Data processing: measuring – calibrating – or testing – Measurement system – Dimensional determination
Reexamination Certificate
2000-12-07
2003-06-10
Barlow, John (Department: 2863)
Data processing: measuring, calibrating, or testing
Measurement system
Dimensional determination
Reexamination Certificate
active
06577985
ABSTRACT:
TECHNICAL FIELD OF THE INVENTION
This application relates generally to encoding systems using truth table logic, and in particular to encoding of, for example, detected array states into a digital representation having improved reliability and predictability.
BACKGROUND OF THE INVENTION
The present invention includes a sensor array encoding system that is specifically designed for implementation on the absolute position detector described in commonly-assigned, co-pending U.S. patent application “ABSOLUTE POSITION DETECTOR INTERPRETING ABNORMAL STATES,” Ser. No. 09/595,368, filed Jun. 15, 2000 (hereafter “APD interpreting Abnormal States”). It will be appreciated, however, that the inventive encoding system disclosed herein is much broader in concept, and is in no way limited to a specific implementation such as is described in APD interpreting Abnormal States. The inventive encoding system has many other applications, including in encoding other suitable sensor array implementations in which abnormal states are accepted rather than rejected.
Encoding systems, such as Binary code or Gray code, for example, are known in the art.
FIG. 1A
illustrates a conventional linear electrical contact encoder using Gray code. Encoder plate
10
has a conductive encoding pattern
11
equivalent to Gray code etched or clad to an insulating material
12
. The conductive pattern is normally connected to the power supply ground
13
while a positive current is applied to each of the contact brushes or pins
14
A through
14
D. As the encoder plate
10
or the contact pins
14
A through
14
D are moved, if any one pin
14
contacts a conductive area
11
of the plate
10
, a one (low true) output is generated in the output truth table
15
. While a pin
14
A through
14
D is contacting the insulating portion
12
of the plate
10
, a zero output is generated in the output table
15
. The four contact pins
14
A through
14
D and the encoding plate
10
as illustrated on
FIG. 1A
will be understood to generate sixteen distinct codes for each of the sixteen steps as shown in the truth table
15
.
Referring now to
FIG. 1B
, Gray code, as with most other conventional encoder codes, must then be decoded into a code recognizable by computers or other devices such as a displays or printers. Since computers are based on the binary numbering system, a conversion to binary code is usually the most logical. Encoded data (Gray code) from the encoder is input to the decoder
16
and the outputs of the decoder is binary code
17
as shown in the truth tables
18
of FIG.
1
C. It will be understood on
FIG. 1C
that exemplary use is made of four bits of binary code represented by four bits of Gray code to encode the sixteen steps (notated in hexadecimal) in the encoder depicted in FIG.
1
A. This exemplary use of four-bit binary code relating to a hexadecimal number of steps will be continued, although it will be understood that the conventional encoding and decoding principles described with reference to
FIGS. 1A through 1C
may be scaled larger or smaller.
Referring now to
FIG. 2A
, it is also known in the art that the linear encoding pattern
11
as shown on
FIG. 1A
may be represented as a circular pattern
21
on endless loop encoder disk
20
. The contact pins
24
are shown at the datum zero position
25
with the least significant bit (LSB) at the outer edge and the most significant bit (MSB) at the enter-most ring of the disk. This encoded disk will produce the same Gray code as shown in the truth table
15
of FIG.
1
A and may be decoded to binary using a decoder
16
as depicted in
FIG. 1B
to yield the truth tables
18
of FIG.
1
C.
Although useful to illustrate a principle, electrical contact encoders are seldom used in today's industrial applications due to numerous performance-related problems. These types of encoders have a finite life due to mechanical wear between the contact plates and brushes or pins. Reliability of a contact device is affected by possible contamination and corrosion which could isolate a contact area. Design and construction require extremely tight tolerances to prevent ambiguous states as discussed below. Some materials used in construction of a contact encoder can generate piezoelectric noise under shock or vibration.
Further, encoders such as illustrated on
FIGS. 1A and 2A
can generate invalid codes resulting from ambiguous states when an unwanted contact is made or is not made relative to another contact. For example, the end of a conductive plate, whether rotary or linear, is often a site of premature mechanical wear. This wear generates imprecise contact. Moreover, rotary encoder devices such as illustrated in
FIG. 2A
are particularly prone to this problem with ambiguous states.
FIG. 2B
shows an area
26
likely to be affected by ambiguous states resulting from an ambiguous contact position. The problem is amplified in multi-turn shaft encoders deployed as geared disks in a hierarchical relationship, where the gears produce a cumulative error from the least significant digit to the most significant.
By examining an example of a conventional multi-turn encoder
27
depicted in
FIG. 2C
, the problem will become more apparent. The multi-turn encoder
27
has four encoder disks A through D, each comprising large gears
28
A through
28
D, and further comprising encoding patterns
21
analogous to those depicted on
FIGS. 2A and 2B
(encoding patterns not illustrated on FIG.
2
C). Large gears
28
A through
28
D mesh with small gears
29
A through
29
D. The ratio between large gears
28
A through
28
D and small gears
29
A through
29
D is a function of desired multiplication (M). Since computers use binary code, it is generally preferable to user a binary number as the multiplier. Using an exemplary value of 16 (or 2
4
) for M, each large gear
28
A through
28
D needs to have 16×N teeth, where N is the number of teeth on each small gear
29
A through
29
D. Using an exemplary value of 8 (or 2
3
) for N, then each large gear
28
A through
28
D has 16×8=128 teeth. Now, with a gear ratio of 16:1 it will be understood that second encoder disk B turns one revolution for each sixteen revolutions of the first encoder disk A, and so on. Thus fourth encoder disk D turns one revolution for each 4096 revolutions of the first encoder disk A. Giving each encoder disk A through D on
FIG. 2C
an encoding pattern identical to the encoding pattern
21
depicted on
FIGS. 2A and 2B
, each encoder disk A through D monitors for 16 steps 0 through F (as notated in hexadecimal). The total number of steps capable of being monitored by multi-turn encoder
27
on
FIG. 2C
is thus 4096×16=65,536 (or 216).
Rotary encoders incur ambiguous state errors primarily because, with reference to
FIGS. 2A and 2B
, the different contact pins
24
travel a significantly different distance around in one revolution of the disk. The difference in distance of travel is the difference in circumference of two paths around the disk. Further, the greatest difference is between least significant bit (LSB) and most significant bit (MSB). For example, if the radius of the MSB track in
FIGS. 2A and 2B
is 0.2″ from the center of the shaft, and the radius of the LSB track is 1″ from the center of the shaft, the difference in circumferential travel is a factor of over 5, meaning that the MSB is moving over five times slower than the LSB. Referring again to
FIG. 2D
, any movement of the encoder disk A will produce {fraction (1/4095)} of that movement in encoder disk D. If the cumulative error of the three sets of gears between the disk is greater than {fraction (1/4095)} (about 0.088 degrees) of one revolution of the input shaft then a whole bit error occurs at the most significant digit. This equates to a maximum tolerance of less than 0.029 degrees per gear. Such tight manufacturing tolerances are hard to maintain, and variances from tolerance tend to create ambiguous state errors.
As noted above, ambiguous state errors are also caused by ambiguou
Barlow John
Bhat Aditya
EIM Company, Inc.
Vinson & Elkins L.L.P.
LandOfFree
Scalable code absolute logic function (SCALF) encoder does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Scalable code absolute logic function (SCALF) encoder, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Scalable code absolute logic function (SCALF) encoder will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3095778