Professional Documents
Culture Documents
Digital System Design
Digital System Design
Semiconductor
SEMICONDUCTOR MATERIALS
Semiconductors are a group of materials having conductivities between those of metals and insulators.
Silicon is by far the most common semiconductor used in integrated circuits and will be emphasized to a great
extent.
Gallium arsenide is one of the more common of the compound semiconductors. Its good optical properties
make it useful in optical devices. GaAs is also used in specialized applications in which, for example, high
speed is required.
Semiconductor
TYPES OF SOLIDS
Amorphous, polycrystalline, and single crystals are the three general types of solids.
The single-crystal regions are called grains and are separated from one another by grain boundaries.
Single-crystal materials, ideally, have a high degree of order, or regular geometric periodicity,
throughout the entire volume of the material.
The advantage of a single-crystal material is that, in general, its electrical properties are superior
Semiconductor
SPACE LATTICES
A representative unit, or a group of atoms, is repeated at regular
intervals in each of the three dimensions to form the single crystal. The
periodic arrangement of atoms in the crystal is called the lattice.
Basic Crystal Structures
Figure shows the simple cubic, body-centered cubic, and face-centered
cubic structures.
Semiconductor
THE DIAMOND STRUCTURE
Silicon is referred to as a group IV element and has a diamond crystal structure.
Germanium is also a group IV element and has the same diamond structure.
Semiconductor
Atomic Bonding
The interaction that occurs between atoms to form a solid and to reach the minimum total energy
depends on the type of atom or atoms involved.
The type of bond, or interaction, between atoms, then, depends on the particular atom or atoms in the
crystal. If there is not a strong bond between atoms, they will not “stick together” to create a solid.
Semiconductor
IMPERFECTIONS AND IMPURITIES IN SOLIDS
• In a real crystal, the lattice is not perfect, but contains imperfections or defects; that is, the perfect
geometric periodicity is disrupted in some manner.
• Imperfections tend to alter the electrical properties of a material and, in some cases, electrical
parameters can be dominated by these defects or impurities.
Semiconductor
Imperfections in Solids
• One type of imperfection that all crystals have in common is atomic thermal vibration.
• The atoms in a crystal, however, have a certain thermal energy, which is a function of temperature.
The thermal energy causes the atoms to vibrate in a random manner about an equilibrium lattice
point.
• This random thermal motion causes the distance between atoms to randomly fluctuate, slightly
disrupting the perfect geometric arrangement of atoms. This imperfection, called lattice
vibrations, affects some electrical parameters.
Semiconductor
Imperfections in Solids
• Second type of defect is called a point defect.
• In an ideal single-crystal lattice, the atoms are arranged in a perfect periodic arrangement.
However, in a real crystal, an atom may be missing from a particular lattice site. This defect is
referred to as a vacancy.
• In another situation, an atom may be located between lattice sites. This defect is referred to as an
interstitial.
• In the case of vacancy and interstitial defects, not only is the perfect geometric arrangement of
atoms broken but also the ideal chemical bonding between atoms is disrupted, which tends to
change the electrical properties of the material.
Semiconductor
Imperfections in Solids
• The point defects involve single atoms or single-atom locations. In forming single-crystal
materials, more complex defects may occur.
• A line defect, for example, occurs when an entire row of atoms is missing from its normal lattice
site. This defect is referred to as a line dislocation
Semiconductor
Impurities in Solids
• Foreign atoms, or impurity atoms, may be present in a crystal lattice. Impurity atoms may be
located at normal lattice sites, in which case they are called substitutional impurities.
• Impurity atoms may also be located between normal sites, in which case they are called interstitial
impurities.
• Some impurities, such as oxygen in silicon, tend to be essentially inert; however, other impurities,
such as gold or phosphorus in silicon, can drastically alter the electrical properties of the material.
Semiconductor
Impurities in Solids
• The technique of adding impurity atoms to a semiconductor material in order to change its
conductivity is called doping. There are two general methods of doping: impurity diffusion and
ion implantation.
Diffusion Process
• Impurity diffusion occurs when a semiconductor crystal is placed in a high temperature (1000ºC)
gaseous atmosphere containing the desired impurity atom.
• At this high temperature, many of the crystal atoms can randomly move in and out of their single-
crystal lattice sites. Vacancies may be created by this random motion so that impurity atoms can
move through the lattice by hopping from one vacancy to another.
• Impurity diffusion is the process by which impurity particles move from a region of high
concentration near the surface to a region of lower concentration within the crystal.
• When the temperature decreases, the impurity atoms become permanently frozen into the
substitutional lattice sites.
• Diffusion of various impurities into selected regions of a semiconductor allows us to fabricate
complex electronic circuits in a single semiconductor crystal.
Semiconductor
Ion Implementation
• Ion implantation generally takes place at a lower temperature than diffusion.
• A beam of impurity ions is accelerated to kinetic energies in the range of 50 keV or greater and
then directed to the surface of the semiconductor.
• The high-energy impurity ions enter the crystal and come to rest at some average depth from the
surface.
• One advantage of ion implantation is that controlled numbers of impurity atoms can be introduced
into specific regions of the crystal.
• A disadvantage of this technique is that the incident impurity atoms collide with the crystal atoms,
causing lattice-displacement damage.
• The lattice damage can be removed by thermal annealing, in which the temperature of the crystal
is raised for a short time. Thermal annealing is a required step after implantation.
Semiconductor
Ion Implementation
• Ion implantation generally takes place at a lower temperature than diffusion.
• A beam of impurity ions is accelerated to kinetic energies in the range of 50 keV or greater and
then directed to the surface of the semiconductor.
• The high-energy impurity ions enter the crystal and come to rest at some average depth from the
surface.
• One advantage of ion implantation is that controlled numbers of impurity atoms can be introduced
into specific regions of the crystal.
• A disadvantage of this technique is that the incident impurity atoms collide with the crystal atoms,
causing lattice-displacement damage.
• The lattice damage can be removed by thermal annealing, in which the temperature of the crystal
is raised for a short time. Thermal annealing is a required step after implantation.
ELECTRICAL CONDUCTION IN SOLIDS
The Energy Band and the Bond Model
• Figure shows a two dimensional representation of the covalent bonding in a single-crystal silicon
lattice at T= 0K
• The splitting of the discrete silicon energy states into bands of allowed energies as the silicon
crystal is formed is shown in Fig.
ELECTRICAL CONDUCTION IN SOLIDS
The Energy Band and the Bond Model
As the temperature increases above 0 K, a few valence band electrons may gain enough thermal
energy to break the covalent bond and jump into the conduction band.
• The semiconductor is neutrally charged. This means that, as the negatively charged electron
breaks away from its covalent bonding position, a positively charged “empty state” is created in
the original covalent bonding position in the valence band.
• As the temperature further increases, more covalent bonds are broken, more electrons jump to the
conduction band, and more positive “empty states” are created in the valence band.
ELECTRICAL CONDUCTION IN SOLIDS
The Energy Band and the Bond Model
• The energy states in the valence band are completely full and the states in the conduction band are
empty.
• For T > 0 K, in which some electrons have gained enough energy to jump to the conduction band
and have left empty states in the valence band. At this point there is no external forces are applied
so the electron and “empty state” distributions are symmetrical with k .
ELECTRICAL CONDUCTION IN SOLIDS
The Energy Band and the Bond Model
Each crystal has its own energy-band structure.
• If an electric field is applied, there are no particles to move, so
there will be no current in Fig. (a).
• Figure (b) shows another allowed energy band whose energy states
are completely full of electrons. A completely full energy band will
also not give rise to a current.
• A material that has energy bands either completely empty or
completely full is an insulator.
• The resistivity of an insulator is very large or, conversely, the
conductivity of an insulator is very small.
• There are essentially no charged particles that can contribute to a
drift current.
• Figure (c) shows a simplified energy-band diagram of an insulator.
The bandgap energy Eg of an insulator is usually on the order of
3.5 to 6 eV or larger, so that at room temperature.
ELECTRICAL CONDUCTION IN SOLIDS
The Energy Band and the Bond Model
The number density N (E) is the number of particles per unit volume per unit energy and the function
g (E) is the number of quantum states per unit volume per unit energy.
The function fF (E) is called the Fermi–Dirac distribution or probability function and gives the
probability that a quantum state at the energy E will be occupied by an electron.
energy-band diagrams for degenerately doped (a) n-type and (b) p-type
semiconductors.
Degenerate and Nondegenerate Semiconductors
Energy-band diagram at T=0 K for (a) n-type and (b) p-type semiconductors.
PN-Junction
• The entire semiconductor is a single-crystal material in which one region is doped with acceptor
impurity atoms to form the p region and the adjacent region is doped with donor atoms to form the n
region.
• The interface separating the n and p regions is referred to as the metallurgical junction.
• Initially, at the metallurgical junction, there is a very large density gradient in both electron and
hole concentrations.
• Majority carrier electrons in the n region will begin diffusing into the p region, and majority
carrier holes in the p region will begin diffusing into the n region.
PN-Junction
• If we assume there are no external connections to the semiconductor, then this diffusion process
cannot continue indefinitely.
• As electrons diffuse from the n region, positively charged donor atoms are left behind. Similarly,
as holes diffuse from the p region, they uncover negatively charged acceptor atoms.
• The net positive and negative charges in the n and p regions induce an electric field in the region
near the metallurgical junction, in the direction from the positive to the negative charge, or from
the n to the p region.
PN-Junction
• The net positively and negatively charged regions are shown in Figure.
• These two regions are referred to as the space charge region. Essentially all electrons and holes
are swept out of the space charge region by the electric field. Since the space charge region is
depleted of any mobile charge, this region is also referred to as the depletion region.
PN-Junction
Zero Applied Bias
Built-in Potential Barrier
• If we assume that no voltage is applied across the pn junction, then the junction is in thermal
equilibrium—the Fermi energy level is constant throughout the entire system.
• Figure shows the energy-band diagram for the pn junction in thermal equilibrium.
• The conduction and valance band energies must bend as we go through the space charge region,
since the relative position of the conduction and valence bands with respect to the Fermi energy
changes between p and n regions.
PN-Junction
Built-in Potential Barrier
• Electrons in the conduction band of the n region see a potential barrier in trying to move into the
conduction band of the p region. This potential barrier is referred to as the built-in potential
barrier and is denoted by Vbi.
• The built-in potential barrier maintains equilibrium between majority carrier electrons in the n
region and minority carrier electrons in the p region, and also between majority carrier holes in the
p region and minority carrier holes in the n region.
• This potential difference across the junction cannot be measured with a voltmeter because new
potential barriers will be formed between the probes and the semiconductor that will cancel Vbi.
The potential Vbi maintains equilibrium, so no current is produced by this voltage.
PN-Junction
PN-Junction
Electric Field
• An electric field is created in the depletion region by the
separation of positive and negative space charge densities.
• We will assume that the space charge region abruptly ends in
then region at x=xn and abruptly ends in the p region at x=-xp.
where φ(x) is the electric potential, E(x) is the electric field, ρ(x)
is the volume charge density, and εs is the permittivity of the
semiconductor
PN-Junction
Electric Field
From Figure, the charge densities are
Above equation states that the number of negative charges per unit area in the p region is equal to the
number of positive charges per unit area in the n region.
PN-Junction
Electric Field
•
The Two-terminal MOS Structure
Energy Band Diagram
• Figure shows a MOS capacitor with a p-type semiconductor substrate.
• The top metal gate is at a negative voltage with respect to the semiconductor substrate. A negative
charge will exist on the top metal plate and an electric field will be induced.
• If the electric field were to penetrate into the semiconductor, the majority carrier holes would
experience a force toward the oxide– semiconductor interface.
• An accumulation layer of holes at the oxide–semiconductor junction corresponds to the positive
charge on the bottom “plate” of the MOS capacitor.
The Two-terminal MOS Structure
Energy Band Diagram
• Figure shows the same MOS capacitor in which the polarity of the applied voltage is reversed.
• A positive charge now exists on the top metal plate and the induced electric field is in the opposite
direction as shown. If the electric field penetrates the semiconductor in this case, majority carrier
holes will experience a force away from the oxide–semiconductor interface.
• As the holes are pushed away from the interface, a negative space charge region is created because
of the fixed ionized acceptor atoms. The negative charge in the induced depletion region
corresponds to the negative charge on the bottom “plate” of the MOS capacitor.
The Two-terminal MOS Structure
Energy Band Diagram
Figure shows the ideal case when zero bias is applied across the MOS device. The energy bands in the
semiconductor are flat indicating no net charge exists in the semiconductor. This condition is known as
flat band.
The Two-terminal MOS Structure
Energy Band Diagram
• Figure shows the energy-band diagram for the case when a negative bias is applied to the gate.
(Remember that positive electron energy is plotted “upward” and positive voltage is plotted
“downward.”)
• The valence-band edge is closer to the Fermi level at the oxide–semiconductor interface than in the
bulk material, which implies that there is an accumulation of holes.
• The semiconductor surface appears to be more p-type than the bulk material. The Fermi level is a
constant in the semiconductor since the MOS system is in thermal equilibrium and there is no
current through the oxide.
The Two-terminal MOS Structure
Energy Band Diagram
• Figure shows the energy-band diagram of the MOS system when a positive voltage is applied to the
gate.
• The conduction- and valence-band edges bend as shown in the figure, indicating a space charge
region similar to that in a pn junction.
• The conduction band and intrinsic Fermi levels move closer to the Fermi level. The induced space
charge width is xd.
The Two-terminal MOS Structure
Energy Band Diagram
• A larger positive voltage is applied to the top metal gate of the MOS capacitor. In this case the
induced electric field to increase in magnitude and the corresponding positive and negative charges
on the MOS capacitor to increase.
• A larger negative charge in the MOS capacitor implies a larger induced space charge region and
more band bending.
• The conduction band at the surface is now close to the Fermi level, whereas the valence band is
close to the Fermi level in the bulk semiconductor.
• This result implies that the surface in the semiconductor adjacent to the oxide–semiconductor
interface is n type.
The Two-terminal MOS Structure
Work Function Differences
• Vacuum level be the minimum energy required
for an electron to free itself from the material.
• The work function is the minimum amount of
energy needed to eject an electron from the
surface of a metal.
• The difference between the vacuum level and the
Fermi level (E0-Ef) is called the work function,
and the difference between the vacuum level and
the conduction band edge (E0-Ec) is called the
electron affinity.
• The electron affinity is fixed for a semiconductor
material, while the work function varies with the
position of the fermi level (function of doping
type and concentration).
The Two-terminal MOS Structure
Work Function Differences
If a gate voltage is applied, the potential drop across the oxide and the surface potential will change. We
can then write
The Two-terminal MOS Structure
Flat-Band Voltage
There is zero net charge in the semiconductor, and we can assume that an equivalent fixed surface
charge density exists in the oxide. The charge density on the metal is Q’m , and from charge neutrality
we have
The Two-terminal MOS Structure
The Two-terminal MOS Structure
The Two-terminal MOS Structure
Threshold Voltage
The Two-terminal MOS Structure
The Two-terminal MOS Structure
The Two-terminal MOS Structure
Threshold Voltage (PMOS)
The MOSFET Transistor
• The MOSFET is a four terminal device. The voltage applied to the gate terminal determines if and
how much current flows between the source and the drain ports.
• The body represents the fourth terminal of the transistor. Its function is secondary as it only serves
to modulate the device characteristics and parameters.
• The transistor can be considered to be a switch. When a voltage is applied to the gate that is larger
than a given value called the threshold voltage VT, a conducting channel is formed between drain
and source.
• The conductivity of the channel is modulated by the gate voltage the larger the voltage difference
between gate and source, the smaller the resistance of the conducting channel and the larger the
current.
• When the gate voltage is lower than the threshold, no such channel exists, and the switch is
considered open.
The MOSFET Transistor
The MOSFET Transistor
• The NMOS transistor consists of n+ drain and source regions, embedded in a p-type substrate. The
current is carried by electrons moving through an n-type channel between source and drain.
• In pn-junction diode, where current is carried by both holes and electrons.
• MOS devices can also be made by using an n-type substrate and p+ drain and source regions. In
such a transistor, current is carried by holes moving through a p-type channel. The device is called a
p-channel MOS, or PMOS transistor.
• In a complementary MOS technology (CMOS), both devices are present.
The MOSFET Transistor
The MOSFET Transistor
The MOSFET Transistor
The MOSFET Transistor
Resistive Operation
Assume now that VGS > VT and that a small voltage, VDS, is applied between drain and source. The
voltage difference causes a current ID to flow from drain to source
Integrating the equation over the length of the channel L yields the voltage-current relation of the
transistor.
The product of the process transconductance and the (W/L) ratio of an (NMOS) transistor is called the
gain factor kn of the device.
For smaller values of VDS, the quadratic factor in above Eq. can be ignored, and we observe a linear
dependence between VDS and ID.
The MOSFET Transistor
Saturation Operation
• As the value of the drain-source voltage is further increased, the assumption that the channel
voltage is larger than the threshold all along the channel ceases to hold.
• This happens when VGS - V(x) < VT. At that point, the induced charge is zero, and the conducting
channel disappears or is pinched off.
k is a measure of the degree of velocity saturation, since VDS/L can be interpreted as the average field
in the channel.
In case of long-channel devices (large values of L) or small values of VDS, k approaches 1
Above Eq. simplifies to the traditional current equation for the resistive operation mode.
For short-channel devices, k is smaller than 1, which means that the delivered current is smaller than
what would be normally expected.
The MOSFET Transistor
Velocity Saturation
When increasing the drain-source voltage, the electrical field in the channel will ultimately reach the
critical value, and the carriers at the drain become velocity saturated.
The saturation drain voltage VDSAT can be calculated by equating the current at the drain to the current
given by above Eq. for VDS = VDSAT.
The MOSFET Transistor
Velocity Saturation
The MOSFET Transistor
Velocity Saturation
• The velocity saturates abruptly at ξc, and is approximated by the following expression:
• The drain-source voltage VDSAT at which the critical electrical field is reached and velocity
saturation comes into play is constant
The gate-capacitance
components are
nonlinear and
varying with the
operating voltages
The MOSFET Transistor
Junction Capacitances
The MOSFET Transistor
Capacitive Device Model
Example
Example
The CMOS Inverter
The CMOS Inverter
The CMOS Inverter
For a dc operating points to be valid, the currents through the NMOS and PMOS devices must be equal. Graphically, this
means that the dc points must be located at the intersection of corresponding load lines
The CMOS Inverter
The CMOS Inverter
The CMOS Inverter
The CMOS Inverter
The CMOS Inverter
The CMOS Inverter
The CMOS Inverter
The CMOS Inverter
The CMOS Inverter
The CMOS Inverter
The CMOS Inverter
Switching Threshold
• The switching threshold, VM, is defined as the point where Vin = Vout. Its value can be obtained
graphically from the intersection of the VTC with the line given by Vin = Vout.
• In this region, both PMOS and NMOS are always saturated, since VDS = VGS. An analytical
expression for VM is obtained by equating the currents through the transistors
The CMOS Inverter
Switching Threshold
The CMOS Inverter
Switching Threshold
Changing the switching threshold by a considerable amount is however not easy, to move the threshold
to 1.5 V requires a transistor ratio of 11, and further increases are prohibitively expensive.
The CMOS Inverter
Noise Margins
Since on-chip interconnects are generally prone to signal noise, the output signal of the first inverter will
be perturbed during transmission.
The CMOS Inverter
Noise Margins
• The voltage level at the input of the second inverter will be either larger or smaller than VOL.
• If the input voltage of the second inverter is smaller than VOL' this signal will be interpreted correctly
as a logic "0" input by the second inverter.
• On the other hand, if the input voltage becomes larger than VIL as a result of noise, then it may not be
interpreted correctly by the inverter.
• Thus, we conclude that VIL is the maximum allowable voltage at the input of the second inverter,
which is low enough to ensure a logic "1" output.
The CMOS Inverter
Noise Margins
• Now consider the signal transmission from the output of the second inverter to the input of the third
inverter, assuming that the second inverter produces an output voltage level of VOH.
• Due to the noise interference, If the input voltage of the third inverter is larger than VOH, this signal
will be interpreted correctly as a logic " 1 " input by the third inverter.
• If the voltage level drops below VIH due to noise, however, the input cannot be interpreted as a logic
"1.“ Consequently, VIH is the minimum allowable voltage at the input of the third inverter which is
high enough to ensure a logic "0" output.
The CMOS Inverter
Noise Margins
VIH is the minimum allowable voltage at the input of the third inverter which is high enough to ensure a
logic "0" output.
The CMOS Inverter
Noise Margins
The CMOS Inverter
Device Variations
Impact of device variations on static CMOS inverter VTC. The “good” device has a smaller oxide
thickness (- 3nm), a smaller length (-25 nm), a higher width (+30 nm), and a smaller threshold (-60
mV). The opposite is true for the “bad” transistor.
The CMOS Inverter
Scaling the Supply Voltage
The CMOS Inverter
Performance of CMOS inverter: Dynamic Behaviour
The propagation delay of the CMOS inverter is determined by the time it takes to charge and discharge
the load capacitor CL through the PMOS and NMOS transistors, respectively.
Wiring Capacitance Cw
The capacitance due to the wiring depends upon the length and width of the connecting wires, and is a
function of the distance of the fanout from the driving gate and the number of fanout gates.
The rise time and the fall time of the output signal
are defined as the time required for the voltage to
change from its 10% level to its 90% level (or vice
versa).
The CMOS Inverter
Performance of CMOS inverter: Dynamic Behaviour
• Reduce CL: Remember that three major factors contribute to the load capacitance: the internal
diffusion capacitance of the gate itself, the interconnect capacitance, and the fanout.
• Increase the W/L ratio of the transistors: This is the most powerful and effective performance
optimization tool in the hands of the designer.
Increasing the transistor size also raises the diffusion capacitance and hence CL. In fact, once
the intrinsic capacitance (i.e. the diffusion capacitance) starts to dominate the extrinsic load
formed by wiring and fanout, increasing the gate size does not longer help in reducing the
delay, and only makes the gate larger in area. This effect is called “self-loading”.
• Increase VDD. The delay of a gate can be modulated by modifying the supply voltage.
The CMOS Inverter
Power Consumption
• During the high-to-low transition, this capacitor is discharged, and the stored energy is dissipated in
the NMOS transistor
The CMOS Inverter
Power Consumption
Energy Calculations:
The values of the energy EVDD, taken from the supply during the transition,
as well as the energy EC, stored on the capacitor at the end of the transition,
can be derived by integrating the instantaneous power over the period of
interest.
During the discharge phase, the charge is removed from the capacitor, and its
energy is dissipated in the NMOS device.
The CMOS Inverter
Power Consumption
To compute the power consumption, we have to take into account how often the device is switched. If
the gate is switched on and off f 0→1 times per second, the power consumption equals
The CMOS Inverter
Power Dissipation Due to Direct-Path Currents
• In actual designs, the assumption of the zero rise and fall times of the input wave forms is not
correct. During the input switching, NMOS and the PMOS transistors are conducting
simultaneously.
• Under the (reasonable) assumption that the resulting current spikes can be approximated as
triangles and that the inverter is symmetrical in its rising and falling responses, we can compute the
energy consumed per switching period,
The CMOS Inverter
Static Power Dissipation
Ideally, the static current of the CMOS inverter is equal to zero, as the PMOS and NMOS devices are
never on simultaneously in steady-state operation.
There is, unfortunately, a leakage current flowing through the reverse-biased diode junctions of the
transistors, located between the source or drain and the substrate.
The CMOS Inverter
Static Power Dissipation
However, be aware that the junction leakage currents are caused by thermally generated carriers. Their
value increases with increasing junction temperature, and this occurs in an exponential fashion. At
85°C (a common junction temperature limit for commercial hardware), the leakage currents increase
by a factor of 60 over their room-temperature value.
The CMOS Inverter
Combinational Circuits
Static CMOS Design
• The static CMOS style is really an extension of the static CMOS inverter to multiple inputs.
• The primary advantage of the CMOS structure is robustness (i.e, low sensitivity to noise), good
performance, and low power consumption with no static power dissipation.
• The complementary CMOS circuit style falls under a broad class of logic circuits called static
circuits in which at every point in time (except during the switching transients), each gate output is
connected to either VDD or Vss via a low-resistance path.
Static CMOS Design
Complementary CMOS
Threshold voltage of transistor M2 will be higher than transistor M1 due to the body effect.
Static CMOS Design
Static CMOS Design