Section 6142

421
6.14 Dynamic Behavior of MOS Logic Gates
CGD2 V
DD
ML2
CGD4 V
DD
CDB4
ML4
CDB2
CSB2
CGD1
CSB4
CGD3
CDB1
vI
vO
CW
MS1
MS3
CGS3
CGS1
CDB3
(a)
VDD
VDD
ML2
vI
ML4
vO
MS1
MS3
CO
(b)
Figure 6.42 (a) Capacitances associated with an inverter pair. (b) Lumped-load capacitance model
for inverters.
6.14.2 Dynamic Response of the NMOS Inverter
with a Resistive Load

Figure 6.43 shows the circuit from our earlier discussion of the inverter with a resistive load. For
hand analysis, the logic input signal is represented by an ideal step function, and we now calculate
the rise time, fall time, and delay times for this inverter.
Calculation of tr and P L H
For analysis of the rise time, assume that the input and output voltages have reached their steadystate levels for t < 0: v I = VH = 5 V and v O = VL = 0.25 V. At t = 0, the input drops
from v I = 5 V to v I = 0.25 V. Because the gate-source voltage of the switching transistor
drops below VT N S , the MOS transistor abruptly stops conducting. The output then charges from
v O = VL = 0.25 V to v O = VH = VD D = 5 V. In this case, the waveform is that of the simple
RC network formed by the load resistor R and the load capacitor C. Using our knowledge of
single-time constant circuits:

t
t
v O (t) = VF (VF VI ) exp
= VF V exp
RC
RC
(6.49)
422
Chapter 6
Introduction to Digital Electronics
VDD = 5 V
VDD = 5 V
R
vO (0+) = 0.25 V
vO
vI
MS
Off
vI
vO
+5 V
0.25 V
+5 V
0.25 V
0V
VOL
(a)
(b)
Figure 6.43 Model for rise time in resistively loaded inverter.

where VF is the final value of the capacitor voltage, VI is the initial capacitor voltage, and
V = (VF VI ) is the change in the capacitor voltage. For the inverter in Fig. 6.43, VF = 5.0 V,
VI = 0.25 V, and V = 4.75 V.
The rise time is determined by the difference between the time t1 when v O (t1 ) = VI + 0.1 V
and the time t2 when v O (t2 ) = VI + 0.9 V . Using Eq. (6.49),

VI + 0.1 V = VF V exp

VI + 0.9 V = VF V exp
t1
RC
t2
RC

yields
t1 = RC ln 0.9
(6.50)
yields
t2 = RC ln 0.1
(6.51)
and
tr = t2 t1 = RC ln 9 = 2.2RC
(6.52)
The delay time P L H is determined by v O ( P L H ) = VI + 0.5 V , which yields

P L H = RC ln 0.5 = 0.69RC
(6.53)
Note that these expressions apply only to the simple RC network.

Equations (6.52) and (6.53) represent the classical expressions for the rise time and propagation delay for an RC network. Similar analyses show that t f = 2.2RC and t P H L = 0.69RC.
Remember that these expressions only apply to the simple RC network.
DESIGN
NOTE
The rise and fall times and propagation delays for an RC network are given by
tr = t f = 2.2RC
t P L H = t P H L = 0.69RC
423
Exercise: Find the tr and PL H for the resistively loaded inverter with C = 0.2 pF and
R = 95 k.
Answers: 41.8 ns; 13.1 ns
Exercise: Derive expressions for the fall time and high-to-low propagation delay for an
RC network.
Answers: t f = 2.2RC; tP H L = 0.69RC
Calculation of P H L and t f
Now consider the other switching situation, with v I = VL = 0.25 V and v O = VH = 5 V, as
displayed in Fig. 6.44. At t = 0, the input abruptly changes from v I = 0.25 V to v I = 5 V. At
t = 0+ , M S has vG S = 5 V and v DS = 5 V, so it conducts heavily and discharges the capacitance
until the value of v O reaches VL .
VDD = 5 V
R
iR
iC
vO (0 +) = 5 V
vO
iD
vI
MS
vI = 5 V
C
vI
vO
+5 V
MS
+5 V
0.25 V
t
0V
(a)
(b)
(c)
t
0
0
(d)
Figure 6.44 Simplified circuit for determining t f and P H L . N I (0+ ) = VH = VD D .

Figure 6.45 shows the currents i R and i D in the load resistor and switching transistor as a
function of v O during the transition between VH and VL . The current available to discharge the
capacitor C is the difference in these two currents:
iC = i D i R
Because the load element is a linear resistor, the current in the resistor increases linearly as
v O goes from VH to VL . However, when M S first turns on, a large drain current occurs, rapidly
discharging the load capacitance C. VL is reached when the current through the capacitor becomes
zero and i R = i D . Note that the drain current is much greater than the current in the resistor for
most of the period of time corresponding to P H L . This leads to values of P H L and t f that are
much shorter than P L H and tr associated with the rising output waveform. This behavior is
characteristic of single channel (NMOS or PMOS) logic circuits. Another way to visualize this
difference is to remember that the on-resistance of the MOS transistor must be much smaller
than R in order to force VO L to be a low value. Thus, the apparent time constant for the falling
waveform will be much smaller than that of the rising waveform.
Chapter 6
500 A
VGS = 5 V
400 A
Current
424
iD
300 A
200 A
iC = iD iR
100 A
iR
0A
0V
1.0 V
2.0 V
3.0 V
vO
4.0 V
5.0 V
6.0 V
Figure 6.45 Drain current and resistor current versus v O .

Calculation of t f and P H L is more complicated here than for the case of the resistor charging
the capacitive load because the NMOS transistor changes regions of operation during the output
voltage transition. Thus, the differential equation that models the VH to VL transition changes at
the point at which the transistor changes operating regions.
First, let us simplify our model for the circuit. From Fig. 6.45, we can see that i D i R
except for v O very near VL . Therefore, the current through the resistor will be neglected so that
we can assume that all the drain current of the NMOS transistor is available to discharge the load
capacitance, as in Fig. 6.44(b). The input signal v I is assumed to be a step function changing to
v I = 5 V at t = 0. At t = 0, the output voltage VC on the capacitor is VH = VD D = 5 V, and the
gate voltage is forced to VG = 5 V.
The graph in Fig. 6.46 shows the important instants in time that need to be considered. At
time t1 the output has dropped by 10% of the logic swing V , and time t4 is the time at which
vO
VH
Saturation regiontriode region

transition
0.1 V
V90%
V50%
V = VH VL
V10%
0.1 V
VL
0
t1
t2
t3
t4
Figure 6.46 Times needed for calculation of P H L and t f for the inverter. Fall time t f = t4 t1 ;
propagation delay P H L = t3 .
425
the output has dropped by 90% of V . Thus, t f = t4 t1 . At t3 , the output is at the 50%
point, given by V50% = (VH + VL )/2, so P H L = t3 . Time t2 is also very important. At this
point v O = VD D VT N S , and this is the time at which the transistor changes from saturation
region operation to triode region operation. Thus, the differential equation that models the circuit
behavior changes at this point. This regional approach is used so often that it is referred to as
piecewise analysis.
Piecewise Analysis of P H L
Let us first focus on calculation of P H L and then calculate t f . At t = 0+ , the NMOS transistor
in Fig. 6.44(b) is operating in the saturation region, and the capacitor current is described by
KS
dvC
(vG S VT N S )2 = C
2
dt
with vC (0+ ) = VH
(6.54)
in which vG S = VH and VT N S are both constant. Thus, the drain current is constant, and the
capacitor discharges at a constant rate until the MOSFET enters the linear region of operation at
time t2 , when vC = vG S VT N S . The MOSFET enters the linear region after the capacitor voltage
drops by one threshold voltage. For these values, the time t2 required for the transistor to reach
the linear region is
t2 =
2C VT N S
VT N S
= 2RonS C
2
K S (VH VT N S )
(VH VT N S )
(6.55)
for
RonS =
1
K S (VH VT N S )
which represents the equivalent on-resistance of the NMOS switching transistor, with vG S = VD D
and v DS = 0.
Once the transistor enters the triode region, the equation characterizing the discharge changes
to

vC
dvC
K S vG S VT N S
vC = C
(6.56)
2
dt
because the v DS = vC for the MOSFET. Rearranging this equation with vG S = VH and integrating
yields
V3
V2
dvC
=
(2(VH VT N S ) vC )vC
t3
t2
KS
dt
2C
(6.57)
in which the limits of integration are

V2 = vC (t2 ) = VH VT N S
and
V3 = vC (t3 ) = 0.5(VH + VL )
The solution to this equation may be found using:

dx
1
= ln
(a x)x
a
x a
x

(6.58)
426
Chapter 6
Using Eq. (6.58), t3 t2 can be found to be

V3 2(VH VT N S )
V2
C
t3 t2 =
ln
K S (VH VT N S )
V3
V2 2(VH VT N S )

VH VT N S
t3 t2 = RonS C ln 4
1
V H + VL
(6.59)
(6.60)
The propagation time P H L is just equal to t3 and is given by

P H L

2VT N S
VH VT N S
1 +
= t3 = (t3 t2 ) + t2 = RonS C ln 4
V H + VL
VH VT N S
(6.61)
This is an extremely useful equation. Not only does it describe the behavior of the NMOS circuit,
but we will also see later that Eq. (6.61) characterizes the delay behavior of CMOS logic gates,
which form todays most widely used logic family.
Piecewise Analysis of t f
Fall time t f can be written as:
t f = t4 t1 = (t4 t2 ) (t2 t1 )
During the time interval t1 t2 , the MOSFET is saturated, and the current discharging the capacitor
is constant. Therefore, using
t = C
t2 t1 = C
V
I
(VH 0.1 V ) (VH VT N S )
VT N S 0.1V
= 2RonS C
Ks
VH VT N S
(VH VT N S )2
2
(6.62)
During the interval t2 to t4 , the MOSFET is operating in the triode region, and the circuit is
described by

V4
V2
dvC
=
(2(VH VT N S ) vC )vC

t2
t4
KS
dt
2C
(6.63)
based on Eqs. (6.56) and (6.57) in which the limits of integration are now defined by
V2 = vC (t2 ) = VH VT N S
and
V4 = vC (t3 ) = VH 0.9 V
(6.64)
Using the results from (6.58) and (6.59),

t4 t2 = RonS C ln
VH 2VT N S + 0.9 V
VH 0.9 V

(6.65)
and our estimate for the fall time is given by

VH 2VT N S + 0.9 V
VT N S 0.1 V
t f = t4 t1 = RonS C ln
+2
VH 0.9 V
VH VT N S
(6.66)

Section 6142

Uploaded by

Copyright:

Available Formats

Section 6142

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Section 6142

Uploaded by

Copyright:

Available Formats

421

6.14 Dynamic Behavior of MOS Logic Gates

6.14.2 Dynamic Response of the NMOS Inverter

with a Resistive Load

Introduction to Digital Electronics

Figure 6.43 Model for rise time in resistively loaded inverter.

The delay time P L H is determined by v O ( P L H ) = VI + 0.5 V , which yields

Note that these expressions apply only to the simple RC network.

6.14 Dynamic Behavior of MOS Logic Gates

Answers: 41.8 ns; 13.1 ns

Answers: t f = 2.2RC; tP H L = 0.69RC

Figure 6.44 Simplified circuit for determining t f and P H L . N I (0+ ) = VH = VD D .

Introduction to Digital Electronics

Figure 6.45 Drain current and resistor current versus v O .

Saturation regiontriode region

6.14 Dynamic Behavior of MOS Logic Gates

in which the limits of integration are

The solution to this equation may be found using:

Introduction to Digital Electronics

Using Eq. (6.58), t3 t2 can be found to be

The propagation time P H L is just equal to t3 and is given by

Using the results from (6.58) and (6.59),

and our estimate for the fall time is given by

You might also like