Note: Descriptions are shown in the official language in which they were submitted.
CA 02579505 2007-03-13
LASER SCANNING SYSTEM AND SCANNING METHOD FOR
SCANNING BAR CODES
This is a division of co-pending Canadian Patent
Application No. 2,161,357 filed on October 25, 1995.
BACKGROUND OF THE INVENTION
This invention relates generally to the design of
scanning systems for reading bar code symbols or similar
indicia, and more particularly, to scanning both
one-dimensional and two-dimensional bar code symbols
- 1 -
CA 02579505 2007-03-13
automatically. Most conventional optical scanning systems
can read either one-dimensional or two-dimensional bar code
symbols. A bar code symbol is a coded pattern of indicia
having a series of variable-width bars separated by
variable-width spaces, the bars and spaces having different
light-reflecting characteristics. One example of a
one-dimensional bar code is the UPC/EAN code currently in
use for identifying articles and other information. One
example of a two-dimensional, or stacked, bar code is the
PDF417 bar code described in U.S. Patent No. 5,159,639.
Most scanning systems, or scanners, generate a beam of
light which reflects off a bar code symbol so the scanning
system can receive the reflected light. The system then
transforms that reflected light into electrical signals,
and decodes those electrical signals to extract the
information embedded in the bar code symbol. Scanning
systems of this general type are described in U.S. Patent
Nos. 4,251,798; 4,360,798; 4,369,361; 4,387,297; 4,409,470;
and 4,460,120, all of which have been assigned to Symbol
Technologies, Inc.
Because both one-dimensional symbols and
two-dimensional symbols are currently being used, it would
be simpler and more efficient if a single scanning system
could not only distinguish a bar code symbol from other
markings, such as text, but also decode the symbol whether
_ 2 -
CA 02579505 2007-03-13
it is one-dimensional or two-dimensional. This operation
is particularly important when the bar code symbol is
skewed relative to the scanning patterns of the scanner.
An additional problem for such scanning systems occurs
when decoding two-dimensional bar code symbols. These
symbols do not all have the same height, so the scanning
system must expand its scanning pattern to cover the entire
two-dimensional symbol. Some conventional systems do this
but sometimes cover areas outside of the symbol. Although
using such a large pattern does not affect the accuracy of
the scanner, it is inefficient. The portions of the
scanning pattern which lay outside the bar code symbol are
useless, and scanning these areas slows down the scanning
operation. In addition, forcing the scanning pattern to be
too large reduces the accuracy of decoding the
two-dimensional bar code symbol.
Building a system to overcome these problems is not
only difficult, it is complicated by an additional concern.
Scanners should not become any larger for ergonometric and
economic reasons. Thus, more powerful and flexible
scanning devices must be compact.
Another concern is speed. The additional processin.g
needed for increased efficiency and flexibility must not
come at the expense of speed. That processing must
therefore proceed quickly and efficiently.
Yet another concern is the need to ensure that the
- 3 -
CA 02579505 2007-03-13
different scanner subsystems communicate with each other
effectively as they become more robust. Scanning systems
typically have different subsystems, such as the scanning
engine, the optical sensor, and the decoder. The
interfaces between these different subsystems must support
the required processing power and allow one to improve one
part a scanning system without redesigning other systems.
To obtain a scanner with all these features is very
difficult indeed. The fast-increasing use of bar codes,
however, demands that scanning systems have increasing
flexibility, robustness, and efficiency.
It is therefore an object of this invention to provide
a scanner capable of differentiating between
one-dimensional and two-dimensional bar codes, and of
decoding them automatically and appropriately even if they
are not initially aligned with the scanner.
It is also an object of the invention to ensure that
the scanning pattern can precess to change the location of
the scan lines.
it is another object of the invention to adjust the
height of the scanning patterns automatically to ensure
that the scanning pattern covers the entire symbol without
extending outside the symbol.
A further object of the invention is to provide a fast
but compact scanning engine to generate scanning patterns
for both one-dimensional and two-dimensional bar codes.
- 4 -
CA 02579505 2007-03-13
It is also an object of the invention to have such a
scanning engine adapt itself to the different
characteristics of the scanning elements.
It is a further object of the present invention to
have such a powerful and flexible scanning engine
communicate with the remainder of the scanner over a robust
and flexible interface.
SUMMARY OF INVENTION
To achieve these objects, the present invention
provides a system for examining the optical reflections to
determine whether a sensed target is a bar code symbol and,
if so, whether that symbol is a one-dimensional or
two-dimensional code. If the symbol is a two-dimensional
code, the present invention aligns the two-dimensional
scanning pattern with the symbol and expands the pattern to
the top and bottom edges of the symbol.
In addition, the present invention provides a
microprocessor-controlled scan engine that uses a coil both
to drive a scanning element and to pick up feedback signals
representing the motion of the scanning element. The scan
engine also uses different circuit techniques to avoid
degrading the system, and has a powerful interface to
decoding and control logic of the scanner. The scan engine
may be set to ensure that two-dimensional scanning patterns
- 5 -
CA 02579505 2007-03-13
precess to move the scan lines at different portions of the
bar code symbols to avoid any gaps that may otherwise
occur.
Specifically, a control circuit for driving a coil to
produce a desired magnetic field according to this
invention comprises a pulse width modulation regulator,
coupled to the coil, for causing a desired current to flow
through the coil in accordance with an analog drive signal;
a digital-to-analog converter, coupled to the pulse width
modulation circuit, for creating the analog drive signal
from a digital drive signal; and a controller, coupled to
the digital-to-analog converter, for generating the digital
drive signal.
A method for driving a coil according to this
invention comprises the steps of generating a desired
current signal to flow through the coil in accordance with
an analog drive signal; creating the analog drive signal
from a digital drive signal; and generating the digital
drive signal.
A control circuit according to this invention for
i driving a scanning element to produce a desired scanning
pattern comprises a coil producing a magnetic field to
control the movement of the scanning element; a pulse width
modulation regulator, coupled to the coil, for causing a
desired current to flow through the coil in accordance with
an analog drive signal; a digital-to-analog converter,
- 6 -
CA 02579505 2007-03-13
coupled to the pulse width modulation circuit, for creating
the analog drive signal from a digital drive signal; and a
controller, coupled to the digital-to-analog converter, for
generating the digital drive signal.
A control system according to this invention for
controlling the optical scanning of an object comprises a
scan control section; a detector section coupled to the
scan control section; and an interface channel between the
detector section and the scan control section for carrying
commands in a defined format from the detector section to
the scan control section, and for carrying messages in the
defined format from the scan control section to the
detector section. The scan control section includes a
scanning element for causing a light beam to move relative
to the object, a processor for receiving external commands,
for generating internal control signals from the external
commands, and for forming messages; and a scanning element
driver for energizing the scanning element to cause the
light beam to move in a predetermined pattern in response
to the internal scanning control signals. The detector
section includes a light beam scanner directing the light
beam toward the scanning element; a detector mounted to
receive portions of the light beam reflected from the
object and to generate electrical signals representing the
received, reflected light beam; and central control means
for receiving the electrical signals, for forming the
- 7 -
CA 02579505 2007-03-13
external commands, and for receiving the messages.
A scanner according to this invention for scanning a
light beam across a two-dimensional pattern comprises a
light source for creating the light beam; a scanning
element
for moving the light beam in the two-dimensional pattern;
x-drivers for causing the scanning element to move the
light beam in a first direction at a first frequency in
response to x-driver signals; y-drivers for causing the
scanning element to move the light beam in a second
direction at a second frequency in response to y-driver
signals, the first and second directions being orthogonal
to each other; and scanning control circuitry for
generating the x-driver signals and the y-driver signals at
the first and second frequencies that are not integer
multiples of each other, thereby causing the
two-dimensional pattern to precess.
A bar code reader according to this invention for
reading a bar code symbol having a defined boundary
comprises a light beam scanner for directing a light beam
toward a target in a predetermined pattern; a detector for
receiving portions of the light beam reflected from the
target and generating electrical signals representing the
received, reflected light beam; identifier means for
determining whether the target is a bar code symbol; and
feedback means, responsive to the electric signals, for
- 8 -
CA 02579505 2007-03-13
controlling the scanner to conform the shape of the
predetermined pattern to the boundary of the target if it
is a bar code symbol.
BRIEF DESCRIPTION OF DRAWINGS
The accompanying drawings provide a further
understanding of the invention and, together with the
description, explain the principles of the invention.
Fig. 1 is a simplified diagrammatic representation of
one embodiment of a laser scanning system according to this
invention;
Fig. 2 is a diagram of a conventional raster scanning
pattern;
Fig. 3 is a flow chart of the preferred procedure for
finding, identifying, and decoding bar codes according to
this invention;
Figs. 4a-4d are representations of raster scanning
beams and bar codes;
Fig. 5 shows a raster scanning pattern traversing a
one-dimensional bar code that is skewed with respect to the
scanning pattern;
Figs. 6a-6d are representations of a raster scanning
pattern traversing a two-dimensional bar code that is
initially skewed with respect to the scanning pattern;
Fig. 7 is a high level diagram showing raster pattern
control according to this invention;
- 9 -
CA 02579505 2007-03-13
Fig. 8 is a front view of an outline of a symbol that
is skewed with respect to the scanning pattern;
Fig. 9 is a flow chart of a procedure for implementing
a full raster scan of a two-dimensional symbol;
Fig. 10 is a block diagram of a scanning system
showing the control of a scanner in a full raster mode;
Fig. 11 is a diagram of a circuit to determine whether
a scanned target is a bar code;
Fig. 12 is a block diagram of a scan engine;
Fig. 13 is a diagram showing the preferred
construction of a PWM regulator used in the scan engine of
Fig. 12;
Fig. 14 is a diagram showing switches used in the PWM
regulator in Fig. 13;
Fig. 15a is a diagram of the frequency response for an
RASE and mylar support which can be used with this
invention;
Fig. 15b is a diagram showing the magnitude and phase
variations as a function of frequency for the RASE;
Fig. 16 is a diagram of two sine waves to show coarse
amplitude adjustment;
Fig. 17 is a diagram of two traces to show fine
amplitude adjustment;
Fig. 18 is a graph showing the changes in the resonant
frequency of an RASE with temperature;
Fig. 19 is a diagram illustrating how the sum of the
- 10 -
CA 02579505 2007-03-13
sine waves for the digital drive signal shown in Fig. 12 is
formed;
Fig. 20 is a flow chart of the operation for motor
control of the scan engine in Fig. 12;
Fig. 21 is a flow chart of the procedure for
generating raster scans;
Figs. 22a-22c are different perspective views of a
preferred assembly for the scan engine shown in Fig. 12;
Fig. 23a-23c shows a scanning pattern and it
precesses;
Fig. 24 is a block diagram of the scan engine in Fig.
12 and its interfaces with other subsystems;
Fig. 25 is a diagram of the preferred formats for
commands and messages exchanged with the scan engine shown
in Fig. 12;
Fig. 26 is a list of the commands and messages
exchanged using the format of Fig. 25;
Fig. 27 is a block diagram of a circuit for changing
the height of a raster scan pattern;
Fig. 28 shows the variation of one of the signals of
Fig. 27 as a function of time; and
Fig. 29 shows a block diagram of an amplitude control
circuit used in Fig. 27.
DESCRIPTION OF THE PREFERRED IMPLEMENTATIONS
The following description of the preferred
- 11 -
CA 02579505 2007-03-13
implementations of this invention do not describe all
possible implementations. In the description, like
reference numerals in different figures refer to like
parts, unless indicated otherwise.
A. Overview
Unless the context or specific instructions indicate
otherwise, the terms "symbol" and "bar code" should be
construed broadly in this specification and the following
claims. For example, those terms cover any number of
patterns having alternating bars and spaces, including
those of various widths, and one-dimensional or
two-dimensional graphic patterns other than those
specifically mentioned.
The present invention relates to scanning systems
including those that can automatically initiate and
terminate scanning of a target. Some scanning systems with
this automatic capability use a manually-operated trigger
to initiate scanning of the target, such as U.S. Patent
No. 4,387,297 describes. Although the trigger is important
for many applications, some applications benefit from other
techniques, and this invention includes such techniques.
Fig. 1 shows a highly simplified embodiment of a bar
code scanner 100 that may be constructed according to the
principles of the present invention. Although Fig. 1 shows
scanner 100 as hand-held, the invention does not require
that the scanner be in this form. For example, the scanner
- 12 -
CA 02579505 2007-03-13
could include a desktop workstation or have some other type
of stationary architecture. Scanner 100 may also function
as a portable computer terminal and include a keyboard 148
and a display 149, such as described in U.S. Patent No.
4,409,470.
Hand-held scanner 100 of Fig. 1 has the style
described generally in U.S. Patent Nos. 4,760,248 or
4,896,026, both assigned to Symbol Technologies, Inc.
Scanner 100 also has a similar design to the bar code
reader commercially available as part number LS 8100 or LS
2000 from Symbol Technologies, Inc.
A user aims scanner 100 at bar code symbol 170 without
physically touching it. Typically, scanner 100 operates
several inches from the bar code symbol being read.
To construct scanner 100, the ordinary-skilled artisan
may refer to U.S. Patent Nos. 4,387,297; 4,409,470;
4,760,248; 4,896,026; and 4,387,298. To assist in the
understanding of the claimed invention, however, the major
features of scanner 100 are described below.
Scanner 100 is preferably gun-shaped in a housing 155
having a pistol-grip handle 153. A movable trigger 154 on
handle 153 allows a user to activate a light beam 151 and
associated detector circuitry when the user has pointed
scanner 100 at a symbol 170.
Housing 155, which is preferably made of lightweight
- 13 -
CA 02579505 2007-03-13
= plastic, contains laser light source 146 (which may be a
semiconductor laser diode or other light source), lens 157,
partially-silvered mirror 147, detector 158, oscillating
mirror 159, motor 160, power source (battery) 162, and
signal processing and control circuitry 165. Circuitry 165
includes CPU 140 and decoding and control electronics 142
on a printed circuit board 161.
When a user activates scanner 100 by pulling
trigger 154, light source 146 generates light beam 151
along the axis of lens 157. Lens 157, which is not
necessary in all embodiments, may be a single lens or a
multiple lens system. After passing through lens 157, beam
151 passes through partially-silvered mirror 147 and, if
desired, other lenses or beam-shaping structures. Beam 151
then strikes oscillating mirror 159 driven by scanning
motor 160, which together direct beam 151 in a scanning
pattern. Preferably, motor 160 also starts when the user
pulls trigger 154.
If light beam 151 is invisible, the optical system may
include an aiming light parallel to beam 151 to help the
user aim scanner 100. The aiming light is a visible beam
of light that may either be fixed or follow light beam 151.
Mirror 159 directs light beam 151 through a light-
transmissive window 156 and across bar code symbol 170 in
some predetermined pattern, such as a linear raster scan
pattern. Fig. 2 is a diagram depicting the pattern of a
- 14 -
CA 02579505 2007-03-13
known linear raster scanner. The two-dimensional scanning
pattern can be generated by displacing in the vertical or y
direction a one-dimensional or linear scan line driven in
the horizontal or x direction. U.S. Patent 4,387,297
explains a technique for forming the scanning pattern in
Fig. 2.
Symbol 170 can be a one-dimensional bar code, such as
the standard UPC/EAN code, or a two-dimensional bar code,
such as PDF417 described in U.S. Patent No. 5,159,639. In
addition, as explained above, symbol 170 can be any other
acceptable symbol carrying information to be decoded.
Light beam 152 is the light from beam 151 reflected
off symbol 170. Beam 152 returns to scanner 100 along a
path parallel to, or at times coincident with, beam 151.
Beam 152 thus reflects off mirror 159 and strikes
partially-silvered mirror 147. Mirror 147 reflects some of
beam 152 onto a light-responsive detector 158 that converts
light 152 into electrical signals.
The electrical signals then pass into signal
processing and control circuitry 165 to be processed and
decoded to extract the information represented by the bar
code. Signal processing and control circuitry 165 also
controls the operation of motor 160 to adjust the scanning
pattern and provide other control.
B. Adaptive Scanning
1. Identifying a bar code symbol
- 15 -
CA 02579505 2007-03-13
The scanner of this invention addresses two concerns.
The first is to ensure that the target being scanned is a
bar code. The second is to identify the type of bar code.
The purpose of ensuring that the scanned target is a
bar code is obvious. The purpose of identifying the type
of bar code is to adjust the scan pattern to improve
detection.
To accomplish both purposes, a system according to the
present invention samples light reflected from a portion of
a target and analyzes those samples. The system first
determines whether the target is a bar code symbol. If so,
the system next determines whether the bar code symbol is
one-dimensional or two-dimensional. If the symbol is
one-dimensional, the system decodes the signals received
from the scan. If the bar code is two-dimensional, the
system ensures the scanning pattern is properly oriented
and then begins to expand the scanning pattern to cover the
entire code.
Fig. 3 is a flow chart 300 indicating the preferred
procedure for finding, identifying, and decoding bar codes.
First, scanner 100 generates a narrow (i.e., small vertical
displacement) scanning pattern (Step 305), and then
scanner 100 takes an initial scan. (Step 310)
In the preferred embodiment, the user presses
trigger 154 to begin this scanning operation. Pressing the
trigger causes scanner 100 to produce a narrow scanning
- 16 -
CA 02579505 2007-03-13
pattern which is easy for a user to aim and direct toward a
target.
Figs. 4a-4d show a sequence of views of a target and a
raster scanning pattern. Fig. 4a is a highly simplified
schematic representation of the initial stage of operation
when scanner 100 has generated the narrow scanning
pattern 410, but the user has not yet properly placed
pattern 410 at target bar code symbol 420.
Fig. 4b shows the next stage of the operation, when
user has properly placed the narrow scanning pattern 410
over two-dimensional bar code 420 (or one-dimensional bar
code 430). Of course, Fig. 4b assumes the target is a bar
code. The target could also be some other mark, in which
case the system remains in the initial scan mode. (Step
310)
Once the scanning pattern is properly placed over the
target, scanner 100 ensures that the scanned target is a
bar code symbol. (Step 320) There are several
conventional techniques for making this determination.
These techniques are not the exclusive methods for
distinguishing bar codes, nor are they mutually exclusive.
One or more techniques may be used together.
One technique involves analyzing the spatial variation
of the areas of different light reflectivity to determine
whether the reflected light has characteristics expected
from a bar code symbol. Signal processing and control
- 17 -
CA 02579505 2007-03-13
circuitry 165 would typically be programmed to perform
these analyses.
A second technique compares the length of the portions
of low light reflectivity to those of high reflectivity.
If the ratio of those lengths, is within a preset range,
the system determines that it has found a bar code.
A third technique counts the number of transitions in
a given time period between portions of different light
reflectivity. That count can characterize the reflected
light as some predetermined pattern, such as a generic bar
code symbol, a class of bar code symbols, or even a
specific bar code symbol.
A fourth technique compares the electrical signals
generated from one scan with that from one or more
subsequent scans. If the successive scans yield identical
or nearly identical signals, the system concludes that it
is viewing a bar code with bars and spaces of uniform
widths. A variation of this technique compares several
scans to determine whether successive scans differ, but
have similar groupings. If so, the target is likely a two-
dimensional bar code.
A fifth technique is to try to decode the scans. If
the decode is not successful, the system concludes that the
target is not a recognizable bar code.
If the target is not a bar code symbol, the system
remains in the initial scan mode. (Step 310) At that
- 18 -
CA 02579505 2007-03-13
point, the user may move scanner 100 closer to or farther
from the target to account for the possibility that the
target lay outside the working range of scanner 100. The
user can also seek a new target.
If the target is a bar code symbol (Step 320),
scanner 100 tries to determine whether the bar code is
one-dimensional or two-dimensional. (Step 330) There are
several ways to make this determination as well. One way
is to try to decode a row and make the determination on the
basis of the decoded information. Another way is to use an
intelligent sensing algorithm, such as attempting a decode,
to determine whether the sensed portion is from a
one-dimensional or a two-dimensional bar code.
In addition, although flow chart 300 shows that the
determination of whether a target is a bar code is separate
from the determination of the bar code's type, the
operations need not be separate. For example, both
determinations may take place during the same operation.
After determining the type of bar code, the scanner
follows different paths for decoding the bar code. If the
symbol is in a one-dimensional bar code, the scanning is
virtually complete. Scanner 100 attempts to decode the
code without altering the height or width of the scanning
pattern. (Step 340) If scanner 100 is successful (Step
345), it sends the decoded data out for further processing.
(Step 350) If not, scanner 100 takes additional scans
- 19 -
CA 02579505 2007-03-13
= (Step 310) until it successfully decodes the symbol. (Step
345) Scanner 100 may also be programmed to stop after too
many unsuccessful decodes or after too much time has
passed.
If the symbol is in a two-dimensional bar code
(Step 330), additional scanning must take place. First,
scanner 100 must make sure the scanning pattern is properly
aligned with the bar code. (Step 355).
Orientation is not a problem for one-dimensional bar
codes. Fig. 5 depicts the raster scanning pattern 510
traversing a one-dimensional bar code 520 skewed with
respect to the direction of the scan lines. The scan lines
of a pattern do not need to be orthogonal to a
one-dimensional bar code's vertical bars because one or
more of the scan lines still traverse the same sequence of
bars and spaces.
The situation for two-dimensional bar codes is more
difficult. Figs. 6a, 6b, 6c and 6d are pictorial
representations of the raster scanning pattern traversing a
two-dimensional bar code 600 originally skewed with respect
to the scanning pattern.
Fig. 6a contains a highly simplified schematic
representation of the initial raster scanning pattern
skewed or misaligned with respect to two-dimensional bar
code 600. Scanner 100 preferably determines this condition
by noting when a scan line crosses a row. For example, the
- 20 -
CA 02579505 2007-03-13
PDF417 code uses different codes in different rows
(actually there are three different codes that repeat).
When the code changes during the decode of a scan line,
scanner 100 detects that the scan lines are not aligned.
To read the skewed two-dimensional bar codes,
scanner 100 can use several different techniques, such as
those described in copending U.S. Patent Number 5,414,250.
Once scanner 100 determines that the scanning pattern is
misaligned, it can begin to reorient, as shown in Fig. 6b.
If further analysis reveals that symbol 600 is still
skewed relative to the scanning pattern, the system can
continue to reorient the scanning pattern until it finally
aligns with symbol 600, as shown in Fig. 6c. When
reorientation is complete, the entire bar code can be read
by using a fully aligned and height-adjusted scanning
pattern, as shown in Fig. 6d.
After aligning the scanning pattern with the bar code
symbol, scanner 100 enters the full raster mode to increase
the height of the scanning pattern. (Step 360) This is
done to decode the entire symbol. The term "full raster
mode" signifies the process of controlling mirror 159 and
motor 160 to change the height, and even width, of the
raster scanning pattern. During full raster mode, the
pattern height and width usually increase in stages, as
- 21 -
CA 02579505 2007-03-13
Figs. 4c and 4d demonstrate.
Fig. 4c is a highly simplified schematic
representation of a intermediate stage in the operation of
scanner 100 during the full raster mode. As the height of
the scanning pattern 412 increases, scanner 100 reads the
bar code rows covered by the pattern to decode the pattern
and to determine when to stop increasing the pattern
height. This continues until the entire symbol is read.
Fig. 4d is a highly simplified schematic
representation of the final stage of operation. At this
point, scanning pattern 413 covers the entire bar code 420
as well as areas outside of the code 420.
When scanning in the full raster mode, the pattern
height increases but the number of scan lines that sweep
the bar code symbol does not change. Instead, the height
and width of the scan pattern increase, and the angles
between adjoining scan lines also increase as Figs. 4b-4d
show. The most effective alignment of the laser scan
pattern occurs when each scan line crosses exactly one row
in the two-dimensional symbol.
Once the size of the pattern is set (and, preferably,
as the size is increasing), scanner 100 decodes the two-
dimensional symbol. (Step 370) If scanner 100
successfully decodes the symbol (Step 380), it transmits
the decoded data and-either narrows the scanning beam or
turns it off. (Step 390) If scanner 100 does not
- 22 -
CA 02579505 2007-03-13
successfully decode the symbol (Step 380), scanner 100
continues scanning until it achieves a successful decode or
until a predetermined amount of time has elapsed. The
predetermined amount of time is typically on the order of
three seconds, which is regarded as sufficient time for an
operator to sight a symbol and obtain a successful decode.
2. Scanning height adjustment
One drawback,to the operation just described is that
in full raster mode, the pattern will sometimes cover more
than just the target bar code. For example, Figs. 4d and
6d show several scan lines extending beyond the top and
bottom edges of the two-dimensional symbol. This
unnecessary scanning slows the entire scanning procedure.
To solve this problem, the present invention minimizes the
amount a given pattern extends beyond a symbol and
decreases the scan time by limiting the area of a bar code
symbol and surrounding area covered by the scan pattern.
There are at least two techniques to accomplish this
goal. The first technique works for codes such as the
PDF417 code that embed information about the size of the
symbol in the rows of the code. The second technique works
for other types of codes.
a. PDF417 mode
For codes such as the PDF417 code, information
acquired from the initial scan pattern allows the system to
determine the number of rows of the symbol above and below
- 23 -
CA 02579505 2007-03-13
the initial scan pattern in the second stage. This allows
the system to find an appropriate new scan pattern and
control the rate of opening the scan angle to reach the new
pattern, and send the proper control signals to the
scanner.
Fig. 7 is a high level diagram showing raster pattern
control according to this invention. A raster pattern
control processor 700, which can be part of control
circuitry 165, controls a scan engine 710, which can also
be part of circuitry 165, to read symbol 720, shown by a
side view. Scan engine 710, located a distance d (not
shown) from symbol 720, emits an initial raster scan
pattern open to a vertical angle z.
When the raster control processor decodes the result
of the scan at angle z, it determines that the initial scan
pattern crosses f rows of the two-dimensional symbol. If
the scan pattern is centered, the number of rows crossed
above the middle row m of the pattern is y, which equals
f/2.
Next, raster control processor 700 decodes information
acquired from the initial scan of symbol 720 to determine
one-half of the maximum number of rows, Y, of the entire
symbol 720. If the bar code is encoded in the PDF417
format, the first two columns in each row contain
information about the height of the entire symbol.
From this information, processor 700 can determine the
- 24 -
CA 02579505 2007-03-13
scan angle Z required to read the entire symbol 720 without
reading areas above or below symbol 720. This can be done
as follows:
d = y/(tan (z));
Z= tan-L (Y/d), or tan-1 ((Y/y) * tan(z) ).
From this value of Z, processor 700 determines the
rate R to open the angle of scan to the angle Z. The rate
R is important because processor 700 is preferably decoding
symbol 720 as the pattern expands, and the rate of
expansion must accommodate this operation. For example, if
symbol 720 is large, the rate should be low. If symbol 720
is small, the rate should be fast. Another reason for
controlling the rate of the y-direction expansion is that
the scanner decodes as it expands.
As the scanning pattern expands in the full raster
mode, the scanned data is most useful at the top and bottom
of the pattern because the scanner has already read the
center areas of the bar code.
To determine the rate R of y-direction expansion,
processor 700 first finds the angle differential a as
a = Z-z.
Then, raster control processor 700 determines the rate R as
follows:
R = a/(r * Y),
where r is the time scan engine requires to read a single
- 25 -
CA 02579505 2007-03-13
row.
Control processor 700 then sends commands to scan
engine 710 to direct it to open to a maximum scan angle Z
at a rate R. A preferred implementation of this
communication is explained below in the section entitled
"Scan engine interface."
In response to the commands, scan engine 710 opens the
scan pattern to angle Z to include all of the rows of the
symbol 720 without including areas above and below the
symbol. This saves scanning time.
This procedure for accelerating scanning applies even
when the initial scan pattern is skewed relative to the
symbol, which is typical. Fig. 8 shows a front view of the
outline of a symbol 800 with portions omitted for clarity.
In Fig. 8, the initial scan lines 810 are skewed from
rows 820 of symbol 800 by an angle
The formula for determining the maximum scan angle Z
and the opening rate R when there is no skew angle also
works when there is a skew angle because the effect of the
skew angle cancels out. Although a nonzero skew angle will
increase y, the number of rows initially scanned above the
middle row m, by a factor of cos -, the maximum number of
rows Y decreases by the same factor. These two effects
cancel out.
The procedure described with regard to Fig. 7 assumes
that the user has positioned the scan line approximately at
- 26 -
CA 02579505 2007-03-13
the vertical midpoint of the bar code. If the user does
not accurately position the scan line at the midpoint,
processor 700 will know this from information in the PDF417
code. The code in each row contains information about the
row's number, which processor 700 can use to make
corrections as appropriate.
b. Alternative mode
If the two-dimensional bar code symbol does not have
embedded information to allow the procedure described in
relationship to Figs. 7 and 8, the present invention
contemplates a different procedure shown in Fig. 9 and
using the scanning system 1000 shown in Fig. 10. Fig. 10
depicts the control of scanner 1010 in the full raster
mode.
Flow chart 900 in Fig. 9 begins by decoding a scan.
(Step 910) This is done by having scanner 1010 begin a
raster pattern across a bar code symbol and receive the
reflected light from the symbol. From that reflected
light, scanner 1010 generates electrical signals
representing the widths of the bar and space pattern.
Actually, scanner 1010 sends two types of signals:
the Start of Scan (SOS) and Digital Bar/space Pattern
(DBP). The SOS signal is a square waveform which changes
levels at the start of each scan, so it is a logical 0 for
scans in one direction and a logical 1 for scans in the
other direction. The DBP signal is a digital waveform
- 27 -
CA 02579505 2007-03-13
consisting of logical 1 and logical 0 pulses whose
durations represent the widths of the bar and space pattern
of the bar code being scanned.
Next, the scanned signals are conditioned. (Step 920)
Fig. 10 actually shows this step as occurring in two
smaller steps. First, scanner interface 1020 measures the
durations of each pulse by counting the number of pulses of
a timer that occur during each bar and space.
Scanner interface 1020 then records those values (DBP
counts) for each bar or space, and uses the SOS signal to
group the DBP counts for each scan. Scanner interface 1020
sends these grouped counts to decoder/scan control 1030 as
a Bar Space Sequence.
Next, the conditioned signals are analyzed to
determine whether they represent bar code information or
some other type of information. (Step 930) In the
embodiment shown in Fig. 10, decoder/scan control 1030
makes this determination by looking at the DBP counts. For
example, a large DBP count indicates a white or black space
that is too large to be a bar or space in a bar code
symbol. Alternatively, decoder/scan control 1030 may sense
that the number of elements in one scan symbol differs
greatly from the number of elements determined from scans
taken inside the symbol. Other techniques are possible as
well. For example, the inability of the decoder/scan
control to recognize any characters might indicate that the
- 28 -
CA 02579505 2007-03-13
current scan did not cross the bar code.
If the conditioned signals represent a bar code
pattern (Step 940), then the scanning pattern increases
(Step 950), and the process repeats. If not, the pattern
is reduced to its previous value (Step 960) and raster mode
is finished. (Step 970)
In Fig. 10, decoder/scan control 1030 determines
whether scanner 1010 must enlarge or reduce the raster
pattern. To change the size of the raster pattern,
decoder/scan control 1030 sends digital control signals to
digital-to-analog converters 1040 and 1050 to provide the X
and Y driving signals, respectively, for scanner 1010.
Instead of simply causing the D/A converters 1040 and
1050 to change voltages, system 1000 can use more
sophisticated control techniques to provide for better
control over the behavior of the scanner raster pattern.
Some of these are described below in the section entitled
"Scanning Control."
There is a note of caution, however. Any technique to
control the raster pattern should have parameters that can
be tied to the mechanical properties of the scanner. These
parameters allow for a smoother change in the size of the
raster pattern, and the proper rate of change enables the
scanner 1010 to respond smoothly to the voltage changes
received from the decoder/scan control 1030. The proper
rate and smoothness of change can eliminate the flicker
- 29 -
CA 02579505 2007-03-13
that annoys some users.
The length of the scan and the rate at which it is
adjusted depend on the amount of white space in the scan
data before and after useful bar code data is decoded.
Decoder/scan control 1030 can determine how much the raster
has exceeded the edge of the bar code symbol in the y-
direction by counting the number of scans from which no
data can be decoded. For example, a large white space can
be detected by observing a large DBP count. The scanner
1010 then makes adjustments, as described above, based on
the data values received.
Although the size of scanning patterns can adapt to
changes in the distance between the scanner and the bar
codes, the change in size should not take place too
quickly. The rate of y-direction expansion should depend
on the number of rows in a label and the label's height.
To teach the scanner operator the correct range and
orientation of the scanner in order to read symbols quickly
and accurately, a feedback signal 1060 (an audible "beep"
or a visual indicator) may be used when a symbol is
detected in range. Fig. 10 shows this signal 1060 as
connected to decoder/scan control 1030. In one embodiment,
an LED blinks slowly when there is poor alignment and
accelerates proportionately as the alignment improves.
3. Bar code detection and identification
circuitry
- 30 -
CA 02579505 2007-03-13
One of the key steps in flow chart 300 shown in Fig. 3
is to determine whether the scanned target was a bar code.
Although there are several ways to make this determination,
Fig. 11 shows a diagram of a circuit 1100 to do so. This
circuit is also disclosed in U.S. Patent 5,229,596.
Circuit 1100 receives a signal 1105 from
amplifier/digitizer 1116, and that signal enters the input
of an inverter 1131. The output of inverter 1131 connects
to the anode of a diode 1132, and the cathode of diode 1132
connects to an RC circuit formed by resistor 1120 connected
in series with a parallel circuit of resistor 1121 and
capacitor 1134. Diode 1132 prevents capacitor 1134 from
discharging into the output of inverter 1131.
Resistor 1121 and capacitor 1134 also connect between
a ground potential and one input 1135 of an open collector
output comparator 1137. The second input 1138 of
comparator 1137 is a threshold level. That threshold level
is also the potential of an intermediate node of a voltage
divider formed by the series connection of resistors 1123
and 1124 between supply voltage V and ground.
The output of comparator 1137 is a "laser enable"
signal 1141 that indicates the target is a bar code=.
Comparator 1137's output also feeds back to the voltage
divider's intermediate node 1150, and thus to input 1138,
via resistor 1125, which has a value R3. This feedback
- 31 -
CA 02579505 2007-03-13
provides a hysteresis effect to the comparison operation.
As Fig. 11 shows, resistor 1120 has a value R1,
resistor 1121 has a value R2, resistor 1123 has a value R4,
resistor 1124 has a value R5, and capacitor 1134 has a
value C. R2 is much larger than Ri.
Circuit 1100 actually examines the lengths of
different portions of signal 1105. When that signal is
low, indicating the presence of a bar, the output of
inverter 1131 is high and charges capacitor 1134 with a
time constant of approximately R1*C, since R2 is so much
larger than R1.
For a space, signal 1105 is high and thus the output
of amplifier 1131 is low. This discharges capacitor 1134
through resistor 1121 since the diode 1132 prevents
discharging through resistor 1120. The time constant R2*C
is much greater than the time constant R1*C, so the circuit
requires a longer space to cancel the effect of a bar.
After the circuit 1100 examines several bars and
spaces from a bar code having a typical density, capacitor
1134 develops a voltage that exceeds the threshold level of
comparator 1137. This activates laser enable signal 1141.
This activation also drives the open collector output
of comparator 1137 low, dropping the threshold voltage at
input 1138. This behavior causes hysteresis to prevent
minor voltage changes on capacitor 1134 by bars, spaces,
and quiet zones from disabling laser enable signal 1141.
- 32 -
CA 02579505 2007-03-13
It might appear as if circuit 1100 would also trigger
if amplifier/digitizer 1116 produced a detected signal
representing a long black bar. This is not a problem,
however, because in the preferred embodiment,
amplifier/digitizer 1116 functions as a high pass filter to
prevent generating long signals. For example,
amplifier/digitizer circuit 1116 could produce only short
pulses of known durations for black bars, so the signal for
a long black bar would be the same as a shorter one.
Circuit 1100 is sufficiently flexible to be used,
along with software executed by CPU 140 (Fig. 1) for other
purposes. For example, circuit 1100 can help discriminate
a bar code from text or other graphics. To do so, circuit
1100 and CPU 140 exploit the uniform width of bar codes as
compared to text which has differing widths. Because of
this characteristic, different scans through different
slices of a bar code pattern will likely yield similar
results. On the other hand, different scans through
different portions of text will yield different results.
Scan lines spaced sufficiently close together may even be
used to distinguish two-dimensional bar codes from graphics
because of the uniform width of the bars and spaces in
two-dimensional bar codes.
A raster scan pattern works well with this technique
by automatically moving the scans perpendicularly. This
guarantees that successive scan lines cross parallel slices
- 33 -
CA 02579505 2007-03-13
= of the scanned pattern.
Furthermore, greater accuracy may be obtained by
controlling certain operational parameters, such as
horizontal and vertical scan angles, in response to the
type of bar code determined to be scanned. This is useful
for a single system to decode both one-dimensional and
two-dimensional bar codes.
Scanning according to this invention is not limited to
raster-type scanning. Individual control of X-axis and Y-
axis allows the system to provide a scan pattern of any
desired shape. For example, using only the X-axis controls
generates a linear scan line at the symbol. Driving the X-
axis and Y-axis controls at uniform rates of speed causes a
raster-type scan pattern having a set of generally parallel
scan lines. Driving the X-axis and Y-axis scans at
sinusoidally varying rates generates an omnidirectional
Lissajous-type scan pattern.
Information on scanning control may be found in U.S.
Patent No. 4,387,297, as well as U.S. Patent No. 5,168,149.
One last advantage of using the circuitry in Fig. 11
arises from the signal processing it performs without using
other scanner resources. This technique reduces the amount
of processing CPU 140 must perform, and thus reduces the
system's latency when reading a bar code symbol.
- 34 -
CA 02579505 2007-03-13
C. Scanning Control
1. Scanning element control
Fig. 12 shows the details of a scan engine 1200 in
accordance with this invention that is both flexible and
efficient. Scan engine 1200 can be used with the scanner
subsystems described in this specification or with
different subsystems.
As Fig. 12 shows, scan engine 1200 has three major
components: controller 1210, digital-to-analog (D/A)
converter 1220, and PWM regulator 1230. PWM regulator 1230
controls a motor coil 1240 that drives a scanning element
(not shown) and provides feedback to controller 1210 via
amplifier 1250.
Controller 1210, which is preferably a PIC16C71
manufactured by Microchip Technology, Inc., forms and sends
a digital drive signal 1215 to D/A converter 1220. Digital
drive signal 1215 is actually a series of 7-bit digital
numbers corresponding to voltage values of a signal to
drive coil 1240. As explained in greater detail below,
digital drive signal 1215 is preferably the sum of two sine
waves, one to drive the horizontal or X-axis deflection,
and one to drive the vertical or Y-axis deflection.
D/A converter 1220 generates analog linear control
voltage signal 1225 from digital drive signal 1215, and
provides analog drive signal 1225 as an input to PWM
regulator 1230. D/A converter 1220 appears conceptually as
- 35 -
CA 02579505 2007-03-13
a ladder-type converter because such converters can be made
very compact. Other converter designs will work as well,
however.
PWM regulator 1230, which includes comparator 1231,
creates a current drive signal 1235 from analog drive
signal 1225, and sends current drive signal 1235 through
coil 1240. The shape of current drive signal 1235
resembles that of analog voltage drive signal 1225, so in
one sense, PWM regulator 1230 acts as a voltage-to-current
converter.
Coil 1240 creates a magnetic field proportional to
current drive signal 1235. That magnetic field causes a
RASE (Resonant Asymmetric Scan Element) to oscillate
horizontally and a mylar support of the RASE to oscillate
vertically. In this manner, the RASE and its support act
as scanning mirror 159 (Fig. 1) does. The RASE and its
support are described in U.S. Patent No. 5,280,165.
The RASE has a high Q allowing it to continue
oscillating even after the driving magnetic field stops.
The preferred implementation takes advantage of this
feature by periodically stopping the active driving of coil
1240. The RASE continues oscillating, however, and'coil
1240 then acts as a passive sensing element picking up
information about the amplitude and the phase of the moving
RASE.
- 36 -
CA 02579505 2007-03-13
Amplifier 1250, which is connected as a buffer,
converts the signals from coil 1240 to amplitude and phase
feedback signal 1255, an input to controller 1210.
Controller 1210 uses feedback signal 1255 to ensure that it
is properly driving coil 1240.
A temperature sensor 1260 provides a second input,
temperature signal 1265, to controller 1210. Signal 1265
represents the ambient temperature proximate to scan
engine 1200.
Sensor 1260 measures the ambient temperature for two
reasons. One is to calibrate engine 1200 to account for
changes due to temperature. The other purpose is to alert
controller 1210 to stop operations if the ambient
temperature is outside of a safe operating range.
A third input for controller 1210 is the two-wire
(clock plus data) serial ZIF interface 1270. As will be
explained in greater detail below in the section entitled
"Scan engine interface," this interface allows controller
1210 to receive commands from the rest of the scanner and
send information to the scanner.
The remaining input to controller 1210 is an external
clock 1280. Clock 1280 generates a signal at the frequency
required by the preferred embodiment of controller 1210.
In addition to ZIF interface 1270 and digital drive
signal 1215, controller 1210 also sends out two other
signals. One is a "Listen" signal 1234 to tell PWM
- 37 -
CA 02579505 2007-03-13
regulator 1230 to stop generating current drive signal
1235, thereby turning coil 1240 into a pickup device.
A second signal sent by controller 1210 is the Start
of Scan or SOS signal 1290. As explained above, the SOS
signal indicates when and in what direction the RASE is
being driven horizontally.
In the preferred embodiment, controller 1210 contains
a microprocessor 1211, a look up ROM 1212, an internal
timer 1213, and an analog-to-digital (A/D) converter 1214.
Microprocessor 1211 includes an arithmetic logic unit and
internal memory. ROM 1212, which is preferably
programmable, provides tables of data and routines to
control the operations of microprocessor 1211. Timer 1213
is an interrupt timer that microprocessor 1211 uses to
control real-time operations. A/D converter 1214 converts
the analog temperature, amplitude, and phase information
into a digital format for microprocessor 1211.
When operating in its normal mode, controller 1210
produces seven-bit digital drive signal 1215 representing
the sum of two sine waves. One is approximately 290 Hz and
the other is approximately 15 Hz. The 290 Hz signal is
used to drive the X-axis deflection of the RASE, and the 15
Hz signal is used to drive the Y-axis deflection of the
mylar support.
Controller 1210 can change digital drive signal 1215
as needed. For example, controller 1210 constantly
- 38 -
CA 02579505 2007-03-13
monitors feedback signal 1255 to determine whether to
change the amplitude or frequency of digital drive signal
1215. In addition, controller 1210 can produce a signal
with a different shape or different frequencies to
accommodate other equipment.
D/A converter 1220 continuously converts digital drive
signal 1215 to analog drive signal 1225. An analog drive
signal allows scan engine 1200 to operate accurately at
higher frequencies. To understand why requires an
understanding of PWM regulator 1230.
PWM regulator 1230 receives analog drive signal 1225
at a non-inverting input of comparator 1231 and receives a
current feedback signal 1232 at the inverting input of
comparator 1231. Current sensor 1242 generates current
feedback signal 1232 as a voltage signal corresponding to
current drive signal 1235. Constructing PWM regulator 1230
in this manner forces regulator 1230 to make current drive
signal 1235 track analog drive signal 1225.
This design of PWM regulator 1230 has two advantages
over a conventional voltage drive circuit. First, voltage
drivers require a very large voltage supply to counteract
the back end of coil 1240. PWM regulator 1230 does not
because it is a current driver.
Second, a voltage driver can reduce the high Q of the
RASE by forcing through coil 1240 signals that are not at
the coil's resonant frequency. The preferred embodiment of
- 39 -
CA 02579505 2007-03-13
regulator 1230 does not reduce the Q of the RASE because
the feedback component ensures that the RASE is driven at
the resonant frequency.
Fig. 13 shows the preferred construction of PWM
regulator 1230 in greater detail. Analog drive signal
1225, shown as Vo, enters the non-inverting input of
operation comparator 1231 via resistor 1301. The inverting
input of comparator 1231 connects to a reference voltage,
Vref, via resistor 1310.
The output of open collector comparator 1231, which is
pulled up to control voltage Vcc via resistor 1320,
produces a PWM signal that inverter 1330 uses to form an
inverted PWM signal. The PWM signal controls a switch
1340, and the inverted PWM signal controls a switch 1345.
Switches 1340 and 1345 connect to opposite sides of coil
1240 to form a~H bridge.
Fig. 14 shows that switches 1340 and 1345 are
preferably transistors 1440 and 1455, respectively.
Transistors 1440 and 1455 act as switches because they are
in one of two states: saturated or nonconducting. Using
transistors in this manner reduces the rds (resistance from
drain to source) loss, and is more efficient than using
transistors in their active regions as linear drives. The
PWM regulator shown in Figs. 13 and 14 is about twice as
efficient as a linear drive. Moreover, this PWM regulator
produces a switching frequency greater than 100 KHz,
- 40 -
CA 02579505 2007-03-13
thereby minimizing the switching frequency current
components.in the motor coil.
Switches 1340 and 1345 alternately switch on and off
to create a high-frequency PWM pulse train across coil
1240. The inductance of coil 1240 integrates this pulse
train to produce in coil 1240 a piecewise linear current
proportional to the D.C. component of the pulse train.
This time-varying current approximates the summed sine
waveform of analog signal 1225.
Fig. 13 shows that in the preferred embodiment,
coil 1240 is split into two serially-connected sections,
1347 and 1348 separated by current sensing resistor Rs
1242. One side of Rs 1242 connects through resistor 1302
to the non-inverting input of amplifier 1231, and the other
side of sensing resistor 1242 connects through resistor
1312 to the inverting input of amplifier 1231. The sizes
of the resistors are chosen so that the current through
coil 1240, iL, approximately equals (Vo-Vref)/Rs.
Coil 1240 generates a magnetic field with the same
shape as current drive signal 1235, the sum of the 290 Hz
and 15 Hz sine waves. The RASE mechanically filters out
the lower frequency portion of the magnetic signal, and the
mylar support mechanically filters out the higher frequency
portion.
Fig. 15a shows the frequency response for the RASE and
its mylar support to demonstrate the mechanical filtering.
- 41 -
CA 02579505 2007-03-13
As is apparent, the mylar support will only respond to the
lower frequency signals, and the RASE will only respond to
signals in a narrow band around 280Hz.
Fig. 15b demonstrates the high Q (about 500) of a RASE
whose resonant frequency is 280Hz. The values for the
magnitude and phase differ markedly as a function of
frequency.
To drive the RASE with such a high Q requires careful
feedback monitoring. Microprocessor 1211 performs such
monitoring by extracting from the digitized feedback signal
maximum amplitude and phase (or frequency) information.
The preferred mode for feedback monitoring involves taking
periodic measurements at the same relative phase of digital
drive signal 1215. In the preferred implementation,
microprocessor 1211 takes a feedback measurement once every
19h cycles of the x-axis drive, and times the measurements
to begin at zero-crossing and end at the maximum amplitude
or quarter-cycle point. If the measured value differs from
the intended value, controller 1210 is out of phase with
the RASE, and microprocessor 1211 changes the frequency to
correct this condition.
To correct the amplitude and frequency values of
signal 1215, microprocessor 1211 uses several values set
during calibration at the factory. This is the preferred
time to insert baseline values into ROM 1212.
Of course the values stored in ROM 1212 at calibration
- 42 -
CA 02579505 2007-03-13
may not always remain accurate. Changes in temperature,
the RASE, or other factors might require microprocessor
1211 to generate different values for driving signal 1215.
This is why microprocessor 1211 makes adjustments based on
feedback. To make adjustments, microprocessor 1211
obtains from ROM 1212 two sets of amplitude values for
X-axis deflection and one frequency (or sine wave period)
setting. In the preferred embodiment there is no Y-axis
adjustment, just a set of amplitude values for X-axis
adjustment. To correct Y-axis deflection, however, ROM
1212 could contain an additional set of amplitude values.
For the X-axis deflection, ROM 1212 includes baseline
values representing the maximum deflection, and coarse
amplitude adjustment values. The coarse adjustment values
are five percent of the corresponding baseline values. In
the preferred embodiment, each set of values for the high
frequency or X-axis signal has sixteen entries, and each
set of values for the low frequency or Y-axis signal has
twenty-six entries. In the preferred embodiment, the
values represent successive values for sine waves.
The frequency value tells microprocessor 1211 how
often to send out new values for the digital drive signal
1215. For example, for a 300 Hz sine wave, microprocessor
1211 would have to send out a complete set of values every
1/300 Hz or 3.33 ms. If a complete set included sixteen
values, microprocessor 1211 would need to provide a new
- 43 -
CA 02579505 2007-03-13
value every 3.33ms/16, or 208 us. Preferably, interrupt
timer 1213 is set to the value necessary to generate the
digital drive signal 1215 at the desired frequency.
Microprocessor 1211 makes frequency adjustments by
loading a new value into interrupt timer 1213. Amplitude
adjustments are not very complicated either.
Microprocessor 1211 forms new values for signal 1215 by
subtracting values using coarse adjustment values from the
baseline and then using a fine adjustment technique.
Fig. 16 illustrates how coarse adjustments are made.
Trace 1610 contains several baseline amplitude values, and
trace 1620 contains several coarse adjustment values to be
subtracted from the baseline values. For example, to form
an amplitude signal fifteen percent lower than the
baseline, shown as trace 1630, microprocessor 1211
subtracts the coarse adjustment values (each at five
percent of the corresponding baseline value) from the
baseline values three times.
Fig. 17 shows how fine adjustments are made in the
preferred implementation. Fine adjustment also requires
repeated subtraction, but instead of using a percentage of
baseline values, microprocessor 1211 uses a fixed offset.
The offset is positive when the baseline value is positive,
negative when the baseline value is negative, and zero when
the baseline value is zero. Fine adjustment transforms
trace 1710 into trace 1720. The fine adjustment tends to
- 44 -
CA 02579505 2007-03-13
distort the sine waveform slightly, but with little
noticeable effect.
In addition, during calibration, microprocessor 1211
uses the ambient temperature signal 1265 to adjust the
frequency of the RASE according to the temperature. Fig.
18 shows a typical relationship between temperature in
degrees centigrade and resonant frequency in Hz for a RASE.
Because the RASE has such a high Q value, it is important
to ensure proper temperature compensation.
In the preferred embodiment, microprocessor 1211 forms
the values for the higher frequency sine wave as described
above, by determining a new voltage value whenever
interrupt timer 1213 generates an interrupt signal.
Microprocessor 2111 forms a new value for the lower
frequency sine wave by accessing ROM 1212 after obtaining
some predetermined number of new values for the higher
frequency signal. For example, to form a 15 Hz signal from
a table with twenty-six values, microprocessor 1211
retrieves a new value for the lower frequency signal every
.75 cycles of a 300 Hz higher frequency signal.
Fig. 19 demonstrates the formation of a high frequency
sine wave for a 300 Hz signal with sixteen values, and a
low frequency sine wave with a new value every .75 cycles
(or every twelve values) of the high frequency sine wave.
As explained above, for a 300 Hz signal with sixteen
values, microprocessor 1211 would receive an interrupt
- 45 -
CA 02579505 2007-03-13
= every 208 ps. Fig. 19 also shows the relationship between
the SOS signal and digital drive signal 1215. The level of
the SOS signal indicates the polarity of the drive signal
1215, and thus the direction of the scan.
This basic understanding of the key components of scan
engine 1200 permits an overview of the entire operation of
scan engine 1200. Fig. 20 contains a flow diagram 2000
showing the operations necessary for motor control.
First, the procedure is started either initially or
after some sort of reset. (Step 2010) Next,
microprocessor 1211 reads the input port from the serial
interface 1270 to see what commands to perform.
(Step 2015) The first command will likely set some initial
values.
Microprocessor 1211 then measures the temperature.
(Step 2020) If the temperature is too hot, then
controller 1200 sends a LASER OFF signal to shut down the
scanner. Otherwise, the procedure continues, and
microprocessor 1211 sets the frequency (i.e., period) for
the X-axis. (Step 2025)
Next, the X-axis scan starts by providing a maximum
input or "kick." (Step 2030) Microprocessor 1211 then
measures the maximum amplitude for the X-axis. (Step 2035)
The "kick" continues until the maximum X-axis amplitude
reaches the baseline value. At this point, the Y-axis
signal deflection starts. (Step 2040)
- 46 -
CA 02579505 2007-03-13
Next, microprocessor 1211 reads the control port again
to get the next command. (Step 2045) If the command is to
modify the X or Y amplitude, then the modification is
accomplished (Step 2050), and the control port is queried
again for another command. (Step 2045)
If the command indicates that the symbol is two
dimensional, microprocessor 1211 opens the Y-axis to
increase the height of the scanning pattern. (Step 2055)
After doing so, microprocessor 1211 modifies the X or Y
amplitudes to reflect the change in the scanning pattern.
(Step 2060) Afterwards, microprocessor 1211 reads the
control port for another command. (Step 2065)
If the command indicates that the scan pattern needs
to be modified again, then microprocessor 1211 complies
(Step 2060), and reads the control port for the next
command. (Step 2065)
At this point, or during an earlier read of the
control port (Step 2045), the controller 1210 may receive a
command indicating a decode to check the scan engine 1200.
In response, microprocessor 1211 brakes the RASE
(Step 2070), and waits for the next operation. Preferably,
microprocessor 1211 brakes the RASE by driving it 180
degrees out of phase.
Fig. 21 contains a flow diagram 2100 showing the steps
for generating the proper raster scans. After starting
(Step 2110), microprocessor 1211 sends the summed sine wave
- 47 -
CA 02579505 2007-03-13
signals to D/A converter 1220. (Step 2115) This will
initiate or continue scanning.
Next, microprocessor 1211 determines whether it needs
to change the amplitude or frequency values in response to
a command from the control port. (Step 2120) If so,
microprocessor 1211 modifies the appropriate registers for
frequency, X-axis amplitude, or Y-axis amplitude.
(Step 2125)
After modifying the registers, or if no change is
required, microprocessor 1211 retrieves the next baseline
value for the X-axis (Step 2130), and gets the
corresponding coarse and fine adjustment values
(collectively X'). (Step 2135)
Then a loop begins in which the X' adjustment values
are repeatedly subtracted from the X baseline value
(Step 2140) a number of times equal to a stored value
"xcntr." (Step 2145) Xcntr indicates how much the
baseline value needs to be modified. Next, microprocessor
1211 performs a similar set of operations for the Y-axis
values. As explained above, this operation occurs every
.75 cycles of the X-axis sine wave in the preferred
embodiment. Similar to the X-axis loop, microprocessor
1211 retrieves the baseline value for the Y-axis
(Step 2150) and the corresponding Y fine and coarse
adjustment values Y'. (Step 2155) Next, microprocessor
subtracts the Y' adjustment values (Step 2160) from the
- 48 -
CA 02579505 2007-03-13
baseline Y value a number of times equal to "ycntr," a
stored value similar to xcntr. (Step 2165)
After determining the adjusted X and Y values,
microprocessor 1211 adds those values together.
(Step 2170) Then it waits for a timer interrupt.
(Step 2175)
When the interrupt occurs, microprocessor 1211 reloads
a timer value into an X-axis frequency control register.
The timer value is either a coarse frequency adjustment,
which is preferably the value used between all but two
sample points, or a fine frequency value, which is
preferably the value used between two sample points to make
minor frequency adjustments. (Step 2180) The entire
process then repeats with the sum of the adjusted X-axis
and Y-axis amplitude values being sent out to D/A converter
1220. (Step 2115)
To complete the understanding of this invention, it is
useful to know the specific variables stored in ROM 1212 in
the preferred implementation. During initial programming
at the factory, technicians preferably assign a value of 30
for maximum Y amplitude variable. Microprocessor 1211 in
scan engine 1200 determines coarse adjustment and fine
adjustment from the value.
The technician enters the following 8-bit words into a
calibration table location in the ROM 1212.
Ymax: Maximum Y amplitude (largest scanning
- 49 -
CA 02579505 2007-03-13
pattern height)
Ymin: Minimum Y amplitude (smallest scanning
pattern height)
Xdrive: X amplitude setting used at startup to
generate coarse and fine adjustment values
Fcrs: Coarse frequency value (distance between all
sample points except for adjustment points)
FFine: Fine frequency value (used during adjustment
points)
Phase: Phase reference value for x axis frequency
control (value use to find maximum value for
phase measurement)
Fback: Amplitude reference used when examining the
feedback signal
Kcount: Number of X-axis cycles to apply the Kick.
This is the only section of the program space of
controller 1200 that changes for each scanner.
The calibration table area in ROM 1212 is large enough
to support reprogramming. Preferably each scanner's
calibration table can be programmed up to four times. To
erase old table values, microprocessor 1211 overwrites the
table space having those values with NOP instructions
(0016). It then writes the new calibration table into new,
previously reserved memory locations of ROM 1212. When
microprocessor 1211 accesses the calibration table, it
skips the NOP instructions and increments a program counter
- 50 -
CA 02579505 2007-03-13
until it finds the new table entries.
Usually, a technician need only reprogram the
calibration table when a motor is replaced during servicing
or if the alignment operator in the factory needs to
correct motor alignment errors. In the factory, the scan
engine is placed on an alignment fixture that automatically
aligns the motor to produce calibration table values.
After a reset, microcontroller 1211 loads the
calibration tables from ROM 1212 into a RAM in
microprocessor 1211. As the section below entitled "Scan
engine interface" explains, the RAM locations can be
modified by using the Calibration Mode to effect "on the
fly" scan pattern changes. In most cases, however, the
scan engine interface allows dynamic scan pattern changes
by having the decoder specify amplitudes and opening rates
by through commands such as OPEN Y. The normal product
commands do not change calibration values, however, and are
not part of the Calibration Mode. Scan engine 1200 product
commands into control variables microprocessor 1211 uses to
manipulate baseline values.
As is apparent from the foregoing description, the
design of scan engine 1200 makes its scanning operation
powerful and efficient. Another advantage of scan
engine 1200 is that it requires very little room. Fig. 22a
shows a perspective view of a preferred assembly containing
a laser 2210, the RASE mirror 2220, the mylar assembly
- 51 -
CA 02579505 2007-03-13
2225, and motor coil 1240.
Fig. 22b shows a different perspective view from below
the view in Fig. 22a, again showing the laser 2210, with
the scan engine drive printed circuit board (PCB) 2250
mounted below laser 2210 and coil 1240. The scan engine
drive PCB 2250, which contains the electronics shown in
Fig. 12 has controller 1210 as its largest element.
Fig. 22c shows the relationship of the detector 2220
and its PCB 2260. The decoder circuitry is preferably
located on a different PCB at another portion of the
scanner.
2. Precession
Another feature of the preferred embodiment that is
not readily apparent is the reason for choosing the lower
frequency at 15 Hz. After all, Fig. 15a suggests several
different frequencies would provide an adequate response.
The reason for choosing 15 Hz is that the X-axis scanning
frequency of 290 Hz is not an integer multiple of 15 Hz.
The actual ratio is about 19.5.
When the X-axis scanning frequency is not an integer
multiple of the Y-axis scanning frequency, the scanning
pattern will tend to precess, or roll, because the
movements along the two axes do not begin simultaneously.
Although some may find such precession a disadvantage, it
can actually be helpful in scanning different parts of a
bar code symbol.
- 52 -
CA 02579505 2007-03-13
For example, a bar code may have some portions that
are difficult to read or the rows may be closely packed.
In that case, precession of the scanning pattern ensures
that successive scans traverse different parts of the
symbol and provide more accurate readings,
Fig. 23a-23c shows successive scanning patterns during
precession. As the patterns demonstrate, the scans
traverse different paths through the pattern to provide
better overall coverage.
In the preferred implementation, the precession, whose
rate depends on the ratio of the X-axis and Y-axis
frequencies, is designed into each scanner. Proper
programming of ROM 1212, such as during calibration, can
change or even eliminate precession.
3. Scan engine interface
As the previous discussions suggest, scan engine 1200
must communicate with other portions of scanner 100. Fig.
24 is a block diagram showing the interfaces between scan
engine 1200 decoder 2410, digitizer 2420, and optical
detector 2430. In Fig. 24, optical detector 2430 receives
the reflected light and outputs electrical signals
reflecting the levels of received light. Digitizer 2420
forms the DBP signals discussed above from those electrical
signals.
Decoder 2410 represents the circuitry for both
decoding the bar code data and controlling the operation of
- 53 -
CA 02579505 2007-03-13
scanner 100. As Fig. 24 shows, decoder 2410 receives SOS
signal 1290 from scan engine 1200, and exchanges commands
and data with microprocessor 1211 via ZIF serial interface
1270. Decoder 2410 also receives DBP data from digitizer
2420 and sends automatic gain control signals back to
detector 2430 to ensure high quality resolution.
Most of the signals and interfaces shown in Fig. 24
have been described above. An important one which has not
is the ZIF serial interface 1270. The interface is very
powerful because it allows scan engine 1200 to operate with
the rest of scanner 100 and produce the advantages listed
in the prior description.
Fig. 25 shows a preferred format for the commands and
messages exchanged over ZIF interface 1270. The commands
and messages are sent in eight-bit bytes 2500. Each byte
has two four-bit nibbles, shown as nibble 2510 and nibble
2520. Nibble 2510 preferably carries the command and
message identifier, and nibble 2520 carries any data
required by the message of the command.
Fig. 26 shows a list of the commands and messages
exchanged over ZIF serial interface 1270 between scan
engine 1200 and decoder 2410 in the preferred embodiment.
There are eight commands from decoder 2410 to scan
engine 1200 and five messages from scan engine 1200 to
decoder 2410.
The CALIBRATION command is principally used in the
- 54 -
CA 02579505 2007-03-13
factory to calibrate scan engine 1200. In the preferred
embodiment, microprocessor 1211 only responds to the
CALIBRATION command if the SOS signal 1290 is low. If not,
microprocessor 1211 will ignore this command.
The specific purpose of calibrating is to change or
read the RAM in microprocessor 1211. Calibration provides
a first approximation of the scanning element parameters
that are later adjusted on feedback.
Nibble 2520 in the CALIBR.ATION command can indicate
that the RAM mode should be read (0001), that the RAM mode
should be written (0000), or that the calibration mode
should be exited (0010). If the scan engine is already in
the calibration mode, additional modes, such as read, write
or exit, can be set without holding SOS signal 1290 low.
To change the contents of a RAM address, decoder 2410
places scan engine 1200 in the Write RAM Calibration mode
and sends it to the desired RAM address (one byte). In
response, scan engine 1200 echoes the address to decoder
2410, which then checks whether the echo was correct.
Next, decoder 2410 sends data for that RAM address, and
scan engine 1200 echoes the data for decoder 2410 to check.
To read a RAM location, decoder 2410 places scan
engine 1200 in the Read RAM Calibration mode and sends the
desired address of the RAM (one byte). In response, scan
engine 1200 sends back the data from the indicated RAM
address.
- 55 -
CA 02579505 2007-03-13
These two mechanisms allow decoder 2410 to change the
parameters of scan engine 1200 dynamically. For example,
this is how decoder 2410 could change the X-axis and Y-axis
amplitudes during scanning.
The AIM SLAB command tells scan engine 1200 to place
the scan element into a predefined pattern for aiming or
initial acquisition. This command does not control the
laser, however.
The AIM DOT command tells scan engine 1200 to place
the scan element into a predefined pattern for dot aiming
in the sunlight. This command also does not control the
laser.
The SET Y ANGLE command tells scan engine 1200 the
maximum angle to open in the Y direction in response to an
OPEN Y command later issued. This command has data in
nibble 2520 identifying one of sixteen possible angles.
The OPEN Y command tells scan engine 1200 to open the
Y raster pattern at one of sixteen possible opening rates.
The rate is encoded in nibble 2520. If the OPEN Y command
is sent without a SET Y ANGLE command, scan engine 1200
will select as a default the maximum angle. If another
angle is desired, the OPEN Y command should immediately
follow the SET Y ANGLE command.
The CLOSE Y command tells scan engine 1200 to close
the Y raster pattern. The rate is encoded in nibble 2520
and can have one of sixteen possible values.
- 56 -
CA 02579505 2007-03-13
The BRAKE command tells scan engine 1200 to apply
braking to the X motion, the Y motion, or both. Parameters
in nibble 2520 indicate which motion should be braked.
The STATUS REQUEST command asks for status from scan
engine 1200. The scanner responds with the current status
of the scan element and the laser.
The OPEN Y DONE message is a status message that scan
engine 1200 has reached the maximum programmed Y opening.
This message only follows an OPEN Y command.
The CLOSE Y DONE message is a status message sent when
scanning engine 1200 finishes closing the raster in the Y
direction. It only follows a CLOSE Y command.
The STATUS message is scan enqine 1200's response to a
STATUS command. It indicates the status of the laser and
the scan element in nibble 2520.
The RESET DONE message is sent following a successful
power up, and indicates that scan engine 1200 may now
receive commands on ZIF serial interface 1270. Nibble 2520
contains the version number of the scan engine software
executed by microprocessor 1211.
4. Simplified scanning height control
Certain aspects of the inventions described above do
not require the use of a scan engine such as scan engine
1200. For example, the change in scanning height can be
accomplished with much simpler circuitry that lacks the
flexibility and robustness of scanner engine 1200. Fig. 27
- 57 -
CA 02579505 2007-03-13
shows one type of circuit 2700.
Circuit 2700 receives as an input signal 2710
indicating that the bar code being sensed is a
two-dimensional bar code. Signal 2710 feeds an amplitude
control circuit 2720 generating a control signal V,, to
increase the height of a scanning pattern when the symbol
being sensed is encoded in a two-dimensional bar code.
Fig. 28 shows V, as a function of time. V, remains
at an initial amplitude V1 until scanning with a larger
pattern is initiated at time to. At that time, V,, climbs
linearly until time tl, when the pattern has expanded to
its maximum height. At that time, Vc is at voltage VZ.
In circuit 2700, multiplier 2740 mixes V,, with the
signal from Y-axis oscillator 2730. Y-axis driver 2750
then uses the mixed signal to control the Y-axis scanning
element 2760, such as the mylar mount for the RASE.
Circuit 2700 therefore operates for all modes. When
scanner 100 is not scanning, or when it is scanning
one-dimensional bar codes, V,, remains at its initial value
V1. When scanner 100 is scanning two-dimensional bar
codes, circuit 2700 increases Vc to generate a larger
pattern.
Fig. 29 shows a preferred embodiment of amplitude
control circuit 2720 to generate Vc. Signal 2710 controls
analog switch 2910 connected in parallel across a charging
capacitor 2920 having a value C1.
- 58 -
CA 02579505 2007-03-13
One side of both capacitor 2920 and switch 2910
connects to the inverting terminal of operation amplifier
2930 and to one side of variable resistor 2940. The other
side of resistor 2940 connects to ground. The value of
resistor 2940 is Rs. The other side of capacitor 2920 and
switch 2910 connects through resistor 2950 to the output of
operational amplifier 2930. The value of resistor 2950 is
R9.
The non-inverting input of operational amplifier 2930
connects to the junction of a voltage divider formed by
resistors 2960 and 2965 having values R. and R7,
respectively. Resistor 2960 also connects to a supply
voltage Vcc, and resistor 2965 also connects to ground.
This circuit sets a voltage Vj at the junction between
resistors 2960 and 2965.
The junction of switch 2910, capacitor 2920, and
resistor 2950 also connects to a zener diode 2970 and forms
the output voltage signal Vo. Zener diode 2970, which has
a breakdown voltage of VZ, also connects to ground. The
value of resistor 2950, R9, is chosen to limit the current
in zener diode 2970 to safe levels.
Output voltage Vo appears at one end of
potentiometer 2980, whose value is Rio. The other end of
potentiometer 2980 is ground, and the control voltage V,
appears at potentiometer 2980's wiper arm.
When the scanner is not scanning a two-dimensional bar
- 59 -
CA 02579505 2007-03-13
code symbol, signal 2710 is LOW and switch 2910 is closed.
This discharges capacitor 2920 and forces Vo to equal V,.
That, in turn, sets V, to a constant value.
When scanning a two-dimensional bar code, signal 2710
is HIGH, switch 2910 is open. This charges capacitor 2920
at a rate set by Vj, R8, and C1. Circuit 2720 then operates
as an integrator, causing the voltage V,, to grow linearly
until reaching the breakdown voltage V. of zener diode
2970. At that point V,, will rise no further. Control
voltage Vc follows suit. Voltage Vo remains at voltage VZ
until switch 2910 closes, discharging capacitor 2920
rapidly and forcing Vo to decrease to V,.
VI. CONCLUSION
The foregoing embodiments and implementations of this
invention are not intended to limit the invention to the
details shown. Instead, various modifications and
structural changes may be made without departing from the
spirit of the present invention.
The foregoing description fully reveals the gist of
the present invention that others can readily adapt it for
various applications without omitting features that, from
the standpoint of prior art, fairly constitute essential
characteristics of the generic or specific aspects of this
invention. Therefore, such adaptations should and are
intended to be comprehended within the meaning and range of
equivalence of the following claims.
- 60 -
