WO1999034572A2

WO1999034572A2 - Method and apparatus for representing an analog waveform as digital messages

Info

Publication number: WO1999034572A2
Application number: PCT/US1998/027592
Authority: WO
Inventors: Lance Ong
Original assignee: Lance Ong
Priority date: 1997-12-24
Filing date: 1998-12-22
Publication date: 1999-07-08
Also published as: WO1999034572A3; EP1072097A2; AU2014299A; US6020840A; US6177898B1

Abstract

A method and apparatus for converting an analog waveform into digital messages. The digital messages describe a change in amplitude of the analog waveform (55), the polarity or direction of the change in amplitude (58), and the elapsed time for the change in amplitude (57). Each digital message is created upon the occurrence of a change in amplitude which exceeds a predetermined or predefined threshold voltage (53).

Description

METHOD AND APPARATUS FOR REPRESENTING AN ANALOG WAVEFORM AS DIGITAL MESSAGES

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates generally to a method and

apparatus for encoding and decoding a waveform, and, in

particular, to a method and apparatus for representing a waveform

as digital messages which are formed independently of a specific

sampling rate.

2. Prior Art

In a conventional system for converting an analog signal into

a digital signal, a reference clock is used to generate a periodic train

of pulses. Upon each pulse, the analog signal is sampled and an

analog to digital conversion is carried out.

At the end of each conversion a binary code is produced

which is proportional to the amplitude of the sampled analog signal.

The conversion process is repeated within each tick of the clock

resulting in a staircase-like amplitude approximation of the original

waveform. The accuracy of this reconstruction depends on a

number of factors including the rate at which the waveform is

sampled, the time required to complete each A/D conversion and the frequency content of the original signal.

In most conventional systems a sampling speed or rate is

selected and fixed, as determined by the reference clock, for use

during the entire AID conversion of the original waveform. In

selecting the optimum sampling frequency, a sampling theorem

commonly known as the "Nyquist Theorem'^' is most often used.

According to the Nyquist Theorem, if a waveform is sampled at a

speed or rate that is approximately greater than twice the highest

frequency component present in the waveform then in theory it is

possible to accurately reconstruct the original waveform from these

samples.

As might be expected, in the case of high frequency analog

waveforms, A/D conversion methods which employ a Nyquist

Theorem based sampling rate tend to generate a high number of

data samples. It is preferable, whenever possible, to minimize the

total number of samples necessary to reconstruct a waveform.

Fewer data samples require a shorter period of time to transmit and

require smaller memory reserves for storing and saving the

samples.

A high number of samples is particularly undesirable when only a fraction of the samples are needed to accurately reconstruct

the original waveform. For example, in an analog waveform

having a combination of low and high frequency components it is

not necessary and in fact undesirable to sample the low frequency

components at the sampling speed necessary for accurately

reproducing the high frequency components. Using conventional

techniques the entire waveform would be sampled at a sampling

rate best suited for reproducing the high frequency components of

the multi-component waveform. This would result in an

unnecessarily high number of data samples being generated.

In an effort to address these undesirable consequences of

employing a fixed conventionally derived sampling rate, there are

known in the prior art a number of systems and methods which

incorporate a frequency monitoring stage for dynamically varying

the sampling rate according to the changing frequencies of a

waveform. Upon a change in the frequency component of a

sampled waveform, the dynamic stage is designed to vary the

sampling rate to a rate which is more appropriate for reconstructing

the particular frequency component being sampled. Depending on

the waveform, the sample rate may be increased for high frequency components or decreased for low frequency components. As a

result of dynamically controlling the sampling rate the conversion

is made more efficient and the number of samples taken for

accurately reproducing the waveform is optimized.

Podalak U.S. Patent No. 4,763,207 is an example of a

system and apparatus where varying the sampling rate in

accordance with the changes in frequency of the sampled waveform

is proposed. Although Podalak represents a satisfactory approach

for dealing with the potential consequences of utilizing a fixed

sampling rate derived using conventional sampling theorems, the

Podalak system proposes the use of expensive hardware in the form

of a logic analyzer to determine the optimum sampling frequency.

In addition to being prohibitively expensive, the Podalak system

continues to implement an analog waveform encoding approach

which is tied to a sampling frequency.

Kitamura US Patent No. 4,370,643 discloses yet another

system where the sampling frequency may be adjusted to reflect

changes in the frequency of the waveform being sampled.

Although Kitamura proposes an apparatus and method which can

be implemented at a far less cost than that of Podalak, Kitamura remains a method and apparatus which reflects a marriage to the

conventional way of thinking in waveform reproduction.

Although the prior art apparatuses and methods have been

effective in reducing the total number of data samples necessary to

reproduce a waveform of varying frequency, these methods and

apparatuses nevertheless represent only an extension or adaptation

of conventional methods. That is, although they manage to

circumvent some of the drawbacks of the prior art systems, they

continue to rely on the basic theory underlying the prior art

systems, namely, A/D conversion of a waveform utilizing an

"ideal" sampling rate. As a result, the converted analog waveform

data is still represented as a series of amplitude points

corresponding to the sampled original waveform which must, at

some point, be passed through a smoothing filter in order to

approximately reconstruct the original waveform.

Furthermore, the problems common to conventional A/D

conversion techniques which rely on a sampling rate including

quantization errors and aliasing have not been overcome. As is

known in the art, quantization noise and aliasing can significantly

deteriorate the quality and accuracy of the reproduced waveform. Thus a need continues to exist for a method and apparatus

for compressively approximating an analog waveform which does

not rely on a fixed or varying sampling rate.

Still a further need exists for a system and apparatus which

can be used to compressively approximate an analog waveform

without the need for an inordinate amount of memory for storing

derived samples.

The object of the present invention, like the prior art

systems, is to encode or represent an original analog waveform as

a series of digital messages which can be used to precisely

reconstruct the original analog waveform. It is expected that the

benefits of the present invention will extend to applications in the

field of Digital Audio, data transmission over low to medium

bandwidth networks such as the Internet and to the field of data

acquisition in general.

To that end, it is a general object of the present invention to

provide a method and apparatus that has intrinsic advantages over

conventional conversion techniques in that there is no one to one

relationship between the number of data samples and a system or

reference clock and as a consequence does not have a fixed data rate per second. Instead, the number of sample data dynamically

fluctuates based on the frequency and amplitude content of an

original analog waveform.

It is another object of the present invention to provide a

method and apparatus which is not based on the sampling theorem

or variations thereof.

It is a further object of the present invention to provide a

method and apparatus which automatically adjusts to varying

frequency components in a sampled analog waveform.

It is yet another object of the present invention to provide a

method and apparatus which performs dynamic data compression

of the previously sampled data.

It is yet a further object of the present invention to provide

a method and apparatus which permits the reconstruction of an

original waveform from the sample data without using smoothing

circuitry.

SUMMARY OF THE INVENTION

In accordance with the stated objectives and other objectives

which will hereinafter become apparent, there is provided a method

and apparatus for representing an analog waveform as a plurality of digital messages describing a segment or sample of the analog

waveform at a particular moment in time. In particular, each digital

message marks the detection by the apparatus of a certain minimum

change in the amplitude of the original analog waveform. The

minimum change in amplitude needed to trigger the creation of a

digital message is a variable value which is selected by the user of

the apparatus prior to converting or encoding a waveform. The

preselected value is fixed for the entire encoding operation and

represents a threshold voltage which must be exceeded before a

conversion of a segment of the analog waveform is performed.

Upon each detected occurrence of a change in amplitude

which is substantially equivalent to the value of the predefined

threshold voltage, a digital message describing that particular

segment of the waveform is generated. Each generated digital

message may contain information regarding the magnitude of the

change in amplitude, the elapsed time or duration of that detected

change in amplitude, and the polarity or direction of that detected

change in amplitude.

After the full waveform conversion is completed, the

resulting plurality of digital messages may be immediately converted to reproduce the original analog waveform, transmitted

to a remote location for reproduction of the original analog

waveform at a remote location, or stored for later use.

In accordance with the above, the apparatus of the present

invention comprises a means for establishing and fixing a threshold

voltage, and a means for detecting a change in the amplitude of the

sampled waveform which is substantially equivalent to the

predefined threshold voltage ("minimum change in amplitude").

Upon the detection of a minimum change in amplitude, the

detecting means generates a signal which indicates that said

minimum change in amplitude has occurred and identifies the

polarity of that detected minimum change in amplitude. The

apparatus further comprises a clock or timer means for measuring

the elapsed time of a minimum amplitude change in the waveform

and a microprocessor which is responsive to the signal generated by

the detecting means and clock or timer means for creating a digital

message.

In a preferred embodiment, the detecting means includes a

means responsive to said signal for sampling and holding said

amplitude change of said waveform for use in detecting subsequent minimum changes in amplitude. The detecting means further

comprising a comparator means for detecting the occurrence of a

minimum amplitude change and the polarity of such change.

Also in accordance with the objects of the present invention,

there is further provided a method and apparatus for compressively

representing an analog waveform as a plurality of new or final data

messages. The new or final data messages represent a combination

of sequential data messages having substantially equivalent elapsed

times. In effect the combination of sequential data messages having

substantially equivalent elapsedtimes represents a slope integration

along selected segments of the analog waveform. The slope

integration or compression reduces the total number of data

messages necessary to accurately recreated the analog waveform.

There is also a method and apparatus for reproducing an

analog waveform which has been represented as a plurality of data

words wherein each of said data words includes a data field

identifying a minimum change in amplitude, a data field indicating

a polarity for said minimum change in amplitude and a data field

indicating the elapsed time of said minimum change in amplitude.

The apparatus comprises a microprocessor for retrieving each of the data words from a storage medium, a means for generating a

reference clock coupled to to the microprocessor and a voltage

source coupled to the microprocessor and responsive thereto for

outputting an analog waveform in accordance with the information

in stored data fields of the retrieved data words.

BRIEF DESCRIPTION OF THE DRAWINGS

The above and related objects, features and advantages of the

present invention will be more fully understood by reference to the

following detailed description of the presently preferred, albeit

illustrative, embodiments of the present invention when taken in

conjunction with the accompanying drawings, wherein:

FIG. 1 is a schematic block diagram of an apparatus for

encoding a waveform as digital messages in accordance with the

present invention.

FIG. 2 is a schematic block diagram of an apparatus for

reproducing an original waveform from digital messages derived

from an apparatus and method in accordance with the present

invention

FIG. 3 shows a representation of a waveform having

indicated thereon positions where compressed digital messages would likely be created in accordance with the present invention;

FIG. 4 is a segment of a representational waveform

indicating a derived data point corresponding to a flat line condition

in the waveform,

FIG. 5 is a flowchart description of a microprocessor

operation; and

FIG. 6 is a flowchart description of a microprocessor

operation.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

Referring now to the drawings and in particular to Fig. 1

thereof, therein illustrated is an apparatus including electronic

circuitry, a microprocessor and software for encoding an analog

waveform as a series of digital messages in accordance with a

preferred embodiment of the present invention. In Fig. 2, there is

shown a counterpart apparatus which may be utilized to reproduce

the original analog waveform by decoding the digital messages

generated by the apparatus of Fig 1. Although not specifically

shown in the Figures, the apparatuses of Figs. 1 and 2 may be

combined into an encoding/decoding system which may utilize a

single memory and single microprocessor to process a waveform. As shown in Fig 1. the schematic block diagram illustrates

an encoding apparatus, generally designated by the reference

numeral 10, in accordance with a preferred embodiment. The

encoding apparatus 10 is generally comprised of six components or

stages acting on an input analog waveform to convert the waveform

into a series of digital messages representing the movement of the

waveform in real time. At the input stage, the encoding apparatus

10 preferably includes a low noise input gain stage 14 for

amplifying the peak to peak voltage of the input waveform; a

detecting means generally designated by the reference numeral 18;

a voltage source generally designated by the reference numeral 22;

a timer/counter 24, a high frequency clock source 26, a

microprocessor 28; and optionally a storage memory 32 for

archiving the digital messages describing a particular analog

waveform.

The encoding apparatus 10 performs an analog to digital

conversion of a waveform essentially by breaking the input

waveform into segments representing a certain minimum change in

amplitude of the waveform. Upon every detection of a minimum

change in amplitude in the waveform, the creation of a digital message is triggered. The minimum change in amplitude for

triggering the creation of a digital message is preferably a value in

the microvolt range which is selected prior to beginning a

waveform encoding operation and remains fixed throughout.

In selecting the minimum change in amplitude, it may be

necessary to first determine the noise components of the waveform

to be sampled. In this manner a trigger level may be selected which

is high enough to avoid the creation of digital messages that merely

represent changes in amplitude corresponding to noise present in

the original waveform. Once a minimum value is selected, the

value is input and fixed by means of the voltage source 22 which is

preferably a digitally controlled voltage source or potentiometer.

In the preferred embodiment, the voltage source 22 is connected

between the detection means 18 and microprocessor 28.

Briefly, when the amplitude of the input waveform changes

a value that is substantially equivalent to the threshold voltage

established by voltage source 22, a minimum change in amplitude

has occurred which is detected by apparatus 10 triggering the

formation of a descriptive digital message. Substantially at the

same time a measurement of the elapsed time for the minimum change in amplitude is made and the polarity or direction of the

minimum change in amplitude is detected. The elapsed time

information and polarity are stored as part of the same descriptive

digital message.

As the digital message preferably describes three aspects or

characteristics of a tracked segment of the waveform, the preferred

format of the digital message is a 32 bit data word wherein bit 32

represents the polarity or direction of the minimum change in

amplitude, bits 25-31 represent a multiplier of the minimum change

in amplitude and bits 1-24 are used to store the measured elapsed

time of the minimum change in amplitude. It should be noted that

as the minimum change in amplitude is fixed, a non-compressed

data word will typically have a multiplier of one.

In accordance with a preferred embodiment of the invention,

an analog waveform to be converted or encoded is first connected

to the positive input of the gain stage 14 which as is shown may be

a conventional operational amplifier 32 having negative feedback

through feedback resistor Rf. The amplified or conditioned signal

is then fed to an input side of the detecting means 18 which

preferably includes a sample and hold circuit 34 and comparator circuit generally designated by reference numeral 36.

In particular, the output of the input gain stage 14 is coupled

to the to the input of the sample and hold circuit 34, the positive

node of a first comparator 40 and the negative node of a second

comparator 42. As indicated further in Fig. 1 , voltage source 22 is

electrically connected to each of comparators 40, 42 for use in

detecting the occurrence of a minimum change in amplitude.

In operation, one of comparators 40, 42 may generate a

signal at each occurrence of a change in amplitude of the waveform

which is substantially equivalent to the level supplied by the

voltage source 22. In the event comparator 40 is triggered, a

positive minimum change in amplitude has occurred; and in the

event comparator 42 is triggered, a negative minimum change in

amplitude has occurred.

The outputs of comparators 40, 42 are fed into separate

interrupts of microprocessor 28 via connections 44, 46 for

distinguishingbetween the positive and negative minimum changes

in amplitude. Upon the firing of one of comparators 40 or 42, the

corresponding interrupt receives an input allowing the

microprocessor 28 to identify the direction of the minimum change in amplitude and further causing the microprocessor 28 to enable

the sample and hold circuit 34.

Once enabled, the sample and hold circuit 34 captures the

current value of the conditioned waveform which is being fed from

the output of the amplifier 14. The held or captured value is

substantially simultaneously fed from the output of the sample and

hold circuit 34 to the negative node of comparator 40 and the

positive node of comparator 42 via connection . This configuration

permits the apparatus 10 to dynamically detect each subsequent

occurrence of a minimum change in amplitude as measured

between the positive and negative nodes of comparators 40, 42.

The process is repeated until the waveform conversion is

completed.

As previously mentioned, upon each occurrence of a

minimum change in amplitude a measurement of the elapsed time

for the minimum change in amplitude is performed and the time

value is included in bits 1-24 of the data word. In the apparatus 10,

the timer/counter 24 in combination with the high speed clock

source 26 are used to measure the elapsed time.

As shown, the timer/counter 24 is coupled to the high speed clock source 26 which is preferably a quartz crystal oscillator. The

timer 24 is preferably comprised of a package of three-eight bit

timers which permit up to a 24 bit countdown. Conceivably, with

such an arrangement, the reference clock can be configured to clock

the occurrences of minimum changes in amplitude at a rate of 16

MHZ.

Generally, in deciding on the speed of the reference clock or

time base to be used in measuring the time between the start and

end of a minimum change in amplitude, the clock speed must be

sufficiently high to mark the time of a voltage transition event

accurately enough to prevent skewing in the reproduced waveform.

Skewing may occur if the speed or frequency of the reference clock

is selected such that when a detectable change in amplitude occurs,

it occurs between ticks or pulses of the reference clock with too

much room on either side of the transition. As a result a detectable

change in amplitude is associated with the nearest clock tick which

may lead to an inaccurate reconstruction of the original waveform.

For example, in the case of an audio waveform which has a

frequency response of 20Khz, an appropriate reference clock

frequency might be lOMhz or 500 times the highest frequency response. In this manner, the occurrence of detectable changes in

amplitude may be associated with a more precise point in time so

as to minimize skewing.

Although a reference clock is utilized, the apparatus 10 does

not have a fixed data rate per second. The data rate and quantity

fluctuate depending on the frequency and amplitude content of the

particular waveform fed into apparatus 10. Low frequency

waveforms have a much longer periodic wave and, consequently,

a change in amplitude takes longer to reach each detectable

threshold point. This gives an output which intrinsically exhibits

a data compression characteristic such that in moments of silence

a digital message may only occur once every 24 bit period. The

highest concentration of data will occur when the waveform

contains a high frequency content of a high amplitude.

Data compression may be enhanced or supplemented by

performing a slope integration along a section of a sampled

waveform. In accordance with a method of the present invention,

data words having a common slope as determined by the minimum

change in amplitude, which is fixed, and the elapsed time for the

change in amplitude, which is variable, may be combined into a single new data word representing all sequential data words along

a common slope. In Fig. 3, a sample waveform is shown having

new data words W 1 - W20. Data word W4, for example, represents

a combination or slope integration of the individual data messages

which were generated to describe the segment of the waveform

between W3 and W4.

In a preferred embodiment, slope integration is

accomplished by performing a comparison of the elapsed times of

sequential data words. In the instance where sequential data words

have an equivalent elapsed time, within a given tolerance, the

sequential data words are combined to form a new or final data

word. The new or final data word describes a total elapsed time

and further describes the total change in amplitude of the waveform

during that elapsed time by a multiplier representing the total

number of occurrences of a minimum change in amplitude. Thus,

for example, if three sequential data words are combined to form a

new or final data word, the multiplier representing the minimum

change in amplitude would be three.

An additional operation which may be performed on the data

words is useful in determining the start of a minimum change in amplitude after a period where no signal has been detected,

otherwise known as a flat line condition. The operation essentially

derives a data word representing a data point along the waveform

which has not been detected but rather calculated and interposed

based on the information contained in data words created prior to

and immediately after the calculated data point. In Fig. 4, the

derived data point is represented as point Pla located between

points PI and P2. As shown in Fig. 4, if a data message

representing point Pla is not inserted as a point the recreated

waveform would resemble more closely the waveform represented

by the dotted line extended between the points PI and P2.

The operations of slope integration and flat line tracking will

become more apparent upon an understanding of the

microprocessor 28 operations during a waveform conversion.

Accordingly the remainder of the operation of the apparatus 10 and

method of the present invention will be described in connection

with the operation of the microprocessor 28.

Briefly describing the function of the microprocessor 28, the

microprocessor 28 monitors interrupts generated by either of the

two comparators 40, 42 and the timer 24. At the start of a conversion and recording process the following occurs: the

microprocessor initializes the components, which includes

establishing the threshold voltage for the conversion as fixed by

voltage source 22, setting the sample and hold circuit to a 0-volt

level, loading the timer 24 with a divisor value to achieve a desired

reference clock rate and zeroing the software counters for clocking

the elapsed time of a minimum change in amplitude.

Following initialization, the microprocessor is ready to

generate a digital message. To that end the timer 24 begins to run

and upon each timer countdown, the timer generates an interrupt

via line 50 to the microprocessor which upon request increments a

timer variable. The timer variable preferably located in the

microprocessor 28 tracks the elapsed time for every occurrence of

a detected minimum change in amplitude. If the timer variable

reaches a maximum value without there having occurred a

minimum change in amplitude, the microprocessor 28 generates a

data word having a multiplier value of zero indicating that a

minimum change in amplitude has not occurred within a maximum

period of time as limited by the 24 bit size of the data field.

The value of the timer variable per occurrence of a detected minimum change in amplitude is temporarily stored in a work

buffer by the microprocessor. The elapsed time value is retrieved

and written to a data word together with the polarity and multiplier

field information. Upon the writing of the data word a flag is set

indicating that the data word is ready for further processing by the

microprocessor 28.

The microprocessor 28 is preferably equipped with at least

two buffers (not shown) which are used to temporarily store the

created data words. Initially data words are stored in a first buffer

(Buffer 1) and following a slope integration or flat line tracking

operation are stored in a secondary buffer (Buffer 2). The data

words stored in the secondary buffer are the final data words which

contain the information for reproducing the original waveform.

The typical operation of a main routine of the

microprocessor is best understood in connection with Fig. 5 which

includes a description of the operations of slope integration and flat

line tracking. In a first steps SI and S2, each newly created data

word is detected as indicated by a data ready signal and processed

by the microprocessor 28. In step S3, the elapsed time, Tval, of the

newly created data word is compared to the elapsed time, Orgval, of an immediately prior data word held in the first buffer, if any. If

the elapsed times Tval and Orgval are not equivalent within a plus

and minus tolerance, a final data word having a new Tval and

multiplier is generated and stored in the secondary buffer.

As represented in steps S4 to S9, in generating the final data

word the data words then present in the first buffer are totaled. This

total is written to the multiplier field of the final data word. The

Tval of the final data word is derived by aggregating the Tval for

each data word then present in the first buffer. The aggregate Tval

is then written to the Tval of the final data word. The final data

word including the correct polarity is written to the secondary

buffer.

Following this operation, the first buffer is reset and the

newly created data word is placed as the first entry in the first

buffer, the buffer address counter is incremented and Tval is set

equal to Orgval. The microprocessor 28 is now ready to receive and

process the next newly created data word.

As indicated in steps S3 and S14, if Tval of a newly created

data word is equivalent to Orgval within a certain positive and

negative window, the newly created data word is checked to determine its location in the first buffer. If the newly created data

word is in any location in the first buffer other than the second

position, the newly created data word is written to the first buffer

and the buffer address counter is incremented by 1. The

microprocessor 28 is again ready to process the next newly created

data word.

Steps S 15 to S20 are performed if the newly created data

word is in the second position in the buffer. In accordance with

steps S15-S20, the Tval for the newly created data word is

compared against the Tval for the last stored data word in buffer 2.

The Tval for the buffer 2 data word is determined by dividing the

total elapsed time field by the multiplier field. If Tval of the newly

created data word is smaller than Tval for the buffer 2 data word

and not a multiple thereof then a next final data word is written to

the secondary buffer having a polarity of the buffer 2 data word, a

multiplier of 1, and a Tval equal to that of the Tval for the newly

created data word. The routine then subtracts Tval of the newly

created data word from Tval of the buffer 2 data and changes the

Tval of the buffer. This modified final data word of buffer 2 is

written to the secondary buffer. The newly created data word is placed in the first buffer and the buffer address counter is

incremented by 1. The microprocessor 28 is ready for the next

newly created data word.

At step SI 5, if the Tval of the newly created data word is

larger than Tval of the buffer 2 data word, the microprocessor

writes the newly created data word to the first buffer, buffer 1 , and

increments the buffer address counter by 1. The microprocessor 28

is ready for the next newly created data word.

The microprocessor 28 will continue its routine until each

data word is processed. At the end of the final data processing

words stored in the secondary buffer may be transferred to the

storage memory 32. Once in the storage memory 32, the data

words may be retrieved by a decoding apparatus in order to

reproduce the original waveform.

Referring now to the schematic block diagram of Fig. 2,

therein illustrated is a decoding apparatus, generally designated by

the reference numeral 60, for reconstructing or reproducing an

original analog waveform from the information contained in the

plurality of digital messages in accordance with a preferred

embodiment. The digital messages are preferably in the format already described herein, namely, a 32 bit data word wherein bit 32

represents the polarity or direction of the minimum change in

amplitude, bits 25-31 represent a multiplier of the minimum change

in amplitude and bits 1-24 are used to store the measured elapsed

time of the minimum change in amplitude.

The decoding apparatus 60 is generally comprised of four

components or stages acting to retrieve and convert a plurality of

sequentially stored data words into a precise reconstruction of the

original analog waveform. The decoding apparatus 60 is comprised

of a voltage source generally designated by the reference numeral

62, preferably a digital potentiometer, a high frequency clock

source 63 and timer/counter package 64 to give a programmable

time base or reference clock of flexible duration, a microprocessor

66, and an output gain stage 68. The digital messages are retrieved

from a storage memory 70 or some type of external memory device

which has not been shown in the figures.

In general, the decoding or D/A conversion is conducted in

an inverse manner to the encoding or A/D conversion. The timer

64 is loaded with a clock value for the conversion. Initially the

clock value should be the same as that used in the A/D conversion. Like the timer of the encoding apparatus 10. the timer 64 may

comprise three eight bit timers for establishing the time base or

reference clock of apparatus 60.

In reconstructing the original waveform from the recorded

digital messages or data words, the microprocessor 66 which, as

shown in Fig. 2, is coupled to the digital potentiometer 62 and

timer 64 is first initialized. Initializing the microprocessor 66

includes loading the timer 64 with the appropriate divisor for the

desired reference clock and zeroing the software counters.

Following initialization the microprocessor 66 executes a

fetch cycle and retrieves a first data word from memory 70. The

microprocessor 66 multiplies the threshold voltage by the value

stored in the multiplier field of the retrieved data word to derive a

full voltage range. The full voltage range is used to calculate the

increments of voltage change relative to the total elapsed time

contained in the time field of the retrieved data word. The

incremental voltage change is then applied to the digital voltage

source in a manner proportional to the total elapsed time as

indicated by the retrieved data word. This process is repeated for

every retrieved data word. To improve the quality of the output waveform, each data

word retrieved from memory 70 is further subdivided into yet

smaller data words by taking advantage of the generally higher

grain or output resolution of the digital potentiometer 62. As

indicated in steps SI to S9 of the flowchart of Fig. 6, the full

voltage range of each retrieved data word is divided by the

maximum resolution of one increment of the digital potentiometer

62. This gives the number of step divisions available for a smooth

output transition over time in accordance with the generally higher

resolution of the digital potentiometer 62.

Next, in step S4, the elapsed time represented in bits 1-24 of

the retrieved data word is divided by the total number of step

divisions to create a string of new data words that represent one

step of the digital potentiometer 62 that is to occur at averaged

divisions of time points across the total elapsed time of the retrieved

data word.

The elapsed time of each of the newly created string of data

words is loaded into a timer variable. The timer 64 which is

connected to an interrupt or IRQ pin on the microprocessor 66

generates a signal at each count down of the timer 64. At each timer 64 interrupt, the timer variable is decremented until the value

reaches zero. When the timer variable reaches zero the routine

increments or decrements the level of the digital potentiometer by

one unit of resolution. It then reloads the timer variable with the

next new data word and follows the same procedure until all the

new data words have been transmitted and output via gain stage 68

which is coupled to the output of the digital potentiometer 62. This

process is repeated for each of the remaining data words stored in

memory 70 until the full original waveform has been recreated.

By recreating the analog waveform in this manner, it is

possible to eliminate the need for the output smoothing filters

otherwise necessary to reconstruct an analog waveform using

conventional technologies.

In order to optimize the amount of data, it is possible to

insert realtime messages which alter the response of the Digital to

Analog converter to allow for additional compression. For

example, a realtime message may be used to change the value of the

trigger or threshold voltage to a value different than the startup

value or to inversely change it back. In this manner the recreation

of a particular segment of the waveform may be adjusted to coincide with the number of data words which have been created to

describe that particular segment of the original waveform.

Now that the preferred embodiments of the present invention

have been shown and described in detail, various modifications and

improvements thereon will become readily apparent to those skilled

in the art. Accordingly, the spirit and scope of the present invention

is to be construed broadly and limited only by the appended claims,

and not by the foregoing specification.

Claims

WHAT IS CLAIMED IS:

1. A method of representing a waveform as a plurality of

digital messages, each digital message describing a segment of said

waveform at a particular moment in time, comprising the steps of:

selecting and fixing a threshold voltage for use in triggering

the creation of said each digital message;

detecting in said analog waveform the occurrence of an

amplitude change which is substantially equivalentto said threshold

voltage;

detecting the polarity of said change in amplitude;

measuring an elapsed time corresponding to the duration of

said occurrence of an amplitude change which is substantially

equivalent to said reference voltage;

generating a plurality of digital messages, each said digital

message having information corresponding to said change in

amplitude, the polarity of said change in amplitude and the elapsed

time for said change in amplitude.

2. The method of claim 1 wherein said generated data

messages are sequentially stored in a storage memory.

3. The method of claim 1 further comprising the step of amplifying said waveform prior to representing said waveform as

said plurality of digital messages.

4. The method of claim 1 wherein said detecting step further

comprises the step of initiating a sampling and holding of the

amplitude of said analog waveform upon the detection of a first

said change in amplitude substantially equivalent to said reference

voltage.

5. The method of claim 4 wherein said detecting step further

comprises the step of comparing said analog waveform to a

sampled and held amplitude for detecting each subsequent

occurrence of a change in amplitude substantially equivalent to said

reference voltage.

6. The method of claim 1 wherein said digital message

comprises a thirty two bit data word.

7. The method of claim 6 wherein bits 1-24 represent said

elapsed time, bits 25-31 represent a multiplier field for indicating

said change in amplitude, and bit 32identifies said polarity of said

change in amplitude.

8. The method of claim 1 further comprising the step of

dynamically compressing the total number of said plurality of digital messages necessary for reconstructingsaid analog waveform

wherein the information of at least two digital messages of said

plurality of digital messages is combined to form a final digital

message.

9. The method of claim 8 wherein said at least two digital

messages are sequential and wherein an elapsed time for each of

said at least two digital messages is compared, said at least two

digital messages being combined when said elapsed times are

substantially equivalent.

10. The method of claim 8 wherein said step of dynamically

compressing the total number of said plurality of digital messages

is initiated upon the detection of a digital message of said plurality

of digital messages having an elapsed time which is not

substantially equivalent to an elapsed time of an immediately

preceding digital message of said plurality of digital messages.

11. An apparatus for representing an analog waveform as

digital messages, each digital message formed upon the detection

of a certain minimum change in amplitude in said analog waveform

comprising;

means for establishing a threshold voltage; means coupled to said threshold voltage for detecting a

certain minimum change in the amplitude of said waveform, said

certain minimum change in amplitude having a value substantially

equivalent to said threshold voltage, said means for detecting

including a means for detecting the polarity of said certain

minimum change in amplitude, said detecting means producing a

signal indicating that said certain minimum change in amplitude

has occurred and identifying the polarity of said certain minimum

change in amplitude;

a means for measuring an elapsed time of each said detected

certain minimum change in amplitude;

a microprocessor coupled to said means for measuring an

elapsed time and said detecting means, said microprocessor being

responsive to outputs from said means for measuring an elapsed

time and said detecting means for generating said digital messages.

12. The apparatus of claim 11 wherein each of said digital

messages includes a field identifying said certain minimum change

in amplitude, a field identifying said polarity of said certain

minimum change in amplitude and a field identifying said elapsed

time of said certain minimum change in amplitude.

13. The apparatus of claim 1 1 wherein said detecting means

comprises a means for sampling and holding an amplitude value of

said waveform for use in detecting each subsequent certain

minimum change in amplitude in said analog waveform.

14. The apparatus of claim 13 wherein said detecting means

further comprises a pair of analog comparators coupled to said

sampling and holding means, each of said comparators comparing

said analog waveform against said amplitude value held in said

means for sampling and holding.

15. The apparatus of claim 11 wherein said means for

establishing a threshold voltage comprises a digitally controlled

voltage source.

16. The apparatus of claim 11 wherein said means for

measuring said elapsed time of said certain minimum change in

amplitude comprises a clock source coupled to three 8-bit timers.

17. The apparatus of claim 11 further comprising a memory

coupled to said microprocessor for storing said digital messages.

18. The apparatus of claim 11 further comprising a gain

stage having an output coupled to said detecting means, for

amplifying the amplitude of said waveform.

19. A method of compressively representing an analog

waveform as aggregate data words representing a combination of

substantially alike segments of said analog waveform, comprising

the steps of:

fixing a threshold voltage for use in triggering the formation

of a first plurality of data words;

detecting in said analog waveform each occurrence of a

certain minimum change in amplitude which is substantially

equivalent to said threshold voltage and detecting the polarity of

said certain minimum change in amplitude;

measuring an elapsed time corresponding to the duration of

said detected certain minimum change in amplitude;

generating said first plurality of data words, each data word

of said first plurality of data words having a data field

corresponding to said detected certain minimum change in

amplitude, a data field corresponding to said polarity of said

detected certain minimum change in amplitude and a data field

corresponding to said elapsed time of said detected certain

minimum change in amplitude

comparing the elapsed time of consecutive data words of said first plurality of data words:

combining said consecutive data words when said elapsed

times are substantially equivalent, each of said aggregate data

words being used to represent said combination of consecutive data

words having substantially equivalent elapsed times.

20. The method of claim 19 further comprising the step of

detecting in said waveform when a minimum change in amplitude

has not occurred within a predetermined duration of time and

generating a data word having information indicating that said

minimum change in amplitude has not occurred within said

predetermined duration of time.

21. The method of claim 20 wherein said duration of time

is determined by a byte size of said elapsed time data field.

22. The method of claim 19 further comprising the step of

detecting an occurrence of said certain minimum change in

amplitude immediately following a condition where substantially

no change in amplitude has occurred for a duration of time at least

as long as said detected occurrence of said certain minimum change

in amplitude.

23. The method of claim 22 further comprising the step of generating a data word indicating that no change in amplitude has

occurred for said duration of time immediately preceding said

detected occurrence of said certain minimum change in amplitude.

24. An apparatus for reproducing an analog waveform

which has been represented as a plurality of data words wherein

each of said data words includes a data field corresponding to a

certain change in amplitude, a data field corresponding to a polarity

for said certain change in amplitude and a data field corresponding

to an elapsed time of said certain change in amplitude comprising:

a microprocessor for retrieving each of a plurality of data

words from a storage medium;

a means for generating a reference clock coupled to said

microprocessor;

a voltage source coupled to said microprocessor and

responsive thereto for outputting said analog waveform in

accordance with the information in said data fields.

25. The apparatus of claim 24 wherein said voltage source

comprises a digitally controlled voltage source.

26. The apparatus of claim 24 wherein said means for

generating a reference clock comprises a clock source coupled to three 8-bit timers.

27. A method of recreating an analog waveform from a

series of stored digital data words encoded with information

representing a certain change in amplitude of said analog

waveform, said certain change in amplitude being represented as a

multiplier of a threshold voltage used to trigger the writing of each

of said stored digital words in an encoding operation, said stored

digital words further encoded with information corresponding to

the total elapsed time for said certain change in amplitude and the

polarity of said certain change in amplitude, comprising the steps

of:

sequentially retrieving each of a series of stored digital data

words;

deriving for each of said retrieved digital data words a full

voltage range for a certain change in amplitude;

calculating increments of voltage change for said certain

change in amplitude relative to said total elapsed time for each said

retrieved digital data word;

applying each saidcalculatedincremental voltage change for

each said retrieved digital data word to a digitally controlled voltage source for outputting said waveform;

outputting said waveform in accordance with each said

calculated incremental voltage change for each retrieved data word.

28. The method of claim 27 wherein said full voltage range

is derived by multiplying said multiplier by said threshold voltage.

29. The method of claim 27 further comprising the step of

dividing said full voltage range for each retrieved digital data word

by a maximum resolution of one increment of said digitally

controlled voltage source, the resultant value giving a number of

step divisions for a smooth output transition over time in

accordance with said maximum resolution of said digitally

controlled voltage source.

30. The method of claim 29 wherein said total elapsed time

of each of said retrieved digital data word is divided by said step

divisions to create a plurality of new data words each representing

one step of said digitally controlled voltage source.

31. A method of recreating an analog waveform from a

series of stored digital data words encoded in an encoding operation

with information representing a certain change in amplitude of said

analog waveform, said certain change in amplitude being represented as a multiplier of a threshold voltage used to trigger the

writing of each of said stored digital words in the encoding

operation, said stored digital words further encoded with

information corresponding to the total elapsed time for said certain

change in amplitude and the polarity of said certain change in

amplitude, comprising the steps of:

sequentially retrieving each of a series of stored digital data

words;

deriving a full voltage range for a certain change in

amplitude for each of said retrieved digital data words by

multiplying a multiplier by said threshold voltage;

generating for each of said retrieved digital data word a

plurality of new data words, each new data word representing an

incremental change to be applied to a digital voltage source for

outputting said analog waveform;

counting down an elapsed time for each said new data word

and upon said count down decrementing or incrementing the output

level of said digital voltage source.

32. The method of claim 1 wherein said plurality of new

data words are generated based on a division of said full voltage range by a value corresponding to the maximum resolution of one

increment of said digital voltage source, said resultant value giving

a number of step divisions.

33. The method of claim 32 wherein said step divisions are

divided into said total elapsed time for said certain change in

amplitude, said resultant value giving the elapsed time for each data

word of said plurality of new data words.

34. The method of claim 31 wherein said digital voltage

source comprises a digital potentiometer.