U.S. patent application number 12/359933 was filed with the patent office on 2010-07-29 for level shifter flip-flop.
This patent application is currently assigned to Sun Microsystems, Inc.. Invention is credited to Jason M. Hart, Robert P. Masleid.
Application Number | 20100188119 12/359933 |
Document ID | / |
Family ID | 42353678 |
Filed Date | 2010-07-29 |
United States Patent
Application |
20100188119 |
Kind Code |
A1 |
Masleid; Robert P. ; et
al. |
July 29, 2010 |
LEVEL SHIFTER FLIP-FLOP
Abstract
A flip-flop or other state circuit that includes level-shifting
functionality. In connection with a flip-flop, embodiments include
an inverter circuit element that has a data input line as its input
and a data complement line as its output. The inverter resides in
voltage domain that is lower than the voltage domain associated
with remainder of the flip-flop.
Inventors: |
Masleid; Robert P.; (Mounte
Sereno, CA) ; Hart; Jason M.; (Hayden, ID) |
Correspondence
Address: |
Polsinelli Shughart PC;on behalf of Sun Microsystems, Inc.
700 West 47th Street, Suite 1000
Kansas City
MO
64112
US
|
Assignee: |
Sun Microsystems, Inc.
Santa Clara
CA
|
Family ID: |
42353678 |
Appl. No.: |
12/359933 |
Filed: |
January 26, 2009 |
Current U.S.
Class: |
326/63 |
Current CPC
Class: |
H03K 3/35625 20130101;
H03K 3/356139 20130101 |
Class at
Publication: |
326/63 |
International
Class: |
H03K 19/0175 20060101
H03K019/0175 |
Claims
1. A flip-flop, comprising: a data input line carrying a logical
value represented by a first voltage range, the first voltage range
between a reference voltage and a first power supply rail; a data
complement line connected to the data input line through an
inverter, wherein the inverter is connected to the first power
supply rail; first and second transistors, the first transistor
connected to the data input line and the second transistor
connected to the data complement line; first and second pre-charge
nodes, the first pre-charge node connected to the first transistor
and the second pre-charge node connected to the second transistor;
a first keeper transistor connected to the first transistor,
wherein the ratio of the width of the first transistor to the width
of the keeper transistor is between 8 to 1 and 4 to 1; a clock line
operable to trigger a capture of the logical value carried by the
data input to form a captured logical value; and a data output line
carrying the captured logical value represented by a second voltage
range, the second voltage range between the reference voltage and a
second power supply rail, wherein the capture of the logical value
includes a discharge of one of the first and second pre-charge
nodes by a path to ground through one of the first and second
transistors by one of the data input and data complement lines;
wherein the second power supply rail is greater than the first
power supply rail.
2. (canceled)
3. (canceled)
4. The flip-flop of claim 1, wherein during a pre-charge phase, the
first and second pre-charge nodes are charged to a voltage of
approximately the second power supply rail.
5. The flip-flop of claim 1, wherein the first power supply rail is
greater than a threshold voltage of the first and second
transistors.
6. The flip-flop of claim 1, wherein the ratio of the width of the
first transistor to the width of the first keeper transistor is 5
to 1.
7. The flip-flop of claim 1, wherein the width of the first
transistor is 0.6 microns.
8. The flip-flop of claim 1, wherein the data input line is
associated with a first integrated circuit region that is powered
by the first power supply rail, and the data output line is
associated with a second integrated circuit region that is powered
by the second power supply region.
9. The flip-flop of claim 8, wherein the flip-flop is located at a
boundary between the first integrated circuit region and the second
integrated circuit region.
10. A method of shifting voltage levels between voltage domains of
a semiconductor device, comprising: receiving an input voltage at
an input to a circuit stage from a first voltage domain defined by
a first power supply rail, the first voltage being representative
of a logical value, the input having a data input line connected to
a first transistor and a data complement line connected to the data
input line through an inverter and connected to a second
transistor, the inverter being connected to the first power supply
rail, wherein the first transistor is connected to a first
pre-charge node and the second transistor is connected to a second
pre-charge node; latching the logical value at the circuit stage
including discharging of one of the first and second pre-charge
nodes by a path to ground through one of the first and second
transistors based on the input voltage; driving at least one of the
pre-charged nodes with at least one of the first and second
transistors, the first and second transistors sized to compensate
for a drive strength associated with the input voltage from the
first voltage domain; and presenting an output voltage at an output
of the circuit stage in a second voltage domain, the second voltage
domain higher than the first voltage domain, the output voltage
being representative of the latched logical value.
11. The level shifting method of claim 10, wherein latching the
logical value includes pre-charging the first and second pre-charge
nodes.
12. (canceled)
13. (canceled)
14. (canceled)
15. A flip-flop, comprising: means for carrying an input logical
value represented by a first voltage range, the first voltage range
between a reference voltage and a first power supply rail, the
means for carrying including a data line and a data complement
line; means for triggering a capture of the input logical value
carried by the means for carrying including discharging one of a
first and second pre-charge nodes based on the data line and the
data complement line, wherein the capture results in a captured
logical value; means for driving at least one of the first and
second pre-charge nodes with one of the data line and the data
complement line, the means for driving compensating for a drive
strength of the signal received in the first voltage range; and
means for carrying the captured logical value represented by a
second voltage range, the second voltage range between the
reference voltage and a second power supply rail, wherein the
second power supply rail is greater than the first power supply
rail.
16. The flip-flop of claim 15, further comprising: means for
inverting the data line to form the data complement line, wherein
the means for inverting is connected to the first power supply
rail.
17. The flip-flop of claim 16, further comprising: first and second
means for holding a charge, the first means for holding a charge
connected to the first pre-charge node and the second means for
holding a charge connected to the second pre-charge node.
18. The flip-flop of claim 17, wherein during a pre-charge phase,
the first and second pre-charge nodes are charged to a voltage of
approximately the second power supply rail.
19. The flip-flop of claim 17, wherein the first power supply rail
is greater than a threshold voltage of the first and second means
for driving.
20. The flip-flop of claim 17, wherein the means for carrying an
input logical value is associated with an integrated circuit region
that is powered by the first power supply rail, and the means for
carrying an output logical value is associated with an integrated
circuit region that is powered by the second power supply
region.
21. The flip-flop of claim 1, further comprising: a second keeper
transistor connected to the second transistor, wherein the ratio of
the width of the second transistor to the width of the second
keeper transistor is between 8 to 1 and 4 to 1.
22. The flip-flop of claim 21, wherein the ratio of the width of
the second transistor to the width of the second keeper transistor
is 5 to 1.
23. The level shifting method of claim 10, wherein the ratio of the
width of the first transistor to the width of a keeper transistor
connected to the first pre-charge node is between 8 to 1 and 4 to
1.
24. The level shifting method of claim 23, wherein the ratio of the
width of the first transistor to the width of the keeper transistor
is 5 to 1.
Description
TECHNICAL FIELD
[0001] Aspects of the present invention are generally directed to
circuits, and particularly to circuits that shift voltage levels
between voltage domain of an integrated circuit.
BACKGROUND
[0002] In the field of a Very Large Scale Integration (VLSI)
design, a method of reducing power or improving power efficiency is
to modify the voltage at which circuits operate. Due to power
constraints or other considerations, an integrated circuit may
include more than one voltage region. In accordance with certain
designs, an integrated circuit's digital circuits may operate in
one voltage domain and the integrated circuit's SRAM cells may
operate in a different voltage domain. This division of a
integrated circuit may occur due to the difference in operating
conditions in a digital circuit versus an SRAM cell. In particular,
digital circuits typically use a relatively high amount of power
and have a high leakage rate. Under these conditions, it may be
preferable to operate at a relatively low voltage. In contrast,
SRAM cells typically are resistant to leakage and do not use a
relatively high amount of power. Moreover, SRAM cells tend to fail
at low voltages. Accordingly, under these conditions, it is may be
preferable to operate at a higher voltage. Operating an integrated
circuit at two or more voltages may require shifting between the
two levels. In particular, it may be necessary for a signal to
travel between voltage domains. However, a signal that is output
from a circuit stage in one voltage domain may not be capable of
driving a circuit stage in another voltage domain. Accordingly,
transitioning between voltage domains sometimes requires passing
the signal through a level shifter.
[0003] Typically, transitioning from a high voltage region to a low
voltage region may be done without a level shifter. Specifically, a
signal defined in a higher voltage region may be used to drive a
gate or other circuit element that resides in a lower voltage
region. In contrast, transitioning from a lower voltage region to a
higher voltage region may be problematic. While it may be possible
to assert a logical low value at an input, it may not be possible
to assert a logical high value at a gate input. In particular,
because the two voltage regions have a common ground, asserting a
logic low, which corresponds or substantially corresponds to
ground, results in a valid logic low. In contrast, asserting a
logical high value from a lower voltage domain may not result in a
voltage level that is within the voltage range needed in the higher
voltage region.
[0004] The fundamental nature of logic gates contributes to the
difficulty in transitioning from a lower voltage domain to a higher
voltage domain. A typical logic gate, such as those implemented
with Complementary Metal Oxide Semi-conductor (CMOS) processes,
include a pull-up transistor that provides a path to the power
supply rail and a pull-down transistor that provides a path to
ground. A logic input value that results in a logical high value at
the output is one in which the pull-up transistor is on and the
pull-down transistor is off. Similarly, a logic value that results
in a logical low at the output is one in which the pull-up
transistor is off and the pull-down transistor is on. In order for
the pull-up and pull-down transistors to be fully turned on and
off, the input voltages should range between ground and the power
supply rail. Intermediate voltages between these values often
result in indeterminate values or erroneous voltages at the output.
Accordingly, asserting a high logic value from a lower voltage
domain to a logic gate in a higher voltage domain may result in
errors in the logic gate output. In particular, the pull-up
transistor may not be fully on, the pull-down transistor may not be
fully off, or both. Accordingly, in transitioning from a low
voltage domain to a high voltage domain, it may be necessary to add
a circuit stage that is devoted to facilitating the transition.
SUMMARY
[0005] Aspects of the invention are directed to a flip-flop or
other state circuit that includes level-shifting functionality. In
connection with a flip-flop, implementations conforming with
various aspects of the present invention include an inverter
circuit element that has a data input line as its input and a data
complement line as its output. The inverter resides in a voltage
domain that is lower than the voltage domain associated with
remainder of the flip-flop. The flip-flop may additionally include
pre-charge nodes that are selectively discharged based on the
logical value present on the data and data complement lines.
[0006] An embodiment is directed to a flip-flop, including a data
input line carrying a logical value represented by a first voltage
range, the first voltage range between a reference voltage and a
first power supply rail; a clock line operable to trigger a capture
of the logical value carried by the data input to form a captured
logical value; and a data output line carrying the captured logical
value represented by a second voltage range, the second voltage
range between the reference voltage and a second power supply rail;
wherein the second power supply rail is greater than the first
power supply rail.
[0007] Another embodiment is directed to a level-shifting method,
comprising: receiving an input voltage from a first voltage domain,
the first voltage being representative of a logical value; latching
the logical value; and presenting an output voltage in a second
voltage domain, the second voltage domain higher than the first
voltage domain, the output voltage being representative of the
latched logical value.
[0008] Another embodiment is directed to a flip-flop, including
means for carrying an input logical value represented by a first
voltage range, the first voltage range between a reference voltage
and a first power supply rail; means for triggering a capture of
the input logical value carried by the means for carrying, wherein
the capture results in a captured logical value; and means for
carrying the captured logical value represented by a second voltage
range, the second voltage range between the reference voltage and a
second power supply rail; wherein the second power supply rail is
greater than the first power supply rail.
BRIEF DESCRIPTION OF THE DRAWINGS
[0009] FIG. 1 is a schematic illustration of a dynamic circuit
having a generic logic section;
[0010] FIG. 2A through 2C are schematic illustrations of example
logic sections for the dynamic circuit shown in FIG. 1;
[0011] FIG. 3 is a schematic illustration of a level shifter
flip-flop in accordance with implementations discussed herein;
[0012] FIG. 4 is a schematic illustration of an inverter portion of
the level shifter flip-flop shown in FIG. 3;
[0013] FIG. 5A through 5C are schematic illustrations of flip-flop
level shifter orientations;
[0014] FIG. 6 is a flow chart illustrating operations in accordance
with the method in accordance with the implementation discussed
herein; and
[0015] FIG. 7 is a flow chart showing further operations in
accordance with the method in accordance with the implementations
discussed herein.
DETAILED DESCRIPTION
[0016] Implementations discussed herein are directed to a circuit
element, such as a flip-flop, that includes a level shifting
functionality. Embodiments are directed to circuit designs that, in
addition to the circuit's logic functionality, allow the circuit to
function as a level shifter. More particularly, implementations
discussed herein avoid adding a dedicated circuit stage for the
purpose of level shifting. Avoiding adding a dedicated level
shifter may have a number of advantages for the circuit designer.
For instance, power consumption may be lowered, additional circuit
delay may avoided and/or circuit area may be preserved.
[0017] Implementations discussed herein are directed to shifting
between voltage domains within an integrated circuit. More
specifically, implementations are directed to passing a signal or
otherwise transitioning from a lower voltage domain to a higher
voltage domain. As used herein, a voltage domain refers to a
portion of an integrated circuit that operates at a certain
voltage. More particularly, power is provided to circuit elements
within the voltage domain by a power supply rail that defines a
maximum voltage. As used herein, two voltage domains V.sub.DD and
V.sub.CS are discussed. For the purposes of this disclosure,
V.sub.DD is defined as being a lower voltage domain than V.sub.CS.
More particularly, the power supply rail associated with the
V.sub.DD voltage domain is lower than the power supply rail defined
or associated with the V.sub.CS voltage domain.
[0018] Implementations discussed herein are directed to a circuit
stage or element that accomplishes a logical function as well as
shifts between voltage domains. In accordance with implementations
discussed herein, characteristics of dynamic circuits are exploited
to accomplish level shifting. Generally, the operation of a dynamic
circuit includes a pre-charge phase and an evaluate phase. An
example dynamic circuit is shown in FIG. 1 and is generally
referred to with reference numeral 100. The circuit 100 includes a
pre-charge transistor 104 and an evaluate transistor 108. The
pre-charge transistor 104 and the evaluate transistor 108 are each
supplied with a clock 112 connected to their respective gates. When
the clock is low, the pre-charge transistor 104 is on and the
evaluate transistor 108 is off. In this pre-charge phase, an
internal node 116 is charged by the pre-charge transistor 104 to a
high voltage.
[0019] During the evaluate phase, the internal node 116 is
selectively discharged based on whether or not a logic condition
associated with logic section 124 is or is not satisfied. The logic
section 124 is illustrated generically in FIG. 1, but may include
transistors arrangements that define logic conditions such as, for
example, AND, NAND, NOR, and OR. The logic section 124 may conduct
or otherwise provide a path to ground during the evaluate phase
when the associated logic condition is satisfied or otherwise
evaluates as TRUE. Example logic sections 124 are shown in FIGS.
2A-2C.
[0020] As shown in FIG. 2A, logic 124 may include a 2-OR gate. In
this instance, the input to the logic section 124 includes two
input lines 122a, b. Each input line 122a, b may be connected to a
transistor 204a, b. When either or both input lines 122a, b are
high, one or more of the transistors 204a, b will be on. Here, the
logic section's OR logic condition is satisfied and the logic
section 124 evaluates as TRUE. More particularly, a conduction path
exists through the logic section 124, between the node 116 and the
evaluate transistor 108.
[0021] The path to ground may be completed when the clock 112 is
high during the evaluate phase. When the clock 112 is high, the
evaluate transistor 108 is on and the pre-charge transistor 104 is
off. When the evaluate transistor 108 is on and the logic condition
of the logic section 124 is satisfied, a path to ground exists at
the internal node 116. Accordingly, the charge that was established
at the internal node 116 during the pre-charge phase may be
discharged. Alternatively, if the logic condition of the logic
section 124 is not satisfied, no path to ground exists at the
internal node 116. In this way, the internal node 116 is initially
set to a high or logical "one" value in the pre-charge phase and
conditionally discharged to a low or logical "zero" value depending
on the inputs to the logic 124 during the evaluate phase.
[0022] In addition to the 2-OR logic function implemented by the
circuit shown in FIG. 2A, other logic conditions may implemented at
logic section 124. For example, as shown in FIG. 2B, a logic
section 124 may include a single gate 208. In this instance, the
input to the logic 124 includes a single line 122c. With a high
value at the logic input 122c, the transistor 208 will be on. With
transistor 208 on, a path to ground will exist at the internal node
116 during the evaluate phase. As shown in FIG. 2C, a logic section
124 may implement an AND function. Here, the input to logic section
124 includes two input lines 122d,e. Each input line 122d,e may be
connected to a transistor 212a,b. With both input lines 122d,e
high, both transistors 212a,b will be on. In this way, an AND
function is implemented by the logic section 124.
[0023] Referring again to FIG. 1, even when the logical condition
at logic 124 is not satisfied, a small leakage current may be
present. Accordingly the logic value established by the pre-charge
at internal node 116 during the pre-charge phase may be degraded by
this leakage current. Accordingly, the circuit 100 may include a
keeper transistor 136 adapted to counteract this leakage current.
Moreover, the value at the internal node 116 may additionally be
inverted by an inverter 128 and provided as the circuit output at
output line 132.
[0024] The dynamic circuit shown in FIG. 1 also provides level
shifting to transition from a lower voltage domain to a higher
voltage domain. In particular, the input or inputs 122 at logic 124
may be provided from a lower voltage region VDD and result in a
valid output in a higher voltage region V.sub.CS. A logical value
is presented at output 132 based on the selective discharge of
internal node 116. In presenting the logical value at output 132,
node 116 is discharged or not discharged depending on whether or
not the logic condition of logic section 124 is satisfied. Logic
section 124 may be provided with inputs 122 from lower voltage
region and operate properly to selectively discharge internal node
116. In this way, the circuit 100 may receive inputs may from a
lower voltage region and produce outputs in a higher voltage
region.
[0025] To further illustrate the level shifting functionality of
the circuit 100 shown in FIG. 1, consider the case where logic
section 124 is implemented as the 2-OR logic function shown in FIG.
2A. With a logical low value presented at both inputs 122, neither
transistor 204 is on. Accordingly, the 2-OR logic function is not
satisfied and the charge at internal node 116, which was
established during the pre-charge phase, is not discharged. As this
charge maintains the output at a valid logic level in the V.sub.CS
domain, the circuit operates properly. With a logical high value
presented at either or both of the inputs 122, one or more of the
transistors 204 is on. Accordingly, the 2-OR logic function is
satisfied and the charge at internal node 116, which was
established during the pre-charge phase, is discharged. In
discharging the internal node 116, the voltage level at input to
logic 124 need not rise to the value that exists at the power
supply rail 120. Because the value at the output 132 is maintained
by a charge that is pre-established at internal node 116 and is not
actively driven to a high state, it is possible to pull the node
116 low through logic 124. Here, all that is needed is for the
voltage level at input 122 to exceed the threshold value of the
logic transistors associated with logic 124.
[0026] The circuit 100 shown in FIG. 1 may be used to accomplish
two tasks. Firstly, the circuit 100 may perform the logic function
specified by logic 124. Additionally, the circuit 100 may be used
to transition or shift from a lower voltage domain to a higher
voltage domain. The circuit 100 shown in FIG. 1 is convenient in
the instance the circuit design or layout specifies that a logic
gate is to be located at the boundary between voltage regions or
domains. However, in other instances, a logic gate may not be
needed at the boundary between a lower voltage domain and a higher
voltage domain. In particular, it may be the case that a state
element such as a flip-flop may be needed at this point.
[0027] Typically, if a flip-flop or other state element is needed
at the boundary between voltage domains, an additional circuit
stage must be added. This is due to the fact that an input to a
flip-flop must be fully driven between ground and the power supply
rail of the flip-flop in order to avoid internal indeterminate
states. More particularly, the input to a conventional flip-flop is
used to switch both a pull-up and a pull-down transistor.
Accordingly, a logical high value that is less than the power
supply rail may result in an indeterminate value due to the fact
that this input value is insufficient to completely turn on and off
the pull-up and pull-down transistors. Accordingly, the voltage
levels may need to be explicitly shifted from the lower voltage
domain to the higher voltage domain prior to being received at the
flip-flop that resides in the higher level voltage domain.
[0028] An explicit level shifting circuit may be used in the
instance that a flip-flop is needed at a transition from a lower
voltage region to a higher voltage region. For example, the circuit
100 may be used with the logic section 124 shown in FIG. 2B. Here,
the logical value present at the input 122 will be passed to the
output 132. In passing the logic value, the voltage domain will
shift from the lower level voltage domain V.sub.DD to the higher
level voltage domain V.sub.CS. The logic value present at the
output 132 may then be presented to the flip-flop that resides in
the higher level voltage domain V.sub.CS. Here, the circuit will
operate as needed, however, an additional circuit stage will be
added in order to accomplish the level shifting. Implementations
discussed herein are directed to avoiding adding this additional
circuit stage. More particularly, implementations discussed herein
are directed to a flip-flop that includes a level shifting
function.
[0029] Referring now to FIG. 3, a level shifter flip-flop is
generally identified with the reference number 300. The flip-flop
300 generally resides in higher level voltage domain V.sub.CS and
includes portions that are connected to the power supply rail
associated with a lower voltage domain V.sub.DD. By connecting
portions of the flip-flop 300 to the lower supply rail and by
exploiting characteristics of dynamic circuits, the flip-flop 300
accomplishes both its flip-flop function and a level shifting
function.
[0030] The flip-flop 300 generally includes a master latch 304 and
a slave latch 308. The master latch 304 and the slave latch 308
operate together to capture data present on the input 312 during
the rising edge of the clock 316. More particularly, the logical
value at the input 312 is transparent through the master latch 304
during a particular clock phase. At a clock edge, the logical value
present at the input 312 is captured by the slave latch 308 and
maintained at output 320 during the remainder of the clock 116
cycle.
[0031] In capturing data present at the input 312, the master latch
304 includes pre-charge transistors 324 and evaluate transistor
328. While the clock 316 is low, the pre-charge transistors 324
pre-charge internal nodes 332a and 332b to a high voltage.
Depending on the logical value present at input 312, one of the
internal nodes 332a or 332b will be pulled low during the evaluate
phase. In this regard, the path to ground for internal nodes 332a
and 332b includes differential pair transistors 336a and 336b.
Transistor 336a is supplied directly with the data line 312.
Transistor 336b is supplied with the inverse of the data input 312.
This inverted line is referred to herein as data complement 340. As
transistors 336a and 336b are provided with opposite values, one or
the other will be on. When the clock line is high, the evaluate
transistor 328 is on. During this phase, depending on which of the
differential pair transistors 336a and 336b is on, one of the
internal nodes of 332a and 332b will discharge. Following the
discharge of one of the internal nodes 332a or 332b, the master
latch 304 will drive the slave latch 308 through the operation of
pull-up transistor 344 and pull-down transistor 348. Specifically,
depending on the input at 312, the node 352 will be driven high or
low. In driving the node 352 low, the pull-up transistor 344 will
be off and the pull-down transistor 348 will be on. In driving the
node 352 high, the pull-up transistor 344 will be on and the
pull-down transistor 348 will be off. In this way, the logical
value present at data input 312 will be present or transparent to
the node 352 and thereby available to be captured by the slave
latch 308.
[0032] While the flip-flop 300 may perform this state function, the
flip-flop 300 may additionally shift between levels or voltage
domains. In particular, data 312 and data complement 340 may be
supplied from a lower level voltage domain V.sub.DD and the circuit
300 will operate with the expected logical or state function. In
this regard, the data complement value 340 is provided by inverting
the data value 312 in the lower level voltage domain V.sub.DD. This
is accomplished with an inverter 356 that has data 312 as an input
and produces data complement 340 as an output.
[0033] Turning now to FIG. 4, an illustration of the inverter 356
is shown. The inverter 356 includes pull-up transistor 404 and
pull-down transistor 408. Notably, here the pull-up transistor 404
is connected to the lower level power supply rail V.sub.DD. With a
low logic level at input 312, the pull-up transistor 404 is on and
the pull-down transistor 408 is off. Here, the output 304 is pulled
up to the power supply rail V.sub.DD with a logical low value at
the input 312. With a high logic value at the input 312, the
pull-up transistor 404 is off and the pull-down transistor 408 is
on. Here, the output 304 is pulled down to ground. As shown in FIG.
3, data complement 304 is provided to differential pair transistor
336B. As the inverter 356 is disposed in lower level voltage domain
V.sub.DD, a logical values are presented to the differential pair
transistors 336A and 336B with voltage levels associated with the
lower level voltage domain V.sub.DD.
[0034] The inverter 356 is a component of a flip-flop. In one
arrangement, the inverter 356, specifically the pull-up transistor
404, is connected to the power supply rail of the higher voltage
domain V.sub.CS. In this arrangement, an undesirable situation
would arise if the input 312 were to be driven by logic values
associated with the lower level voltage region V.sub.DD. In
particular, the high voltage level, which when presented at data
input 312, would be insufficient to adequately establish a stable
output at data complement 340. In particular, the pull-up
transistor 404 would not be completely off and the pull-down
transistor 404 would not be completely on. In this conventional
arrangement, an indeterminate value may exist at data complement
340 and thus at the gate of differential pair transistor 336B.
Accordingly, improper operation of the flip-flop 300 may
result.
[0035] As shown in FIG. 4, implementations discussed herein include
positioning inverter 356 in the lower voltage domain V.sub.CS. In
this way, the inverter 356 will operate properly when driven by the
voltage levels associated with the lower level voltage region
V.sub.DD. Accordingly, data 312 and data complement 340 are
presented to the differential transistor pair 336a and 336b with
voltage levels associated with the lower level voltage region
V.sub.DD. These lower level voltage levels may then be used to
selectively discharge the internal nodes 332a and 332b through the
operation of the differential pair transistors 336a and 336b. In
discharging a particular node 332a or 332b, an input or gate
voltage at the differential pair transistors 336a and 336b only
needs to be in excess of the transistor's threshold voltage and
does not need to reach the power supply rail V.sub.CS. This is due
to the fact that the internal nodes 332a and 332b are not actively
driven high, but instead are pre-charged to a high value during a
portion of the clock 316 cycle. Accordingly, the differential pair
transistors 336a and 336b may be gated by logic values within the
lower level voltage domain V.sub.DD.
[0036] In accordance with implementations discussed herein, the
differential pair transistors 336a and 336b may be sized
appropriately to compensate for the lower drive strength associated
with gate values provided from a lower level voltage region
V.sub.DD. In one particular example, the width of the differential
pair transistors may be between approximately 0.8 microns and 0.4
microns. In another example, the width of the differential pair
transistors is approximately 0.6 microns. The ratio of the width of
a differential pair transistor 336a to the width of a keeper
transistor may be between 8 to 1 and 4 to 1. In yet another
example, the ratio of the width of the differential pair transistor
to the width of the keeper transistor is 5 to 1.
[0037] In one arrangement, a flip-flop resides in a single voltage
domain and is supplied with two power lines. In this arrangement, a
flip-flop is provided with a ground power line and with a power
supply rail line. A flip-flop in accordance with implementations
discussed herein is provided with three power supply lines,
specifically, a ground power supply line, a higher power supply
rail (V.sub.CS) and a lower power supply rail (V.sub.DD).
[0038] Turning now to FIG. 5A, a schematic illustration of an
integrated circuit 500 is shown. In accordance with implementations
discussed herein, the flip-flop 300 may be considered or
implemented as a single integrated circuit cell 504. In considering
flip-flop 300 as a single cell, it should be appreciated that the
cell 504 is provided with two power supply lines and a ground line
as shown in FIG. 5A. FIG. 5B shows an alternative implementation
wherein the flip-flop 300 is considered as two separate cells, each
supplied with a separate power supply rail. In particular, a first
cell 508 is provided with the lower level power supply rail
V.sub.DD and a second cell 512 is supplied with the higher level
power supply rail V.sub.CS. In the implementation shown in FIG. 5B,
the cell 508 includes the inverter 356, while the cell 512 includes
the remainder of the flip-flop 300.
[0039] Regardless of the manner in which the flip-flop 300 is
viewed, it should be appreciated that the flip-flop accomplishes
its conventional logical function as well as a level shifting
function. Moreover, in accomplishing these two functions, no
additional circuit elements are added in addition to those
typically associated with a conventional flip-flop. In this way, no
additional circuit stages or elements are added which would
otherwise lead to unwanted circuit delays or chip area usage. With
reference to FIG. 5B, it should be appreciated that the first cell
508 does not include any components that are added in addition to
the normal components of a conventional flip-flop. In particular,
the associated inverter 356 is part of the normal delay path of a
conventional flip-flop.
[0040] In accordance with implementations discussed herein, the
flip-flop 300 is associated with a cell that is disposed on a
boundary between voltage regions. As shown in FIG. 5C, the circuit
cell 516 is disposed between lower level voltage region V.sub.DD
(520), and higher level voltage region V.sub.CS (524). In arranging
the circuit cell 516 in this position, an undesirable configuration
is avoided in which power supply rails typically associated with
the V.sub.DD region are extended into the V.sub.CS region.
Likewise, power supply rails associated with the V.sub.CS region
are not extended into the V.sub.DD region.
[0041] Turning now to FIG. 6, reference is made to a flow chart
generally identified with reference number 600. The method
illustrated by flow chart 600 summarizes an operation of a
flip-flop. With specific reference to the flip-flop 300 shown in
FIG. 3, at step 604, an input voltage is received at input 312. The
input is received at input 312 from a first voltage region. The
first voltage region may be a section or region of an integrated
circuit (IC) containing circuits, such the IC's digital circuits,
that operate at a relatively low voltage. The flip-flop 300 may be
disposed on or near a border between the first voltage region and
another voltage region operating a higher voltage. The input
received at input 312 is representative of a logical value. As
signals pass between these voltage regions the voltages that
represent logical values may need to shift. Generally, a voltage
range substantially close to ground is representative of a logical
zero, whereas a voltage range substantially close to the power
supply rail V.sub.DD is representative of a logical one. After a
voltage representative a logical value is received at operation
604, operation 608 may be executed.
[0042] At operation 608 the logical value present at the input 312
is captured or latched by the flip-flop 300. This may include
capturing the logical value 312 at a clock 316 edge. In capturing a
logical value at a clock edge, the flip-flop may include a master
latch 304 that is associated with a slave latch 308. After
operation 608, operation 612 may be executed.
[0043] At operation 612 an output voltage is presented at the
output 320. The output presented at 320 is in the voltage domain
V.sub.CS and is representative of the logic value that was present
at data input 312. At the logical output 320, the logical value is
presented in a voltage domain that is higher than the voltage
domain present at the input 312. Accordingly, the method
illustrated by flow chart 600 accomplishes two functions. Through
the operation of one circuit stage, a logical value is stored and
level shift occurs.
[0044] FIG. 7 includes a flow-chart generally identified with
reference number 700. Flow chart 700 illustrates a method including
further aspects of latching a logic value in accordance with
implementation as discussed herein. In this regard, flow chart 700
may be considered an elaboration of operation 608, shown in FIG. 6.
Initially, at operation 704, first and second internal nodes 332a
and 332b of flip-flop 300 are pre-charged. Pre-charging the
internal nodes 332a and 332b may occur during a low portion of the
clock 316. The pre-charge transistors 324 may pull the nodes up to
the power supply rail V.sub.CS. After operation 704, operation 708
may be executed.
[0045] At operation 708, the differential pair transistors 336A and
336B may be supplied with data 312 and data complement 340 from a
lower level voltage domain V.sub.DD. Supplying data complement 340,
may include an inverter 356 that is associated or otherwise resides
in the lower level voltage domain V.sub.DD. The inverter 356 may be
provided with data 312 at its input and may provide data complement
340 at its output. Both data 312 and data complement 340 are
logical values that are presented in the lower level voltage domain
V.sub.DD. After operation 708, operation 712 may follow.
[0046] At operation 712, the differential pair transistors 336A and
336B may evaluate their respective inputs. More particularly, when
the clock 316 is high, the pre-charge transistors 324 are off and
the evaluate transistor 328 is on. With the evaluate transistor 328
on, a path to ground is provided for either of the differential
pair transistors 336a and 336b depending on the value of data 312
and data complement 340. In turning on either of the differential
pair transistors 336a or 336b, the gate voltages may only need to
exceed the threshold voltages associated with the differential pair
transistors. Such a voltage is sufficient to turn the transistors
on and discharge one of the internal nodes 332a or 332b. In this
way, the master latch 304 operates as expected.
[0047] The foregoing merely illustrates certain principles of
aspects of the invention with reference to circuit implementations
that conform to inventive concepts. Various modifications and
alterations to the described embodiments will be apparent to those
skilled in the art in view of the teachings herein. It will thus be
appreciated that those skilled in the art will be able to devise
numerous systems, arrangements and methods which, although not
explicitly shown or described herein, embody the principles of the
invention and are thus within the spirit and scope of the present
invention. From the above description and drawings, it will be
understood by those of ordinary skill in the art that the
particular embodiments shown and described are for purposes of
illustration only and are not intended to limit the scope of the
present invention. References to details of particular embodiments
are not intended to limit the scope of the invention.
* * * * *