This document discusses BiCMOS and DCFL logic circuitry as alternatives to traditional CMOS. It provides details on the operation and advantages/disadvantages of BiCMOS inverters, NAND gates, and other circuits. DCFL using gallium arsenide is also examined and compared to CMOS. While BiCMOS and DCFL provide higher speeds, they have higher fabrication complexity and costs compared to CMOS. Examples of applications where the extra speed justifies the costs are mentioned.
2. University of Connecticut 225
BiCMOS - Best of Both Worlds?
n CMOS circuitry exhibits very low power dissipation, but
n Bipolar logic achieves higher speed and current drive capability.
n BiCMOS achieves low standby dissipation like CMOS, but high
speed and current drive capability like TTL and ECL.
n The disadvantage of BiCMOS is fabrication complexity (up to 30
masking steps, compared to about 20 for bipolar logic or
CMOS). This translates into higher cost and longer design
cycles.
n Notable examples of the BiCMOS technology are the Intel P6
(a.k.a. Pentium Pro) which appeared in 1996, and its successor
the P7.
$
3. University of Connecticut 226
BiCMOS Inverter
n P1 and N1 perform the logic
function.
n QP and QO are low-
impedance output drivers.
n N2 and N3 remove base
charge from the bipolar
transistors during switching.
QP
QO
P1
N3
N1
N2
VOUT
VIN
VDD
5. University of Connecticut 228
BiCMOS Inverter VTC
• The BiCMOS inverter shown
here exhibits reduced logic
swing (VDD - 2VBEA) compared
to CMOS (VDD).
• Reduction of the supply
voltage will make this problem
more severe.
0.0
1.0
2.0
3.0
0.0 1.0 2.0 3.0
VIN
VOUT
CMOS
BiCMOS
V V
K A V V V
V V
DD
T
F BEA
=
= =
= =
3 3
40 1
50 0 7
2
.
/
.
µ
β
6. University of Connecticut 229
BiCMOS NAND Gate
QP
QO
PA
NB3
NA1
N2
VOUT
VA
VDD
VB
PB
NB1
NA3
With both inputs high:
With VA high, VB low:
7. University of Connecticut 230
How Fast is BICMOS?
n For highly-capacitive off-chip
loads, fast switching is possible
due to the high current driving
capability of the bipolar
transistors. The speed is limited
by the parasitic capacitances of
the QP, which must be driven by
the P1 - N3 CMOS circuit.
n For on-chip loads presenting very
little capacitance, BiCMOS offers
no advantage if
CL < CBCP
n BiCMOS integrated circuits are
really CMOS on the inside!
QP
QO
P1
N3
N1
N2
VOUT
VIN
VDD
8. University of Connecticut 231
BiCMOS Applications
n Modern BiCMOS, invented by Intel, hit the market in 1992.
n Ever-increasing clock frequencies on motherboards of PC’s and
workstations may require that the VLSI / ULSI chips be made in
BiCMOS. (Witness the Intel, AMD, and Cyrix µP chips.)
n Central Processing Units (CPU’s) of “minisupercomputers” can
be implemented in BiCMOS, with packing density and
dissipation advantages over ECL. (e.g., the Cray Research
“Baby Cray” J916 Computer)
n TTL will soldier on in motherboard SSI and MSI applications,
where BiCMOS does not boast an advantage.
n But … the BiCMOS party may be over when supply
voltages drop below 1.8 V. BJT’s have a fixed turn-on
voltage; MOSFET thresholds can be reduced to at least
0.3V for room temperature operation.
9. University of Connecticut 232
The Problem with BiCMOS
n For standard BiCMOS, the logic swing is VDD - 2VBEA.
n Supply voltages are continually being reduced, because
n When VDD is reduced to 1.8V, standard BiCMOS will provide a
logic swing of only 0.4V; this isn’t acceptable! We can provide
shunt elements which increase the voltage swing of BiCMOS,
but …
n Turning off the BJT’s isn’t the answer! If the supply voltage is
1.8V, the BJT’s can only conduct for
n In this case the BJT’s can not effectively boost the switching
speed.
P C V
L DD
≈ 2
0 7 11
. .
V V V
OUT
≤ ≤
10. University of Connecticut 233
Full-Rail BiCMOS Inverter w/
Resistive Shunts
QP
QO
P1
N1
VOUT
VIN
VDD
R1
R2
• This BiCMOS design provides a
rail-to-rail voltage swing.
• For VOUT < VBEA, N1 and R2
conduct, bringing VOL all the way to
0.
• For VBEA < VOUT < VDD - VBEA, one
or both BJT’s conducts.
• For VDD - VBEA < VOUT, P1 and R1
conduct, bringing VOH all the way to
VDD.
• It is not practical to fabricate this
circuit with resistors, but a similar
circuit can be made using an active
shunt for QO.
11. University of Connecticut 234
BiCMOS Inverter w/ Active Shunt
QP
QO
P1
N1
VOUT
VIN
VDD • This BiCMOS design provides a
voltage swing of VDD - VBEA.
• For VOUT < VBEA, N3 and N2
conduct, bringing VOL all the way to
0.
• For VBEA < VOUT < VDD - VBEA, one
or both BJT’s conducts.
• The base-emitter junction of QP is
not shunted, so VOH = VDD - VBEA.
N2
N3
12. University of Connecticut 235
Full Rail BiCMOS Inverter w/
Paralleled CMOS Output
QP
QO
P1
N3
N1
N2
VOUT
VIN
VDD
NO
PO
• The parallel CMOS
inverter provides rail-to-rail
operation.
• For VOUT < VBEA, NO
conducts, bringing VOL all
the way to 0.
• For VBEA < VOUT < VDD -
VBEA, one or both BJT’s
conducts.
• For VDD - VBEA < VOUT, PO
conducts, bringing VOH all
the way to VDD.
14. University of Connecticut 237
CMOS - Single Stage
VDD = 1.8V
VIN
VT = -0.6V
2.2µ m/ 0.5µ m
VT = 0.6V
0.9µ m/ 0.5µ m
tP =
VOUT
CL
2
'
2
'
/
200
/
80
100
V
A
k
V
A
k
Angstroms
t
N
P
OX
µ
µ
=
=
=
=
A
15. University of Connecticut 238
CMOS - Single Stage / 50pF
VDD = 1.8V
VIN
VT = -0.6V
2.2µ m/ 0.5µ m
VT = 0.6V
0.9µ m/ 0.5µ m
VOUT
50pF
=
=
N
P
K
K
=
P
t
16. University of Connecticut 239
CMOS - Three Stages / 50pF
VDD = 1.8V
VIN
2.2/ 0.5
VOUT
50pF
=
P
t
0.9/ 0.5
11/ 0.5
4.5/ 0.5
55/ 0.5
22/ 0.5
=
=
=
1
1
1
P
L
t
C
K
=
=
=
2
2
2
P
L
t
C
K
=
=
=
3
3
3
P
L
t
C
K
17. University of Connecticut 240
CMOS - Six Stages / 50pF
VDD = 1.8V
VIN
2.2/ 0.5
0.9/ 0.5
11/ 0.5
4.5/ 0.5
55/ 0.5
22/ 0.5
=
=
=
1
1
1
P
L
t
C
K
=
=
=
2
2
2
P
L
t
C
K
=
=
=
3
3
3
P
L
t
C
K
WIRED
TO THE
NEXT
PAGE!
18. University of Connecticut 241
CMOS - Six Stages / 50pF
VDD = 1.8V
275/ 0.5
VOUT
50pF
=
P
t
110/ 0.5
1375/ 0.5
550/ 0.5
6875/ 0.5
2750/ 0.5
=
=
=
5
5
5
P
L
t
C
K
=
=
=
6
6
6
P
L
t
C
K
WIRED
FROM THE
PREVIOUS
PAGE!
=
=
=
4
4
4
P
L
t
C
K
20. University of Connecticut 243
DCFL Inverter
VOUT
VIN
VDD
• DCFL gates are similar to NMOS
circuits, but are implemented with GaAs
MESFET’s rather than Si MOSFET’s.
• The advantage of DCFL is speed - it
is up to 3 times faster than CMOS.
• The disadvantages of DCFL are
fabrication complexity and cost.
• GaAs 75 mm wafer - $100
• Si 200 mm wafer - $10
• Si 300 mm wafers - coming soon!
• GaAs technology is less
established compared to Si
technology, and the fabrication of
enhancement type MESFET’s is
difficult.
21. University of Connecticut 244
DCFL Inverter - Basic Operation
VOUT
VIN
VDD
NL
NO
VIN = LOW.
VIN = HIGH.
22. University of Connecticut 245
DCFL NOR Gate
VOUT
VB
VDD
NL
NOB
VA NOA
VA = VB = VOL.
VA = VDD or VB = VDD.
DCFL NAND gates are not practical due to restrictions imposed on VDD, VOL,
and the enhancement device threshold voltages.
23. University of Connecticut 246
Buffered DCFL NOR Gate
VB
VDD
VA
The added source follower provides a low-impedance output driver for off-chip
loads.
VOUT
VB
VA
24. University of Connecticut 247
DCFL Characteristics
Compare the 1999 state-of-the art for GaAs DCFL and Si CMOS:
GaAs DCFL vs. Si CMOS: 0.25 µm technology
propagation delay
dissipation
SRAM embedded in VLSI
GaAs DCFL
35 ps
30 µW (DC)
32 kB
Si CMOS
75 ps
1 µW / MHz
128 kB
• GaAs exhibits higher electron mobility than Si.
• Due to the GaAs electron velocity characteristic, DCFL can operate
at a reduced supply voltage without a penalty in switching speed.
25. University of Connecticut 248
DCFL Applications
n For a given minimum linewidth, GaAs DCFL circuitry is about 2
to 3 times faster than Si CMOS because of the difference in
electron mobilities.
n The extra speed comes at a premium, because GaAs
technology is less developed and DCFL is expensive.
n DCFL applications are at the high end, where the extra cost can
be justified. Examples are the Cray Y-MP and the Vitesse
Semiconductor GaAs microprocessor, which boasts 1.2 M
transistors [see Ira Deyhimy, “Gallium Arsenide Joins the Giants,” IEEE
Spectrum, pp. 33-40, February 1995].
n At the present time, the area of fastest growth for GaAs DCFL is
communications.
n A factor of three isn’t much, though, when you consider the rapid
advancement of Si CMOS / BiCMOS technology.