Coded data generation or conversion – Analog to or from digital conversion – Digital to analog conversion
Reexamination Certificate
1999-05-28
2001-08-28
Tokar, Michael (Department: 2819)
Coded data generation or conversion
Analog to or from digital conversion
Digital to analog conversion
C341S145000, C341S152000, C341S053000
Reexamination Certificate
active
06281822
ABSTRACT:
CROSS-REFERENCE TO RELATED APPLICATIONS
The following applications, assigned to the assignee of the current invention, are being filed concurrently, contain material related to the subject matter of this application, and are incorporated herein by reference:
by J. McDonough, entitled “Method of and Apparatus for Generating Data Sequences for Use in Communications,” Ser. No. 09/322,373, currently pending;
by D. Chen et al., entitled “Device and Method for Generating Clock Signals From a Single Reference Frequency Signal and for Synchronizing Data Signals with a Generated Clock,” Ser. No. 09/322,282, currently pending;
by J. McDonough et al., entitled “Method and Apparatus for Controlling System Timing with Use of a Master Timer,” Ser. No. 09/321,697, currently pending; and
by J. McDonough et al., entitled “Device and Method for Maintaining Time Synchronous with a Network Master Time,” Ser. No. 09/322,240, currently pending.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to circuits such as pulse density modulators, which represent a multi-bit digital input signal with a serial binary output comprising a corresponding sequence of fixed amplitude pulses. More particularly, the invention concerns such a system that uses novel circuitry to generate a binary output with the advantage that the pulses that are evenly spaced over time.
2. Description of the Related Art
Many circuits today have need for both digital and analog circuitry. Digital circuitry provides fast computation, parallel input/output, low power, and other benefits. Analog circuitry is nonetheless useful and sometimes essential in many applications, such as transmitting electromagnetic signals, controlling motors, applying power to loads, etc. To concurrently reap the benefits of both analog and digital circuitry, some circuits employ appropriate digital-to-analog converters and analog-to-digital converters. Today's market offers digital/analog converters with many different operating theories, speeds, and other specifications.
Known digital-to-analog chips provide significant advantages, primarily in the area of speed. Nonetheless, circuit designers still lament certain features of known digital-to-analog converter chips. For example, digital-to-analog converter chips consume a significant amount of power, which can be problematic for hand-held telephones and other applications that receive limited electrical power from a compact battery. Also, when incorporated into custom-designed integrated circuits, digital-to-analog converter chips typically occupy substantial size and thereby impinge upon the circuit designer's goal of producing produce the smallest possible overall circuit. Digital-to-analog circuits may also be prone to defects, since analog components are represented by narrow integrated circuit etchings with unforgiving tolerances. Finally, the inclusion of a digital-to-analog converter increases the cost and complexity of the manufacturing process, which then requires one manufacturing run to lay down digital components and a separate manufacturing run to lay down analog components.
One proposed solution to the limitations of known digital-to-analog converter chips is the pulse density modulation (“PDM”) circuit. Many PDM circuits operate by receiving a multi-bit parallel digital input signal, and providing a corresponding serial binary output signal. The binary, digital output signal is then treated as an analog signal, which is averaged, smoothed, or otherwise filtered. The amplitude of this analog output signal is proportional to the multi-bit parallel digital input signal, and therefore constitutes an analog representation of the digital input signal.
One known PDM circuit operates by producing a single-pulse binary signal whose duty cycle is proportional to the multi-bit parallel input signal. This PDM is said to use pulse width modulation, or “PWM.” The pulse width modulated output signal is fed to a low pass filter, creating an analog output signal that is roughly proportional to the original, multi-bit digital input signal.
FIG. 1
shows a circuit
100
including an exemplary pulse width modulation system that utilizes a counter
100
, comparator
102
, and an analog low pass filter
104
. The comparator
102
compares the output of the counter
100
to a digital input. When the input is larger or equal to the output of the counter, the comparator outputs a binary “one.” Otherwise, the comparators output is a binary “zero.” The low pass filter
104
processes the binary output, thus creating the final analog output signal. As explained below, one drawback of this approach is that the comparators binary “one” outputs are not evenly distributed over time.
FIGS. 2B-2I
show serial binary outputs corresponding to inputs of zero through seven, respectively, from a pulse width modulation system utilizing the reference clock signal of FIG.
2
A. When used in this manner, pulse width modulation has certain drawbacks. Chiefly, the approach of varying the duty cycle according to input signal has the effect of lumping binary output pulses together for larger multi-bit digital input signals.
FIG. 2B
shows the serial binary output corresponding to a digital input of “zero.” One multi-bit binary input is represented in one output period, such as the period
201
. The output of
FIG. 2B
represents a single output pulse, and there is no problem with signal lumping. However,
FIG. 2F
shows the binary output corresponding to a multi-bit parallel digital input of “four.” The output signal includes a contiguous high signal
200
(binary “1”), which actually represents five output pulses together. After the high signal
200
, there is a contiguous low signal
202
(binary “0”). Due to the grouping high and low signals in this way, the analog output signal exhibits overshoots. This is because, upon applying the sequential binary output signal to an analog filter, the resulting analog signal overshoots in the realm of
200
, and undershoots in the realm of
202
, as shown by the superimposed analog signal
204
.
A variation of the foregoing circuit outputs a binary “one” when the input is strictly larger than the output of the counter. Using strictly larger instead of “greater than or equal to” results in a difference of one pulse per period. For example, a digital input of zero in a system that uses a comparator that is strictly larger than the output of the counter will not output any “ones.” However, a system that uses a comparator that is “greater or equal to” zero will have one digital “one” per period. This embodiment, like the previous example, has the drawback of unevenly distributing binary “ones” in the output signal.
As shown in
FIG. 3
, one solution to the foregoing problems of pulse width modulation was proposed in U.S. Pat. No. 5,337,338, Sutton et al., issued on Aug. 9, 1994. This approach uses a PDM
300
that is nearly identical to the circuit
100
of FIG.
1
. The PDM
300
includes a counter
302
, comparator
304
, and analog low pass filter
306
. However, unlike the circuit
100
(FIG.
1
), the circuit
300
reverses the order of coupling
308
. Namely, the mapping between most significant bits (MSBs) and least significant bits (LSBs) is swapped in the coupling between the counter
302
and the comparator
304
.
FIGS. 4B-4I
show serial binary outputs corresponding to inputs of zero through seven, respectively, utilizing the clock signal of
FIG. 4A. A
single period of the binary output is identified by
402
. The purported advantage of this system is that, within a single period of the resulting serial binary output signal, like bits are not lumped together as much as the outputs of
FIGS. 2B-2I
. Nonetheless, as discovered by the present inventor, the output pulses are still not evenly spaced in
FIGS. 4B-4I
. For example, the representation of a “two” input has three “one” pulses in a period, and includes a non-symmetrical blank space
404
. When filtered into an analog signal, these blank spots and lumps manifest themselves as signal oversh
Brady III W. James
Dot Wireless, Inc.
Hernandez Pedro P.
Paik Steven S.
Telecky , Jr. Frederick J.
LandOfFree
Pulse density modulator with improved pulse distribution does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Pulse density modulator with improved pulse distribution, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Pulse density modulator with improved pulse distribution will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2465025