Beruflich Dokumente
Kultur Dokumente
I0
ep
h+
np and pn low
I0 small
I0
ep
n
h+
I0
1. Increase
2. Decrease
3. Remain the same
V
ep
n
h+
If np and pn higher
|I0| larger
I0
V
ep
Test: True-False
n
h+
I
If we only increase pn
then |I0| will still increase.
I0
V
h+
p
n
e-
Hole injector
Take pn diode
h+
p+p
e-
If
Ip pno (eeV/kT-1)
In npo (eeV/kT-1)
Since NA >> ND
np << pn
Ip >> In
Thus:
A forward biased p+n diode is a good hole injector
n
h+
e-
ep
If W large, then?
W
1.
2.
3.
4.
Recombination of excess holes will occur and excess will be 0 at end of layer
Recombination of excess holes will occur and excess will be large at end of layer
No recombination of excess holes will occur.
Recombination of excess electrons will occur and excess will be np0 at end of layer
Thus:
A forward biased p+n diode is a good hole injector
dpn
n
e- L
p
W
n
h+ x
ep
If W large holes
recombine
Excess hole
concentration reduces
exponentially in W to
some small value.
dpn
n
e- L
p
n
h+ x
ep
W
1.
2.
3.
Thus:
A forward biased p+n diode is a good hole injector
dpn
n
e- L
p
W
Reduce W
n
h+ x
ep
if W large holes
recombine
Since gradient of dpn @
x=W is zero, hole
diffusion current is also
zero
BJT p+np
E: emitter
B: base
C: collector
EB
E
IC
p+
IE
IC
B
n
VVBC
W < Lp
IE
VBC
pno
npo
pno
npo
Double junction
npo
npo
dp( x, t )
2dp( x, t ) dp( x, t )
Dp
2
t
p
x
Steady state
d 2dp( x) dp( x)
Dp
0
2
p
dx
dp( x) C1 exp
C2 exp
Lp
Lp
C1=0
C20
n+ Si
ohmic contact
device insulation
n+ Si
p+ Si
p Si
p+ Si
n Si
p Si
p-substrate
n-well for collector
IE
p+
n
holes
IB
IC
e- gain, reverse bias
holes
IE
IC
IB
IB
ICB0
Recombination
e- loss
IB = IB + IB ICB0
Electrostatically neutral
IE = Ip
t transit time
t < p
h+
e-
Wb << Lp
recombine with
Based on the given timescales, holes can pass through the narrow base
before a supplied electron recombines with one hole: ic/ib = p/t
The electron supply from the base contact controls the forward bias to
ensure charge neutrality!
Injection of carriers
Wish list:
E IEp B
C
IEp>>IEn
IC
or g = IEp/(IEn + IEp) 1
g: emitter injection efficiency
equilibrium
IC IEp
or B= IC/IEp 1
eVBE>0
B: base transport factor
IEn
or a= IC/IE 1
x
h+ Wb < Lp
a: current transfer ratio
IB IEn + (1-B) IEp
No amplification!
thus b= IC/IB = a/(1-a)
b: current amplification factor
Amplification!
ICB0 ignored
EB
p+
IE
IC
B
n
VV
BC
p
W < Lp
Forward biased p+n
junction is a hole injector
EB
p+
V
IE IB=IB+IB I C VBC
B
n
W < Lp
Forward biased p+n
junction is a hole injector
Review 2
Amplification?
IB = IB + IB ICB0
Recombination only case: IB, ICB0 negligible
ic/ib = p/t
b = p/t
With
Currents?
In order to calculate currents in pn junctions, knowledge of
the variation of the minority carrier concentration is
required in each layer.
The current flowing through the base will be determined by
the excess carrier distribution in the base region.
Simple to calculate when the short diode approximation is
used: this means linear variations of the minority carrier
distributions in all regions of the transistor. (recombination
neglected)
Complex when recombination in the base is also taken into
account: then exponential based minority carrier
concentration in base.
p(x)
dp(x)
DpE
Without recombination
pn0
pn0
00
With recombination
DpC
Wb
DpE
DpC
0
Wb
DpE DpC
DpE
A
Wb
Wb
B DpE DpC DpE
DpC
0
dp( x) DpE 1
W
b
Note: no recombination
Wb x
Collector current: Ip
Diffusion current: I p eADp
ddp( x)
dx
ddp( x)
DpE
dx
Wb
Hole current:
Wb
Wb
eVEB
kT
Look at emitter: In
minority carrier density gradient in the emitter
dn(x)
Dnp
n p0 Dn p
xe
Dn p
xe
B Dn p
Dn p
x Dn p
dn( x)
xe
0
xe
x
0
Base current: In
ddn( x)
Diffusion current: I n eADn
dx
Base current:I B
In
eADn n p0 e
eVEB
kT
xe
The base contact has to re-supply only the electrons that are
escaping from the base via the base-emitter junction since no
recombination IB=0 and no reverse bias electron injection
into base ICB0=0.
Emitter current
The emitter current is the total current flowing through the
base emitter contact since IE=IC+IB (current continuity)
Emitter current:
Dn n p0 D p pn0
I E I n I p eA
WB
xe
I C I p D p pn0 xe
Current gain: b
I B I n Dn n p0Wb
eVEB
kT
dc(x)
IE = IpEB + InEB
DpE
IC = IpBC + InBC
DnE
IC IpBC = IpEB
IE = IB
+ IC
IB = IE
- IC
IB = InEB
-Xe
DpC
DnC
Wb Xc
dc(x)
DpE
DnE
-Xe
-LpE
DpC
DnC LpC
Wb < LnB
x
Xc
DpE
DnE
-Xe
-LpE
0
Emitter
Collector
DpC
DnC LpC
Wb < LnB
x
Xc
DpE
DpC
Wb
DpE
dp(x) = C3 e-x/Lp
Boundary condition at BC
junction cannot be guaranteed
DpC LnB
Wb
http://www.ecse.rpi.edu/~schubert/Course-ECSE-2210-Microelectronics-Technology-2010/
dc(x)
IE = IpEB + InEB
DpE
IC = IpBC + InBC
IC IpBC
DnE
IE = IB
+ IC
IB = IE
- IC
-Xe
-LpE
DpC
DnC LpC
Wb < LnB
x
Xc
dp(x)
DpE
Starting point:
Assume IE=IEp & IBC0= 0
DpC 0
Wb
Then IE = Ip(x=0)
x
and IC = Ip(x=Wb)
IB=IE - IC =IB
dp(x)
DpE
IE = Ip(x=0)
DpC
IC = Ip(x=Wb)
Wb
Hyperbolic functions
BE
p+
IE
IC
Bn
VVBC
Depletion width
changes with VBC
D p pn0
W
Wbb
eVBE
kT
ideal
VA
-vCE
IB
Conclusions
Characteristics of bipolar transistors are based on
diffusion of minority carriers in the base.
Diffusion is based on excess carrier concentrations:
dp(x)
Transistor switching
Ic
p-type material
n-type material
iC
iC
ECC
RS
es
ib higher
RL
iB
iE
ECC /RL
ib
On
es
-vCE
ECC
Off
iC
iC
ic=biB
RL
iB
ECC
RS
es
Es
-Es
t
iE
-vCE
iC
ic=biB
iC
RL
ECC
RS
es
Es
-Es
iE
-vCE
iC
icbiB
iC
RL
ECC
RS
es
Es
-Es
iE
-vCE
Switching cycle
iB
IB
iC
Switch to ON
IBEs/RS
Switch OFF
RL
iB
ECC
RS
DpE
QB
es
Es
-Es
-IB
DpE
iE
Qs
iC
DpE
t1 ts
ECC /RL
iC
IC
ECC
-vCE
t2
ts
tsd
ICECC/RL
-pn
dp
t2
ts
t1
DpC
0
t0
Wb
Saturation
dp
DpE
DpE
dp
VBC>0
DpC
-pn
x
Wb
x
Wb
Currents - review.
forward active mode
dc(x)
IE = IpEB + InEB
DpE
IC = IpBC + InBC
IC IpBC
DnE
IE = IB
+ IC
IB = IE
- IC
-Xe
-LpE
DpC
DnC LpC
Wb < LnB
x
Xc
iB
iB
IB
IBEs/RS
-IB
DpE
QB
es
Es
-Es
DpE
iE
Qs
ECC
-vCE
With IB>ICmax/b
Over-saturation
ECC
RS
Switch to ON
DpE
t1 ts
iC
IC
t2
-pno
dp
t2
ts
t1
DpC
t0
Wb
ICmaxECC/RL
iB
Switch OFF
-IB-Es/RS
ECC
RS
DpE
QB
es
Es
-Es
iB
IB
DpE
iE
Qs
ECC
-vCE
iC
IC
t2 ts
tsd
t3 t4
ICECC/RL
-pno
dp
t2
ts
t3
t4
DpC
0
Wb
ON switching
OFF=0ON
RL
C
p
RS
e(t)
B n
p
vbc
ECC
veb
E
iC
dpnB(x)
E I
B
IB
IB
ICsat
QB
IBp
Qsat
IB
IB
0
WB x
t<0
t0
veb= 0ON0.7V
E-p
B-n
tsat
i (t )
dQB (t ) QB (t )
dt
p
t
QB I B p 1 exp
p
&
RS
+E>>0.7V
E
IB
RS
QB
x WB
e A dn
x 0
tsat
pB
( x)dx
t<tsat
iC (t )
QB (t )
t
I B p
1 exp
t
p
ttsat
iC (t )
ECC
I Csat
RL
Driving off
Time to turn the BJT OFF is determined by:
1) The degree of over-saturation (BC junction)
2) The off-switching of the emitter-base diode
ib
IB
ib
IB
CASE 1: OFF=IB=0
0N (saturation)OFF
Qb
IB p
CASE 2: OFF=-IB
0N (saturation)OFF
-IB
Qb
IB p
Qs
Qs = IC t
iC
IC
tsd
-IB p
iC
IC
tsd
OFF switching
RS
e(t)
B n
p
vbc
ECC
veb
E
dpnB(x)
E t<0
IB
t0 IB=0
iC tsd
B
ICsat
QB
IBp
tsd
Qsat
WB x
t<0
RS
IB
tsd
B-n
E=0V
E
0
RS
tsd
i (t )
dQB (t ) QB (t )
dt
p
t
QB (t ) I B p exp
p
&
QB
tsd
x WB
e A dn pB ( x)dx
x 0
t<tsd
iC (t )
ECC
I Csat
RL
ttsd
iC (t )
QB (t )
I B p
t
exp
p
RS
e(t)
B n
p
ECC
veb
E t<0
IB
t0
vbc
dpnB(x)
iC
tsd
ICsat
-IB
QB
IBp
tsd
tsd
Qsat
WB x
tsd
tsd
t<0
veb= 0.7V (ON)-E
E-p
B-n
i (t )
dQB (t ) QB (t )
dt
p
RS
IB
-E
E
RS
t
QB (t ) I B p 2 exp 1
p
&
QB
x WB
e A dn pB ( x)dx
x 0
t<tsd
iC (t )
ECC
I Csat
RL
ttsd
iC (t )
QB (t )
t
I B p
2 exp 1
t
p
iC tsd
ICsat
iC
t<tsd
E
iC (t ) CC I Csat
RL
tsd
iC (t )
QB (t )
I B p
t<tsd
iC (t )
ICsat
ttsd
tsd
ECC
I Csat
RL
ttsd
t
exp
p
iC (t )
QB (t )
tsd
t
I B p
2 exp 1
t
p
t
t sd
ECC I B p
iC (t sd ) I Csat
exp
RL
t
p
I B p
t sd p ln
Csat t
iC (t sd ) I Csat
ECC I B p
RL
t
t
2 exp sd
I B p
t sd p ln
1 1 I B p
I Csat t
2 2 I Csat t
shorter delay
Transients
Turn-on: off to saturation
iC
IC
ICECC/RL
t
ts
Time to saturation
ON switching
QB
OFF=0ON
IBp
RL
t
QB I B p 1 exp
p
Qsat
C
RS
e(t)
p
B n
p
vbc
ECC
tsat
veb
iC
iC (t )
ICsat
ttsat
t<tsat
iC (t )
QB (t ) I B p
t
t
ECC
I Csat
RL
t
1 exp
tsat
t=tsat
iC (t sat )
t
I B p
1 exp sat I Csat
p
t
Transients
Turn-on: off to saturation
iC
IC
ICECC/RL
t
ts
Transients
Turn-off: saturation to off
Storage delay time: tsd
iC
IC
IC ECC/RL
toff ts
iC tsd
ICsat
t<tsd
E
iC (t ) CC I Csat
RL
ttsd
iC (t )
tsd
iC (t sd ) I Csat
QB (t )
I B p
t
exp
p
t
ECC I B p
exp sd
RL
t
p
I B p
t sd p ln
I Csat t
Transients
Turn-off: saturation to off
Storage delay time: tsd
iC
IC
IC ECC/RL
toff ts
Determined by
EB diode
Transients
Turn-on: off to saturation
iC
IC
ICECC/RL
IC ECC/RL
t
ts
toff ts
Determined by
EB diode
Solution to dilemma
The Schottky diode clamp
C
I
B
B
metal
0.3 0.7
pn diode
Schottky diode
iB
ECC
RS
es
iE
iC
dp
DpE
dp
DpE
DpC
Wb x
IEN
ICN
Wb x
IEI
DpC
ICI
Wb x
negative
IE = IEN + IEI
IC = ICN + ICI
Ebers-Moll equations
IE = IEN + IEI
IC = ICN + ICI
Diode currents
Ebers-Moll equations
IE = IEN + IEI
IC = ICN + ICI
Collected currents
IEI = aI ICI
ICN = aN IEN
Ebers-Moll equations
IE = IES (eeVEB/kT 1) aI ICS (eeVCB/kT 1)
IC = aN IES (eeVEB/kT 1) ICS (eeVCB/kT 1)
Or:
eVEB/kT
eV
/kT
EB
IIEE == aaII IICC +
(1a
a
)
I
(e
+ IEO (eN I ES1)eV /kT 1)
/kT (e CB
IICC == aaNN IIEE -- (1aCB
1)
I) ICS1)
ICOa(eNeV
ICO
General equivalent circuit based on diode circuit
Equivalent circuit
IE = aI IC + IEO (eeVEB/kT 1)
IC = aN IE - ICO (eeVCB/kT 1)
C
IC
IE
IB
B
Valid for all biasing modes
Active
IE
IC
IB
IE
IC
IB
B
IE = -(1-aN) IES
IC = (1-aI) ICS
IC = IC0 + aN IE
IC
IE
Small!
IC0, IE=0
-VCB
Now
Amplification and maximum operation
frequency
SMALL SIGNAL equivalent circuit
Cj,BC
Cj,BE
vbe
C
npn
Cd,BE
Rp
gmvbe
Ro
Vbe
I C I 0 exp
VT
I C
IC
gm
Vbe VT
Cj,BC
Cj,BE
Cd,BE
Rp
gmvbe
Ro
I C
Vbe
Vbe b
Rp
I B
gm
Cj,BC
Cj,BE
Cd,BE
Rp
gmvbe
Ro
Cd , BE
dQb
dVbe
Cj,BE
Depletion capacitance
Cd,BE
Diffusion capacitance
See SG on pn-diode
Cd , BE
IE
B
VT
Cj,BC
Cj,BE
Cd,BE
Rp
gmvbe
Ro
Base-collector capacitance
Cj,BC
Depletion capacitance
Miller capacitance: feedback between B & C
Cj,BC
Cj,BE
Cd,BE
Rp
gmvbe
Ro
Output resistance
iC
ideal
Vce Vcb
VA
Vce VA
R0
I C
IC
IB
-vCE
Cj,BC
Cj,BE
Cd,BE
Rp
gmvbe
Ro
ic g m vbe
b
ib
ib
1 j C j , BE Cd , BE Rp
Circuit analysis
ib
Cj,BC
Cj,BE
vbe
Max gain
C
Cd,BE
Rp
gmvbe
Ro
Transit frequency fT
Small signal current gain=1
h fe 1
fT
2p C
j , BE
b
j , BE
Cd , BE
Cd , BE Rp
1
2p
C j , BEVT
IE
Transit frequency fT
Base transit time
Qb
B t
I C
B
IC I p
Wb2
B
2D
for p+n
D D p ( pnp), Dn (npn)
Note: this approach ignores delay caused by BC junction (see 3rd year)
ic
vbe
rbe
bib
rce
or
gmvbe
ie
ic
ie
Cdif
CjE
re
aie
CjC
rc
Conclusion
Delays in BJTs are a result of the storage of
minority carriers.
Main delay in common BJTs is due to the
base transit time t.