Lecture 26

OUTLINE
The BJT (cont’d)
• Ideal transistor analysis
• Narrow base and narrow emitter
• Ebers-Moll model
• Base-width modulation

Reading: Pierret 11.1-11.2; Hu 8.2-8.6

 Notation (PNP BJT)

NE  NAE NB  NDB NC  NAC

DE  DN DB  DP DC  DN
E  n B  p  C  n
L E  LN L B  LP LC  LN
nE0  np0 = ni2/NE pB0  pn0  ni2/NB nC0  np0  ni2/NC

EE130/230A Fall 2013 Lecture 26, Slide 2 R. F. Pierret, Semiconductor Device Fundamentals, Fig. 11.1
 “Game Plan” for I-V Derivation
• Solve the minority-carrier diffusion equation in each quasi-neutral
region to obtain excess minority-carrier profiles
– different set of boundary conditions for each region

• Find minority-carrier diffusion currents at depletion region edges

dnE dp B
I En   qAD E dx" I Ep   qAD
B dx
x " 0 x 0

dnC dp B
I  qAD I Cp   qAD
• Add hole Cn
& electronCcomponents
dx ' x ' 0
together  B dx
terminal x W
currents

EE130/230A Fall 2013 Lecture 26, Slide 3

Emitter Region Analysis
d 2 n E
• Diffusion equation: 0  DE dx "2
 nEE

• General solution: nE ( x" )  A1e  x"/ LE  A2 e x"/ LE

• Boundary conditions: nE ( x"  )  0

nE ( x"  0)  nE 0 (e qVEB / kT  1)

• Solution: nE ( x" )  nE 0 (e qVEB / kT  1)e  x"/ LE

nE
I En   qADE ddx "  qA DLEE nE 0 (e qVEB / kT  1)
x " 0

EE130/230A Fall 2013 Lecture 26, Slide 4

Collector Region Analysis
d 2 nC
• Diffusion equation: 0  DC dx '2
 nCC

• General solution: nC ( x' )  A1e  x '/ LC  A2 e x '/ LC

• Boundary conditions: nC ( x'  )  0

nC ( x'  0)  nC 0 (e qVCB / kT  1)

• Solution: nC ( x' )  nC 0 (e qVCB / kT  1)e  x '/ LC

I Cn  qADC ddxn'C  qA DLCC nC 0 (e qVCB / kT  1)

x ' 0

EE130/230A Fall 2013 Lecture 26, Slide 5

Base Region Analysis
d 2 p B
• Diffusion equation: 0  DB dx 2
 pBB

• General solution: pB ( x)  A1e  x / LB  A2 e x / LB

• Boundary conditions: pB (0)  pB 0 (e qVEB / kT  1)

pB (W )  pB 0 (e qVCB / kT  1)

• Solution: p B ( x )  p B 0 ( e qVEB / kT
 1)  e ( W  x ) / LB  e  ( W  x ) / LB
eW / LB  e W / LB

 pB 0 ( e qVCB / kT
 1)  e x / LB  e  x / LB
eW / LB  e W / LB

EE130/230A Fall 2013 Lecture 26, Slide 6
 Since sinh     e  e  
2

we can write

pB ( x)  pB 0 (e qVEB / kT
 1)  e ( W  x ) / LB  e  ( W  x ) / LB
eW / LB  e W / LB

 p B 0 (e qVCB / kT
 1)  e x / LB  e  x / LB
eW / LB  e W / LB

sinh   W  x  LB 
pB ( x)  pB 0 (e qVEB / kT  1)
sinh  W LB 
as

sinh  x LB 
 pB 0 (e qVCB / kT  1)
sinh  W LB 
EE130/230A Fall 2013 Lecture 26, Slide 7
 d d  e  e   e  e 
sinh        cosh  
d d  2  2

sinh  W  x  LB  sinh x LB 
pB ( x)  pB 0 (e qVEB / kT  1)  p B 0 (e qVCB / kT
 1)
sinh LB 
W sinh W LB 

dp B
I Ep   qAD B dx
x 0

 qA DB
LB p cosh(W / LB )
B 0 sinh(W / LB ) (e qVEB / kT
 1)  1
sinh(W / LB )
e qVCB / kT

1

I Cp   qADB ddxpB
x W

 qA DB
LB p  1
B 0 sinh(W / LB ) (e qVEB / kT
 1) 
cosh(W / LB )
sinh(W / LB )
e qVCB / kT
1 
EE130/230A Fall 2013 Lecture 26, Slide 8
 BJT Terminal Currents
• We know:
DE
I En  qA LE nE 0 (e qVEB / kT  1)

I Ep  qA DB
LB
pB 0  cosh(W / L B )
sinh(W / L B )
( e qVEB / kT
 1)  1
sinh(W / L B )
e qVCB / kT
1  
I Cp  qA DB
LB p  1
B 0 sinh(W / LB ) (e qVEB / kT
 1) 
cosh(W / LB )
sinh(W / LB )
e qVCB / kT
1 
DC
I Cn  qA LC
nC 0 (e qVCB / kT  1)

• Therefore:
I E  qA DE
LE
nE 0 
DB
LB
pB 0
cosh(W / L B )
sinh(W / L B )
(e qV EB / kT
 1)   DB
LB

pB 0 sinh(W1 / LB ) e qVCB / kT  1 
 qA  (e   e 
DB DC cosh(W / LB )
IC LB pB 0 1
sinh(W / LB )
qVEB / kT
 1)  LC nC 0  DLBB pB 0 sinh(W / LB )
qVCB / kT
1
EE130/230A Fall 2013 Lecture 26, Slide 9
 BJT with Narrow Base
• In practice, we make W << LB to achieve high current gain.
Then, since
sinh      for   1
cosh     1  2
2 for   1

we have:

pB ( x)  pB 0 (e qVEB / kT  1)1  Wx 

 pB 0 (e qVCB / kT  1) Wx 

EE130/230A Fall 2013 Lecture 26, Slide 10 R. F. Pierret, Semiconductor Device Fundamentals, Fig. 11.2
 BJT Performance Parameters
1
  ni E 2 D N W
Assumptions:
1 E B
ni B 2 DB N E LE
• emitter junction forward
biased, collector junction
1 reverse biased
T 
1 1
2  
W 2
LB
• W << LB
Replace LE with WE’ if WE’ << LE
1
 dc 
1
ni E 2 D N W
E B
ni B 2 DB N E LE
 1
2  
W 2
LB

1
 dc  ni E 2 D N W
E B
ni B 2 DB N E LE
 1
2  
W 2
LB

EE130/230A Fall 2013 Lecture 26, Slide 11

 Ebers-Moll Model

increasing

(npn) or VEC (pnp)

C. C. Hu, Modern Semiconductor Devices for Integrated Circuits, Figure 8-2

The Ebers-Moll model is a large-signal equivalent circuit which

describes both the active and saturation regions of BJT operation.
•Use this model to calculate IB and IC given VBE and VBC

EE130/230A Fall 2013 Lecture 26, Slide 12

I E  qA  DE
LE
nE 0 
DB
LB
pB 0
cosh(W / L B )
sinh(W / L B )
(e qV EB / kT
 1)   DB
LB

pB 0 sinh(W1 / LB ) e qVCB / kT  1 
 qA  (e  e  1 
DB DC DB cosh(W / L B )
IC LB
p 1
B 0 sinh(W / L B )
qV EB / kT
 1)  LC
nC 0  LB
pB 0 sinh(W / L B )
qVCB / kT

If only VEB is applied (VCB = 0):

V EB V CB
I E  I F 0 ( e qVEB / kT  1) IB

I C   F I F 0 ( e qVEB / kT  1)
I B  1   F  I F 0 ( e qVEB / kT  1) E B C
IC

If only VCB is applied (VEB = 0): : aR : reverse common base gain

aF : forward common base gain
I C   I R 0 (e qVCB / kT  1)
I E   R I R 0 (e qVCB / kT  1)
Reciprocity relationship:
I B  I R 0 (1   R )(e qVCB / kT
 1) DB pB 0
 F I F 0   R I R 0  qA
LB sinh(W / LB )
EE130/230A Fall 2013 Lecture 26, Slide 13
 In the general case, both VEB and VCB are non-zero:

I C   F I F 0 (e qVEB / kT
 1)  I R 0 ( e qVCB / kT
 1)
IC: C-B diode current + fraction of E-B diode current that makes it to the C-B junction

I E  I F 0 (e qVEB / kT  1)   R I R 0 (e qVCB / kT  1)
IE: E-B diode current + fraction of C-B diode current that makes it to the E-B junction
Large-signal equivalent circuit for a pnp BJT

R. F. Pierret, Semiconductor Device Fundamentals, Fig. 11.3

EE130/230A Fall 2013 Lecture 26, Slide 14

 Base-Width Modulation
Common Emitter Configuration, Active Mode Operation
W
IE IC
P N P IC 1
  dc 
 
+
niE 2 DE N B W
IB  12 W 2
n 2 D B N E LE
iB
LB

2
niB DB N E LE
+
VEB
  2
niE DE N BW
pB(x) IC
 
p B 0 e qVEB / kT  1
(VCB=0)

x VEC
0 W(VBC)
EE130/230A Fall 2013 Lecture 26, Slide 15
Ways to Reduce Base-Width Modulation
1. Increase the base width, W

2. Increase the base dopant concentration NB

3. Decrease the collector dopant concentration NC

Which of the above is the most acceptable action?

EE130/230A Fall 2013 Lecture 26, Slide 16

 Early Voltage, VA
1
 I C  V
Output resistance: r0     A
 VEC  IC

A large VA (i.e. a large ro ) is desirable

IC
IB3

IB2

IB1

0 VEC
VA

EE130/230A Fall 2013 Lecture 26, Slide 17

 Derivation of Formula for VA
dI C I IC
Output conductance: g 0   C
dVEC VA
 VA 
g0

dI C dI C
VEC  VEB  VBC so g o   for fixed VEB
dVEC dVBC

dI C dW dI C  dxnC  where xnC is the width of the

go        collector-junction depletion region
dW dVBC dW  dVBC  on the base side
xnC

P+ N P

EE130/230A Fall 2013 Lecture 26, Slide 18

 I C  qA  DB
LB

pB 0 sinh(W1 / LB ) (e qVEB / kT  1)   DC
LC
nC 0 
DB
LB
pB 0
cosh(W / L B )
sinh(W / L B )
 e qVCB / kT

1

IC 
qAni2 DB qVEB / kT
WN B
e 
1 
dI C
dW
qAni2 DB qVEB / kT
 2
W NB
e 
1  
IC
W

dQdepC d (qN B xnC ) dxnC
C JC    qN B
dVBC dVBC dVBC
dxnC C JC
 
dVBC qN B

IC IC IC qN BW
VA    
g 0 dI C  dxnC   I C   C JC  C JC
          
dW  dVBC   W   qN B 
EE130/230A Fall 2013 Lecture 26, Slide 19
Summary: BJT Performance Requirements
• High gain (dc >> 1)
 One-sided emitter junction, so emitter efficiency  1
• Emitter doped much more heavily than base (NE >> NB)

 Narrow base, so base transport factor T  1

• Quasi-neutral base width << minority-carrier diffusion length (W << LB)

• IC determined only by IB (IC  function of VCE,VCB)

 One-sided collector junction, so quasi-neutral base width W does
not change drastically with changes in VCE (VCB)
• Based doped more heavily than collector (NB > NC)
(W = WB – xnEB – xnCB for PNP BJT)

EE130/230A Fall 2013 Lecture 26, Slide 20