Microwave

Solid-state
Devices 9
9.1 INTRODUCTION
Before the 1960s, vacuum tube technology was widely used in electronics, which increased the size
and cost of the entire system. Later, the development in semiconductor technology led to the growth of
solid-state devices, which were more reliable and had a longer life. These devices generally require low
power and are very compact and lighter in weight.
In the recent past, tremendous research activities have taken place for the development of better,
low-noise, high-frequency, and greater-bandwidth components to achieve lesser switching time and
better performance characteristics. In this chapter, several microwave solid-state devices, including
two terminal devices such as transferred electron devices (Gunn diodes), avalanche transit time devices
(IMPATT, TRAPATT, and BARITT diodes), tunnel diodes, and three terminal devices such as bipolar
junction transistors and field effect transistors (MESFETs and HEMTs) are discussed.This chapter also
discusses the PIN diode, Schottky diode, Varactor diode, parametric amplifier, step-recovery diode, and
crystal diode in detail.

9.2 NEGATIVE RESISTANCE PHENOMENON

There are various microwave solid state devices which show negative resistance characteristics. The
real part of their impedance is negative over a range of frequencies. In a positive resistance, the current
through the resistor and the voltage across it are in phase. The voltage drop across positive resistance
is positive, and the power that is given by the product of voltage (V) and current (I) is dissipated in the
resistor. In a negative resistance, the current and voltage are out of phase by 180°. Therefore, the volt-
age drop across it is negative, and a negative power is generated by the power supply associated with
the negative resistance. In other words, negative resistance generates or supplies power (e.g. active solid
state devices) to the external circuits and positive resistance absorbs power (e.g. passive solid state
devices).
However, the reason for exhibiting negative resistance varies from one device to another. For instance,
in tunnel diode the negative resistance is due to heavy doping where the majority carries tunnel through
the thin barrier. In an IMPATT, it is due to delay, which causes the current to lag behind the voltage;
whereas in TEDs, it is due to differential negative electron mobility characteristic in materials such as
GaAs and InP.

Chapter 09.indd 9.1 1/3/2014 10:39:23 AM

 9.2 | Microwave and Radar Engineering

9.3 CLASSIFICATION OF SOLID-STATE DEVICES

Solid-state devices can be classified into two categories depending on the number of terminals these
device have. They are diodes and transistors. Depending on their principle of operation, the two termi-
nal diodes are further classified as shown in Figure 9.1(a).Transferred Electron Devices (TED), tunnel
diodes, and Avalanche Transit Time Devices (ATD) show negative resistance. The three terminal devices
are also further divided into two types as shown in Figure 9.1(b).Microwave BJTs are bipolar devices,
whereas microwave FETs are unipolar devices.

DIODES

Metal-
Tunnel PIN
TEDs ATDs Varicaps semiconductor
diodes diodes
junction diodes
(a)

Transistor

Bipolar microwave Unipolar microwave

BJTs FETs

MESFETs HEMTs

(b)

Figure 9.1 (a) Two terminal devices; (b) Three terminal devices

9.4 APPLICATIONS OF SOLID-STATE DEVICES

Microwave solid-state devices are becoming more important. These devices have been invented for
various applications in microwave frequency regions such as frequency multiplication, signal detection,
attenuation, generation of oscillation, switching, phase shifting, amplitude limiting, and amplification.
Some of their applications are as follows:
• As microwave generators
• As amplifiers in satellite communications and also in other space applications
• As transmitters for millimeter communication systems
• In radio transmitters, such as CW Doppler radar
• In broadband linear amplifiers and in low-power amplifiers
• As a pumping source for parametric amplifiers
• In transponders
• In both combinational and sequential logic circuits
• In microwave receivers

Chapter 09.indd 9.2 1/3/2014 10:39:24 AM

 Microwave Solid-state Devices | 9.3

9.5 TRANSFERRED ELECTRON DEVICES (TEDS)

Transferred Electron Devices (TEDs) are one of the important microwave devices. They are bulk devices
that have no junction or gates as compared with microwave transistors, which operate with either junc-
tion or gates. Transferred electron devices are fabricated with compound semiconductor materials (e.g.
gallium arsenide (GaAs), cadmium telluride (CdTe) and indium phosphide (InP)). They have two energy
regions or valleys in the conduction band. The TEDs show a transferred electron effect in which transfer
of electrons takes place from lower valley to upper valley in the conduction band. Most of the electrons
will be in lower valley at low electric fields. When the electric field strength is increased to higher values,
most of the electrons will be transferred into high-energy bands. In the higher-energy bands, the effective
mass of electrons is larger than in lower-energy bands. Since the conductivity is directly proportional to
mobility, the higher-energy band has lower conductivity than the lower-energy band. Hence, conductivity
decreases with an increase in electric field strength. Thus, the current decreases with an increase in volt-
age, showing negative resistance. Gunn diode is an example of TEDs, and its theory is better explained
in the next section.
A few differences between microwave transistors and TEDs are as follows:
• First, TEDs do not have junctions or gates as in the case of transistors.
• TEDs with smaller physical dimensions have a limited power output. In order to get a reason-
able power output, the physical dimensions are to be made large as compared with a microwave
transistor.
• TEDs are made of compound semiconductors from group III–V and II–VI elements (of peri-
odic table) such as gallium arsenide (GaAs), indium phosphide (InP), and cadmium telluride
(CdTe); whereas the majority of transistors are made from elements such as germanium or
silicon.

9.6 GUNN DIODE

In 1963, Gunn observed a periodic variation of current passing through an N-type GaAs semiconductor.
He found the following when a DC bias voltage is applied to the contacts of N-type GaAs or InP:
• Current first rises linearly from zero.
• Then, it begins to oscillate when a certain threshold is reached.
• The time period of oscillation is equal to the travel time of electrons from cathode to anode.
This is known as Gunn effect or bulk effect. The device that shows Gunn effect is known as Gunn diode.
Gunn diodes are usually fabricated using N-type semiconductor materials (eg. GaAs, and InP); so, they
should be associated with electrons rather than with holes. Gunn diode operations do not depend on
junction properties. Even though it has no junction, it is called a diode, because it has two terminals
(anode and cathode) attached to it. It is an active two-terminal solid-state device and is mainly used as a
local oscillator in the microwave frequency range of 1 to 100 GHz. Gunn effect can be explained on the
basis of domain formation, two-valley theory of Ridley–Watkins–Hilsum (RWH), or the transfer electron
mechanism.

9.6.1 Operation and Characteristics of Gunn Diode

The Gunn diode is made up of a single piece of N-doped semiconductor material (e.g. GaAs, InP) with
two thinner N+-doped layer contacts on opposite ends. The two N+ layers are required to connect the
anode and cathode leads.

Chapter 09.indd 9.3 1/3/2014 10:39:25 AM

 9.4 | Microwave and Radar Engineering

Current Eth
density
(J )

Ev
V
− +

+ N +
N E0 N
0.32 V/μm Electric Field (E)
(a) (b)
Figure 9.2 (a) Three-layer structure of Gunn diode; (b) J/E curve of a Gunn diode

When a DC voltage (V) is applied to the two terminals, an electric field (Eo) will be established across
the
 piece of GaAs (just as in a resistor) (Figure 9.2(a)). Figure 9.2(b) shows how the current density
( ) through the material varies with the electric field (Eo) across it. We know that the drift velocity (vd),
current density, and electric field have the following relations:

vd μ E
 
J nq E
Initially, when the electric field is increased, from the above relations it can be seen that the drift
velocity and current density increase. Thus, the current density increases with an increase in the electric
field, resulting in a positive resistance. This continues till the electric field reaches a value known as
threshold value Eth (corresponding threshold voltage Vth). When the electric field is increased beyond
the threshold value Eth, the Gunn effect takes place and the current density decreases, causing the device
to exhibit negative resistance. This behavior is due to domain formation, which will be explained in the
next section. This will continue till the field reaches a value known as Ev (corresponding valley voltage).
When the voltage is increased beyond Ev, the current density increases. Thus, the device again exhibits
positive resistance.
Negative resistance region: In common the current first rises linearly from zero with increasing electric
field, however there is a region between the threshold electric field and valley electric field, where the
current decreases as the electric field is increased. This is called the negative resistance region. The
dynamic resistance, r, in this voltage range is given by
r = dV/dI, r < 0.

9.6.2 Domain Formation

When bias voltage (V) is applied across the diode terminals, an electric field (E0) is established across
the GaAs piece. As E0 increases, electron drift velocity increases up to a certain potential (Eth). This is
because the velocity (vd)is proportional to the applied electric field. However, with a further increase in
voltage, the mobility decreases, which causes the velocity to decrease, thereby causing the electrons to
slow down. This slowing down results in a traffic jam, and more electrons pileup to form a charge layer
or domain. This charge layer produces an electric field (E) such that it decreases the original field (E0)
on the left and increases the original field (E0) to the right (Figure 9.3). As a result, the charge bunch
gets pushed toward the anode on the right, forming a current pulse.
When this pulse moves to the anode, E0 goes back to the original value, resulting in another traffic
jam and another pulse formation, which causes oscillation.

Chapter 09.indd 9.4 1/3/2014 10:39:25 AM

 Microwave Solid-state Devices | 9.5

V
− +

+ E E +
N N
Cathode anode

Domain or charge layer

Figure 9.3 Domain formation in a Gunn diode

In this procedure, N-type GaAs releases the power that it has acquired during the domain formation
in the direction of the anode. If conditions are appropriate, this release of power at anode will be utilized
by the noise energy of suitable frequency existing in the Gunn diode configuration and gets amplified.
The amplification of noise energy results in sustained oscillations. The frequency of oscillation depends
on the length of the GaAs piece and concentration of electrons.

9.6.3 RWH Theory or Two-valley Theory of Gunn Diode

The Gunn diode, which is made of n-doped semiconductor material (e.g. GaAs or InP), is characterized
by having two valleys in their conduction bands with different mobility. The two-valley model is also
called the Ridley–Watkins–Hilsum(RWH) theory. There are two regions in the conduction band of the
N-type GaAs. These conduction band regions are known as upper valley and lower valley. The energy
band and transfer of electrons between the two valleys are shown in Figures 9.4 and 9.5, respectively.
The following can be noted about GaAs:
• In N-type GaAs, the valence band is filled with electrons, and the conduction band is partly filled.
• The forbidden energy gap between the valence band and the conduction band is about 1.43 eV.
• Electrons in the lower valley (n1) exhibit a small effective mass (m1) and very high mobility, m1.
• Electrons in the upper valley (n2) exhibit a large effective mass (m2) and very low mobility, m2.
• The two valleys are separated by a small energy gap, ΔE, of approximately 0.36 eV.
• At a low electric field, electrons remain in the lower valley and material behaves ohmically.
When the electric field reaches a certain threshold value Eth, the electrons will be swept from the
lower valley to the upper valley. If the rate of transfer of electrons from the lower to the upper
valley is very high, the current will decrease with an increase in voltage. This leads to a decrease
in the average electron mobility, m with an increase in the field in a bulk semiconductor, thus
resulting in equivalent negative resistance. This manifests itself as a bulk negative differential
resistance.
• The average electron mobility, m is given by

μ = ( 1 μ1 + n2 μ2 )/( n1 + n2 ) (9.1)

where n1 = electron density in lower valley

n2 = electron density in upper valley
μ1 = mobility of electron in lower valley
μ2 = mobility of electron in upper valley
Total carrier concentration = n1 + n2

Chapter 09.indd 9.5 1/3/2014 10:39:25 AM

 9.6 | Microwave and Radar Engineering

Energy (E) Conduction

m1* m2*
m1* m 2*
0.36 eV

1.43 eV
GaAs-N

Valance

Wave number (k)

k

Figure 9.4 Energy band diagram for GaAs

J
Negative
resistance
region

E1 Eth Ev E

Figure 9.5 Inter-valley transfer

where m1* = effective mass of electron in lower valley
m2* = effective mass of electron in upper valley
The inter-valley transfer of electrons with regard to the Eth and E is illustrated in Figure. 9.5. Three
different
 ranges of electric field strength may be considered.The current (I1) and the current density
( 1 ) corresponding to the electric field < ( ) where all the electrons occupy the lower-energy
valley, is given by
   ⎛  I ⎞
J1 1en
e 1 E I1 eeA Aμ1n1 E J= ⎟
⎝ A⎠
   
The current (I2) and the current density ( 2 ) corresponding to the electric field E th < E < E v , where
the total number of electrons are distributed between the lower and upper valleys, is given by
  
J 2 e( n1 1 + n2 μ2 ) E , I 2 e( n1 1 + n2 μ2 ) AE
  
The current (I3) and the current density ( 3 ) corresponding to the electric field E E v , where all
electrons occupy the upper-energy valley, is given by
  
J 3 e μ 2 n2 E , I 3 eeA 2 n2 E

Chapter 09.indd 9.6 1/3/2014 10:39:26 AM

 Microwave Solid-state Devices | 9.7

 
where ( ) = current density, ( ) = electric field, A = area of cross-section of device, and e = charge
of electrons.
Performance characteristics: Gunn diode operates at 10-12 V of typical biased voltage, 300 mA of
bias current, and the power output is 100 mW in the X band. The typical frequency range that it can
operate is from 4 GHz to 94 GHz, efficiency is 10-20 %, CW power output is 250 mW at 15 GHz, and
pulsed power is 5 W at 12 GHz.

9.6.4 Equivalent Circuit of Gunn Diode

The scale of the parasitic impedances accredited to the package element decreases with the package
dimension. The equivalent circuit of Gunn diode is shown in Figure 9.6(a).

RS LP

RG CG CP

DIODE ELEMENT PACKAGE ELEMENT

CP = Package Capacitance
LP = Bond Lead Inductance
RS = Series Resistance
RG = Gunn Diode Resistance
CG = Gunn Diode Capacitance
(a) (b)

Figure 9.6 (a) Equivalent circuit Symbol; (b) Symbol

As a result, at lower microwave frequencies the stronger package styles are normally precise and for the
higher microwave frequency the smaller parasitic impedance packages are suggested. Its symbol is as
shown in Figure 9.6(b).

9.6.5 Basic Modes of Operation

Depending on the device characteristics and external circuitry, the Gunn diode can be made to oscillate
in any one of the following four frequency modes:
• Gunn oscillation mode
• Limited space charge accumulation (LSA) mode
• Stable amplification mode
• Bias circuit oscillation mode
The modes of operations are classified based on the condition

ε s vd
no l >
e μn

Chapter 09.indd 9.7 1/3/2014 10:39:28 AM

 9.8 | Microwave and Radar Engineering

vd = electron drift velocity

ε s = semiconductor dielectric permittivity
μn = electron mobility
e = electron charge
no = doping concentration
l = device length
9.6.5.1 Gunn oscillation modes
This mode has the following features:
• The electric field is greater than the threshold.
• The product of frequency and length is about 107cm/s.
• The product of doping and length is more than 1012 /cm2.
• Due to the cyclic arrangement of either accumulation layer or the high voltage domain the Gunn
diode is unstable.
• There are three types of Gunn oscillation mode. They are
• Transit time domain mode
• Delayed domain mode
• Quenched domain mode.
Transit time domain mode: Transit time domain oscillation is the basic mode and is not depend on the
external circuit. When a domain is quenched at the anode the current peaks are obtained. Then, another
is nucleated near the cathode. At any time, the entire electric field across the device is above the thresh-
old electric field. The frequency is given by
f vd / l since domain transit time is t = l/vd
where vd is the domain or drift velocity = fl = 1 × 107 cm/s.
l is the effective length, and fl is the product of frequency and length
The oscillation time (to) is equal to the transit time (t ).It is a low-power (< 2 W), low-efficiency
mode, and the operating frequency is between 1 GHz and 18 GHz.
Delayed domain mode: This is also called an inhibited mode. In this mode, the domain is collected
when the dc bias is less than the threshold electric field (Eth). The next domain can only be formed
when the field again reaches the threshold. The oscillation period is greater than the transit time of the
critical field (Eth), that is, τ o > τ . The frequency of oscillation is determined by the resonant circuit.
The efficiency of this mode is limited to 20%. The drift velocity (vd) or (f × l)lies between 106 cm/s
and 107 cm/s.
Quenched domain mode: In this mode, the domain collapses before it reaches the anode; that is, it is
quenched before it is collected, hence the name quenched mode. As shown in Figure 9.7, the bias electric
field reduces below the sustaining electric field (Es) in the negative half cycle. When the bias electric
field swings back again more than the threshold electric field, another domain is nucleated and the pro-
cedure repeats. The operating frequencies are higher than the transit time frequency (Figure 9.7). The
maximum efficiency of this mode is 13%. The drift velocity (vd) or (f × l) = 107 cm/s.
9.6.5.2 Limited Space charge Accumulation (LSA) mode
This mode operates by using a high-Q resonant cavity with current pulses from the Gunn diode. The
Gunn diode is placed in a resonant cavity, which is tuned to a frequency of the LSA mode (f0), so that
the circuit operates like a negative resistance device and the domains do not have enough time to form.
In this mode, the device can be biased to several times higher than Eth. When the input field increase

Chapter 09.indd 9.8 1/3/2014 10:39:29 AM

 Microwave Solid-state Devices | 9.9

v t
Transit-time
mode dc bias
(tto = t )
Eth
Es
0
Delayed
mode dc bias
(tto > t )
Eth
Es
0

Quenched
mode dc bias
(tto < t ) Eth
Es
0
LSA
mode dc bias
(tto < t )
Eth
to = 3t d
Es
0

Figure 9.7 Various modes of Gunn diode

more than the threshold electric field, the device remains in the negative resistance region. Oscillation
time is obtained by the external circuit, which is given by
L
τ o = 2π LC + (9.2)
R(Vb / Vth )
whereto is period in seconds
L is inductance in Henry
C is capacitance in farad
R is resistance in ohm
Vth is threshold potential
Vb is bias voltage
Diode
RF Choke current

+
V BIAS −R C L
−

DC
Isolation

Figure 9.8 Gunn diode in LSA mode

Chapter 09.indd 9.9 1/3/2014 10:39:29 AM

 9.10 | Microwave and Radar Engineering

Figure 9.8 shows the LSA mode Gunn oscillator simplified circuit. Gunn oscillator with LSA mode
consists of Gunn diode which is shunted by either an LC tank circuit or a tuned cavity that behaves like
a tank circuit. The condition for resonant oscillation is given by
(−G) = 1/−R ≥ conductance represented by circuit losses (G0).
Where G = Negative conductance
−G ≥ G0 (9.3)
Output power of LSA mode Pout = h I0 V0
where h = conversion parameter of material
I0 = operating current
V0 = operating voltage
Advantages of LSA mode are as follows:
• Operating frequency is not limited by transit time effect. Therefore, device length can be made
larger and the device can sustain higher voltage.
• This mode gives high-pulsed output power (> 100 W) with a high frequency (100 GHz). However,
the power decreases with frequency.

EXAMPLE PROBLEM 9.1

A Gunn diode with 3.5 V critical threshold potential and 25 ohms resistance is connected to 3 V dc.
If the diode is connected to a tank circuit where L = 0.1 nH and C = 2.5 pH, what is the period of
oscillation? Also find the frequency of oscillation.
Solution
Given Vb = 3 V, Vth = 3.5 V, R = 25 ohms, L = 0.1 nH, and C = 2.5 pH
L
The period of oscillation is τ o = 2π LC +
R(Vb / Vth )
0 1 × 10 −9
= 2π 2.5 × 0.1 × 10 −21 + = 9.9 × 10 −11 + 0.466 × 10 −11 s = 1.366 picoseconds
25(3/ 3.5)
1
The frequency of oscillation is f = 1/τ o = × 1011 Hz = 7.32 × 1012 Hz
1.366
9.6.5.3 Stable amplification mode
This mode is defined as the region where the frequency times length is about 107cm/s and the doping
time length is between 1011cm−2 and 1012 cm−2.
9.6.5.4 Bias circulation mode
In this mode, the product of frequency and length is very less. This mode occurs along with Gunn or
LSA oscillations. When a bulk diode is biased to the threshold, the average current suddenly drops as
Gunn oscillations begin. The drop in current at the threshold can lead to oscillations in the bias circuit
that are typically from 1 kHz to 100 MHz.
9.6.6 Applications of Gunn Diode
1. It is used in Radar transmitters as low power oscillator (e.g. CW Doppler Radar, Police Radar).
2. Gunn diode oscillators with pulsed signal are used in transponders in industry telemetry systems and
for air traffic control.
3. It is used in broadband linear amplifier.
4. It is used in sequential logic circuits and fast combinational circuits

Chapter 09.indd 9.10 1/3/2014 10:39:30 AM

 Microwave Solid-state Devices | 9.11

5. It is used in microwave receivers as low and medium power oscillator.

6. It is used in parametric amplifier as pump sources.

EXAMPLE PROBLEM 9.2

A Gunn diode has the following parameters: electron drift velocity, vd = 2.5 × 105m/s, negative electron
mobility, μ n = 0.015 m2/Vs, relative dielectric constant, and e r = 13.1 . Determine the criterion for clas-
sifying the modes of operation.
Solution
The data given are vd = 2.5 × 105 m/s, μ n = 0.015 m2/Vs, and er = 13.1
ε s vd ε εv
We have no l > ⇒ n0 l > 0 r d ( ε = ε 0 ε r )
e μn e μn
8.85 × 10 −12 × 13.1 × 2 5 × 105 2 9 × 10 −5
> > > 1.2 × 1016
1.6 × 10 −19 × 0.015 2.4 × 10 −21
This means that the product of the doping concentration and the device length should be
no l > 1.2 × 1016 /m 2

EXAMPLE PROBLEM 9.3

Determine the conductivity of n-type GaAs Gunn diode if
Electron density n = 1018 cm−3, electron density at lower valley nl = 1010 cm−3
Electron density at upper valley nu = 108 cm−3, temperature T = 300o K
Solution
Given that electron density, n = 1018 cm−3, nl = 1010 cm−3,nu = 108 cm−3, and temperature T = 300o K.
We know that
μl = 8000 cm2/v.sec = 8000 × 10 −4 m/v.s, μu = 180 cm2/v.sec = 180 × 10 −4 m/v.s
We have conductivity,
−19
σ ( l μl + nu μu ) = 1.. × (8000 × 10 −4 ×
×10 80 × 10 −4 × 1014 ) = .28 10 −3 
1016 + 180

EXAMPLE PROBLEM 9.4

A GaAs Gunn diode has an active region of 10 micrometers. If the electron drift velocity is 105 m/sec,
calculate the natural frequency and the threshold voltage. The critical electric field is 3 kv/
v cm.
Solution
The given data are vd = 105 m/sec = 107 cm/sec, l = 10 μm = 10 × 10 −4 cm, and Ec f = ?, V = ?
vd 107
We have natural frequency, f = ∴f 10
f = 10 GHz
l 10 × 10 −4
Critical voltage, V = l × Ec
i.e., Critical voltage = Active length × Critical electric field
∴V = 10 × 10 −4 × 3 × 103 = 3 V
∴ Critical voltage, V =3 V

Chapter 09.indd 9.11 1/3/2014 10:39:30 AM

 9.12 | Microwave and Radar Engineering

9.7 TUNNEL DIODES

In a conventional PN junction diode, the concentration of impurities is in the order of one part in 108.
When the impurity concentration is increased to 1 part in 103, the conventional PN diode becomes the
tunnel diode.
Principle of Operation: In a tunnel diode, the impurity concentration is greatly increased as compared
with the PN diode. This reduces the width of the depletion layer to the order of hundred angstroms.
According to quantum mechanics, if the barrier is less than 3A0, there is a large probability that particles
will tunnel through the potential barrier. This is known as tunneling.
The operation can be best explained by considering energy band diagram of P-type and N-type mate-
rials. Due to high impurity concentration, there are many holes in valence band of P-type material and
many electrons in conduction band of N-type material. Now, if no voltage is applied, the alignment of
valence band and conduction band of P-type and N-type is as shown in Figure 9.9 (a). Hence, there is
no flow of current. Now, if some forward voltage is applied the energy levels of N-region move upward
and that of P-region move downward relatively. Therefore, the electrons on the right side of the potential
barrier are precisely opposite to the holes on the left side and the corresponding energy band diagram
is as shown in Figure 9.9 (b). This results in flow of current, whose magnitude gradually increase as
forward voltage increase till V = Vp which is known as peak voltage.
At the peak, as shown in Figure 9.10, the voltage energy levels on either side are in perfect align-
ment. As the voltage is increased beyond Vp, the energy levels of the N region move further upward
and those of the P region move further downward, causing a condition as shown in Figure 9.9 (c). In
this condition, some of the N-region electron energy levels are opposite to the band gap and some
are opposite to the holes (or empty states). This results in a decrease of current flow; this condition
continues till V = Vv, called valley voltage. At this voltage, current (I = Iv) is minimum. If the voltage
is further increased as shown in Figure 9.9 (d), then all electrons in the N-region are opposite to the
band gap and no currents flows. The corresponding point in the Figure 9.10 is the Vv. If the voltage
continues to increase further as shown in Figure 9.9 (e), the diode shows characteristics of a normal
PN diode. Thus, between peak voltage and valley voltage, the tunnel diode shows negative resistance
characteristics. The volt-ampere characteristic curve of a tunnel diode is as shown in Figure 9.10.
The V-I characteristic of a tunnel diode exhibits the current as multi-valued and the voltage as a
single-valued.

Conduction band edge Conduction band

N-Type N-Type
Band gap Band gap region
region
Electron Electron
holes holes or
energy level energy level E ec
Electr
tron
onss
Elec
Electr
tron
onss empty states
P-
Re
Region
Reg n P-
P
Region
R e on
eg
Valence
Va
a en
e ce
ce ban
b
band
nd e
edge
d e
dg Valence
Va
Vaeencce
ebband
nd

Distance Distance

Figure 9.9 (a) Energy bands in unbiased Figure 9.9 (b) Electron energy levels
tunnel diode of N-region are equal to the empty states
level on P region.

Chapter 09.indd 9.12 1/3/2014 10:39:32 AM

 Microwave Solid-state Devices | 9.13

Conduction band N-Type N-Type

Conduction band
region region
Band gap
Electron Elecctron
o s Band gap E ec
Ele tron
trrons
holes Electron
energy level energy level holes
P-
P
Region
R egion
i P-
P
Valence
V a e ce band
ba d Reg
Region
Reg
eg on
o Valence
V alen
e ce band
ban d

Distance Distance

Figure 9.9 (c) Some N-region electrons Figure 9.9 (d) All N-region electrons
energy level are opposite to “band gap,” and energy level are opposite to band gap
some are opposite to holes

N-Type
Conduction band region

Band gap E ec
Ele tr
t on
o s
Electron
energy level
Empty states
or holes
P-
P
Region
R egio
gion Valence
Va ence b
band
and
nd

Distance

Figure 9.9 (e) Tunnel diode functions as normal diode

(Peak)
Ip
Forward current

(Valley)
Iv
O Vp VV V
Forward voltage

Figure 9.10 VI characteristics of a tunnel diode

The equivalent circuit of a tunnel diode is as shown in Figure 9.11. Rn is negative resistance. From the
volt-ampere characteristics,
∂i 1
−g = =− (9.4)
∂v Rn

Chapter 09.indd 9.13 1/3/2014 10:39:33 AM

 9.14 | Microwave and Radar Engineering

Rs is series ohmic resistance; Ls is series inductance that depends on lead length and packaging circuit of
the diode; and C is junction capacitance.

Rs Ls
Tunnel diode

Zin C −R
Rn

Figure 9.11 Equivalent circuit of tunnel diode and its symbol

The input impedance of the circuit is

Rn ⎡ ω Rn2C ⎤
Zin Rs − + j ⎢ Ls − 2 ⎥
(9.5)
( + (ω RnC ) )
2
⎣ ( + (ω RnC ) ) ⎦
The cutoff frequency is obtained by putting Re (Zin) = 0 and is given by

1 Rn
fc = −1 (9.6)
2π RnC Rs

Self-resonance frequency is obtained by putting Img (Zin) = 0 and is given by

1 Rn2C
fr = −1 (9.7)
2π RnC Ls

Applications
• It is used as a high-speed switch. The switching times of the order of nanoseconds are obtained.
• As a logic memory storage device
• As a microwave oscillator
• In a relaxation oscillator
• The tunnel diode can be used in bistable, a stable, and monostable circuits. If the operating point
is made to vary over the entire range of the voltage, the circuit is in bistable mode. If the operating
point is between Vp and Vv, the circuit is a stable. If the operating point is between 0 and Vp, the
circuit is monostable.
Some important characteristics of a tunnel diode are as follows:
• Doping is very high, usually 1000 times of normal diodes.
• It has zero breakdown voltage.
• It has a dynamic negative resistance region, because when we increase its voltage, the current will
decrease beyond the peak voltage.
• For an ideal tunnel diode, Vp or Ip should be very large.
• It behaves similar to both an amplifier and an oscillator.
• In reverse bias, it behaves similar to a good conductor.

Chapter 09.indd 9.14 1/3/2014 10:39:35 AM

 Microwave Solid-state Devices | 9.15

9.8 AVALANCHE TRANSIT TIME DEVICES

It is possible to realize a microwave diode that exhibits negative resistance, by providing a delay between
voltage and current through the material. In 1958 Read had proposed that there must be a phase delay of
more than 90° between RF input voltage and avalanching current if the total voltage go above breakdown
voltage in a diode due to the RF input voltage. Such devices are called avalanche transit time devices.
At microwave frequencies, avalanche transit time devices produce negative resistance by using carrier
impact ionization and transit time in the high voltage region of a semiconductor junction. There are dis-
tinct modes of avalanche oscillators, out of which the following three are discussed:
1. IMPATT: Impact Ionization Avalanche Transit Time diode
2. TRAPATT: Trapped Plasma Avalanche Triggered Transit diode
3. BARITT: Barrier injection transit time diode

9.9 IMPATT DIODE

IMPATT diode is abbreviated as impact-ionization avalanche transit time diode. It is an active solid
state device that operates by a reverse bias adequate to cause avalanche breakdown. This is a high-power
diode and a very powerful microwave source that is used in high-frequency electronics and microwave
devices. They may be operated at frequencies up to about 350 GHz when manufactured with silicon.
The IMPATT diode exhibits a dynamic negative resistance that is required for microwave oscillation and
amplification applications. This is due to the following two reasons:
Impact Ionization avalanche effect: This causes the carrier current to lag behind the ac voltage by 90
degrees.
Transit time effect: This causes a further time delay and causes the external current to lag behind the
ac voltage by a further 90 degrees.
The summation of delay involved in generating avalanche current multiplication along with delay
due to transit time through drift space provides the necessary 180° phase difference between the applied
voltage and the resulting current in an IMPATT diode.
These devices can be classified as follows:
• Single drift devices: Devices such as P+NN+, N+PP+, P+NIN+, and N+PIP+ come under this cate-
gory. Consider the P+NN+ device. In this device, when the P+N junction is reverse biased, it causes
an avalanche breakdown. This causes the P+ region to inject electrons into the NN+ region. These
electrons move with a saturated velocity. However, the holes injected from the NN+ region do not
drift. Hence, these are called single drift devices.
• Double drift devices: The example of a double drift device is P+PNN+. In this device, when the
PN junction is biased near an avalanche breakdown, electrons drift along the NN+ region and holes
drift along the PP+ region. Hence, they are called double drift devices.

9.9.1 Principle of Operation of IMPATT Diode

To understand the operation of an IMPATT diode, here we consider the N+PIP+ diode. Let Vb be the
reverse bias breakdown voltage that is applied to the IMPATT device. Assume that a sinusoidal wave-
form V1 sinw t is superimposed on Vdc, resulting in a total device voltage V(t) = (Vdc + V1 sinw t) as shown
in Figure 9.12.

Chapter 09.indd 9.15 1/3/2014 10:39:36 AM

 9.16 | Microwave and Radar Engineering

Vdc V1sinwt

I
N + P (Intrinsic) P+

L
Electric
field
Distance

Voltage

0 p 2p

Hole current
External
Current
electron current
0
T/4

T/4

T t
/

/
T

Figure 9.12 IMPATT diode operation

Initially, the device contains a few thermally generated free electrons. When V(t) > Vb breakdown
occurs at the N+P junction, these electrons gain energy from the applied voltage and knock off electrons
in the valence band to the conduction band. As a result, a new electron hole pair is created. An electron-
hole pair generated because of such impact ionization is called a secondary electron-hole pair. These
secondary electrons again pick up sufficient energy and generate more secondary electron-hole pairs.
Therefore, as long as V(t) > Vb, the number of carriers increases exponentially, even beyond the voltage
maximum irrespective of magnitude of V(t). This is because of sufficient number of secondary electron-
hole pairs presence. This exponential increase continues until the sine wave crosses zero and then drops
exponentially until the sine wave reaches its negative peak. This avalanche current (generated holes) is
injected into the I-region and drifts toward P+ region with saturated velocity along the depletion region.
The electrons move toward the positive terminal. In this way, this current will have a one-quarter period
(T/4) delay or a 90° phase shift with regard to the applied signal voltage.
To achieve the desired 180° phase shift between input voltage and external current, additional T/4
delay is essential. This is made available by the hole drift along the depletion region. It is the property
of semiconductor materials that the drift velocity tends to be constant at high field strengths. Since the
holes move at the constant velocity vd, the device length may be chosen to provide the necessary delay
for a 180° phase shift between the device voltage and current, which is given by
T
l vd (9.8)
4

EXAMPLE PROBLEM 9.5

The carrier drift velocity in silicon semiconductors is vd = 105 m/s. What should be the length of silicon
IMPATT in order to obtain a negative resistance at about 12 GHz?

Chapter 09.indd 9.16 1/3/2014 10:39:36 AM

 Microwave Solid-state Devices | 9.17

Solution
In order to obtain a negative resistance, the carrier drifting through the device should give rise to a phase
shift of p /4, to be added to the avalanche phase shift of p /4. The drift time should, therefore, be T/4.
The device length is given by
T
l vd
4
1
T= = 8.33 × 10 −11 s
12 × 109
8 33 × 10 −11
Therefore, l = 105 ×
4
l = 2 μm

Equivalent Circuit: The IMPATT diode equivalent circuit is as shown in Figure 9.13. It is composed of
two parts that are the avalanche and drift regions, and a loss resistance (Rs). One part of the equivalent
circuit is the avalanche region it consists of a resonant circuit, with an avalanche inductance (La) and a
capacitance (Ca). The avalanche capacitance is given by:

εsS
Ca = (9.9)
wa
Ca Rd Cd Rs
Ld

Drift region
La

Avalanche region

Figure 9.13 Equivalent circuit of IMPATT diode

where, wa is the width of the avalanche region,

ε s = permittivity of the dielectric, and
A is the area.
The IMPATT diode exhibits negative resistance for frequencies higher than the avalanche resonant fre-
quency ( fa ), and is given by

fa = l / LaCa

The second part of the equivalent circuit is the drift region, it consist of a series resonant circuit, with
drift capacitance (Cd), and is given by
εsS
Cd = (9.10)
w wa

where w is total device width.

The operating frequency ( f ) of the IMPATT diode is given by

Chapter 09.indd 9.17 1/3/2014 10:39:36 AM

 9.18 | Microwave and Radar Engineering

f = l/(2t ) = vd /2L
where vd = Carrier drift velocity
L = Length of the drift space charge region
For the condition, f > fa the drift resistance (Rd) is negative. At frequencies more than avalanche
frequency, the avalanche sub-circuit behaves like capacitor. The drift region jointly with the avalanche
capacitance, account for the phase shift. At a specified frequency, the maximum power output of a
simple diode is limited by semiconductor material and the achievable impedance level in microwave
circuits. The most commonly used semiconductors are GaAs and Si. Though GaAs is costly and dif-
ficult to fabricate, it is the most preferred semiconductor because it gives less noise, high efficiency
and maximum operating frequency.
Derivation for the output power and efficiency of IMPATT diode
The maximum voltage applied across the diode is given by
Vm = Eml
where Em → Maximum electric field
l → Depletion length
The breakdown voltage limits the applied voltage Vm; as a result of this, the maximum current is also
limited, that is, the maximum current is given by
ε v ε E A
Im = Jm A = s Em A = s Em A = d s m
τ l
vd s E m A
∴ Im = Jm A = (9.11)
l
The upper limit of the input power is obtained as
vd s E m A
Pm = Vm Im = (Em l) ×
l
∴ Pm = vdes Em2 A (9.12)
εs A
We know that C= (9.13)
l
Substituting Eq. 9.13 in Eq. 9.12,
Pm = vd Em2 Cl
Multiplying and dividing the R.H.S with 2p f, we get
vd Em2 Cl × 2π f E2 v l ⎡ 1 ⎤
Pm = ⇒ Pm = m d ⎢∵ X c = 2π fC ⎥ (9.14)
2π f 2π fX
Xc ⎣ ⎦
Using 2π f τ 1 , we have
1 v ⎡ l ⎤
f = = d ⎢∵ τ = ⎥
2πτ 2π l ⎣ vd ⎦
vd
∵l = (9.15)
2π f
Substituting Eq. 9.15 in Eq. 9.14, we get
Em2 vd2 Em2 vd2
Pm = ∴ Pm f 2 =
4π f X c
2 2
4π 2 X c

Chapter 09.indd 9.18 1/3/2014 10:39:37 AM

 Microwave Solid-state Devices | 9.19

Pac ⎡Va × I a ⎤
Efficiency, η = =⎢ ⎥ (9.16)
Pdc ⎣Vd × I d ⎦

For practical IMPATT diodes, the efficiency is less than 30%.

9.9.2 Characteristics of IMPATT Diode

• IMPATT diode operates in reverse bias. It exhibits negative resistance region due to the impact
avalanche and transit time effects.
• The phase difference between voltage and current is 180°. Here 90° phase delay is due to ava-
lanche effect, and the remaining 90° is due to transit time effect.
• It is a narrow-band amplifier that provides output power in the millimeter-wave frequency range.
• At low frequencies, their power output is inversely proportional to frequency. At high frequencies,
their power output is inversely proportional to the square of frequency.
• They are often used in the design of oscillators and amplifiers when a high output power is
required. They provide higher output power than Gunn diodes.
• They are manufactured in Si, GaAs, and InP. They can be operated up to 350 GHz when manu-
factured in Si.
• These diodes are of low cost, reliable, and compact. They are moderately efficient milliwatt power
sources.
• These are noisier than Gunn diodes. Therefore, they are rarely used for local oscillators in receivers.
Performance characteristics
• Theoretical, h = 30% (< 30% in practice) and 15% for Si, 23% for GaAs
• Frequency: 1 to 300 GHz
• Maximum output power for a single diode: 5W in X band to 6.5 W at 30 GHz
• Several diodes combined: 40 W in X band
• Pulsed powers = 4 kW
Disadvantages of IMPATT diode
• In terms of noise figure an IMPATT diode is not good as in comparison with the TWT amplifier
or Gunn diode oscillator or klystron tube. Because the avalanche is a high noise process, so the
IMPATT is very noisy diode, the value of noise figure is 30 dB.
• In IMPATT diode matching is difficult because of the low value of their negative resistance.
• It is sensitive to operational conditions.
• It has large electronic reactance, which can cause detuning or burn out the device if proper care
is not taken.
Applications
• IMPATT diodes are used as microwave oscillators in microwave generators, in modulated output
oscillators.
• They are used in microwave links, continuous wave radars, and electronic counter measures.
• IMPATT diodes are also used as amplification with negative resistance. In police radars, low
power transmitters, and intrusion alarm devices are used the high-Q IMPATT diodes. In frequency
modulated telecommunication transmitters and continuous wave Doppler radar transmitters are
used the low-Q IMPATT diodes.

Chapter 09.indd 9.19 1/3/2014 10:39:38 AM

 9.20 | Microwave and Radar Engineering

EXAMPLE PROBLEM 9.6

Determine transit time of the carriers and operating frequency of the IMPATT diode. If an IMPATT
diode has drift length of 4 µm and vd =106 m/sec.
Solution
1 vd
Given that l = 4 μm, we have f = =
2τ 2l
where vd = Carrier drift velocity
l = Drift length
l 4 × 10 −6
Transit time, τ [ d 106 / ] τ = 4 × 10 −1122 Sec
vd 106
1 1
Operating frequency, f f = 0.25 × 110012 Hz = 125 GHz
2τ 2 × 4 × 10 12

EXAMPLE PROBLEM 9.7

An IMPATT has the following parameters: pulse operating current is 1A and a pulse operating voltage
is 200 V. The efficiency is 15%. If the pulse width is 0.02 ns and the frequency is 20 GHz, determine the
power output and the duty cycle.
Solution
Given that, Vo = 200 V, Io = 1 A.
Output power, P = η × Pdc
where Pdc = V0 × I0 = 200 × 1= 200 W
P = 0.15 × 200 W = 30 W
−9
Pulse width = 0.02 ns = 0.02 × 10 s, Frequency = 20 GHz = 20 × 109 Hz
1 1
Time, T = = = 5 × 10 −11 s
f 20 × 109
0 02 × 10 −9
Duty cycle = pulse width/total time = =04
5 × 10 −11

9.10 TRAPATT DIODE

IMPATT and Gunn diodes cannot operate at lower frequencies and they operate at frequencies of 3 GHz or
above. In these diodes as it is difficult to increase transit time. A slightly modified structure of the IMPATT
diode that can be used at low frequencies is the TRAPATT diode, where TRAPATT stands for trapped
plasma avalanche triggered transit mode. In 1967,the first TRAPATT diode was produced, which has an
efficiency of 25% and produces 400 W at 1000 MHz. Now a days, efficiencies of 60% to 75% are obtained.
Principle of operation: The TRAPATT diode depicted in Figure 9.14 (a) is a P+NN+ diode that is driven
by a large repetitive pulse of current. Breakdown will occur at one of the junctions. Since the current
drive is very large, a large collection of disassociated electrons and holes known as plasma is generated.
These carriers do not easily recombine. The violent breakdown creates a high electric field shock front
that moves across the N-type drift region. After the passage of shock front, the plasma is located in a low

Chapter 09.indd 9.20 1/3/2014 10:39:39 AM

 Microwave Solid-state Devices | 9.21

field region and is trapped because it takes a long time to clear the drift region of the carriers. When the
plasma has been cleared from the drift region, the cycle will be repeated. The formation of plasma in the
active region increases the transit time that is required for low-frequency operation. Initially, the device
starts oscillating in the IMPATT mode. When the amplitude of oscillations increases, the TRAPATT
mode of oscillation is established.
The operation can be better explained with help of Figure 9.14 (a).When a square current drive is
applied to the diode, the output will be as shown in Figure 9.14 (b). To better explain the electric field
variations, let it be divided into six regions as shown in Figure 9.14 (a).
L

P+ N N+

Plasma formed
Plasma extracted
d
Electric field

Output current

5 6
0.5T 1T
12 3 4
Time
0.5 T 1T
(a) (b)

Figure 9.14 (a) Structure of TRAPATT; (b) Output of TRAPATT

Region 1 (Charging): Initially, only thermally generated minority charge carriers are present, and,
hence, the current pulse charges up the depletion-layer capacitance. Therefore, the output voltage ramps
up as shown in Figure 9.14 (a).
Region 2 (Plasma formation): When the electric field becomes more than what is required to cause a
breakdown, a large number of carriers is generated. Hence, in region 2, internal current exceeds external
current. This results in the formation of a plasma region of a large number of disassociated holes and
electrons that are hard to be recombined. As a result of this, conductivity of the medium increases, volt-
age is reduced to lower values, and the electric field at the depletion region decreases.
Region 3 (Plasma extraction): As the voltage drops significantly, the plasma starts being removed.
However, the process takes a lot of time. As the plasma gets extracted, the current starts decreasing.
As the conductivity is decreasing, the voltage starts building up. In this region, the entire plasma is
removed.
Region 4 (Residual Plasma extraction): A few charge carriers remain on either side of the depletion
layer, resulting in a residual electric field in region 4, which gradually decreases.
Region 5 (Charging): In this region, the diode again behaves similar to a capacitor, resulting in the
electric pattern.
Region 6: The voltage remains constant as the current drops to zero.

Chapter 09.indd 9.21 1/3/2014 10:39:39 AM

 9.22 | Microwave and Radar Engineering

A sharp rise time current pulse, harmonic rich output of TRAPATT diode is shown in Figure 9.14 (b).
This pulse must be applied to a low-pass filter which is connected to the TRAPATT diode at the input
of the waveguide or transmission line. Here the harmonics are not accepted with the filter and thus are
reflected back to the TRAPATT diode to activate the next current pulse.
Performance characteristics of TRAPATT diode
• It can be operated at comparatively low frequencies.
• With five diodes connected in a series, the highest pulse power of 1.2 kW is obtained at 1.1 GHz.
• Operating voltage: 60–150 V
• Frequency: 3 to 50 GHz
• Efficiency: 15 to 40% (8 GHz to 0.5 GHz)
• Continuous power: 1–3 W between 8 GHz and 0.5 GHz
• Noise figure: greater than 30 dB
Advantages
• It is more efficient than the IMPATT diode
• Low power dissipation
Drawbacks
• High noise figure
• Strong harmonics due to short current pulse
Applications
• They are used in low-power Doppler radars, microwave beacon landing systems, phased array
radars, and so on.
Comparison of IMPATT and TRAPATT diodes
IMPATT diode TRAPATT diode
1. It stands for Impact Avalanche and Transit 1. It stands for Trapped Plasma Avalanche Triggered
Time Transit
2. Efficiency of operation is 30% 2. Efficiency is in between 15% and 40%
3. Frequency = 1 to 300 GHz 3. Frequency = 3 to 50 GHz
4. Pulsed power = 4 kW 4. Pulsed powers =1.2 kW at 1.1 GHz
5. It finds applications in microwave oscillators 5. They are used in low-power Doppler radars,
phased array radars, radio altimeters, and so on

9.11 BARITT DIODE

BARITT diode is an acronym for Barrier Injection Transit Time diode. These are improved versions of
IMPATT diodes. The ionization technique of IMPATT makes it very noisy; therefore it is avoided in
BARITT devices. It is replaced by barrier-injected minority carriers generated from forward-biased junc-
tions instead of being generated from impact ionization of the avalanche region as in the IMPATT diode.
Therefore, these devices are less noisy than IMPATT. The negative resistance is due to the drift of injected
holes to the P region. It consists of three regions of semiconductor material, as shown in Figure 9.15.

Chapter 09.indd 9.22 1/3/2014 10:39:40 AM