Mitsubishi Q00JCPU User Manual page 188

3
SEQUENCE PROGRAM CONFIGURATION AND
EXECUTION CONDITIONS
High
Basic
Performance Process
(2) Double-precision floating-point data
Note3.40 Note3.40 Note3.40
Redundant
Note3.40
b52 to b62
Note39
High
Basic
Performance Process
Note3.40 Note3.40 Note3.40
Redundant
Note3.40
3
- 90 3.9 Numeric Values which can be Used in Sequence Programs
3.9.4 Real numbers (Floating-point data)
(a) Internal expression of real number data
The following shows the internal expression of real number data dealt with the
Universal model QCPU.
Using four word deices, the real number data is expressed as follows:
[Sign] 1.[Mantissa]
The bit configuration of internal expression for the real number data is as follows:
b63 b62 b52
to
b63
b52 to b62
Exponent part (11bit)
Sign
Figure 3.72 Bit configuration of real number data
1) Sign
The b63 shows a sign.
0: Positive
1: Negative
2) Exponent part
The "n" of 2n is expressed by b52 to b62, depending on the BIN value in b52
to b62.
7FF 7FE 7FD
H
H
n
Not used 1023 1022
Figure 3.73 Relation between value stored in exponent part and exponent
3) Mantissa:
For a binary number "1.XXXXXX.....", the "XXXXXX....." portion is expressed
by 52 bits (b0 to b51).Note39
The Basic model QCPU, High Performance model QCPU, Process CPU, and Redundant CPU
cannot use the double-precision floating-point data.
Note3.40
(exponent part)
2
b51 b16 b15
to
400 3FF 3FE
H H H
2 1 0
to
b0 to b51
Mantissa (52bit)
3FD 3FC 02
H H H H
-1 -2 -1021
b0
01 00
H H
-1022 Not used