The Bipolar Junction Transistor

19 The Bipolar Junction Transistor n Physical Structure: oxide-isolated, low-voltage, high-frequency design ... typical of the bipolar transistor fou...
21 downloads 0 Views 42KB Size
19 The Bipolar Junction Transistor n

Physical Structure: oxide-isolated, low-voltage, high-frequency design ... typical of the bipolar transistor found in a BiCMOS process, such as the MicroLinear tile array chips used in the laboratory experiments

metal contact to base

, , ,

n+ polysilicon contact to n+ emitter region

p+

n+ buried

field oxide

n

layer

p-type base

,, ,,, ,

,,, ,, A

metal contact to collector

n+

n+ buried

n+ - p - n sandwich (intrinsic npn transistor)

A'

layer

p-type substrate

(a)

,,,,,,,,, ,,,,,,,,, ,,,, , , , , ,,,, ,,,, , , ,,,,,,, ,,,,,,,, , , , (base)

A

p+

p

n + emitter area, AE (intrinsic npn transistor)

(emitter)

edge of n + buried layer

field oxide

A'

n+

(collector)

(b)

EECS 105 Fall 1998 Lecture 19

Circuit Symbol and Terminal Characteristics n

As with MOSFETs, we have two devices that have complementary characterisitcs, in this case the npn transistor and the pnp transistor The direction of the diode arrow indicates whether the central layer (the base) is n or p C B IB

+ VBE

− E

E +

IC + VCE − −IE

(a)

npn normal operation: VCE positive IC positive VBE = 0.7 V IB positive -IE positive n

VEB −

B −IB

C

IE + VEC − −IC

(b)

pnp normal operation: VEC positive -IC positive VEB = 0.7 V -IB positive IE positive

The pnp usually has a very different physical structure ... we will concentrate on the npn and then consider the pnp briefly

EECS 105 Fall 1998 Lecture 19

npn BJT Collector Characteristics n

Similar test circuit as for n-channel MOSFET ... except IB is controlled instead of VBE (for convenience) IC = IC(IB, VCE)

+ V − CE IB

(a) IC (µA) 300

IB = 2.5 µA IB = 2 µA

250 200

IB = 1.5 µA

(saturation)

150

IB = 1 µA

(forward active)

100 IB = 500 nA 50 −3

−2

IB = 0 (cutoff)

−1 1

IB = 1 µA IB = 2 µA

2

3

4

5

6

VCE (V)

−4 (reverse active)

−8

(b)

EECS 105 Fall 1998 Lecture 19

Regions of Operation n

Constant-current region is called forward active ... corresponds to MOSFET saturation region (!?!) IC = βF IB = IS exp[VBE/Vth](1 + VCE / VA) ... (VA)-1 is like λ for MOSFET

n

Constant-voltage region is called saturation ... corresponds to MOSFET triode region V CE ≈ V CE ( sat ) = 0.1V or 0.2 V

n

Cutoff ... corresponds to MOSFET cutoff region

n

Reverse active ... terminal voltages for npn sandwich are flipped so that VCE is negative and VBC = 0.7 V. Only occasionally useful.

Boundary between saturation and forward-active regions: V CE > V CE ( sat )

and

IB > 0

... much easier to apply this test than VDS > VDS(sat)

EECS 105 Fall 1998 Lecture 19

Small-Signal Model of the Forward-Active npn BJT n

Transconductance (same concept as for MOSFET):

gm =

Ebers-Moll (forward-active):

∂i C ∂ v BE

iC = I S e

Q

v BE ⁄ V th

iC

iC

IC + ic IC

Q

slope = gm

IC Q VBE

0.2

0.4

0.6 V

BE

VBE + vbe

vBE

vBE

Evaluating the derivative, we find that

IC  I S  V BE ⁄ V th ------------gm =   e = V V  th th

EECS 105 Fall 1998 Lecture 19

Input Resistance n

The collector current is a function of the base current in the forward-active region (recall IC = βFIB). At the operating point Q, we define ∂i C βo = ∂ iB Q

and so ic = βo ib. (Note that the “DC beta” βF and the small-signal βo are both highly variable from device to device) n

Since the base current is therefore a function of the base-emitter voltage, we define the input resistance rπ as:

–1

rπ =

n

∂i B ∂ v BE

= Q

∂i B ∂ iC

∂i C Q

∂ v BE

Q

1 =  ------ g m β  o

Solving for the input resistance β o V th βo kTβ o r π = ------ = -------------- = -----------gm IC qI C

n

For a high input resistance (often desirable), we need a high current gain or a low DC bias current.

EECS 105 Fall 1998 Lecture 19

Output Resistance n

The Ebers-Moll model has perfect current source behavior in the forward-active region -- actual characteristics show some increase: IC

−VAn

n

VCE

Why? Base width shrinks due to encroachment by base-collector depletion region Approximate model: introduce Early voltage VAn to model increase in iC

Model:

n

iC = IS e

v BE ⁄ Vth 

v CE  1 + --------- V An 

Output resistance: ∂i C –1 ro = ∂ v CE

Q

IC --------≅ V An

EECS 105 Fall 1998 Lecture 19

Complete Small-Signal Model n

Add the depletion capacitance from the base-emitter junction to find the total base-emitter capacitance: Cπ = CjE + Cb C jE =

2C jEo

CjEo is proportional to the emitter-base junction area (AE) n

Depletion capacitance from the base-collector junction: Cµ C µo C µ = ------------------------------------1 + V CB ⁄ φ Bc Cµo is proportional to the base-collector junction area (AC)

n

Depletion capacitance from collector (n+ buried layer) to bulk: Ccs C cso C cs = -----------------------------------1 + V CS ⁄ φ Bs Ccso is proportional to the collector-substrate junction area (AS) base ib +

rb

rc

+ Cπ

vbe

collector ic







gmvπ



ro

+

Ccs substrate

vce

rex −

− emitter

EECS 105 Fall 1998 Lecture 19

Numerical Values of Small-Signal Elements

ib +

base

vbe −

ic collector +

+ vπ

gmvπ



ro



vce −

emitter n

Transconductance:

IC = 100 µA, Vth = 25 mV -->

gm = 4 mS = 4 x 10-3 S

Note: gm varies linearly with collector current and is independent of device geometry, in contrast to the MOSFET

n

Input resistance:

βo = 100, IC = 100 µA, Vth = 25 mV -->

n

rπ = 25 kΩ

Output resistance:

IC = 100 µA, VAn = 35 V -->

ro = 350 kΩ

VAn = Early voltage increases with increasing base width and decreases with decreasing base doping.

EECS 105 Fall 1998 Lecture 19

npn BJT SPICE model Close correspondence to Ebers-Moll and small-signal models Name

Parameter Description

Units

IS

transport saturation current [IS]

Amps

BF

ideal maximum forward beta [βF]

None

VAF

forward Early voltage [VAn]

Volts

BR

ideal maximum reverse beta [βR]

None

RB

zero bias base resistance [rb]

Ohms

RE

emitter resistance [rex]

Ohms

RC

collector resistance [rc]

Ohms

CJE

B-E zero-bias depletion capacitance [CjEo]

Farads

VJE

B-E built-in potential [φBe]

Volts

MJE

B-E junction exponential factor

None

CJC

B-C zero-bias depletion capacitance [Cµo]

Farads

VJC

B-C built-in potential [φBc]

Volts

MJC

B-C junction exponential factor

None

CJS

substrate zero-bias depletion capacitance [Ccso]

Farads

VJS

substrate built-in potential [φBs]

Volts

MJS

substrate junction exponential factor

None

TF

ideal forward transit time [τF]

Seconds

.MODEL MODQN NPN IS=1E-17 BF=100 VAF=25 TF=50P + CJE=8E-15 VJE=0.95 MJE=0.5 CJC=22E-15 VJC=0.79 MJC=0.5 + CJS=41E-15 VJS=0.71 MJS=0.5 RB=250 RC=200 RE=5

EECS 105 Fall 1998 Lecture 19

Suggest Documents