Analog circuits

Joycee Mekie

Autumn 2013

Course outline
Structure L-T-P-C: 3-1-3-5
Contents: Review of basic concepts of
semiconductor devices, Physics of MOS
devices, CMOS amplifiers, Cascodes and
current mirrors, Differential amplifiers, Analog
filters, Frequency response, Feedback and
stability, Analog to digital converter, Digital to
analog converter
Textbooks and References
Behzad Razavi, Fundamentals of
Microelectronics, Wiley, 2008
Sedra and Smith, Microelectronic circuits,
Oxford university press, 2010
Allen P. and Holberg D., CMOS Analog circuit
design, Oxford university press, 3
rd
Ed.
Carusone T., Johns D. and Martin K., Analog
integrated circuit design, 2
nd
Ed., Wiley
Course plan
Theory (70%)
Quiz-1 - 10%
Mid-semester exams 20%
Quiz-2 - 10%
End-semester exams 25%
Assignments 5%
Laboratory (30%)
Continuous evaluation (Quizzes, viva-voce, etc.)
attendance, lab-reports, exams
Project work
Why Analog circuits?
The world around us is Analog
audio, video, sensor information, etc.
Noise removal (at analog stage or digital
stage)
Amplify the signals (to be sampled as proper
1 and 0)
Converting back the signals to Analog domain







Example: Cellular technology
Microelectronics exist in black boxes that
process received and transmitted voice signals
Frequency Up-conversion
Voice is up-converted by multiplying two sinusoids.
When multiplying two sinusoids in time domain, their
spectra are convolved in frequency domain.

Transmitter
Two frequencies are multiplied and radiated by an antenna in (a).
A power amplifier is added in (b) to boost the signal.
Receiver
High frequency is translated to DC by multiplying by f
C
.
A low-noise amplifier is needed for signal boosting
without excessive noise.

Digital camera
Pixel array
Signal processing in a typical system
Analog
processing
Analog-to-Digital
conversion
Digital
processing
and storage
Analog signal
Gain,
Power,
Speed,
.
Noise
distortion
Digital or Analog
Finite rise and fall times of the signal x2(t) raises
several issues. Can such a signal be fed directly to the
flip-flops, latches?
Semiconductors
Charge carriers in semiconductors
Periodic table
This abridged table contains elements with three to five
valence electrons, with Si being the most important.
Silicon
Si has four valence electrons. Therefore, it can form
covalent bonds with four of its neighbors.
When temperature goes up, electrons in the covalent
bond can become free.

Electron-hole Pair interaction
With free electrons breaking off covalent bonds, holes
are generated.
Holes can be filled by absorbing other free electrons,
so effectively there is a flow of charge carriers.

Free electron density at Given
Temperature
E
g
, or bandgap energy determines how much effort is needed to
break off an electron from its covalent bond.
There exists an exponential relationship between the free-electron
density and bandgap energy
For Silicon, Eg = 1.12 eV
3 15 0
3 10 0
3 2 / 3 15
/ 10 54 . 1 ) 600 (
/ 10 08 . 1 ) 300 (
/
2
exp 10 2 . 5
cm electrons K T n
cm electrons K T n
cm electrons
kT
E
T n
i
i
g
i
= =
= =

=
Free electron density at Given
Temperature
3 15 0
3 10 0
3 2 / 3 15
/ 10 54 . 1 ) 600 (
/ 10 08 . 1 ) 300 (
/
2
exp 10 2 . 5
cm electrons K T n
cm electrons K T n
cm electrons
kT
E
T n
i
i
g
i
= =
= =

=
Silicon has 5x10
22
atoms/cm
3
np = n
i
2
Doping (N-type)
Pure Si can be doped with other elements to change its
electrical properties.
For example, if Si is doped with P (phosphorous), then
it has more electrons, or becomes type N (electron).
np = n
i
2


Problem
A piece of crystalline silicon is doped
uniformly with phosphorous atoms. The
doping density is 10
16
atoms/cm
3
. Determine
the electron and hole densities in this material
at the room temperature.

Doping (P-type)
If Si is doped with B (boron), then it has more
holes, or becomes type P.

Summary of charge carriers
Electron and Hole densities
The product of electron and hole densities is
ALWAYS equal to the square of intrinsic electron
density regardless of doping levels.

2
i
n np =
Majority Carriers :
Minority Carriers :
Majority Carriers :
Minority Carriers :

D
i
D
A
i
A
N
n
p
N n
N
n
n
N p
2
2
~
~
~
~
Charge transport mechanism: Drift
The process in which charge particles move because of an
electric field is called drift.
Charge particles will move at a velocity that is proportional
to the electric field.



=
=
E v
E v
n e
p h

Problem
A uniform piece of n-type of Silicon that is 1 m
long senses a voltage of 1V. How much time
does it take to cross the 1-m length?
Current flow
Electric current is calculated as the amount of charge in
v meters that passes thru a cross-section if the charge
travel with a velocity of v m/s.
q n h W v I =
Current flow: Drift
Since velocity is equal to E, drift characteristic is
obtained by substituting V with E in the general
current equation.
The total current density consists of both electrons and
holes.
E p n q
q p E q n E J
q n E J
p n
p n tot
n n
) (


+ =
+ =
=
Problem
How should we select the carrier densities so
that we get equal electron and hole drift
currents?
Velocity saturation
In reality, velocity does not increase linearly with electric
field. It will eventually saturate to a critical value.

E
v
E
v
b
v
bE
sat
sat
0
0
0
0
1
1

+
=
=
+
=
Second charge transport mechanism:
Diffusion
Charge particles move from a region of high concentration
to a region of low concentration. It is analogous to an
every day example of an ink droplet in water.
Example of diffusion

Current flow: Diffusion
Diffusion current is proportional to the gradient of charge
(dn/dx) along the direction of current flow.
Its total current density consists of both electrons and
holes.
dx
dn
qD J
dx
dn
AqD I
n n
n
=
=
) (
dx
dp
D
dx
dn
D q J
dx
dp
qD J
p n tot
p p
=
=
Example: Linear vs non-linear charge
density profile
Linear charge density profile means constant diffusion
current, whereas nonlinear charge density profile
means varying diffusion current
L
N
qD
dx
dn
qD J
n n n
= =
d d
n
n
L
x
L
N qD
dx
dn
qD J

= = exp
Example: Linear vs non-linear charge
density profile
Linear charge density profile means constant diffusion
current, whereas nonlinear charge density profile
means varying diffusion current
L
N
qD
dx
dn
qD J
n n n
= =
d d
n
n
L
x
L
N qD
dx
dn
qD J

= = exp
Operation of MOS Transistors
Operation in Triode region
Operation in Saturation region
I/V characteristics
Metal-Oxide-Semiconductor (MOS)
Capacitor
The MOS structure can be thought of as a parallel-plate
capacitor, with the top plate being the positive plate, oxide
being the dielectric, and Si substrate being the negative plate.
(We are assuming P-substrate.)

Structure and Symbol of MOSFET
This device is symmetric, so either of the n+ regions can be source or drain.

MOS Structure
The gate is formed by polysilicon, and the insulator by
Silicon dioxide.

Formation of channel
Formation of channel
First, the holes are repelled by the positive gate voltage,
leaving behind negative ions and forming a depletion
region.
Formation of channel
Next, electrons are attracted to the interface,
creating a channel (inversion layer).
MOSFET Characteristics
MOSFET Characteristics
MOSFET Characteristics
Problem
Sketch the I
D
V
G
and I
D
V
D
characteristics
for (a) different channel lengths, and (b)
different oxide thickness.
L and t
ox
Dependence
As length increases, on-resistance increases
L and t
ox
Dependence
As t
ox
increases, C decreases, Q decreases (higher on-
resistance)
Voltage-Dependent resistor
The inversion channel of a MOSFET can be seen as a
resistor.
Since the charge density inside the channel depends on the
gate voltage, this resistance is also voltage-dependent.

Effect of W
As the gate width increases, the current increases due
to a decrease in resistance. However, gate capacitance
also increases thus, limiting the speed of the circuit.


Effect of W
An increase in W can be seen as two devices in parallel.

MOSFET structure
Channel Potential Variation
Since theres a channel resistance between drain and source, and if
drain is biased higher than the source, channel potential increases
from source to drain, and the potential between gate and channel
will decrease from source to drain.

Channel Pinch-Off
As the potential difference between drain and gate
becomes more positive, the inversion layer beneath
the interface starts to pinch off around drain.

Channel Pinch-Off
When V
D
V
G
= V
th
, the channel at drain
totally pinches off
Channel Pinch-Off
When V
D
V
G
> V
th
, the channel length starts
to decrease.
Channel Pinch-Off
Once the electrons reach end of channel, they experience
high electric field in the dep. Region surrounding the drain
region, and are swept into the drain terminal.
Charge density at a point
Let x be a point along the channel from source to drain, and
V(x) its potential; the expression above gives the charge
density (per unit length).

| |
TH GS ox
V x V V WC x Q = ) ( ) (
Charge density and current
The current that flows from source to drain (electrons)
is related to the charge density in the channel by the
charge velocity.

Charge density and current
The current that flows from source to drain (electrons)
is related to the charge density in the channel by the
charge velocity.

v Q I =
Drain current
| |
| |
2
) ( 2
2
1
) (
) (
DS DS TH GS ox n D
n TH GS ox D
n
V V V V
L
W
C I
dx
x dV
V x V V WC I
dx
dV
v
=
=
+ =

Parabolic I
D
-V
DS
Relationship
By keeping V
G
constant and varying V
DS
, we obtain a
parabolic relationship.
The maximum current occurs when V
DS
equals to V
GS
- V
TH
.

I
D
-V
DS
for Different Values of V
GS

( )
2
max , TH GS D
V V I
Linear Resistance
At small V
DS
, the transistor can be viewed as a resistor, with the
resistance depending on the gate voltage.
It finds application as an electronic switch.

( )
TH GS ox n
on
V V
L
W
C
R

1
Different Regions of Operation
Triode or saturation?

When the potential difference between gate
and drain is greater than V
TH
, the MOSFET is in
triode region.
When the potential difference between gate
and drain becomes equal to or less than V
TH
,
the MOSFET enters saturation region.

Triode or Saturation?
When the region of operation is not known, a region is
assumed (with an intelligent guess). Then, the final answer
is checked against the assumption.

Problem
Calculate the drain current of M1. Assume C
ox
= 100
A/V
2
and V
TH
= 0.4 V. If the gate voltage increases by
10 mV, what is the change in the drain voltage?

Problem
Determine the value of W/L that places M
1
at
the edge of saturation.
Problem
Calculate the maximum allowable gate voltage
if M1 must remain saturated?
Channel length modulation
The original observation that the current is constant in the saturation
region is not quite correct. The end point of the channel actually moves
toward the source as V
D
increases, increasing I
D
. Therefore, the current in
the saturation region is a weak function of the drain voltage.

( ) ( )
DS TH GS ox n D
V V V
L
W
C I + = 1
2
1
2
and L
For long L, the channel-length modulation effect
is less than that of short L.

Problem
A MOSFET carries a drain current of 1 mA with
V
DS
= 0.5V in saturation. Determine the
change in I
D
if V
DS
rises by 1V and = 0.1V
-1
.
What is the device output impedance?
Doubling of g
m
Due to Doubling W/L
If W/L is doubled, effectively two equivalent
transistors are added in parallel, thus doubling the
current (if V
GS
-V
TH
is constant) and hence g
m
.
Body Effect
As the source potential departs from the bulk
potential, the threshold voltage changes.

( )
F SB F TH TH
V V V | | 2 2
0
+ + =
Large-signal models
Small-signal model
When the bias point is not perturbed significantly, small-signal
model can be used to facilitate calculations.
To represent channel-length modulation, an output resistance is
inserted into the model.
D
o
I
r

1
~
PMOS Transistor
It behaves like an NMOS device with all the
polarities reversed.

Operation of PMOS

To create an inversion layer in the n-type substrate, we must attract holes to the gate
electrode. A p-type channel will be induced, connecting the p+ wells at the drain
and the source.

However, to attract holes toward the gate, the voltage VGS must be sufficiently
negative! The threshold voltage Vt is thus a negative value, so that a channel is
induced only if VGS <Vt (i.e., VGS is more negative than Vt ).

Likewise, we find that we typically get current to flow through this channel by
making the voltage VDS negative. If we make the voltage VDS sufficiently negative,
the p-type induced channel will pinch off.

Note that when VDS is negative, the drain current will flow from the PMOS source,
to the PMOS drain (i.e., exactly opposite that of the NMOS device with a positive
VDS).

I
D
-V
DS
curve for PMOS
Current I
D
flows from source to drain (exactly opposite to that of NMOS)
Extra slides
Einsteins Relation
While the underlying physics behind drift and diffusion currents are
totally different, Einsteins relation provides a mysterious link
between the two.
q
kT D
=

Extra slides: Convolution