Data processing: measuring – calibrating – or testing – Calibration or correction system
Reexamination Certificate
2002-10-09
2004-08-03
Barlow, John (Department: 2863)
Data processing: measuring, calibrating, or testing
Calibration or correction system
C702S097000, C702S158000, C702S106000, C702S159000, C702S161000, C073S001790, C073S001810, C356S498000, C356S496000, C356S003000, C033S706000, C033S707000, C033S700000
Reexamination Certificate
active
06772078
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a linear scale reader.
2. Description of the Prior Art
Conventionally, some of machine tools are provided with, for example, glass linear scales for reading the position of a workpiece. In the linear scale, light is illuminated onto the linear scale and a relative travel distance of a workpiece can be measured using the reflected light.
In some known linear scales, magnetic score marks are inscribed in a metal scale and movement of the scale can be read by means of variations in magnetic field.
Such a linear scale generally is provided with an operation circuit which computes lower measured values obtained by sub-dividing one pitch of the scale, in order to obtain higher precision than a pitch between score marks.
FIG. 4
shows schematically a method of reading the above-mentioned linear scale. Referring to
FIG. 4
, a glass linear scale
10
, which forms a length measuring apparatus, is provided a main scale
10
a
and a sub-scale
10
b.
Each of the scales
10
a
and
10
b
includes a grid having score mark formed at predetermined intervals. A relative movement between the main scale
1
a
and the sub-scale
1
b
is detected by detecting Moire fringes of light passing through the score marks of the two scales.
An optical system
20
illuminates light onto the linear scale
10
. The optical system
20
generally includes a light emitting element
21
and light receiving elements
22
(
a
) and
22
(
b
). The light emitting element
21
illuminates light onto the linear scale
10
. Each of the two light receiving element
22
(
a
) and
22
(
b
) detects a Moire fringes based on translucent light or reflected light of the illuminated light and then converts them into an electrical signal. When the scale travels at a constant rate, a relative movement thereof is output as a sine-wave-like Lissajous' figure.
In this case, by shifting the position of a received light spot, the light receiving element
22
a
produces an A-phase signal while the light receiving element
22
b
produces a B-phase signal, the A-phase signal and the B-phase signal being shifted 90° from each other. Thus, the travel direction of the scale can be detected.
The A-phase signal is supplied to the amplifier
23
a
and the B-phase signal is supplied to the amplifier
23
b
. The amplifier
23
a
,
23
b
is formed of, for example, a high-gain differential amplifier. The amplifier
23
a
converts the A-phase signal into a voltage signal level of about 2 Vp-p and the amplifier
23
b
converts the B-phase signal into a voltage signal level of about 2 Vp-p. The amplifier
23
a
supplies the converted signal to the A/D converter
24
(
a
), which converts an analog signal into a digital signal. The amplifier
23
b
supplies the converted signal to the A/D converter
24
(
b
).
The comparator
25
receives the outputs of the amplifiers
23
a
and
23
b
and then converts them into rectangular waveforms inverted to zero levels. For example, the upper counter
26
receives the B-phase signal, of which the phase is shifted 90°, and then counts, for example, the rising edges of rectangular waves.
When the main scale
10
a
shits from the sub-scale
10
b
by one pitch, the upper counter
26
outputs the measured length data N of upper bits, in which the count value is incremented by 1.
The A/D converter
24
a
,
24
b
samples sine wave detection signals, each representing a relative value of an input sine wave scale, every predetermined phase intervals, and then outputs them as digital values.
In this case, the A/D converter
24
(
a
) samples an A-phase detection signal and then outputs the sampled signal as a digital signal. The A/D converter
24
(
b
) samples a B-phase detection signal and then outputs the sampled signal as a digital signal. As described later, the phase division data of a sine wave signal is computed based on both the sampling values. The phase division data with high precision, obtained by further dividing the upper data, is output. That is, a ROM table
27
which previously stores lower data is read out, with the outputs of the A/D converters
24
a
and
24
b
acting as address signals. Thus, the lower data n, obtained by multiplying one pitch of the scale by a phase division number, is output.
An adder
28
adds the lower measured length data n read out from the ROM table and the upper data N in which one pitch unit of the scale is a measured length value and then supplies its output to a measured length display (not shown).
The measured length display latches and manifests the value of a request signal.
FIG. 5
shows a sine-wave-like A-phase signal iA and a sine-wave-like B-phase signal iB, created from Moire fringes generated when a linear scale is relatively moving at a fixed rate.
Ad represents an upper signal waveform output from the comparator
25
, inverted at the zero level of an A-phase signal iA. Bd represents an upper signal waveform output from the comparator
25
, inverted at the zero level of an B-phase signal iB. In this example, when the scale is moving in one direction, the upper counter
26
produces an addition output of upper bits N at the time the B-phase signal from the comparator
25
rises. When the scale is moving in the opposite direction, a subtraction output from the upper counter is output at the time the B-phase signal falls.
Both the sine-wave Lissajous' figure A supplied from the A/D converter
23
a
and the sine-wave Lissajous' figure B supplied from the A/D converter
23
b
are sampled every predetermined phases, as shown in FIG.
5
. Thus, the lower bit data (n) can be read out from the ROM table, with the sampling data acting as an address signal. As shown in
FIG. 5
, the lower data n takes a value increasing stepwise and linearly every pitch. By adding upper data N, the resolution of the scale to a relative moving distance is improved.
In the above-mentioned linear scale reading method, because upper data N and lower data n are not output in a synchronous mode, an error may occur in the vicinity of a digit-taking-up of upper data (or a digit-taking-down of upper data).
This process will be explained with reference to FIG.
6
.
Referring to
FIG. 6
, n represents lower position data output when a linear scale is relatively moving at a constant rate. N represents upper count data. The measured length display manifests a value (N+n).
When the lower data becomes 99 normally with the sampling timing Sy, Sx, it is read out with the timing at which the upper data is incremented.
The point where the digit of upper data takes up corresponds to the timing where inversion occurs at the zero level of the B-phase signal. However, the point where the A-phase signal or B-phase signal and the zero level cross changes, for example, due to noises induced slightly. Moreover, the cross point may change due to dust adhered to during movement of the scale.
A change of the cross point makes unstable the timing with which the upper count value N increments, as shown in FIG.
6
.
It is now assumed that the lower measured length data is n and that the resolution is 1/100. In the case of n=100, one pitch (0.1 mm) is obtained. In such a case, with the sampling timing S
1
, because the upper data N is 0 and the lower data n is 90 &mgr;m, the relative moving distance is 90 &mgr;m.
However, because taking up, or carry, of the digit of the upper data erroneously speeds up with the sampling timing S
2
, the moving distance to be displayed jumps to 290 &mgr;m when the lower data n is 90. Because the moving distance is actually 190 &mgr;m, the error is very large.
With the sampling timing S
3
, the upper count data N is 2 and the lower data n is 10.
Therefore, the relative moving distance of the scale is 210 &mgr;m. However, the accurate measured length value is 310 &mgr;m with the sampling timing S
3
.
As described above, the conventional linear scale measured length reading method has such a disadvantage. That is, because different circuits create lower data n and upper data N, respecti
Barlow John
Dougherty Anthony
Futaba Corporation
LandOfFree
Linear scale reader does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Linear scale reader, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Linear scale reader will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3358946