Specifications

000100 1101
λ
Incremental
signal
M code
0001
0010
Incremental signal detection
Absolute signal detection
Sin Cos
bit1 2 3 4...
Output signal
- Wavelength: 40, 80 μm, etc.
- Phase angle accuracy:
Approx. 0.1 to 0.2 μm
A sinusoidal signal of approx.
1/400 (52 dB) is required.
Factors that deteriorate
the phase angle
- DC, gain, or phase variations
- Noise
- Harmonics
Example of the measurement of Lissajous figure
Tek
stop
1
200mv ch2 200mv M 100ms A ch2 ∫ 976mvch1
y
X
y
X
An M code is a code string
consisting of two values (0 and 1)
known as M sequence and is a
cyclic code whose period is N =
2n-1. Where “n” is the order of a
primitive polynomial that
generates an M code. In the
scales, it is the number of digits
of M-code bits (value written as
“18 bits” above). There is a
primitive polynomial for each
order, and a code string is serially
created from any initial value. As
the creation and features of
codes, the following shows an
example of the order “8” in the
condition that the number of
digits of a bit = the order of a
primitive polynomial.
Assuming that a creating
polynomial is F(x) = x0 + x2 + x3
+ x4 and the 8-bit initial value is
00000001, x0 + x2 + x3 + x4 = 0
+ 0 + 0 + 0 = 1 (“+” is exclusive
OR and the order and bit are in
reverse, so that a higher-level
order corresponds to a
lower-order bit) is established,
making the initial value 8-bit’s
next data “1.” The following
performs the same logical
calculation for data 00000011 in
which the initial value is moved
by one, to obtain code
00000001100111110101… In
this case, when the code string
created by the order “8” is
observed as 8-bit data (a set of 8
pieces) one by one, the code
strings of 8-bit data can be
obtained as follows:
00000001,
00000011,
00000110,
00001100,
00011001, …
In this case, for M code = cyclic
code, when the sequence is
cyclically replaced in a period (N
= 2n-1), the code word of that
sequence always becomes
different. That is, the same value
of 8-bit data is never presented.
Moreover, the code string is
composed of apparently random
code strings, which are not
arranged in an ascending or
descending order or in other
rules as seen from the example
above.
Configuration of absolute
encoder detection
M codes are non-repetitive
codes in which n-bit M codes are
created by a generating
polynomial of n bits and that
there is no identical code among
2n-1 data. Because codes of 2n
are required for a rotary encoder,
it adopts the 2-track M-code
system that adds 0 (all 0).
- Lamda: 40, 80 microns
- Number of M-code bits: Up to
18 bits
Figure 1 below: Example of 4-bit
codes
Incremental/absolute signal
record reproduction to signal
composition
- Phase angle in λ is calculated
from an incremental signal and
taken as positional information
in λ.
- An address in λ is calculated
from each absolute signal and
an M code is recognized by
M-code sensors (multiple).
The M code is decoded into an
address in λ.
- Both data are added.
- Data format conversion to
communication protocol
- Communication (Figure 2
below)
Incremental signal Absolute signal
An incremental signal is taken
as the source of scale
performance and does not have
absolute positional information
of length longer than the
wavelength.
In magnetic scales, the S and N
poles are regularly arranged at
wavelength intervals.
Automatic consecutive
compensation
Because signals are detected
at a maxmum resolution of 0.01
μm, consistent signal detection
is achieved at high precision
even in case of unmatched
parallelism or in an environment
in which the temperature varies.
Improved signal stability, etc.
by harmonics and DC
cancellation or averaging
effects
Improved signal by the
arrangement of
MR-element patterns
Automatic consecutive
compensation
Compensation
Figure 1 Figure 2
Incremental sensor (MR)
M-code sensor (MR)
Detector circuit
Interpolation
circuit
Reading
compensation
M-Code
to
Binary
Addition
Serial
IF
＊
＊
Serial
communication
Compensation circuit
10 11