Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Quality evaluation
Reexamination Certificate
1999-06-28
2002-07-09
Assouad, Patrick (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Quality evaluation
C702S057000, C702S081000, C702S118000, C714S724000
Reexamination Certificate
active
06418387
ABSTRACT:
CROSS-REFERENCE TO RELATED APPLICATIONS
Not Applicable
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH
Not Applicable
REFERENCE TO MICROFICHE APPENDIX
Not Applicable
FIELD OF THE INVENTION
The present invention relates to methods of and systems for plotting data, and more particularly, to methods of and systems for efficiently plotting data in the form of N dimensional shmoo plots.
BACKGROUND OF THE INVENTION
It is often necessary to characterize the performance of an electrical device with respect to functional parameters associated with the device. This characterization is sometimes referred to as parametric testing. One way to measure the performance of such a device is on a pass/fail basis. A complete parametric test could provide a pass/fail result for every possible combination of all values of the functional parameters in question. This type of test provides users of the device information as to how the device will perform over a broad range of conditions.
For example, a computer memory board designer could utilize data regarding how a dynamic random access memory (hereinafter referred to as DRAM) device performs with respect to various functional parameters, e.g., the supply voltage (V
CC
) into the device, and the device read/write cycle time (i.e., the amount of time it takes to write to and subsequently read from the DRAM). One way to characterize the performance of such a device is to apply a simple read/write test as follows: if a series of test patterns can be written and subsequently read from the device, then the device is deemed to have passed the performance test; otherwise, the device is deemed to have failed the performance test. Assume that the nominal supply voltage of the device is 3.0V and the nominal read/write cycle time is 10 nS. A parametric test may be arranged to evaluate the device performance as V
CC
varies from 2.2V to 3.7V in steps of 0.1V (a total of 16 discrete voltages), and as the read/write cycle time varies from 4.5 to 12 nS in steps of 0.5 nS (for a total of 16 discrete cycle times). Thus, the total number of individual performance tests that must be conducted to evaluate all possible combinations of V
CC
and read/write cycle time is 16×16=256. Each individual combination of test parameters is referred to herein as a test condition.
FIG. 1
shows these 256 pass/fail results in what is known in the art as a ‘shmoo’ (or alternatively, schmoo, smoo, shmu, schmu, etc.) plot.
FIG. 1
is an example of a two dimensional shmoo plot, although in general, a shmoo plot may be N-dimensional, N being an integer greater than 1. Although the term ‘plot’ suggests a graphical depiction of the data (i.e., at most three dimensions), description the term ‘plot’ should be taken as a general representation of data in N-dimensional data space, memory space, etc. In the shmoo plot of
FIG. 1
, an array of 16 rows and 16 columns represents the 256 individual performance tests; each row represents one of the 16 cycle times, and each column represents one of the 16 V
CC
voltages. Thus, the array element in the i
th
row and j
th
column corresponds to the performance test T
ij
in which the i
th
cycle time and the j
th
V
CC
voltage are applied to the device. A darkened array element represents a failed test T
ij
, and a white array element represents a passed test T
ij
.
As
FIG. 1
illustrates, a shmoo plot provides a clear, graphical indication of the performance envelope of a device. An observer can easily comprehend where the device performance transitions from passing to failing as the input parameters vary. However, a significant amount of time is required to perform the tests that form all of the individual array elements of the shmoo plot. In the simple
FIG. 1
example, assuming that each individual test takes 100 &mgr;S to perform, the entire shmoo plot will take 16×16×100 &mgr;S=25.6 mS to complete. In a more realistic parametric test, each variable in the array might be swept over a range of 100 values, and each individual test could take on the order of a second to complete. A shmoo plot corresponding to such a parametric test would take 100×100×1 S=2.78 hours to complete. If the parametric dimensional space increases from two to three (e.g., assume the device is to be further tested over a temperature range of 100 temperature values), the shmoo plot corresponding to such a parametric test would take 100×100×100×1 S=11.6 days to complete. Thus, the amount of time to complete the shmoo plot increases directly with the amount of time per test, and exponentially with the order of the dimensional space.
It is an object of the present invention to substantially overcome the above-identified disadvantages and drawbacks of the prior art.
SUMMARY OF THE INVENTION
The foregoing and other objects are achieved by the invention which in one aspect comprises a method of generating a plot that evinces common result regions of a test as a function of at least one controllable input parameter, where each parameter has a predetermined range and a predetermined resolution. The method defines an overall plot region that is a function of a maximum and a minimum of each predetermined range. The method further subdivides the overall plot region into at least two sub-regions, each of the sub-regions having a sub-region boundary. The method evaluates, for each of the sub-regions, a plurality of boundary test conditions on the sub-region boundary according to the test, so as to assign a test status to each of the plurality of boundary test conditions. For each of the sub-regions with at least a predetermined threshold number of boundary test conditions having a common test status, the method designates that sub-region with the common test status. For each of the sub-regions not having at least the predetermined threshold number of boundary test conditions with a common test status, the method designates that sub-region with an indeterminate status.
In another embodiment of the invention, the method further including the step of recursively performing the subdividing step, the evaluating step and the designating step for each of the sub-regions designated with an unknown, indeterminate result status until all of the sub-regions are designated by a determinate test status.
In another embodiment of the invention, the evaluating step further includes the step of assigning either a pass test status or a fail test status to each of the plurality of boundary test conditions.
In another embodiment of the invention, the sub-region boundary includes a plurality of corner test conditions, and the plurality of boundary test conditions includes the corner test conditions.
In another embodiment of the invention, the predetermined threshold number of boundary test conditions includes all of the corner test conditions.
In another embodiment of the invention, the at least two sub-regions are substantially identical.
In another embodiment of the invention, the plot is an N-dimensional plot that corresponds to N controllable input parameters.
In another embodiment of the invention, the method further includes the step of searching for at least one discrepancy between (i) a test status of a test condition designated by the designating step, and (ii) a test status of the test condition according to the performance test.
In another embodiment of the invention, the method further includes the step of evaluating, via the performance test, those test conditions that have a test status designated by the designating step and are adjacent to a test condition having a test status assigned according to the performance test.
In another embodiment of the invention, the method further includes the step of reassigning the test condition according to the performance test.
In another aspect, the invention comprises a method of generating a shmoo plot as a function of a performance test and a plurality of parameters, each of the parameters having a predetermined parameter range and a predetermined parameter resolution. The met
Assouad Patrick
LTX Corporation
McDermott & Will & Emery
Tsai Carol S. W.
LandOfFree
Method of and system for generating a binary shmoo plot in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of and system for generating a binary shmoo plot in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of and system for generating a binary shmoo plot in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2836950