Method of producing a two-dimensional probability density function (PDF) eye diagram and Bit Error Rate eye arrays Number:7,522,661 from the United States Patent and Trademark Office (PTO) owispatent A method of producing of a two-dimensional probability density function eye diagram and Bit Error Rate eye arrays generates a two-dimensional PDF array of a correlated waveform record of a data pattern under test which is convolved with a two-dimensional probability density function (PDF) array of the uncorrelated jitter and noise in the data pattern under test. The resulting aggregate two dimensional PDF array of the correlated waveform record pattern with uncorrelated jitter and noise is divided into unit intervals and the unit intervals are summed to generated a two-dimensional PDF eye diagram array. The two-dimensional PDF eye diagram is processed to generate a two-dimensional Bit Error Rate eye array.<H probability, dimensional, Method, of, produ, 7522661.html, Patent, pto, 7522661

Title: Method of producing a two-dimensional probability density function (PDF) eye diagram and Bit Error Rate eye arrays

Abstract: A method of producing of a two-dimensional probability density function eye diagram and Bit Error Rate eye arrays generates a two-dimensional PDF array of a correlated waveform record of a data pattern under test which is convolved with a two-dimensional probability density function (PDF) array of the uncorrelated jitter and noise in the data pattern under test. The resulting aggregate two dimensional PDF array of the correlated waveform record pattern with uncorrelated jitter and noise is divided into unit intervals and the unit intervals are summed to generated a two-dimensional PDF eye diagram array. The two-dimensional PDF eye diagram is processed to generate a two-dimensional Bit Error Rate eye array.

Patent Number: 7,522,661 Issued on 04/21/2009 to Nelson, et al.

Inventors: Nelson; Michael A. (Lake Oswego, OR), Carlson; John E. (Portland, OR)
Assignee: Tektronix, Inc. (Beaverton, OR)

Appl. No.: 11/189,955

Filed: July 25, 2005

Related U.S. Patent Documents


Application Number	Filing Date	Patent Number	Issue Date
60591357	Jul., 2004

Current U.S. Class: 375/224 ; 375/150; 375/225; 375/226; 375/228; 375/254; 702/181; 702/66; 702/69

Current International Class: H04B 3/46 (20060101)

Field of Search: 375/224,225,226-228,254,285,296,150,317,371 702/69-70,181,191,66 703/2

References Cited [Referenced By]

U.S. Patent Documents


6275959	August 2001	Ransijn
7275004	September 2007	Casper et al.
7426233	September 2008	Richards et al.
2004/0268190	December 2004	Kossel et al.
2005/0031029	February 2005	Yamaguchi et al.
2005/0111536	May 2005	Cranford et al.
2006/0167652	July 2006	Stimple
2007/0171998	July 2007	Hietala et al.

Primary Examiner: Liu; Shuwang
Assistant Examiner: Singh; Hirdepal
Attorney, Agent or Firm: Bucher; William K.

Parent Case Text

CROSS-REFERENCE TO RELATED APPLICATION

This application claims the benefit of the U.S. Provisional Application No. 60/591,357, filed Jul. 26, 2004.

Claims

What is claimed is:

1. A method of producing a two-dimensional probability density function eye diagram array, for use in a test and measurement instrument comprising the steps of: a) acquiring a correlated waveform record of a data pattern under test containing data samples having horizontal locations and associated vertical amplitudes, said correlated waveform record being acquired via an acquisition portion of said test and measurement instrument; b) calculating a probability density function integration value for the correlated waveform record data pattern; c) dividing the correlated waveform record data pattern into unit intervals; d) defining a common array for each unit interval; e) locating each data sample of the correlated waveform record data pattern within the unit interval arrays; and f) entering the probability density function integration value at each data sample location in the unit interval arrays, g) generating a two-dimensional probability density function array of the correlated waveform record data pattern; h) acquiring an uncorrelated vertical and horizontal waveform records of the data pattern under test at respective zero slope and maximum slope locations in the correlated waveform record data pattern; i) generating a one-dimensional uncorrelated vertical probability density function array of uncorrelated vertical signal content in the uncorrelated vertical waveform record representing the noise in the data pattern under test; j) generating a one-dimensional uncorrelated horizontal probability density function array of uncorrelated horizontal signal content in the uncorrelated horizontal waveform record representing the jitter in the data pattern under test; k) multiplying the one-dimensional uncorrelated horizontal probability density function array by the one-dimensional uncorrelated vertical probability density function array to generate a two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test; l) convolving the two-dimensional probability density function array of the correlated waveform record data pattern with the two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test to generate an aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise; and m) dividing the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise into unit intervals and summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise to generate a two-dimensional probability density function eye diagram array.

2. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 1 wherein the acquiring of the uncorrelated vertical and horizontal waveform records step further comprises the steps of: a) determining the zero slope location in the correlated waveform record data pattern as the minimum of the derivative of the correlated waveform record data pattern; b) determining the maximum slope location in the correlated waveform record data pattern as the maximum of the derivative of the correlated waveform record data pattern; c) periodically sampling the zero slope location in the correlated waveform record data pattern to acquiring multiple sample records for the uncorrelated vertical waveform record of the data pattern under test; and d) periodically sampling the maximum slope location in the correlated waveform record data pattern to acquiring multiple sample records for the uncorrelated horizontal waveform record of the data pattern under test.

3. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 2 wherein the generating of a one-dimensional uncorrelated vertical probability density function array step further comprises the steps of: a) generating a spectra of the vertical signal content in the uncorrelated vertical waveform record of the data pattern under test by performing a Fast Fourier Transform on each acquired multiple sample record wherein the spectra contains random and periodic components of the uncorrelated vertical waveform record of the data pattern under test; and b) integrating the vertical signal content spectra containing the random components of the uncorrelated vertical waveform record of the data pattern under test to determine a standard deviation of the random components representing the random noise in the in the data pattern under test.

4. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 3 wherein the integrating of the vertical signal content spectra containing the random components of the uncorrelated vertical waveform record of the data pattern under test further comprises the steps of: a) identifying the vertical signal content of the periodic components in the spectra of the uncorrelated vertical waveform record of the data pattern under test; and b) removing the vertical signal content of the periodic components in the spectra of the uncorrelated vertical waveform record of the data pattern under test.

5. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 3 wherein the generating of a one-dimensional uncorrelated horizontal probability density function array step further comprises the steps of: a) determining a slew rate for the maximum slope location in the correlated waveform record data pattern; b) converting each amplitude value of the acquired multiple sample records for the uncorrelated horizontal waveform record of the data pattern under test into a time value as a function of the amplitude value divided by the slew rate; c) generating a spectra of the horizontal signal content of the uncorrelated horizontal waveform record of the data pattern under test by performing a Fast Fourier Transform on the converted time values of each acquired multiple sample record wherein the spectra contains random and periodic components of the uncorrelated horizontal waveform record of the data pattern under test; d) integrating the horizontal signal content spectra containing the random components of the uncorrelated horizontal waveform record of the data pattern under test to determine a standard deviation of the random components representing the observed random jitter in the in the data pattern under test; e) determining random jitter in the in the data pattern under test as a function of the observed random jitter, the random noise and the slew rate of the maximum slope location in the correlated waveform record data pattern using the formula: ##EQU00004##

6. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 5 wherein the determining of a slew rate for the maximum slope location in the correlated waveform record data pattern step further comprises the step of calculating a straight line through the acquired multiple record samples for the uncorrelated horizontal waveform record of the data pattern under test.

7. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 5 wherein the integrating of the horizontal signal content spectra containing the random components of the uncorrelated horizontal waveform record of the data pattern under test further comprises the steps of: a) identifying the horizontal signal content of the periodic components in the spectra of the uncorrelated horizontal waveform record of the data pattern under test; and b) removing the horizontal signal content of the periodic components in the spectra of the uncorrelated horizontal waveform record of the data pattern under test.

8. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 5 further comprising the steps of: a) comparing the periodic components in the spectra of the uncorrelated horizontal waveform record to the periodic components in the spectra of the uncorrelated vertical waveform record to identify periodic vertical components in the spectra of the uncorrelated horizontal waveform record; b) removing the identified periodic vertical components in the spectra of the uncorrelated horizontal waveform record from the spectra of the uncorrelated horizontal waveform record; c) generating an amplitude voltage waveform of the vertical signal content spectra containing the periodic components of the uncorrelated vertical waveform record of the data pattern under test by performing an Inverse Fast Fourier Transform on the vertical signal content of the periodic component spectra; d) generating a histogram of the amplitude voltage waveform of the vertical signal content of the periodic component spectra representing the periodic noise in the data pattern under test; e) convolving a Gaussian distribution function having the calculated standard deviation of the random noise and the histogram of the periodic noise to generate the one-dimensional uncorrelated vertical probability density function array of the uncorrelated vertical signal content in the uncorrelated vertical waveform record representing the noise in the data pattern under test; f) generating an amplitude voltage waveform of the horizontal signal content spectra containing the periodic components of the uncorrelated horizontal waveform record of the data pattern under test by performing an Inverse Fast Fourier Transform on the horizontal signal content of the periodic component spectra; g) generating a histogram of the amplitude voltage waveform of the horizontal signal content of the periodic component spectra representing the periodic jitter in the data pattern under test; and h) convolving a Gaussian distribution function having the calculated standard deviation of the random jitter and the histogram of the periodic jitter to generate the one-dimensional uncorrelated horizontal probability density function array of the uncorrelated horizontal waveform record representing the jitter in the data pattern under test.

9. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 1 further comprising the steps of: a) defining a center coordinate in the two-dimensional probability density function eye diagram array; b) defining integrations paths between the center coordinate and each element of the two-dimensional probability density function eye diagram array; c) summing the element values of the two-dimensional probability function eye diagram array intersecting each of the defined integration paths; d) assigning the summed element value for each of the defined integration paths to the respective element of the two-dimensional probability density function eye diagram defining the integration path from the center coordinate; and e) generating a Bit Error Rate eye array from the summed element values.

10. A method of producing a two-dimensional probability density function eye diagram array for use in a test and measurement instrument, comprising the steps of: a) acquiring a correlated waveform record of a data pattern under test, said correlated waveform being acquired by an acquisition portion of said test and measurement instrument; b) dividing the correlated waveform record data pattern into unit intervals; c) generating a one-dimensional array of ideal 1's and 0's logical values for the unit intervals of the correlated waveform record; d) generating a two-dimensional probability density function array of the correlated waveform record data pattern divided into unit intervals; e) acquiring an uncorrelated vertical and horizontal waveform records of the data pattern under test at respective zero slope and maximum slope locations in the correlated waveform record data pattern; f) generating a one-dimensional uncorrelated vertical probability density function array of uncorrelated vertical signal content in the uncorrelated vertical waveform record representing the noise in the data pattern under test; g) generating a one-dimensional uncorrelated horizontal probability density function array of uncorrelated horizontal signal content in the uncorrelated horizontal waveform record representing the jitter in the data pattern under test; h) multiplying the one-dimensional uncorrelated horizontal probability density function array by the one-dimensional uncorrelated vertical probability density function array to generate a two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test; i) convolving the two-dimensional probability density function array of the correlated waveform record data pattern with the two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test to generate an aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise; j) correlating the one-dimensional array of ideal 1's and 0's logical values for the unit intervals of the correlated waveform record with the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise; k) summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise correlated with the ideal 1's logic values to generate a two-dimensional 1's probability density function eye array; l) summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise correlated with the ideal 0's logic values to generate a two-dimensional 0's probability density function eye array; and m) adding the two-dimensional 1's probability density function eye array with the two-dimensional 0's probability density function eye array to generate a two-dimensional probability density function eye diagram array.

11. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 10 wherein the correlated waveform record data pattern contains data samples having horizontal locations and associated vertical amplitudes and the generating of the two-dimensional probability density function of the correlated waveform record data pattern step further comprises the steps of: a) calculate a probability density function integration value for the correlated waveform record data pattern; b) define a common array for each unit interval of the correlated waveform record data pattern; c) locate each data sample of the correlated waveform record data pattern within the unit interval arrays; and d) enter the probability density function integration value at each data sample location in the unit interval arrays.

12. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 10 wherein the acquiring of the uncorrelated vertical and horizontal waveform records step further comprises the steps of: a) determining the zero slope location in the correlated waveform record data pattern as the minimum of the derivative of the correlated waveform record data pattern; b) determining the maximum slope location in the correlated waveform record data pattern as the maximum of the derivative of the correlated waveform record data pattern; c) periodically sampling the zero slope location in the correlated waveform record data pattern to acquiring multiple sample records for the uncorrelated vertical waveform record of the data pattern under test; and d) periodically sampling the maximum slope location in the correlated waveform record data pattern to acquiring multiple sample records for the uncorrelated horizontal waveform record of the data pattern under test.

13. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 12 wherein the generating of a one-dimensional uncorrelated vertical probability density function array step further comprises the steps of: a) generating a spectra of the vertical signal content in the uncorrelated vertical waveform record of the data pattern under test by performing a Fast Fourier Transform on each acquired multiple sample record wherein the spectra contains random and periodic components of the uncorrelated vertical waveform record of the data pattern under test; and b) integrating the vertical signal content spectra containing the random components of the uncorrelated vertical waveform record of the data pattern under test to determine a standard deviation of the random components representing the random noise in the in the data pattern under test.

14. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 13 wherein the integrating of the vertical signal content spectra containing the random components of the uncorrelated vertical waveform record of the data pattern under test further comprises the steps of: a) identifying the vertical signal content of the periodic components in the spectra of the uncorrelated vertical waveform record of the data pattern under test; and b) removing the vertical signal content of the periodic components in the spectra of the uncorrelated vertical waveform record of the data pattern under test.

15. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 13 wherein the generating of a one-dimensional uncorrelated horizontal probability density function array step further comprises the steps of: a) determining a slew rate for the maximum slope location in the correlated waveform record data pattern; b) converting each amplitude value of the acquired multiple sample records for the uncorrelated horizontal waveform record of the data pattern under test into a time value as a function of the amplitude value divided by the slew rate; c) generating a spectra of the horizontal signal content of the uncorrelated horizontal waveform record of the data pattern under test by performing a Fast Fourier Transform on the converted time values of each acquired multiple sample record wherein the spectra contains random and periodic components of the uncorrelated horizontal waveform record of the data pattern under test; d) integrating the horizontal signal content spectra containing the random components of the uncorrelated horizontal waveform record of the data pattern under test to determine a standard deviation of the random components representing the observed random jitter in the in the data pattern under test; e) determining random jitter in the in the data pattern under test as a function of the observed random jitter, the random noise and the slew rate of the maximum slope location in the correlated waveform record data pattern using the formula: ##EQU00005##

16. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 15 wherein the determining of a slew rate for the maximum slope location in the correlated waveform record data pattern step further comprises the step of calculating a straight line through the acquired multiple record samples for the uncorrelated horizontal waveform record of the data pattern under test.

17. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 15 wherein the integrating of the horizontal signal content spectra containing the random components of the uncorrelated horizontal waveform record of the data pattern under test further comprises the steps of: a) identifying the horizontal signal content of the periodic components in the spectra of the uncorrelated horizontal waveform record of the data pattern under test; and b) removing the horizontal signal content of the periodic components in the spectra of the uncorrelated horizontal waveform record of the data pattern under test.

18. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 15 further comprising the steps of: a) comparing the periodic components in the spectra of the uncorrelated horizontal waveform record to the periodic components in the spectra of the uncorrelated vertical waveform record to identify periodic vertical components in the spectra of the uncorrelated horizontal waveform record; b) removing the identified periodic vertical components in the spectra of the uncorrelated horizontal waveform record from the spectra of the uncorrelated horizontal waveform record; c) generating an amplitude voltage waveform of the vertical signal content spectra containing the periodic components of the uncorrelated vertical waveform record of the data pattern under test by performing an Inverse Fast Fourier Transform on the vertical signal content of the periodic component spectra; d) generating a histogram of the amplitude voltage waveform of the vertical signal content of the periodic component spectra representing the periodic noise in the data pattern under test; e) convolving a Gaussian distribution function having the calculated standard deviation of the random noise and the histogram of the periodic noise to generate the one-dimensional uncorrelated vertical probability density function array of the uncorrelated vertical signal content in the uncorrelated vertical waveform record representing the noise in the data pattern under test; f) generating an amplitude voltage waveform of the horizontal signal content spectra containing the periodic components of the uncorrelated horizontal waveform record of the data pattern under test by performing an Inverse Fast Fourier Transform on the horizontal signal content of the periodic component spectra; g) generating a histogram of the amplitude voltage waveform of the horizontal signal content of the periodic component spectra representing the periodic jitter in the data pattern under test; and h) convolving a Gaussian distribution function having the calculated standard deviation of the random jitter and the histogram of the periodic jitter to generate the one-dimensional uncorrelated horizontal probability density function array of the uncorrelated horizontal waveform record representing the jitter in the data pattern under test.

19. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 10 further comprising the steps of: a) defining a center coordinate in the two-dimensional probability density function eye diagram array; b) defining integrations paths between the center coordinate and each element of the two-dimensional probability density function eye diagram array; c) summing the element values of the two-dimensional probability function eye diagram array intersecting each of the defined integration paths; d) assigning the summed element value for each of the defined integration paths to the respective element of the two-dimensional probability density function eye diagram defining the integration path from the center coordinate; and e) generating a Bit Error Rate eye array from the summed element values.

20. A method of producing a two-dimensional Bit Error Rate eye array for use in a test and measurement instrument, comprising the steps of: a) acquiring a correlated waveform record of a data pattern under test, said correlated waveform being acquired using an acquisition portion of said test and measurement instrument; b) dividing the correlated waveform record data pattern into unit intervals; c) generating a one-dimensional array of ideal 1's and 0's logical values for the unit intervals of the correlated waveform record; d) generating a two-dimensional probability density function array of the correlated waveform record data pattern divided into unit intervals; e) acquiring an uncorrelated vertical and horizontal waveform records of the data pattern under test at respective zero slope and maximum slope locations in the correlated waveform record data pattern; f) generating a one-dimensional uncorrelated vertical probability density function array of uncorrelated vertical signal content in the uncorrelated vertical waveform record representing the noise in the data pattern under test; g) generating a one-dimensional uncorrelated horizontal probability density function array of uncorrelated horizontal signal content in the uncorrelated horizontal waveform record representing the jitter in the data pattern under test; h) multiplying the one-dimensional uncorrelated horizontal probability density function array by the one-dimensional uncorrelated vertical probability density function array to generate a two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test; i) convolving the two-dimensional probability density function array of the correlated waveform record data pattern with the two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test to generate an aggregate two-dimensional probability density function of the correlated waveform record pattern with uncorrelated jitter and noise; j) correlating the one-dimensional array of ideal 1's and 0's logical values for the unit intervals of the correlated waveform record with the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise; k) summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise correlated with the ideal 1's logic values to generate a two-dimensional 1's probability density function eye array; l) summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise correlated with the ideal 0's logic values to generate a two-dimensional 0's probability density function eye array; m) integrating the two-dimensional 1's probability density function eye array upward and downward to generate two-dimensional 1's upward and downward integration arrays; n) integrating the two-dimensional 0's probability density function eye array downward and upward to generate two-dimensional 0's downward and upward integration arrays; o) adding the two-dimensional 1's upward integration array with the two-dimensional 0's downward integration array; p) adding the two-dimensional 1's upward and downward integration arrays with the two-dimensional 0's downward and upward integration arrays; and q) dividing each element of the added the two-dimensional 1's upward integration array and two-dimensional 0's downward integration array by the corresponding element of the two-dimensional added 1's upward and downward integration arrays and the two-dimensional 0's downward and upward integration arrays to generate a two-dimensional Bit Error Rate eye array.

21. A method of producing a two-dimensional probability density function eye diagram array for use in a test and measurement instrument, comprising the steps of: a) acquiring a correlated waveform record of a data pattern under test, said correlated waveform being acquired by use of an acquisition portion of said test and measurement instrument; b) dividing the correlated waveform record data pattern into unit intervals; c) generating a one-dimensional array of ideal 1's and 0's logical values for the unit intervals of the correlated waveform record; d) generating a two-dimensional probability density function array of the correlated waveform record data pattern; e) determining the maximum slope location in the correlated waveform record data pattern as the maximum of the derivative of the correlated waveform record data pattern; and f) determining additional slope locations as a percentage of the maximum slope location; g) periodically sampling the slope locations in the correlated waveform record data pattern to acquiring multiple sample records at each slope location for the uncorrelated horizontal waveform record of the data pattern under test; h) determining a slew rate for each of the slope locations in the correlated waveform record data pattern using the multiple sample records for each slope location; i) converting each amplitude value of the acquired multiple sample records for each slope location of the uncorrelated horizontal waveform record of the data pattern under test into time a value as a function of the amplitude value divided by the slew rate; j) generating a spectra of the horizontal signal content of the uncorrelated horizontal waveform record of the data pattern under test at each slope location by performing a Fast Fourier Transform on the converted time values of each acquired multiple sample record wherein the spectra contains random and periodic components of the uncorrelated horizontal waveform record of the data pattern under test; k) integrating the horizontal signal content spectra containing the random components of the uncorrelated horizontal waveform record of the data pattern under test at each slope location to determine standard deviations of the random components representing the observed random jitter in the in the data pattern under test at each slope location; l) determining random jitter and random noise in the in the data pattern under test as a function of the observed random jitters and the slew rates at the slope locations in the correlated waveform record data pattern using the formula: .times..times. ##EQU00006## m) determining periodic jitter and periodic noise in the data pattern under test as a function of the observed random jitters and slew rates at the slope location in the correlated waveform record data pattern using the formula: .times..times. ##EQU00007## n) generating a one-dimensional uncorrelated vertical probability density function array from the calculated random and periodic noise representing the noise in the data pattern under test; p) generating a one-dimensional uncorrelated horizontal probability density function array from the calculated random and periodic jitter representing the jitter in the data pattern under test; q) multiplying the one-dimensional uncorrelated horizontal probability density function array by the one-dimensional uncorrelated vertical probability density function array to generate a two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test; r) convolving the two-dimensional probability density function array of the correlated waveform record data pattern with the two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test to generate an aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise; s) correlating the one-dimensional array of ideal 1's and 0's logical values for the unit intervals of the correlated waveform record with the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise; t) summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise correlated with the ideal 1's logic values to generate a two-dimensional 1's probability density function eye array; u) summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise correlated with the ideal 0's logic values to generate a two-dimensional 0's probability density function eye array; and v) adding the two-dimensional 1's probability density function eye array with the two-dimensional 0's probability density function eye array to generate a two-dimensional probability density function eye diagram array.

22. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 21 further comprising the steps of: a) defining a center coordinate element in the two-dimensional probability density function eye diagram array; b) defining integrations paths between the center coordinate and each element of the two-dimensional probability density function eye diagram array; c) summing the element values of the two-dimensional probability function eye diagram array intersecting each of the defined integration paths; d) assigning the summed element value for each of the defined integration paths to the respective elements of the two-dimensional probability density function eye diagram defining the integration path from the center coordinate; and e) generating a Bit Error Rate eye array from the summed element values.

23. A method of producing a two-dimensional probability density function eye diagram, for use in a test and measurement instrument, comprising the steps of: a) acquiring a correlated waveform record of a data pattern under test, said correlated waveform being acquired via an acquisition portion of said test and measurement instrument; b) dividing the correlated waveform record data pattern into unit intervals; c) generating a one-dimensional array of ideal 1's and 0's logical values for the unit intervals of the correlated waveform record; d) generating a two-dimensional probability density function array of the correlated waveform record data pattern; e) determining the maximum slope location in the correlated waveform record data pattern as the maximum of the derivative of the correlated waveform record data pattern; and f) determining additional slope locations as a percentage of the maximum slope location; g) periodically sampling the slope locations in the correlated waveform record data pattern to acquiring multiple sample records at each slope location for the uncorrelated horizontal waveform record of the data pattern under test; h) determining a slew rate for each of the slope locations in the correlated waveform record data pattern using the multiple sample records for each slope location; i) converting each amplitude value of the acquired multiple sample records for each slope location of the uncorrelated horizontal waveform record of the data pattern under test into time a value as a function of the amplitude value divided by the slew rate; j) generating a spectra of the horizontal signal content of the uncorrelated horizontal waveform record of the data pattern under test at each slope location by performing a Fast Fourier Transform on the converted time values of each acquired multiple sample record wherein the spectra contains random and periodic components of the uncorrelated horizontal waveform record of the data pattern under test; k) integrating the horizontal signal content spectra containing the random components of the uncorrelated horizontal waveform record of the data pattern under test at each slope location to determine standard deviations of the random components representing the observed random jitter in the in the data pattern under test at each slope location; l) determining random jitter and random noise in the in the data pattern under test as a function of the observed random jitters and the slew rates at the slope locations in the correlated waveform record data pattern using the formula: .times..times. ##EQU00008## m) determining periodic jitter and periodic noise in the data pattern under test as a function of the observed random jitters and slew rates at the slope location in the correlated waveform record data pattern using the formula: .times..times. ##EQU00009## n) generating a one-dimensional uncorrelated vertical probability density function array from the calculated random and periodic noise representing the noise in the data pattern under test; p) generating a one-dimensional uncorrelated horizontal probability density function array from the calculated random and periodic jitter representing the jitter in the data pattern under test; q) multiplying the one-dimensional uncorrelated horizontal probability density function array by the one-dimensional uncorrelated vertical probability density function array to generate a two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test; r) convolving the two-dimensional probability density function array of the correlated waveform record data pattern with the two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test to generate an aggregate two- dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise; s) correlating the one-dimensional array of ideal 1's and 0's logical values for the unit intervals of the correlated waveform record with the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise; t) summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise correlated with the ideal 1's logic values to generate a two-dimensional 1's probability density function eye array; u) summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise correlated with the ideal 0's logic values to generate a two-dimensional 0's probability density function eye array; v) integrating the two-dimensional 1's probability density function eye array upward and downward to generate two-dimensional 1's upward and downward integration arrays; w) integrating the two-dimensional 0's probability density function eye array downward and upward to generate two-dimensional 0's downward and upward integration arrays; x) adding the two-dimensional 1's upward integration array with the two-dimensional 0's downward integration array; y) adding the two-dimensional 1's upward and downward integration arrays with the two-dimensional 0's downward and upward integration arrays; and z) dividing each element of the added the two-dimensional 1's upward integration array and two-dimensional 0's downward integration array by the corresponding element of the added two-dimensional 1's upward and downward integration arrays and the two-dimensional 0's downward and upward integration arrays to generate a two-dimensional Bit Error Rate eye array.

24. A method of producing a two-dimensional probability density function eye diagram, for use in a test and measurement instrument, comprising the steps of: a) acquiring a correlated waveform record of a data pattern under test, said correlated waveform being acquired using an acquisition portion of said test and measurement instrument; b) generating a two-dimensional probability density function array of the correlated waveform record data pattern; c) determining the maximum slope location in the correlated waveform record data pattern as the maximum of the derivative of the correlated waveform record data pattern; and d) determining additional slope locations as a percentage of the maximum slope location; e) periodically sampling the slope locations in the correlated waveform record data pattern to acquiring multiple sample records at each slope location for the uncorrelated horizontal waveform record of the data pattern under test; f) determining a slew rate for each of the slope locations in the correlated waveform record data pattern using the multiple sample records for each slope location; g) converting each amplitude value of the acquired multiple sample records for each slope location of the uncorrelated horizontal waveform record of the data pattern under test into time a value as a function of the amplitude value divided by the slew rate; h) generating a spectra of the horizontal signal content of the uncorrelated horizontal waveform record of the data pattern under test at each slope location by performing a Fast Fourier Transform on the converted time values of each acquired multiple sample record wherein the spectra contains random and periodic components of the uncorrelated horizontal waveform record of the data pattern under test; i) integrating the horizontal signal content spectra containing the random components of the uncorrelated horizontal waveform record of the data pattern under test at each slope location to determine standard deviations of the random components representing the observed random jitter in the in the data pattern under test at each slope location; j) determining random jitter and random noise in the in the data pattern under test as a function of the observed random jitters and the slew rates at the slope locations in the correlated waveform record data pattern using the formula: .times..times. ##EQU00010## k) determining periodic jitter and periodic noise in the data pattern under test as a function of the observed random jitters and slew rates at the slope location in the correlated waveform record data pattern using the formula: ##EQU00011## l) generating a one-dimensional uncorrelated vertical probability density function array from the calculated random and periodic noise representing the noise in the data pattern under test; m) generating a one-dimensional uncorrelated horizontal probability density function array from the calculated random and periodic jitter representing the jitter in the data pattern under test; n) multiplying the one-dimensional uncorrelated horizontal probability density function array by the one-dimensional uncorrelated vertical probability density function array to generate a two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test; o) convolving the two-dimensional probability density function array of the correlated waveform record data pattern with the two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test to generate an aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise; and p) dividing the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise into unit intervals and summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise to generate a two-dimensional probability density function eye diagram array.

25. The method of producing a two-dimensional probability density function eye diagram array as recited in claim 24 further comprising the steps of: a) defining a center coordinate element in the two-dimensional probability density function eye diagram array; b) defining integrations paths between the center coordinate and each element of the two-dimensional probability density function eye diagram array; c) summing the element values of the two-dimensional probability function eye diagram array intersecting each of the defined integration paths; d) assigning the summed element value for each of the defined integration paths to the respective elements of the two-dimensional probability density function eye diagram defining the integration path from the center coordinate; and e) generating a Bit Error Rate eye array from the summed element values.

Description

BACKGROUND OF THE INVENTION

The present invention relates generally to the generation of eye and bit error rate diagrams and more particularly to producing a two-dimensional probability density function (PDF) eye diagram and a Bit Error Rate eye arrays for display on a display device.

Digital sampling oscilloscopes acquire waveform records of data patterns under test to generate and display eye diagrams representing unit intervals of the data pattern. The digital sampling oscilloscope has sampling circuitry that acquires very accurate data samples of the data pattern under test. However, the eye diagrams are approximations of the probability of any particular data sample being at a particular location due to the limited number of samples acquired by the sampling oscilloscope. These approximation places limitations on the accuracy of the measurement analysis of the eye diagram. The generation of a Bit Error Rate eye from eye diagram data has limited accuracy. To increase the accuracy of the eye diagram and the resulting measurement analysis requires the acquisition of large quantities of data samples of the data pattern under test. Such acquisitions would take days and months of continuous acquisitions.

A Bit Error Rate Tester (BERT) is another type of measurement instrument that acquires data samples of a data pattern under test. The BERT has specialized circuitry that can acquire extremely large records (giga-samples of data) over a relatively short period of time to generate eye diagrams and Bit Error Rate eyes. However, there are tradeoffs between the extremely fast acquisition of data samples and data sample accuracy. The acquisition circuits in the BERT have inferior flatness and frequency response resulting in acquired data samples having inaccurate amplitude levels. These inaccuracies are translated into eye diagrams and Bit Error Rate eyes. The resulting measurement analysis of the eye diagram and Bit Error Rate eye will have these inaccuracies.

What is needed is a method of producing highly accurate two-dimensional probability (PDF) eye diagram and Bit Error Rate eye arrays from a limited number of samples from an oscilloscope or other type of digitizing instrument. The method should accurately represent the data pattern under test and the vertical and horizontal uncertainties (i.e. noise and jitter) in the data pattern. The vertical and horizontal uncertainties should be a statistically accurate representation of these uncertainties. The statistically accurate vertical and horizontal uncertainties should be combinable with the accurate representation of the data pattern under test to produce the accurate two-dimensional probability (PDF) eye diagram and Bit Error Rate eye arrays for displaying as a PDF eye diagram and Bit Error Rate eye.

SUMMARY OF THE INVENTION

Accordingly, the present invention is a method of producing a probability density function eye diagram and a Bit Error Rate eye arrays. The method convolves a two-dimensional PDF array of a correlated waveform record of a data pattern under test and a statistically correct two-dimensional uncorrelated PDF array of the jitter and noise in the data pattern under test. The two-dimensional PDF array of correlated waveform record data pattern is generated by acquiring a correlated waveform record and dividing the record into array elements defined by the unit interval of the data pattern. A one-dimensional array of ideal 1's and 0's logic levels is generated for the unit intervals of the correlated waveform record. Locations having a maximum and zero slope is determined for the correlated waveform record using the maximum and minimum values of the derivative of the correlated waveform record pattern. Multiple sample records are acquired at the zero slope and the maximum slope locations. A Fast Fourier Transform is applied to each of the multiple sample records of the vertical sample records to generate a vertical spectra of the vertical signal content of the uncorrelated vertical waveform records. The periodic components of the vertical signal content are identified from the vertical spectra and removed and the resulting spectra is integrated to determine the standard deviation which represents the random noise in the data pattern under test.

The slew rate of the maximum slope is determined using the multiple samples records acquired at the maximum slope. The amplitude values in each of the multiple sample records are converted to time values and a Fast Fourier Transform is applied to converted time values for each multiple sample record to generate a horizontal spectra of the horizontal signal content of the uncorrelated horizontal waveform records. The periodic components of the horizontal signal content are identified from the horizontal spectra and removed and the resulting spectra is integrated to determine the standard deviation which represents the observed random jitter in the data pattern under test. The true random jitter is determined as a function of the observed random jitter, random noise and the slew rate.

The periodic components in the spectra of the uncorrelated horizontal waveform record are compared to the periodic components in the spectra of the uncorrelated vertical waveform record to identify periodic vertical components in the spectra of the uncorrelated horizontal waveform record. The identified periodic vertical components in the spectra of the uncorrelated horizontal waveform record are removed from the spectra of the uncorrelated horizontal waveform record. An Inverse Fast Fourier Transform is applied to the horizontal signal content spectra containing the periodic components of the uncorrelated horizontal waveform record of the data pattern under test to generate an amplitude voltage waveform. A histogram is generated of the amplitude voltage waveform that represents the periodic jitter in the data pattern under test. The histogram of the periodic jitter is convolved with a Gaussian distribution function having the calculated standard deviation of the random jitter to generate the one-dimensional uncorrelated horizontal probability density function array of the uncorrelated horizontal waveform record representing the jitter in the data pattern under test.

An Inverse Fast Fourier Transform is applied to the vertical signal content spectra containing the periodic components of the uncorrelated vertical waveform record of the data pattern under test to generate an amplitude voltage waveform. A histogram is generated of the amplitude voltage waveform that represents the periodic noise in the data pattern under test. The histogram of the periodic noise is convolved with a Gaussian distribution function having the calculated standard deviation of the random noise to generate the one-dimensional uncorrelated vertical probability density function array of the uncorrelated vertical waveform record representing the noise in the data pattern under test.

The one-dimensional uncorrelated horizontal probability density function array and the one-dimensional uncorrelated vertical probability density function array are multiplied together to generate a two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test. The two-dimensional probability density function array of the correlated waveform record data pattern is convolved with the two-dimensional probability density function array of the uncorrelated jitter and noise in the data pattern under test to generate an aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise. The aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise is correlated with the one-dimensional array of ideal 1's and 0's logical levels for the unit intervals of the correlated waveform record. A two-dimensional probability density function 1's eye array is generated by summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform pattern with uncorrelated jitter and noise correlated with the ideal 1's logic values. A two-dimensional probability density function 0's eye array is generated by summing the unit intervals of the aggregate two-dimensional probability density function array of the correlated waveform pattern with uncorrelated jitter and noise correlated with the ideal 0's logic values. The two-dimensional probability density function 1's eye array and the two-dimensional probability density function 0's eye array are added together to generate a two dimensional probability density function eye diagram array. Alternately, the aggregate two-dimensional probability density function array of the correlated waveform record pattern with uncorrelated jitter and noise may be divided into unit intervals and the unit intervals summed together to generate a two-dimensional probability density function eye diagram array.

A Bit Error Rate eye array is generated using the two-dimensional probability density function 1's eye array and the two-dimensional probability density function 0's eye array. The two-dimensional probability density function 1's eye array is upward integrated and downward integrated to generate two-dimensional 1's upward and downward integration arrays. The two-dimensional probability density function 0's eye array is downward integrated and upward integrated to generate two-dimensional 0's downward and upward integration arrays. The two-dimensional 1's upward integration array is added to the two-dimensional 0's downward integration array and the two-dimensional 1's upward and downward integration arrays are added with the two-dimensional 0's downward and upward integration arrays. The added two-dimensional 1's upward integration array and two-dimensional 0's downward integration array is divided by the added 1's upward and downward integration arrays and the two-dimensional 0's downward and upward integration arrays to generate a two-dimensional Bit Error Rate eye array. Alternately, the aggregate two-dimensional probability density function eye diagram array can be radially integrated from the center of the PDF eye array to generate a Bit Error Rate eye array.

The objects, advantages and novel features of the present invention are apparent from the following detailed description when read in conjunction with appended claims and attached drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of a representative digitizing instrument implementing the method of producing of a two-dimensional probability density function (PDF) eye array and Bit Error Rate eye array according to the present invention.

FIG. 2 is a set-up screen for entering parameters for the digitizing instrument.

FIG. 3 is a flow chart representing the steps in producing the two-dimensional probability density function (PDF) eye diagram array according to the present invention.

FIG. 4 is a flow chart representing the steps in generating a PDF of the correlated waveform record data pattern.

FIG. 5 is a representation of the PDF of the correlated waveform record data pattern.

FIG. 6 is a flow chart representing the steps in determining the zero slope and the steepest rising edge location in the correlated waveform record data pattern.

FIG. 7 is a flow chart representing the steps in determining the standard deviation of the random noise components of the amplitude spectra of the uncorrelated noise.

FIG. 8 is a flow chart representing the steps in generating in determining the standard deviation of the random jitter components in the horizontal spectra of the uncorrelated jitter.

FIG. 9A and 9B are flow charts representing the steps in generating a two-dimensional PDF of the uncorrelated noise and jitter in the data pattern under test.

FIG. 10 is a visual representation of the two-dimensional PDF array of the uncorrelated noise and jitter in the data pattern under test.

FIG. 11 is a three-dimensional visual representation of the two-dimensional PDF array of the uncorrelated noise and jitter in the data pattern under test.

FIG. 12 is a visual representation of the convolving of the two-dimensional PDF array of the correlated waveform record data pattern with the two-dimensional PDF array of the uncorrelated noise and jitter in the data pattern under test.

FIG. 13 is a display of a PDF eye diagram generated from the aggregate two-dimensional PDF eye diagram array.

FIG. 14 is a display of a Bit Error Rate eye generated from the two-dimensional Bit Error Rate eye array according to the present invention.

FIG. 15 is a flow chart representing the step in generating the Bit Error Rate eye according to the present invention.

FIG. 16 is a visual representation of the effective polar integration paths associated with the two-dimensional PDF eye diagram array for generating the Bit Error Rate eye according to the present invention.

FIGS. 17A and 17B is a flow chart representing the steps in the preferred method of generating a two-dimensional PDF eye diagram array according to the present invention.

FIG. 18A is a representation of a one-dimensional array of logic levels of unit intervals in a correlated waveform record.

FIG. 18B is a representations of the correlated waveform record.

FIG. 18C is a representation of the aggregate two-dimensional PDF of the correlated waveform record with the uncorrelated noise and jitter.

FIG. 19A is a representation of a two-dimensional 1's PDF eye array.

FIG. 19B is a representation of a two-dimensional 0's PDF eye array.

FIG. 20 is simplified hardware circuit for detecting the logic levels of the test signal in a Bit Error Rate Tester.

FIG. 21 is a representation of a portion of the correlated waveform data used as a test signal and the resulting detected output levels of the hardware detecting circuit.

FIG. 22 is a probability function representing a Bit Error Rate eye.

FIG. 23A is a presentation of a portion of the correla