Internal bus termination technique for integrated circuits...

Details Internal bus termination technique for integrated circuits... Internal bus termination technique for integrated circuits...

2002-05-13

2004-03-30

Cho, James H (Department: 2819)

Electronic digital logic circuitry

Signal sensitivity or transmission integrity

Bus or line termination

C326S083000, C327S262000

Reexamination Certificate

active

06714039

ABSTRACT:

FIELD OF THE INVENTION
The present invention pertains generally to integrated circuits, and more particularly to a termination technique for internal integrated circuit busses with local process/voltage/temperature compensation.
BACKGROUND OF THE INVENTION
Integrated circuits communicate with one another using digital signals. In a binary system, the digital state of zero (“0”, logic “low”) is represented by the range of voltages between a minimum voltage V
MIN
(e.g., 0 volts) of the potential voltage range of the signal and a voltage V
LOW
that is low relative to the total range of voltage. The digital state of one (“1”, logic “high”) is represented by the range of voltages between a voltage V
HIGH
that is high relative to the total range of voltages and a maximum voltage V
MAX
(e.g., 1.5 volts) of the potential voltage range of the signal.
Because the signals are actually analog when transmitted over circuit transmission lines, the transition between digital states does not occur instantaneously, but instead occurs over a period of time that is dependent on physical conditions such as the parasitic resistance, inductance, and capacitance of the transmission line. The amount of time it takes to propagate a signal from one end of a transmission line to the other end of the transmission line is known as “propagation delay”. Propagation delay is affected by the characteristic resistance, capacitance, and inductance of the metal used to construct the transmission line; accordingly, since the characteristic resistance, capacitance, and inductance of a line is dependent not only on the chemical and molecular properties of the metal used to form the transmission line, but also on the length, width, and thickness of the metal, propagation delay is therefore directly affected by the length and width of the line.
Functional circuit elements often operate on data that comprises multiple signals. For example, a circuit element may operate on a 16-bit data word. In order for the circuit element to complete the operation, all 16 bits must typically be available before the operation can be performed. Multiple-bit signals are typically transmitted in parallel over a multiple-transmission-line signal bus.
Variations in the lengths and widths of the individual transmission lines of a signal bus can result in non-synchronous arrival times of the individual bit signals at the receiving end of the bus. This phenomenon can result in undesirable signal jitter.
One method of improving signal integrity and arrival time skew of individual signal bits of a multiple-bit data bus is by equalizing the propagation delay between each of the transmission lines in the bus. Equalization may be achieved by inserting an appropriate number of repeaters along the various individual transmission lines such that each line is subject to approximately the same delay. Of course, the mere inclusion of a repeater along the transmission line results in some unavoidable minimum delay time. The minimum delay between the transmission of a signal and its receipt at the receiving end of the transmission line is often referred to as the insertion delay of the transmission line.
Insertion delay is an important property to consider in the design of transmission lines. Insertion delay is especially important when dealing with high frequency circuit operation because a long insertion delay can prevent the signal from being transmitted across a transmission line within a single (or other chosen integer number) clock cycle. If the insertion delay cannot be reduced enough to allow the signal to be transmitted over the line within a single clock cycle, then the signal takes twice (or even more times) as long to be transmitted over the bus, thereby effectively cutting the operating frequency of the bus in half (or less). Accordingly, minimizing the insertion delay is critical.
The insertion of repeaters along the transmission line can be problematic. First, the addition of repeaters along a transmission line essentially generates a subset of shorter transmission lines, each corresponding to a segment of the whole. Because each individual segment of the transmission line is itself merely a shorter transmission line, each segment will therefore exhibit transmission line properties such as propagation delay and signal reflections as well, albeit on a smaller scale.
Second, it is well-known that unless the impedance of a transmission line matches the impedance of the load it drives, a signal transmitted along the line will degrade due to signal reflections at the load. Accordingly, the repeater terminating each individual segment of the transmission line generates additional signal reflections.
Another difficulty in meeting satisfactory transmission line design requirements is that the operating characteristics of transistors such as CMOS transistors, from which integrated circuit signal drivers/receiver/repeaters are typically constructed, change under a variety of conditions, often referred to as process, voltage, temperature (PVT) variations. More specifically, the operating characteristics due to PVT variations may change with variations in manufacturing process (which can result in variations in the length L or width W dimensions of the transistors) as well as with variations in operating conditions such as junction temperature and supply voltage levels. The operating characteristics may also change with variations of voltage differences across the transistor terminals. If inadequate compensation is made for these variations, the output slew rate and output impedance of the driver/receiver/repeater may vary substantially along a particular transmission line as well as from transmission line to transmission line on a chip.
As is known in the art, complementary CMOS inverters are commonly used as repeaters along long transmission lines.
FIG. 1
illustrates a complementary CMOS inverter
10
. As illustrated, the complementary CMOS inverter
10
includes a p-channel field effect transistor (PFET)
12
with a source connected to a high power supply VDD, a gate connected to an inverter input node
16
, and a drain connected to an inverter output node
18
. The complementary CMOS inverter
10
also includes an n-channel field effect transistor (NFET)
14
with a source connected to a low voltage source (e.g., ground), a gate connected to the inverter input node
16
, and a drain connected to the inverter output node
18
.
The transfer characteristic of the CMOS inverter of
FIG. 1
is shown in FIG.
2
. The transfer characteristic plots V
OUT
over V
IN
for the range of values of V
IN
. It is known in the art that if the threshold voltage V
T
for both the PFET and NFET is identical, V
OUT
may be defined as:
 
V
OUT
=V
IN
+V
T
+[(
V
IN
+V
T
)
2
+VDD
(
VDD−
2
V
IN
−2
V
T
)−(
V
IN
−V
T
)
2
]
1/2
,
where V
IN
is the input voltage, and VDD is the power supply voltage. As shown, in the ideal case, the inverter switches when V
IN
=VDD/2.
As shown in
FIG. 2
, the transfer characteristic of the inverter is affected by the proportionality constants &bgr;
p
and &bgr;
n
of the respective PFET
12
and NFET
14
. The proportionality constants &bgr;
p
and &bgr;
n
are dependent on the doping, construction, and dimensions (width W and length L) of the transistor. Accordingly, since the proportionality constants &bgr;
p
and &bgr;
n
may vary from one manufacturing process to another and even between localized areas of the chip within the same process, the transfer characteristic may be different across different chips manufactured in different groups and between localized areas of the same chip.
The transfer characteristic illustrates the effect of the proportionality constant ratio &bgr;
p
/&bgr;
n
on the switching point (also called the “trigger point”) of the inverter
10
. As illustrated in
FIG. 2
, an inverter whose proportionality constant ratio &bgr;
p
/&bgr;
n
is 10 switches at a much lower input voltage V
IN
than an inverter whose proportionality constant rat

Affiliated with

Also associated with

Rating

Internal bus termination technique for integrated circuits... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Internal bus termination technique for integrated circuits..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Internal bus termination technique for integrated circuits... will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3189881