EECS 105 Spring 2004, Lecture 20 EECS 105 Spring 2004, Lecture 20 Prof. J. S.

Smith

Context
This lecture will discuss
Lecture 20: PN diodes (forward z Diode Currents in forward and
bias), small signal model, BJTs reverse bias (6.1-6.3)
z Small signal models for diodes

And if we have some time, we will

Prof J. S. Smith start:
z BJTs (Bipolar Junction Transistors)

Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Reading Lecture Outline

z The midterm covers chapters 1-4, plus phasors,
linear circuits, and Bode plots z Diode applications
z Today we will be covering the material in chapter z Forward currents in diodes
6, PN diodes
z Diode Small Signal Model
z Wednesdays lecture will be review, example
z Diode Charge Storage (6.4.4)
problems, and answers to any last minute questions.
z Types of diodes
z After the midterm, we will be covering chapter 7,
Bipolar Junction Transistors (BJT’s) z The BJT (7.1)
z BJT Physics (7.2)

Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

1
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Diode applications More Diode applications

z Varactor Tuner z High frequency up-converter mixers
z Rectification z Down-conversion mixers
z Envelope Detector (AM demodulator, for instance) z Tunnel diode oscillator
z Phase modulator z GUNN diode oscillator
z Voltage reference/Limiters/Regulators z Light emitting diodes
z Diode Clamp z Laser diodes
z ESD (electrostatic discharge) protection circuits z Light detectors
z Voltage multipliers (doublers, etc) – Solar cells
– PIN diodes
– Avalanche diodes

Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Minority carrier injection Generation and Recombination

z In the last lecture, we discussed the ideal diode
equation, and how it treats the minority carriers as
if they always make it all the way across the
junction, but then disappear once they have gotten
across (we didn’t account for the possibility that
they wandered back, for instance)
z Now, we will more accurately account for the
z The processes which cause the electron and hole
populations to come into equilibrium with each
other are rather slow in silicon, so we can make
some approximations:
z Generation and recombination will be neglected
inside the depletion zone, for one.

transport of the minority carriers. Sidebar:

Conduction band
Including: It is possible to add atoms to silicon which
greatly enhance generation and recombination,
z Diffusion such as gold and copper. The are called traps. Trap, ex: Gold atom

Except for very pure silicon, impurities will

z Recombination/Generation
dominate carrier recombination Valence band

Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

2
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Minority Carriers at Junction Edges

minority carrier concentration at boundaries of depletion The minority carrier concentrations at the edges of
region increase as barrier lowers. This is called minority the depletion region will then be given by:
carrier injection. If we neglect generation and
recombination inside the depletion regions, the number
of carriers function is approximately the density of pn ( x = xn ) = N Ae −q (φ B −VD ) / kT
carriers on the other side of the barrier at the equivalent
energy: n p ( x = − x p ) = N D e − q (φ B −VD ) / kT
p n ( x = xn ) (minority) hole conc. on n-side of barrier
=
p p (x = −x p ) (majority) hole conc. on p-side of barrier Note: NA and ND are the majority carrier concentrations on
= e −( Barrier Energy ) / kT the other side of the junction, with the assumption
that pn << ND and np << NA
p n ( x = xn )
= e − q (φ B −VD ) / kT
NA
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Quasi-Neutrality Diffuse and Recombine

z Since the regions outside the depletion regions are z Once the minority carriers have been injected
going to be very close to electrically neutral (called across the depletion region, they will diffuse, and
quasi-neutrality) the number of majority carriers they will recombine.
They will recombine, because now pn > ni
2
will increase slightly as well (slightly because there z
are usually many more majority carriers than and since recombination is proportional to pn, it
minority carriers anyway) will now cause carriers to recombine at a rate faster
than they are generated.
Carrier Nd z They will also diffuse into the other side, because
concentrations
there are more of them (the minority carriers) at the
Minority carriers edge of the depletion region than there are further
in.
distance
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

3
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Excess injected minority carriers Ambipolar diffusion

qVA
z The diffusion constant for minority carriers is complicated
pn 0 e kT by the fact that the minority carriers are dragged on (or
p side n side pulled!) by the majority carriers, in a mechanism called
qVA
ambipolar diffusion
x
n p 0e kT ⎛ qVA ⎞ −L z There is a small E field which keeps the excess holes and
pn ( x) = pn 0 + pn 0 ⎜ e kT − 1 ⎟ e p
⎝ ⎠ electrons together.
Minority Carrier z The ambipolar diffusion constant for minority carriers
np0
pn 0 Diffusion Length comes out to be the diffusion constant for the other,
-Wp -xp xn Wn
majority carriers, in the cases where the majority carriers
greatly outnumber the minority carriers. In other words,
Once the minority carriers are injected into the other side of the minority holes diffuse with Dn and electrons with Dp!
junction, the minority carrier concentration in the bulk region z The minority diffusion length is a function of the ambipolar
for forward bias is a decaying exponential. diffusion constant for the minority carriers, as well as the
variability in the lifetime, etc. We will just take it to be a
number!
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Steady-State Concentrations Diode Current Densities

Ln , p >> Wn , p qVA
pn 0 e kT qVA
In silicon, it is easy for the diffusion lengths to be dn p n p 0 e kT − n p 0
( x) ≈
p side qVA n side

much longer than the distance to the contacts, if n p 0e kT

dx − x p − (−W p )
none of the diffusing holes and electrons pn 0
recombine until they get to the contactsÆ get a np0 np0 =
ni2
Na
linear concentration gradient. -Wp -xp xn Wn

qVA
pn 0 e kT Under this linear approximation, we can calculate the current due to the holes
diffusing into the N side, and the electrons diffusing into the P side. Notice that
p side n side we are using the electron diffusion constant for the minority holes,
qVA The population is assumed and visa versa.
kT to be the equilibrium
n p 0e population at the contacts
because of the many dn p Dn ⎛ qVA ⎞
traps (defects) there J ndiff = qDn ≈q n p 0 ⎜ e kT − 1⎟
dx x =− x p
Wp ⎝ ⎠
np0
pn 0
dp Dp ⎛ qVA
⎞
J pdiff = − qD p n ≈ −q pn 0 ⎜ 1 − e kT ⎟
-Wp -xp xn Wn dx x = xn Wn ⎝ ⎠
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

4
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Net forward current (short device) Fabrication of IC Diodes

The total current is the sum of the currents carried by
cathode anode
the minority carriers on each side:
p+ p n+
n-well
⎛ Dp Dn ⎞ ⎛ qVA ⎞
p-type
J diff
= qn ⎜ 2
+ ⎟⎟ ⎜ e kT − 1⎟
⎜ N dWn N aW p
i p-type
⎝ ⎠⎝ ⎠

z Start with p-type substrate

z Create n-well to house diode
z p and n+ diffusion regions are the cathode and anode
z N-well must be reverse biased from substrate
z Parasitic resistance due to well resistance
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Diode large signal model Large-Signal Model

z It is inconvenient to use the exponential current
model in a circuit, so a diode is often modeled with
an approximation. One possible large signal model
is: The resistance R is the forward resistance
i (t ) → at large currents
+
The forward voltage drop Vf is about .7 volts
R because that is about where the forward
current goes from negligible to very large!
v (t )
The diode in the model is a perfect diode,
C perfect conductor when forward biased, open
when reverse biased
Vf
The choice of capacitance C depends on
− which is most important, the capacitance under
forward or reverse bias conditions.
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

5
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Diode Small Signal Model Forward Diode Capacitance

z The I-V relation of a diode can be linearized z We have already seen that a reverse biased diode
acts like a capacitor since the depletion region
⎛ q (Vd + vd ) ⎞ qVd qvd
grows and shrinks in response to the applied field.
I D + iD = I S ⎜ e kT − 1⎟ ≈ I S e kT e kT
⎝ ⎠ the junction capacitance in forward bias is given by
x 2 x3 εS
e = 1 + x + + +L
x
Cj = A ≈ 1.4C j 0
2! 3! X dep
z But another charge storage mechanism comes into
⎛ qv ⎞ play in forward bias:
I D + iD ≈ I D ⎜ 1 + d + L ⎟
⎝ kT ⎠ z Minority carriers injected into p and n regions must
qVd
qI D be built up, which takes current×time, and they
ID = ISe kT iD ≈ vd = g d vd
kT must also be extracted as the voltage is lowered.
The small signal model of a diode in forward bias is a resistance in
parallel with a capacitance. In reverse, it is just a capacitance. (the z The effect is additional charge is stored in diode
reverse leakage current is constant, thus no contribution to small signal)
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Charge Storage Symbols

q (Vd + vd )
kT
pn 0 e
p side q (Vd + vd )
n side
n p 0e kT Extra charge
Stored in diode

pn 0
np0
-Wp -xp xn Wn

z Increasing forward bias increases minority charge

density
1 qI d
z A detailed analysis yields: Cd = 2 kT τ T
1
Cd = g dτ T Time to cross junction
(or minority carrier lifetime)
2
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

6
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith
The PIN Diode
z The PIN diode has heavily doped p-type and n-type
regions separated by an intrinsic region. When
reverse biased, it acts like an almost constant
capacitance and when forward biased it behaves as
a variable resistor.
z The built in field stretches over the intrinsic region,
causing minority carriers to be swept out by the
field over a larger volume. It is often used for light
detectors, and some high efficiency solar cells.
Department of EECS University of California, Berkeley
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith
Step Recovery Diode
z In the step-recovery diode the doping level is
gradually decreased as the junction is approached.
z This reduces the switching time since the smaller
amount of stored charge near the junction can be
released more rapidly when changing from forward
to reverse bias.
z The forward current can also be established more
rapidly than in the ordinary junction diode.
z This diode is used in fast switching applications,
such as high frequency mixers
Department of EECS University of California, Berkeley
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith
Schottky Diode
z Metal junction with a (typically) n-type
semiconductor.
z Mainly used in high frequency circuits or high
speed digital circuits.
z There are no holes (minority carriers), so the
conduction quickly stops upon change to reverse
bias.
z Schottky diodes find application as rectifiers for
high frequency signals.
Department of EECS University of California, Berkeley
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith
Zener diodes
z Diodes which are called Zener diodes do not use the Zener
effect (tunneling) but are avalanche breakdown devices.
z With the application of sufficient reverse voltage, a p-n
junction will experience a rapid avalanche breakdown and
conduct current in the reverse direction.
z Under a high electric field, high energy carriers can cause
the generation of more electron hole pairs, and the
subsequent collisions quickly become an avalanche. When
this process is taking place, very small changes in voltage
can cause very large changes in current.
z Zener diodes can be made which break down at precise
voltages from about 4 volts to several hundred volts. The
avalanche breakdown occurs at a particular field strength,
so the high field region just needs to be the correct length
z Avalanche breakdown does not damage the diode as long as
power dissipation limits are not exceeded.
Department of EECS University of California, Berkeley
7
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Bipolar Junction Transistor (BJT) Ideal BJT Structure

IC + IE
z A BJT is physically just two back to back PN
diodes, with three contacts, but the current between VEB
Collector (N) IB + Emitter (P) +
the emitter and the collector is a minority carrier −
current in the base. + VCE VEC
Base (P) Base (N)

z Essentially, a forward biased diode is used to create VBE − IB −

−IE − IC
a minority current, most of which then goes all the Emitter (N) − Collector (P)

way across to the depletion region of another,

reverse biased diode.
z NPN or PNP sandwich (Two back-to-back diodes)
z The geometry can be such that almost all the
current goes across to the second diode, so that the z How does current flow? Minority carriers diffusing
controlling electrode doesn’t have to supply much across the base, which is thin so most go across
of the current, maybe 1:100 to 1:400 z A good BJT satisfies the following IC ≈ − I E
qVBE
I C >> I B IC ≈ I S e kT
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

Actual BJT Cross Section BJT Layout

z Vertical npn sandwich (pnp is usually a lateral structure)

z n+ buried layout is a low resistance contact to collector
z Base width determined by vertical distance between emitter
z Emitter area most important layout parameter
diffusion and base diffusion
z Multi-finger device also possible for reduced base resistance
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

8
EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

BJT Schematic Symbol Simple NPN BJT model

IC = β I B z A simple model for a NPN BJT:
qVBE C
IC ≈ I S e kT
IB +
+ VCE I B (t ) →
VBE − B
−
−IE The arrow on the symbol
shows the controlling diode.
+ βiB (t ) C
VBE (t )
IB +
z Collector current is control by base current linearly, a
typical value would be β = 100 , because only one in 100 − + VCE
electrons would stop in the base instead of making it across
Real diode, not E VBE −
to the collector −IE
z Collector is controlled by base-emitter voltage exponentially
an ideal diode −
Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley

EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith EECS 105 Spring 2004, Lecture 20 Prof. J. S. Smith

BJT Collector Characteristic Collector Characteristics (IB)

z Ground emitter Saturation Region Breakdown

(Low Output Resistance)
z Fix VCE
z Drive base with Linear Increase
Reverse Active
fixed current IB (poor Transistor)
z Measure the
collector current

Forward Active
Region
(Very High Output Resistance)

Department of EECS University of California, Berkeley Department of EECS University of California, Berkeley