = 1 2.29
⁄𝑟𝜋 +𝑠𝐶µ +𝑠𝐶𝜋

Thus, 𝐴𝐼 (𝑠𝑐) = ℎ𝑓𝑒 (𝑠) can be obtained by combining Eqs. (2.28) and (2.29):
𝐼𝑏
𝐼𝑐 = (𝑔𝑚 − 𝑠𝐶𝜇 ) 1
⁄𝑟𝜋 +𝑠𝐶µ +𝑠𝐶𝜋

𝐼𝑐 (𝑔𝑚 − 𝑠𝐶𝜇 )
= 𝐴𝐼 = ℎ𝑓𝑒 =
𝐼𝑏 1⁄ + 𝑠𝐶 + 𝑠𝐶
𝑟𝜋 µ 𝜋

Generally, 𝑔𝑚 >> 𝑤𝐶𝜇 , so 𝑠𝐶𝜇 term can be eliminated

𝑔𝑚
ℎ𝑓𝑒 ≅
1⁄ + 𝑠𝐶 + 𝑠𝐶
𝑟𝜋 µ 𝜋

𝑔𝑚 𝑟𝜋
ℎ𝑓𝑒 ≅ ∴ 𝑔𝑚 𝑟𝜋 = 𝛽𝑜
1 + 𝑠(𝐶𝜋 + 𝐶𝜇 )𝑟𝜋

𝜷𝒐
𝒉𝒇𝒆 = 2.30
𝟏+𝒔(𝑪𝝅 +𝑪𝝁 )𝒓𝝅

𝛽𝑜
ℎ𝑓𝑒 =
1+𝑗 𝑤 (𝐶𝜋 +𝐶𝜇 )𝑟𝜋

22
 𝟏
𝝎𝜷 = 2.31
(𝑪𝝅 +𝑪𝝁 )𝒓𝝅

𝛽𝑜
If 𝒘 = 𝜔𝛽 , 𝑡ℎ𝑒𝑛 ℎ𝑓𝑒 = , 𝑠𝑜 𝜔𝛽 𝑐𝑎𝑙𝑙𝑒𝑑 𝑎𝑠 𝑐𝑢𝑡 − 𝑜𝑓𝑓 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦.
√2

Fig 2.16 Bode plot for |ℎ𝑓𝑒 |

Figure 2.16 shows a Bode plot for|ℎ𝑓𝑒 |. From the –6-dB/octave slope, it follows that the frequency at
which |ℎ𝑓𝑒 |drops to unity, which is called the unity-gain bandwidth.

𝛽𝑜
ℎ𝑓𝑒 =
1 + 𝑗 𝑤 (𝐶𝜋 + 𝐶𝜇 )𝑟𝜋

At 𝒘 = 𝜔 𝑇 , 𝑡ℎ𝑒𝑛 ℎ𝑓𝑒 = 1

𝛽𝑜
1=
1 + 𝑗𝜔𝑇 (𝐶𝜋 + 𝐶𝜇 )𝑟𝜋
𝛽𝑜 1
1= 𝜔 ∴ 𝜔𝛽 =
1 + 𝑗 𝜔𝑇 (𝐶𝜋 + 𝐶𝜇 )𝑟𝜋
𝛽

𝜔𝑇
1+𝑗 = 𝛽𝑜
𝜔𝛽
𝜔𝑇
𝑗 = 𝛽𝑜 − 1
𝜔𝛽

generally, 𝛽𝑜 >> 1
𝜔𝑇
𝑗 = 𝛽𝑜
𝜔𝛽
𝜔𝑇
= 𝛽𝑜
𝜔𝛽

𝝎𝑻 = 𝝎𝜷 𝜷𝒐

23
 1 1
𝜔𝑇 = 𝑔𝑚 𝑟𝜋 ∴ 𝛽𝑜 = 𝑔𝑚 𝑟𝜋 , 𝜔𝛽 =
(𝐶𝜋 + 𝐶𝜇 )𝑟𝜋 (𝐶𝜋 + 𝐶𝜇 )𝑟𝜋
𝑔𝑚
𝜔𝑇 =
𝐶𝜋 + 𝐶𝜇

𝒈𝒎
𝒇𝑻 = 2.32
𝟐𝝅(𝑪𝝅 +𝑪𝝁 )

𝐶𝜋 + 𝐶𝜇 ≅ 𝐶𝜋 ∴ 𝐶𝜋 >> 𝐶𝜇

𝐶𝜋 + 𝐶𝜇 = 𝐶𝜋 = 𝐶𝑑𝑒 + 𝐶𝑗𝑒

𝐶𝜋 + 𝐶𝜇 = 𝐶𝜋 = 𝐶𝑑𝑒 ∴ 𝐶𝑑𝑒 >> 𝐶𝑗𝑒

𝐶𝜋 + 𝐶𝜇 = 𝐶𝜋 = 𝐶𝑑𝑒 = 𝜏𝐹 𝑔𝑚

Sub above value in Eq. 2.32

𝑔𝑚
𝑓𝑇 =
2𝜋𝜏𝐹 𝑔𝑚
𝟏
𝒇𝑻 = 2.33
𝟐 𝝅 𝝉𝑭

Table 2.2 The BJT High-Frequency Model

24
2.9 High-Frequency Response of the Common-Source (CS) Amplifier

Fig 2.17 Common-Source Amplifier

Fig 2.18 Equivalent circuit of Common-Source Amplifier in high frequency

Figure 2.18 shows the high-frequency, equivalent-circuit model of a CS amplifier. It is obtained by

replacing the MOSFET in an amplifier circuit such as that in Fig. 2.17 by its high-frequency, equivalent-
circuit model and all external capacitors can be replaced by short circuit. where 𝑅𝐺 = 𝑅𝐺1 || 𝑅𝐺2. The
midband gain AM can be found from this circuit by setting Cgs and Cgd to zero.

25
 𝑅𝐿′ = 𝑟𝑜 || 𝑅𝐷 || 𝑅𝐿

𝑉𝑜 = 𝐼𝐿 𝑅𝐿′

𝑉𝑜 = − 𝑔𝑚 𝑉𝑔𝑠 𝑅𝐿′ ∴ 𝐼𝐿 = − 𝑔𝑚 𝑉𝑔𝑠

𝑅𝐺
At input side, by using voltage divider rule 𝑉𝑔𝑠 = 𝑅 𝑉𝑠𝑖𝑔
𝐺 +𝑅𝑠𝑖𝑔

Sub 𝑉𝑔𝑠 value in above equation

𝑅𝐺
𝑉𝑜 = − 𝑔𝑚 𝑉 𝑅′
𝑅𝐺 + 𝑅𝑠𝑖𝑔 𝑠𝑖𝑔 𝐿

𝑉𝑜 𝑅𝐺
=− 𝑔 𝑅′
𝑉𝑠𝑖𝑔 𝑅𝐺 + 𝑅𝑠𝑖𝑔 𝑚 𝐿
𝑹𝑮
𝑨𝑴 = − 𝑹 𝒈𝒎 𝑹′𝑳 2.34
𝑮 +𝑹𝒔𝒊𝒈

Midband gain without source 𝐴𝑀 = − 𝑔𝑚 𝑅𝐿′

The equivalent circuit of Fig. 2.18 can be simplified by utilizing Thevenin theorem at the input side and
by combining the three parallel resistances at the output side. The resulting simplified circuit is shown in
Fig. 2.19.

Thevenin’s theorem states that “any linear circuit containing several voltages and resistances can be
replaced by just one single voltage in series with a single resistance .“.

26
 For equivalent voltage

′ 𝑅𝐺
By using voltage divider rule 𝑉𝑠𝑖𝑔 =𝑅 𝑉𝑠𝑖𝑔
𝐺 +𝑅𝑠𝑖𝑔

For equivalent resistance voltage source replaced with short circuit

′
𝑅𝑠𝑖𝑔 = 𝑅𝐺 || 𝑅𝑠𝑖𝑔

𝑅𝐿′ = 𝑟𝑜 || 𝑅𝐷 || 𝑅𝐿

Fig 2.19 the circuit of 2.18 simplified at the input and the output

27
Miller’s Theorem states that any impedance Z connected between input and output (in feedback loop),
that can be distributed to equivalent input impedance [𝑍(1 − 𝐴) ] as well as equivalent output
1
impedance [𝑧 (1 − 𝐴)]. Where A is amplifier gain.

By using Miller’s Theorem, the feedback capacitance 𝐶𝑔𝑑 can be distributed to equivalent input
capacitance 𝐶𝑒𝑞 = 𝐶𝑔𝑑 (1 − (− 𝑔𝑚 𝑅𝐿′ ) = 𝐶𝑔𝑑 (1 + 𝑔𝑚 𝑅𝐿′ ) and equivalent output capacitance 𝐶𝑒𝑞 ′
=
1 1
𝐶𝑔𝑑 (1 − (− 𝑔 ′ ) = 𝐶𝑔𝑑 (1 + ( ).
𝑚 𝑅𝐿 𝑔𝑚 𝑅𝐿′

′ ′
The capacitance 𝐶𝑒𝑞 value is very small compared to 𝐶𝑒𝑞 so 𝐶𝑒𝑞 can be eliminated at output side.
The simplified circuit shown in below figure.

Fig 2.20 The equivalent circuit with Cgd replaced at the input side with the equivalent capacitance Ceq

1
𝐶𝑖𝑛 = 𝐶𝑔𝑠 + 𝐶𝑔𝑑 (1 + ( )
𝑔𝑚 𝑅𝐿′

𝑉𝑜 = 𝐼𝐿 𝑅𝐿′

𝑉𝑜 = − 𝑔𝑚 𝑉𝑔𝑠 𝑅𝐿′ ∴ 𝐼𝐿 = − 𝑔𝑚 𝑉𝑔𝑠

1/𝑠𝐶𝑖𝑛 ′
At input side, by using voltage divider rule 𝑉𝑔𝑠 = 1/𝑠𝐶 ′ 𝑉𝑠𝑖𝑔
𝑖𝑛 +𝑅𝑠𝑖𝑔

1/𝑠𝐶𝑖𝑛 𝑅𝐺 ′
𝑅𝐺
𝑉𝑔𝑠 = 𝑉 ∴ 𝑉𝑠𝑖𝑔 = 𝑉
1/𝑠𝐶𝑖𝑛 + 𝑅𝑠𝑖𝑔 𝑅𝐺 + 𝑅𝑠𝑖𝑔 𝑠𝑖𝑔
′
𝑅𝐺 + 𝑅𝑠𝑖𝑔 𝑠𝑖𝑔

28
Sub 𝑉𝑔𝑠 value in above in 𝑉𝑜 equation

𝑉𝑜 = − 𝑔𝑚 𝑉𝑔𝑠 𝑅𝐿′

1/𝑠𝐶𝑖𝑛 𝑅𝐺
𝑉𝑜 = − 𝑔𝑚 𝑉 𝑅′
1/𝑠𝐶𝑖𝑛 + 𝑅𝑠𝑖𝑔 𝑅𝐺 + 𝑅𝑠𝑖𝑔 𝑠𝑖𝑔 𝐿
′

𝑉𝑜 𝑅𝐺 𝑠𝐶𝑖𝑛
=− 𝑔𝑚 𝑅𝐿′ ′
𝑉𝑠𝑖𝑔 𝑅𝐺 + 𝑅𝑠𝑖𝑔 𝑠𝐶𝑖𝑛 + 𝑅𝑠𝑖𝑔

1/𝑠𝐶𝑖𝑛 𝑅𝐺
𝐴 = 𝐴𝑀 ′ ∴ 𝐴𝑀 = − 𝑅 𝑔𝑚 𝑅𝐿′
1/𝑠𝐶𝑖𝑛 +𝑅𝑠𝑖𝑔 𝐺 +𝑅𝑠𝑖𝑔

1
𝑠𝐶𝑖𝑛
𝐴 = 𝐴𝑀 1+𝑠𝐶𝑖𝑛 𝑅′𝑠𝑖𝑔
𝑠𝐶𝑖𝑛

𝟏
𝑨 = 𝑨𝑴 2.35
𝟏+𝒔𝑪𝒊𝒏 𝑹′𝒔𝒊𝒈

From Eq 2.35, if frequency increases, gain also decreases

Fig 2.21 the frequency-response plot of CS amplifier

1
𝐴 = 𝐴𝑀 ′
1+𝑗 𝑤 𝐶𝑖𝑛 𝑅𝑠𝑖𝑔

𝟏
𝒘𝑯 =
𝑪𝒊𝒏 𝑹′𝒔𝒊𝒈

𝐴𝑀
If 𝑤 = 𝑤𝐻 , 𝑡ℎ𝑒𝑛 𝑔𝑎𝑖𝑛 𝐴 = , 𝑠𝑜 𝑤𝐻 called as cut-off frequency or corner frequency or break
√2
frequency, or pole frequency
𝒘 𝟏
𝒇𝑯 = 𝟐 𝑯𝝅 = ′ 2.36
𝟐 𝝅 𝑪𝒊𝒏 𝑹𝒔𝒊𝒈

′
∴ 𝑅𝑠𝑖𝑔 = 𝑅𝐺 || 𝑅𝑠𝑖𝑔

29
Example 2.3

Find the midband gain 𝐀𝐌 and the upper 3-dB frequency

𝐟𝐇 𝐨𝐟 𝐚 𝐂𝐒 𝐚𝐦𝐩𝐥𝐢𝐟𝐢𝐞𝐫 𝐟𝐞𝐝 𝐰𝐢𝐭𝐡 𝐚 𝐬𝐢𝐠𝐧𝐚𝐥 𝐬𝐨𝐮𝐫𝐜𝐞 having an internal resistance 𝐑 𝐬𝐢𝐠 = 𝟏𝟎𝟎𝐤Ω. The
amplifier has 𝐑 𝐆 =4.7MΩ, 𝐑 𝐃 = 𝐑 𝐋 = 𝟏𝟓𝐊Ω, 𝐠 𝐦 = 𝟏𝐦 𝐀⁄𝐕 , 𝐫𝐨 = 𝟏𝟓𝟎𝐊Ω,𝐂𝐠𝐬 = 𝟏𝐩𝐅, 𝐚𝐧𝐝 𝐂𝐠𝐝 =
𝟎. 𝟒𝐩𝐅. Also, find the frequency of transmission zero.

Solution:

𝑅𝐺
𝐴𝑀 = − 𝑔 𝑅′
𝑅𝐺 + 𝑅𝑠𝑖𝑔 𝑚 𝐿

Where

𝑅′𝐿 = 𝑟0 ||𝑅𝐷 ||𝑅𝐿 = 150||15||15 = 7.14𝐾Ω

𝑔𝑚 𝑅′𝐿 = 1 × 7.14 = 7.14𝑉/𝑉

4.7
𝐴𝑀 = 𝑥 7.14 = −7
4.7 + 0.1

The equivalent capacitance, 𝐶𝑒𝑞 , is found as

𝐶𝑒𝑞 = (1 + 𝑔𝑚 𝑅′ 𝐿 )𝐶𝑔𝑑

= (1 + 7.14) × 0.4 = 3.26𝑝𝐹

The total capacitance 𝐶𝑖𝑛 can be now obtained as

𝐶𝑖𝑛 = 𝐶𝑔𝑠 + 𝐶𝑒𝑞 = 1 + 3.26 = 4.26𝑝𝐹

The Upper 3-dB frequency 𝑓𝐻 is found from

1
𝑓𝐻 =
′
2 𝜋 𝐶𝑖𝑛 𝑅𝑠𝑖𝑔

1
𝑓𝐻 =
2𝜋𝐶𝑖𝑛 (𝑅𝑠𝑖𝑔 ||𝑅𝐺 )

1
=
2𝜋 × 4.26 × 10−12 (0.1||4.7) × 106

=382 kHz

Finally, the transmission zero has a frequency

𝑔𝑚 1 × 10−3
𝑓𝑧 = = = 398 𝑀𝐻𝑧
2𝜋𝐶𝑔𝑑 2𝜋 × 0.4 × 10−12

Which is more than 1000 times higher than 𝑓𝐻

30
2.10 High-Frequency Response of the Common-Emitter (CE) Amplifier

Fig 2.22 The Common-Emitter Amplifier

Figure 2.23 shows the high-frequency equivalent circuit of a CE amplifier. It is obtained by replacing the
BJT in a circuit such as that in Fig. 2.22 with its high-frequency equivalent-circuit model as usual,
eliminating all dc sources and all external capacitors can be replaced by short circuit.

Fig 2.23 Equivalent circuit of Common-Emitter Amplifier in high frequency

The equivalent circuit of Fig. 10.19(a) can be simplified by utilizing Thevenin theorem at the input side
and by combining the three parallel resistances at the output side.

For equivalent voltage

′ 𝑅𝐵
By using voltage divider rule 𝑉𝑠𝑖𝑔 =𝑅 𝑉𝑠𝑖𝑔
𝐵 +𝑅𝑠𝑖𝑔

For equivalent resistance voltage source replaced with short circuit

′
𝑅𝑠𝑖𝑔 = 𝑅𝐵 || 𝑅𝑠𝑖𝑔

31
 =

For equivalent voltage

′′ 𝑟𝜋 ′
By using voltage divider rule 𝑉𝑠𝑖𝑔 =𝑟 ′ 𝑉𝑠𝑖𝑔
𝜋 +𝑟𝑥 +𝑅𝑠𝑖𝑔

′′
𝑟𝜋 𝑅𝐵
𝑉𝑠𝑖𝑔 = 𝑉
𝑟𝜋 + 𝑟𝑥 + 𝑅𝑠𝑖𝑔 𝑅𝐵 + 𝑅𝑠𝑖𝑔 𝑠𝑖𝑔
′

For equivalent resistance voltage source replaced with short circuit

′′ ′
𝑅𝑠𝑖𝑔 = 𝑟𝜋 || (𝑟𝑥 + 𝑅𝑠𝑖𝑔 )

′′
𝑅𝑠𝑖𝑔 = 𝑟𝜋 || (𝑟𝑥 + 𝑅𝐵 || 𝑅𝑠𝑖𝑔 )

𝑅𝐿′ = 𝑟𝑜 || 𝑅𝐶 || 𝑅𝐿

The simplified equivalent circuit of common emitter amplifier shown below figure

32
 Fig 2.24 the circuit of Fig 2.23 simplified at both the input side and the output side

The midband gain AM can be found from this circuit by setting 𝐶𝜋 and 𝐶𝜇 to zero.

𝑉𝑜 = 𝐼𝐿 𝑅𝐿′

𝑉𝑜 = − 𝑔𝑚 𝑉𝜋 𝑅𝐿′ ∴ 𝐼𝐿 = − 𝑔𝑚 𝑉𝜋

𝑟𝜋 𝑅𝐵
At input side, 𝑉𝜋 = 𝑉′′𝑠𝑖𝑔 = 𝑉𝑠𝑖𝑔
𝑟𝜋 +𝑟𝑥 +𝑅′𝑠𝑖𝑔 𝑅𝐵 +𝑅𝑠𝑖𝑔

𝑟𝜋 𝑅𝐵
𝑉𝜋 = 𝑉′′𝑠𝑖𝑔 = 𝑟 𝑉𝑠𝑖𝑔 ′
∴ 𝑅𝑠𝑖𝑔 =
𝜋 +𝑟𝑥 +(𝑅𝐵 || 𝑅𝑠𝑖𝑔 ) 𝑅𝐵 +𝑅𝑠𝑖𝑔

𝑅𝐵 || 𝑅𝑠𝑖𝑔

Sub 𝑉𝜋 value in above 𝑉𝑜 equation

𝑉𝑜 = − 𝑔𝑚 𝑉𝜋 𝑅𝐿′

𝑟𝜋 𝑅𝐵
𝑉𝑜 = − 𝑔𝑚 𝑉𝑠𝑖𝑔 𝑅𝐿′
𝑟𝜋 + 𝑟𝑥 + (𝑅𝐵 || 𝑅𝑠𝑖𝑔 ) 𝑅𝐵 + 𝑅𝑠𝑖𝑔

33
 𝑉𝑜 𝑟𝜋 𝑅𝐵
=−− 𝑔 𝑅𝐿′
𝑉𝑠𝑖𝑔 𝑟𝜋 + 𝑟𝑥 + (𝑅𝐵 || 𝑅𝑠𝑖𝑔 ) 𝑅𝐵 + 𝑅𝑠𝑖𝑔 𝑚

𝑟𝜋 𝑅𝐵
𝐴𝑀 = − − 𝑔𝑚 𝑅𝐿′ 2.37
𝑟𝜋 +𝑟𝑥 +(𝑅𝐵 || 𝑅𝑠𝑖𝑔 ) 𝑅𝐵 +𝑅𝑠𝑖𝑔

Midband gain without source 𝑨𝑴 = − 𝒈𝒎 𝑹′𝑳

Miller’s Theorem states that any impedance Z connected between input and output (in feedback
loop), that can be distributed to equivalent input impedance [𝑍(1 − 𝐴) ] as well as equivalent output
1
impedance [𝑧 (1 − 𝐴)]. Where A is amplifier gain.

By using Miller’s Theorem, the feedback capacitance 𝐶𝑔𝑑 can be distributed to equivalent input
capacitance 𝐶𝑒𝑞 = 𝐶𝜇 (1 − (− 𝑔𝑚 𝑅𝐿′ ) = 𝐶𝜇 (1 + 𝑔𝑚 𝑅𝐿′ ) and equivalent output capacitance 𝐶𝑒𝑞
′
= 𝐶𝜇 (1 −
1 1
( ) = 𝐶𝜇 (1 + ( ).
− 𝑔𝑚 𝑅𝐿′ 𝑔𝑚 𝑅𝐿′

′ ′
The capacitance 𝐶𝑒𝑞 value is very small compared to 𝐶𝑒𝑞 so 𝐶𝑒𝑞 can be eliminated at output side.
The simplified circuit shown in below figure.

Fig 2.24 equivalent circuit with Cμ replaced at the input side with the equivalent capacitance Ceq

𝑪𝒊𝒏 = 𝑪𝝅 + 𝐶𝑒𝑞 = 𝑪𝝅 + 𝐶𝜇 (1 + 𝑔𝑚 𝑅𝐿′ )

𝑉𝑜 = 𝐼𝐿 𝑅𝐿′

𝑉𝑜 = − 𝑔𝑚 𝑉𝜋 𝑅𝐿′ ∴ 𝐼𝐿 = − 𝑔𝑚 𝑉𝜋

34
 1/𝑠𝐶𝑖𝑛
At input side, by using voltage divider rule 𝑉𝜋 = 𝑉′′𝑠𝑖𝑔
1/𝑠𝐶𝑖𝑛 +𝑅′′𝑠𝑖𝑔

1/𝑠𝐶𝑖𝑛 𝑟𝜋 𝑅𝐵 𝑟𝜋 𝑅𝐵
𝑉𝜋 = 𝑉
𝑟𝜋 +𝑟𝑥 +(𝑅𝐵 || 𝑅𝑠𝑖𝑔 ) 𝑅𝐵 +𝑅𝑠𝑖𝑔 𝑠𝑖𝑔
∴ 𝑉′′𝑠𝑖𝑔 = 𝑟 𝑉𝑠𝑖𝑔
1/𝑠𝐶𝑖𝑛 +𝑅′′𝑠𝑖𝑔 𝜋 +𝑟 𝑥 +( 𝑅𝐵 || 𝑅𝑠𝑖𝑔 ) 𝑅𝐵 +𝑅 𝑠𝑖𝑔

Sub 𝑉𝜋 value in above 𝑉𝑜 equation

𝑉𝑜 = − 𝑔𝑚 𝑉𝜋 𝑅𝐿′

1/𝑠𝐶𝑖𝑛 𝑟𝜋 𝑅𝐵
𝑉𝑜 = − 𝑔𝑚 ′′ 𝑉𝑠𝑖𝑔 𝑅𝐿′
1/𝑠𝐶𝑖𝑛 + 𝑅𝑠𝑖𝑔 𝑟𝜋 + 𝑟𝑥 + (𝑅𝐵 || 𝑅𝑠𝑖𝑔 ) 𝑅𝐵 + 𝑅𝑠𝑖𝑔

𝑉𝑜 𝑟𝜋 𝑅𝐵 1/𝑠𝐶𝑖𝑛
= − 𝑔𝑚 𝑅𝐿′ ′′
𝑉𝑠𝑖𝑔 𝑟𝜋 + 𝑟𝑥 + (𝑅𝐵 || 𝑅𝑠𝑖𝑔 ) 𝑅𝐵 + 𝑅𝑠𝑖𝑔 1/𝑠𝐶𝑖𝑛 + 𝑅𝑠𝑖𝑔

1/𝑠𝐶𝑖𝑛 𝑟𝜋 𝑅𝐵
𝑨 = 𝑨𝑴 ′′ ∴ 𝑨𝑴 = − 𝑔𝑚 𝑅′𝐿
1/𝑠𝐶𝑖𝑛 +𝑅𝑠𝑖𝑔 𝑟𝜋 +𝑟𝑥 +(𝑅𝐵 || 𝑅𝑠𝑖𝑔 ) 𝑅𝐵 +𝑅𝑠𝑖𝑔

1
𝑠𝐶𝑖𝑛
𝐴 = 𝐴𝑀 1+𝑠𝐶𝑖𝑛 𝑅′𝑠𝑖𝑔
𝑠𝐶𝑖𝑛

𝟏
𝑨 = 𝑨𝑴 2.38
𝟏+𝒔𝑪𝒊𝒏 𝑹′′
𝒔𝒊𝒈

From Eq 2.38, if frequency increases, gain also decreases

Fig 2.25 the frequency-response plot of CE amplifier

1
𝐴 = 𝐴𝑀 ′′
1+𝑗 𝑤 𝐶𝑖𝑛 𝑅𝑠𝑖𝑔

𝟏
𝒘𝑯 =
𝑪𝒊𝒏 𝑹′𝒔𝒊𝒈
′

𝐴𝑀
If 𝑤 = 𝑤𝐻 , 𝑡ℎ𝑒𝑛 𝑔𝑎𝑖𝑛 𝐴 = , 𝑠𝑜 𝑤𝐻 called as cut-off frequency or corner frequency or break
√2
frequency, or pole frequency

35
 𝒘 𝟏
𝒇𝑯 = 𝟐 𝑯𝝅 = ′′ 2.39
𝟐 𝝅 𝑪𝒊𝒏 𝑹𝒔𝒊𝒈

′′
∴ 𝑅𝑠𝑖𝑔 = 𝑟𝜋 || (𝑟𝑥 + 𝑅𝐵 || 𝑅𝑠𝑖𝑔 )

Example 2.4

Determine the mid band gain and the upper 3-dB frequency of the common emitter amplifier for
the following case:𝑰𝑬 = 𝟏 𝒎𝑨, 𝑹𝑩 = 𝑹𝑩𝟏 ||𝑹𝑩𝟐 = 100𝒌Ω, 𝑹𝒄 = 8KΩ, 𝑹𝒔𝒊𝒈= 5KΩ, 𝑹𝑳 = 𝟓𝑲Ω, 𝜷𝟎 =
𝟏𝟎𝟎, 𝑽𝑨 = 𝟏𝟎𝟎𝑽, 𝑪µ = 𝟏𝒑𝑭 , 𝒇𝑻 = 𝟖𝟎𝟎𝑴𝑯𝒛, 𝒂𝒏𝒅 𝒓𝒙 = 𝟓𝟎Ω. Also ,determine the frequency of the
transmission zero.

SOLUTION:

The transistor is biased at 𝐼𝐶 ≅ 1 𝑚𝐴. Thus, the values of its hybrid-π model parameters are

𝐼 1 𝑚𝐴
𝑔𝑚 = 𝑉𝑐 = 25 𝑚𝑉 = 40𝑚𝐴/V
𝑟

𝛽0 100
𝑟𝑥 = = = 100𝑘Ω
𝑔𝑚 40𝑚𝐴/𝑉
𝑔𝑚
𝜔𝑇 =
𝐶𝜋 + 𝐶𝜇

𝑔𝑚 𝑔𝑚 40 × 10−3
𝐶𝜋 + 𝐶𝜇 = = = = 8𝑝𝐹
𝜔 𝑇 2𝜋 𝑓𝑇 2𝜋 × 800 × 106

𝐶𝜇 = 1𝑝𝐹

𝐶𝜇 = 7𝑝𝐹

𝑟𝑥 = 50Ω

The midband voltage gain is

𝑟𝜋 𝑅𝐵
𝑨𝑴 = − 𝑔𝑚 𝑅′𝐿
𝑟𝜋 + 𝑟𝑥 + (𝑅𝐵 || 𝑅𝑠𝑖𝑔 ) 𝑅𝐵 + 𝑅𝑠𝑖𝑔

Where

𝑅′𝐿 = 𝑟𝑜 ||𝑅𝑐 ||𝑅𝐿

=(100 || 8|| 5)𝐾Ω=3KΩ

𝑉
Thus, 𝑔𝑚 𝑅′𝐿 = 40 × 3 = 120 𝑉

100 2.5
𝐴𝑀 = − 100+5 × 2.5+0.05+(100||5) × 120

𝐴𝑀 = 𝟑𝟗

36
 1
𝑓𝐻 = ′′
2 𝜋 𝐶𝑖𝑛 𝑅𝑠𝑖𝑔

𝑪𝒊𝒏 = 𝑪𝝅 + 𝐶𝑒𝑞 = 𝑪𝝅 + 𝐶𝜇 (1 + 𝑔𝑚 𝑅𝐿′ )

𝑪𝒊𝒏 == 7 + 1(1 + 120) = 128𝑝𝐹

′′
𝑅𝑠𝑖𝑔 = 𝑟𝜋 || (𝑟𝑥 + 𝑅𝐵 || 𝑅𝑠𝑖𝑔 )
′′
𝑅𝑠𝑖𝑔 = 2.5|| [0.05 + (100 || 5)]
′′
𝑅𝑠𝑖𝑔 = 1.65kΩ

1 1
𝑓𝐻 = ′′ = −12
= 𝟕𝟓𝟒 𝒌𝑯𝒛
2 𝜋 𝐶𝑖𝑛 𝑅𝑠𝑖𝑔 2𝜋 × 128 × 10 × 1.65 × 103

𝑔𝑚
𝑓𝑧 = = 𝟔. 𝟑𝟕 𝑮𝑯𝒛
2𝜋 𝐶𝜇

2.11 High-Frequency Response of the Source-Follower (CS) Amplifier

(a)

(b)

37
 (c)

Fig 2.26 (a)A directly coupled source follower without the bias detail; (b) high-frequency equivalent
circuit of the source follower; (c) a simplified version of the equivalent circuit

Figure 2.26(a) shows a source follower without the biasing arrangement. The follower is driven by a
signal source (Vsig, Rsig) and is loaded with a resistance RL and capacitance CL. Replacing the MOSFET
with its hybrid-π equivalent-circuit model results in the equivalent circuit shown in Fig. 2.26(b).

The body-effect generator 𝑔𝑚𝑏 𝑉𝑏𝑠 also included because it plays an important role in determining the
source-follower gain. Also, the capacitances exist between the MOSFET source and ground, such as Csb,
have been lumped into CL. the body terminal B is connected to ground, the voltage 𝑉𝑏𝑠 appears across the
controlled source 𝑔𝑚𝑏 𝑉𝑏𝑠 . Thus, the controlled source 𝑔𝑚𝑏 𝑉𝑏𝑠 can be replaced by a resistance 1/𝑔𝑚𝑏 by
using the source-absorption.

1
𝑅𝐿′ = 𝑟𝑜 || 𝑅𝐿 ||
𝑔𝑚𝑏

𝑽 (𝒔)
Obtaining the Transfer Function 𝑽 𝒐 Analysis of the equivalent circuit in Fig. 2.26 (c) to determine
𝒔𝒊𝒈 (𝒔)
𝑽𝒐 (𝒔)
the transfer function proceeds as follows.
𝑽𝒔𝒊𝒈 (𝒔)

A loop equation at the input yields

𝑉𝑠𝑖𝑔 = 𝐼𝑖 𝑅𝑠𝑖𝑔 + 𝑉𝑔

where Vg can be expressed as

𝑉𝑔 = 𝑉𝑔𝑠 + 𝑉𝑜

Thus,

𝑉𝑠𝑖𝑔 = 𝐼𝑖 𝑅𝑠𝑖𝑔 + 𝑉𝑔𝑠 + 𝑉0 2.40

38
A node equation at G provides

𝐼𝑖 = 𝑠𝐶𝑔𝑑 𝑉𝑔 + 𝑠𝐶𝑔𝑠 𝑉𝑔𝑠

𝐼𝑖 = 𝑠𝐶𝑔𝑑 (𝑉𝑔𝑠 + 𝑉𝑜 ) + 𝑠𝐶𝑔𝑠 𝑉𝑔𝑠 2.41

Sub Eq 2.41 in 2.40

𝑉𝑠𝑖𝑔 = 𝐼𝑖 𝑅𝑠𝑖𝑔 + 𝑉𝑔𝑠 + 𝑉0

𝑉𝑠𝑖𝑔 = (𝑠𝐶𝑔𝑑 (𝑉𝑔𝑠 + 𝑉𝑜 ) + 𝑠𝐶𝑔𝑠 𝑉𝑔𝑠 )𝑅𝑠𝑖𝑔 + 𝑉𝑔𝑠 + 𝑉0

𝑉𝑠𝑖𝑔 = [1 + 𝑠(𝐶𝑔𝑠 + 𝐶𝑔𝑑 )𝑅𝑠𝑖𝑔 ]𝑉𝑔𝑠 + [1 + 𝑠𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 ]𝑉𝑜 2.42

A node equation at S gives

1
(𝑔𝑚 + 𝑠𝐶𝑔𝑠 )𝑉𝑔𝑠 = ( + 𝑠𝐶𝐿 ) 𝑉𝑜
𝑅𝐿′

𝑠𝐶𝑔𝑠 1 + 𝑠𝐶𝐿 𝑅𝐿′

𝑔𝑚 (1 + ) 𝑉𝑔𝑠 = ( ) 𝑉𝑜
𝑔𝑚 𝑅𝐿′

1 1+𝑠𝐶𝐿 𝑅𝐿′
𝑉𝑔𝑠 = ( 𝑠𝐶𝑔𝑠 ) 𝑉𝑜 2.43
𝑔𝑚 𝐿𝑅′ 1+
𝑔𝑚

Substituting this expression of Vgs into Eq. (2.42)

1 (1 + 𝑆𝐶𝐿 𝑅′𝐿 )
𝑉𝑠𝑖𝑔 = {1 + 𝑆(𝐶𝑔𝑠 + 𝐶𝑔𝑑 )𝑅𝑠𝑖𝑔 } 𝑉 + [1 + 𝑆𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 ]𝑉0
𝑔𝑚 𝑅′𝐿 𝐶𝑔𝑠 𝑜
(1 + 𝑆 )
𝑔𝑚
1 (1 + 𝑆𝐶𝐿 𝑅𝐿 ′)𝑉0 1 (1 + 𝑆𝐶𝐿 𝑅𝐿 ′)𝑉0 1 (1 + 𝑆𝐶𝐿 𝑅𝐿 ′)𝑉0
𝑉𝑠𝑖𝑔 = + 𝑆𝐶𝑔𝑠 𝑅𝑠𝑖𝑔 + 𝑆𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 + 𝑉0 + 𝑉0 𝑆𝐶𝑔𝑠 𝑅𝑠𝑖𝑔
𝑔𝑚 𝑅′ 𝐿 𝐶𝑔𝑠 𝑔𝑚 𝑅′ 𝐿 𝐶𝑔𝑠 𝑔𝑚 𝑅′ 𝐿 𝐶𝑔𝑠
1+𝑆 1+𝑆 1+𝑆
𝑔𝑚 𝑔𝑚 𝑔𝑚

1 (1 + 𝑆𝐶𝐿 𝑅𝐿 ′) 1 (1 + 𝑆𝐶𝐿 𝑅𝐿 ′) 1 (1 + 𝑆𝐶𝐿 𝑅𝐿 ′)

𝑉𝑠𝑖𝑔 = [1 + + 𝑆𝐶𝑔𝑠 𝑅𝑠𝑖𝑔 + 𝑆𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 + 𝑆𝐶𝑔𝑠 𝑅𝑠𝑖𝑔 ] 𝑉0
𝑔𝑚 𝑅𝐿 ′ 𝐶𝑔𝑠 𝑔𝑚 𝑅𝐿 ′ 𝐶𝑔𝑠 𝑔𝑚 𝑅𝐿 ′ 𝐶𝑔𝑠
1+𝑆 1+𝑆 1+𝑆
𝑔𝑚 𝑔𝑚 𝑔𝑚

𝑉0 1
=
𝑉𝑠𝑖𝑔 (1 + 𝑆𝐶𝐿 𝑅𝐿 ′) (1 + 𝑆𝐶𝐿 𝑅𝐿 ′) (1 + 𝑆𝐶𝐿 𝑅𝐿 ′)
1+ 𝐶𝑔𝑠 + 𝑆𝐶𝑔𝑠 𝑅𝑠𝑖𝑔 𝐶𝑔𝑠 + 𝑆𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 𝐶𝑔𝑠 + 𝑆𝐶𝑔𝑠 𝑅𝑠𝑖𝑔
𝑔𝑚 𝑅𝐿 ′ (1 + 𝑆 ) 𝑔𝑚 𝑅𝐿 ′ (1 + 𝑆 ) 𝑔𝑚 𝑅𝐿 ′ (1 + 𝑆 )
𝑔𝑚 𝑔𝑚 𝑔𝑚

𝐴𝑉
1
=
𝐶𝑔𝑠 𝐶𝑔𝑠
𝑔𝑚 𝑅′ 2 2
𝐿 (1 + 𝑆 𝑔 ) + (1 + 𝑆𝐶𝐿 𝑅𝐿 ′) + 𝑆𝐶𝑔𝑠 𝑅𝑠𝑖𝑔 +𝑆 𝐶𝐿 𝐶𝑔𝑠 𝑅𝐿 ′𝑅𝑠𝑖𝑔 + 𝑆𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 + 𝑆𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 + 𝑆 𝐶𝐿 𝐶𝑔𝑑 𝑅𝐿 ′𝑅𝑠𝑖𝑔 + 𝑆𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 (1 + 𝑆 𝑔 )
𝑚 𝑚
𝐶𝑔𝑠
𝑔𝑚 𝑅𝐿 ′ (1 + 𝑆 )
𝑔𝑚

39
𝐴𝑉
𝐶𝑔𝑠
𝑔𝑚 𝑅𝐿 ′ (1 + 𝑆 )
𝑔𝑚
=
𝐶𝑔𝑠 𝐶𝑔𝑠
𝑔𝑚 𝑅𝐿 ′ + 𝑆𝑔𝑚 𝑅𝐿 ′ + 1 + 𝑆𝐶𝐿 𝑅𝐿 ′ + 𝑆𝐶𝑔𝑠 𝑅𝑠𝑖𝑔 + 𝑆 2 𝐶𝐿 𝐶𝑔𝑠 𝑅𝐿 ′𝑅𝑠𝑖𝑔 + 𝑆𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 + 𝑆 2 𝐶𝐿 𝐶𝑔𝑑 𝑅𝐿 ′𝑅𝑠𝑖𝑔 + 𝑆𝐶𝑔𝑠 𝑅𝑠𝑖𝑔 𝑔𝑚 𝑅𝐿 ′ + 𝑆 2 𝐶𝑔𝑑 𝑅𝐿 ′𝑔𝑚 𝑅𝑠𝑖𝑔
𝑔𝑚 𝑔𝑚

𝐴𝑉
𝐶𝑔𝑠
𝑔𝑚 𝑅𝐿 ′ (1 + 𝑆 )
𝑔𝑚
=
𝐶𝑔𝑠 𝐶𝑔𝑠
1 + 𝑔𝑚 𝑅𝐿 ′ + 𝑆 [𝑔𝑚 𝑅𝐿 ′ + 𝐶𝐿 𝑅𝐿 ′ + 𝐶𝑔𝑠 𝑅𝑠𝑖𝑔 + 𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 + 𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 𝑔𝑚 𝑅𝐿 ′] + 𝑆 2 [𝐶𝐿 𝐶𝑔𝑠 𝑅𝐿 ′𝑅𝑠𝑖𝑔 + 𝐶𝐿 𝐶𝑔𝑑 𝑅𝐿 ′𝑅𝑠𝑖𝑔 + 𝐶𝑔𝑑 𝑅𝐿 ′𝑔𝑚 𝑅𝑠𝑖𝑔 ]
𝑔𝑚 𝑔𝑚

𝐶𝑔𝑠
𝑔𝑚 𝑅𝐿 ′ (1 + 𝑆
)
𝑔𝑚
𝐴𝑉 =
1 + 𝑔𝑚 𝑅𝐿 ′ + 𝑆[𝐶𝑔𝑠 𝑅𝐿 ′𝐶𝐿 𝑅𝐿 ′ + 𝐶𝑔𝑠 𝑅𝑠𝑖𝑔 + 𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 + 𝐶𝑔𝑑 𝑅𝑠𝑖𝑔 𝑔𝑚 𝑅𝐿 ′] + 𝑆 2 [𝐶𝐿 𝐶𝑔𝑠 𝑅𝐿 ′𝑅𝑠𝑖𝑔 + 𝐶𝐿 𝐶𝑔𝑑 𝑅𝐿 ′𝑅𝑠𝑖𝑔 + 𝐶𝑔𝑑 𝑅𝐿 ′𝑅𝑠𝑖𝑔 𝐶𝑔𝑠 ]

𝐶𝑔𝑠
𝑔𝑚 𝑅 ′ 𝐿 (1+𝑆 )
𝑔𝑚
Av=
1+𝑔𝑚 𝑅′ 𝐿 +𝑆{(𝐶𝑔𝑠 +𝐶𝐿 )𝑅 ′ 𝐿 +(𝐶𝑔𝑠 +𝐶𝑔𝑑 +𝐶𝑔𝑑 𝑔𝑚 𝑅′ 𝐿 )𝑅𝑠1 }+𝑆
2 [{(𝐶 ′
𝑔𝑠 +𝐶𝑔𝑑 )𝐶𝐿 +𝐶𝑔𝑠 𝐶𝑔𝑑 }𝑅𝑠𝑖𝑔 𝑅 𝐿 ]

𝐶𝑔𝑠
𝑔𝑚 𝑅′ 𝐿 (1 + 𝑆
)
𝑔𝑚
𝐴𝑣 =
{(𝐶𝑔𝑠 + 𝐶𝐿 )𝑅′ 𝐿 + (𝐶𝑔𝑠 + 𝐶𝑔𝑑 + 𝐶𝑔𝑑 𝑔𝑚 𝑅′ 𝐿 )𝑅𝑠𝑖𝑔 } {[(𝐶𝑔𝑠 + 𝐶𝑔𝑑 )𝐶𝐿 + 𝐶𝑔𝑠 𝐶𝑔𝑑 ]𝑅𝑠𝑖𝑔 𝑅′ 𝐿 }
1 + 𝑔𝑚 𝑅′ 𝐿 {1 + 𝑠 ′ + 𝑠2 }
1 + 𝑔𝑚 𝑅 𝐿 1 + 𝑔𝑚 𝑅′ 𝐿

𝐶𝑔𝑠
𝑔𝑚 𝑅′ 𝐿 (1 + 𝑆
)
𝑔𝑚
𝐴𝑣 =
1 + 𝑔𝑚 𝑅′ 𝐿 {(𝐶𝑔𝑠 + 𝐶𝐿 )𝑅′ 𝐿 + (𝐶𝑔𝑠 + 𝐶𝑔𝑑 + 𝐶𝑔𝑑 𝑔𝑚 𝑅′ 𝐿 )𝑅𝑠𝑖𝑔 } {[(𝐶𝑔𝑠 + 𝐶𝑔𝑑 )𝐶𝐿 + 𝐶𝑔𝑠 𝐶𝑔𝑑 ]𝑅𝑠𝑖𝑔 𝑅′ 𝐿 }
1 + 𝑠 + 𝑠 2
[ 1 + 𝑔𝑚 𝑅′ 𝐿 1 + 𝑔𝑚 𝑅′ 𝐿 ]

𝑪𝒈𝒔
(𝟏+𝑺 )
𝒈𝒎
𝑨𝒗 = 𝑨𝑴 [ {(𝑪𝒈𝒔 +𝑪𝑳 )𝑹′ 𝑳 +(𝑪𝒈𝒔 +𝑪𝒈𝒅 +𝑪𝒈𝒅 𝒈𝒎𝑹′ 𝑳 )𝑹𝒔𝒊𝒈 } {[(𝑪𝒈𝒔 +𝑪𝒈𝒅 )𝑪𝑳 +𝑪𝒈𝒔 𝑪𝒈𝒅 ]𝑹𝒔𝒊𝒈 𝑹′ 𝑳 }
] 2.44
𝟏+𝒔 ′ +𝒔𝟐
𝟏+𝒈𝒎 𝑹 𝑳 𝟏+𝒈𝒎 𝑹′ 𝑳

𝑔 𝑚 𝑅′ 𝐿
∴ 𝐴𝑀 =
1 + 𝑔 𝑚 𝑅′ 𝐿

𝑠
1+( )
𝑤𝑧
𝐴𝑣 (𝑠) = 𝐴𝑀 (1+𝑏 2)
2.45
1 𝑠+𝑏2 𝑠

Compare Eqs (2.44) and (2.45)

𝒈𝒎
𝒘𝒛 =
𝑪𝒈𝒔

(𝑪𝒈𝒔 + 𝑪𝑳 )𝑹′ 𝑳 + (𝑪𝒈𝒔 + 𝑪𝒈𝒅 + 𝑪𝒈𝒅 𝒈𝒎 𝑹′ 𝑳 )𝑹𝒔𝒊𝒈

𝒃𝟏 =
𝟏 + 𝒈𝒎 𝑹′ 𝑳

[(𝑪𝒈𝒔 + 𝑪𝒈𝒅 )𝑪𝑳 + 𝑪𝒈𝒔 𝑪𝒈𝒅 ]𝑹𝒔𝒊𝒈 𝑹′ 𝑳

𝒃𝟐 =
𝟏 + 𝒈𝒎 𝑹′ 𝑳

40
The two poles of the source follower can be found as the roots of the denominator polynomial 1 + 𝑏1 𝑠 +
𝑏2 𝑠 2 If the poles are real, their frequencies, say 𝑤𝑃1 and 𝑤𝑃2, can be found from

𝑠 𝑠
1 + 𝑏1 𝑠 + 𝑏2 𝑠 2 = (1 + ) (1 + )
𝑤𝑃1 𝑤𝑃2

Now if 𝑤𝑃1<< 𝑤𝑃2 (at least four times larger), a dominant pole exists with frequency 𝑤𝑃1 and the 3-dB
frequency 𝑓𝐻 is given by

𝟏
𝒇𝑯 ≅ 𝒇𝑷𝟏 ≅ 2.46
𝟐 𝝅 𝒃𝟏

If the poles are real but none is dominant, the 3-dB frequency can be determined by using below formula

1
𝑓𝐻 ≅
1 1 2
√ 2 + 2 − 𝑓2
𝑓𝑃1 𝑓𝑃2 𝑧

1
Where 𝑓𝑃1 ≅
2 𝜋 𝑏1

1
𝑓𝑃2 ≅
2 𝜋 𝑏2

1 𝒈𝒎
𝑓𝑧 ≅
2𝜋 𝑪𝒈𝒔

2.12 High-Frequency Response of the Emitter-Follower (CC) Amplifier

Fig 2.27 Emitter-Follower Circuit

• Figure 2.27 shows an emitter-follower circuit with the output signal at the emitter capacitively
coupled to a load.

41
 Fig 2.28 High-frequency equivalent circuit of emitter follower

Figure 2.28 shows the high-frequency equivalent circuit of a CE amplifier. It is obtained by replacing the
BJT in a circuit such as that in Fig. 2.27 with its high-frequency equivalent-circuit model as usual,
eliminating all dc sources and all external capacitors can be replaced by short circuit.

Figure 2.29 Simplified circuit of Figure 2.28

Figure 2.29 shows the rearranging of the circuit. Cμ is tied to ground potential and that 𝑟𝑜 is in parallel
with RE and RL. the impedance 𝑍𝑏′ looking into the base without capacitance Cμ. The current 𝐼𝑏′ entering
the parallel combination of 𝑟𝜋 and 𝐶𝜋 is the same as that coming out of the combination.

𝑅𝐿′ = 𝑟𝑜 || 𝑅𝐿 || 𝑅𝐸

𝑉𝑜 = 𝐼𝐿 𝑅𝐿′

Apply KCL at node E

𝐼𝐿 = 𝐼𝑏′ + 𝑔𝑚 𝑉𝜋

Sub 𝐼𝐿 value in 𝑉𝑜 equation

42
 𝑉𝑜 = 𝐼𝐿 𝑅𝐿′

𝑉𝑜 = (𝐼𝑏′ + 𝑔𝑚 𝑉𝜋 )𝑅𝐿′ 2.47

Voltage Vπ is given by

𝐼𝑏′ 1
𝑉𝜋 = ∴ 𝑦𝜋 = 𝑟 + 𝑠𝐶𝜋
𝑦𝜋 𝜋

𝑉𝑏 𝑉𝜋 +𝑉𝑜
the impedance 𝑍𝑏′ = = ∴ 𝑉𝑏 = 𝑉𝜋 + 𝑉𝑜
𝐼𝑏′ 𝐼𝑏′

𝐼′𝑏
𝑦𝜋
+(𝐼𝑏′ +𝑔𝑚 𝑉𝜋 )𝑅𝐿′ 𝐼′
𝑍𝑏′ = ∴ 𝑉𝜋 = 𝑦𝑏
𝐼𝑏′ 𝜋

𝐼′𝑏 𝐼′
𝑦𝜋
+(𝐼𝑏′ +𝑔𝑚 𝑦𝑏 )𝑅𝐿′
𝜋
𝑍𝑏′ = 𝐼𝑏′

1 1
𝑍𝑏′ = 𝑦 + (1 + 𝑔𝑚 𝑦 ) 𝑅𝐿′
𝜋 𝜋

1 𝑔 𝑅′
𝑍𝑏′ = 𝑦 + 𝑅𝐿′ + 𝑚𝑦 𝐿
𝜋 𝜋

1
𝑍𝑏′ = 𝑦 [1 + 𝑔𝑚 𝑅𝐿′ ] + 𝑅𝐿′
𝜋

1
𝑍𝑏′ = 1 [1 + 𝑔𝑚 𝑅𝐿′ ] + 𝑅𝐿′
+𝑠𝐶𝜋
𝑟𝜋

1
𝑍𝑏′ = 1 𝑠𝐶𝜋 + 𝑅𝐿′
+
𝑟𝜋 [1+𝑔𝑚 𝑅′𝐿 ] [1+𝑔𝑚 𝑅′𝐿 ]

1 1 1
If Z1 and Z2 are in parallel, then equivalent impedance is 𝑍 = 𝑍 + 𝑍
𝑇 1 2

1
𝑍𝑇 = 1 1
+
𝑍1 𝑍2

𝑠𝐶𝜋
Similarly, the term in 𝑍𝑏′ is equal to parallel combination of 𝑟𝜋 [1 + 𝑔𝑚 𝑅𝐿′ ] and
[1+𝑔𝑚 𝑅𝐿′ ]

1
Finally, 𝑍𝑏′ is equal to the series combination of first term 1 𝑠𝐶𝜋 and 𝑅𝐿′
+
𝑟𝜋 ([1+𝑔𝑚 𝑅′𝐿 ]) [1+𝑔𝑚 𝑅′𝐿 ]

43
 𝑍𝑏′ =

Figure 2.30 High-frequency equivalent circuit with effective input base impedance

In many applications, the impedance of rπ(1 + gm R’L) in parallel with Cπ/(1 + gm 𝑅𝐿′ )) is large compared
to 𝑅𝐿′ ). If we neglect 𝑅𝐿′ . Then modified circuit is

𝐶
𝐶𝑖𝑛 = 𝐶𝜇 + (1+ 𝑔𝜋 ′
𝑚 𝑅𝐿

44
Total Capacitance in the above circuit 𝐶𝑇 = 𝐶𝑖𝑛

′
Total resistance seen by 𝐶𝑇 is equal to 𝑅𝑆 || 𝑅𝐵 || 𝑟𝜋 [1 + 𝑔𝑚 𝑅 ]
𝐿

Time constant 𝜏 = 𝐶𝑇 𝑅𝑇 = 𝐶𝑖𝑛 (𝑅𝑆 || 𝑅𝐵 || 𝑟𝜋 [1 + 𝑔𝑚 𝑅𝐿′ ])

𝐶𝜋
𝜏 = (𝐶 𝜇 + ) (𝑅𝑆 || 𝑅𝐵 || 𝑟𝜋 [1 + 𝑔𝑚 𝑅𝐿′ ])
(1 + 𝑔𝑚 𝑅′𝐿

1
The 3 dB frequency or upper cut-off frequency is 𝑓𝐻 = 2𝜋𝜏

𝟏
𝒇𝑯 = 2.48
𝑪𝝅
𝟐 𝝅 (𝑪𝝁 + )(𝑹𝑺 || 𝑹𝑩 || 𝒓𝝅 [𝟏+𝒈𝒎 𝑹′𝑳 ])
(𝟏+ 𝒈𝒎 𝑹′𝑳

𝑉𝑜 = (𝐼𝑏′ + 𝑔𝑚 𝑉𝜋 )𝑅𝐿′

𝐼′ 𝐼′
𝑉𝑜 = (𝐼𝑏′ + 𝑔𝑚 𝑦𝑏 ) 𝑅𝐿′ ∴ 𝑉𝜋 = 𝑦𝑏
𝜋 𝜋

𝑦 +𝑔𝑚
𝑉𝑜 = 𝐼𝑏′ ( 𝜋𝑦 ) 𝑅𝐿′
𝜋

𝑉𝑜 = 𝑉𝜋 (𝑦𝜋 + 𝑔𝑚 )𝑅𝐿′

For frequency of zero 𝑉𝑜 = 0

For 𝑉𝑜 = 0, 𝑉𝜋 ≠ 0, 𝑅′𝐿 ≠ 0 So 𝑦𝜋 + 𝑔𝑚 = 0

45
1
+ 𝑠𝐶𝜋 + 𝑔𝑚 = 0
𝑟𝜋

1
+ 𝑠𝐶𝜋 = −𝑔𝑚
𝑟𝜋

1 + 𝑠𝐶𝜋 𝑟𝜋 = −𝑔𝑚 𝑟𝜋

1 + 𝑠𝐶𝜋 𝑟𝜋 = −𝛽 ∴ 𝑔𝑚 𝑟𝜋 = 𝛽

𝑠𝐶𝜋 𝑟𝜋 = −(1 + 𝛽)

𝑗𝑤𝑧 𝐶𝜋 𝑟𝜋 = −(1 + 𝛽)

𝑤𝑧 𝐶𝜋 𝑟𝜋 = (1 + 𝛽)

(1 + 𝛽)
𝑤𝑧 =
𝐶𝜋 𝑟𝜋

(1 + 𝛽)
𝑓𝑧 =
2 𝜋𝐶𝜋 𝑟𝜋
1
𝑓𝑧 = 2 𝜋𝐶𝜋 𝑟𝜋
(1+𝛽)

𝟏
𝒇𝒛 𝒐𝒓 𝒇𝟎 = 𝒓𝝅 2.49
𝟐 𝝅𝑪𝝅 ( )
(𝟏+𝜷)

• Since rπ/(1 + β) is small, frequency 𝒇𝒛 is usually very high.

Example 2.5

Find the midband gain AM, upper 3-dB frequency fH, Lower 3-dB frequency fL, bandwidth and Unity-Gain
Frequency fT of a CS amplifier fed with a signal source having an internal resistance Rsig = 100 ΚΩ. The
amplifier has RG = 4.7 ΜΩ, RD = RL = 15ΚΩ, RS=10 KΩ, gm = 1 mA/V, ro = 150 ΚΩ, Cgs = 1 pF, and Cgd = 0.4
pF, Cs = 1.75 µF, CC1 = 3.3 nF, CC2 = 0.53µF

46
Solution:
𝑅𝐺
𝐴𝑀 = − 𝑔 𝑅′
𝑅𝐺 + 𝑅𝑠𝑖𝑔 𝑚 𝐿

Where

𝑅′𝐿 = 𝑟0 ||𝑅𝐷 ||𝑅𝐿 = 150||15||15 = 7.14𝐾Ω

𝑔𝑚 𝑅′𝐿 = 1 × 7.14 = 7.14𝑉/𝑉

4.7
𝐴𝑀 = 𝑥 7.14 = −7
4.7 + 0.1
1
1
𝑔𝑚 + 𝑅
𝑆
𝑓𝐿 = 𝑓𝑃2 = 2𝜋 𝐶𝑠
= 𝟏𝟎𝟎 𝑯𝒛

1
𝑓𝐻 =
′
2 𝜋 𝐶𝑖𝑛 𝑅𝑠𝑖𝑔

𝐶𝑖𝑛 = 𝐶𝑔𝑠 + 𝐶𝑒𝑞 𝑅′𝑠𝑖𝑔 = 𝑅𝑠𝑖𝑔 ||𝑅𝐺

The equivalent capacitance, 𝐶𝑒𝑞 , is found as

𝐶𝑒𝑞 = (1 + 𝑔𝑚 𝑅′ 𝐿 )𝐶𝑔𝑑

= (1 + 7.14) × 0.4 = 3.26𝑝𝐹

The total capacitance 𝐶𝑖𝑛 can be now obtained as

𝐶𝑖𝑛 = 𝐶𝑔𝑠 + 𝐶𝑒𝑞 = 1 + 3.26 = 4.26𝑝𝐹

The Upper 3-dB frequency 𝑓𝐻 is found from

1
𝑓𝐻 =
′
2 𝜋 𝐶𝑖𝑛 𝑅𝑠𝑖𝑔

1
𝑓𝐻 =
2𝜋𝐶𝑖𝑛 (𝑅𝑠𝑖𝑔 ||𝑅𝐺 )

1
=
2𝜋 × 4.26 × 10−12 (0.1||4.7) × 106

=382 kHz

Bandwidth = 𝑓𝐻 − 𝑓𝐿 = 382 𝑘𝐻𝑧 − 100 𝐻𝑧 = 𝟑𝟖𝟏. 𝟗 𝒌𝑯𝒛

𝑔𝑚
𝑓𝑇 = = 3.4 GHz
2𝜋 (𝐶𝑔𝑠 +𝐶𝑔𝑑 )

47