Single-Precision Real/Double-Precision Real Data - Mitsubishi Electric MELSEC Q Series Programming Manual

Single-precision real/double-precision real data

Single-precision real/double-precision real data are 32-bit floating-point data used in basic instructions and application
instructions.
For FXCPU, double-precision real data is not supported.
Real number data can be stored only in word devices.
Single-precision real (single-precision floating-point data)
Devices used in lower 16 bits are specified for instructions that use real number data.
Real number data are stored in 32 bits of 'specified device number' and 'specified device number +1'.
M0
EN
s
Var_R100
Two points (32 bits) of R100 and R101 are the target word devices
Floating-point data are represented by two word devices.
[Sign] 1. [Fraction]  2
The following explains the bit configuration and its meaning when floating-point data are internally represented.
b31 b30 to
b31 b23 to b30
Sign Exponent
• Sign: b31 represents a sign.
• 0: Positive
• 1: Negative
• Exponent: b23 to b30 represent n of 2
b23 to b30 FF
H
n
Not used 127
• Mantissa: Each of the 23 bits, b0 to b22, represents the "XXXXXX..." portion when the data is represented in binary,
"1.XXXXXX...".
4 PROGRAM CONFIGURATION
42
4.4 Method for Specifying Data
Transfers real number data
EMOV
ENO
Two points (32 bits) of D0 and D1
d
Var_D0
are the target word devices
[Exponent]
b23 b22 to b16
n
. The values of n are as follows according to BIN values of b23 to b30.
FE FD 81
H H H
126 2
b15 to
b0 to b22
Fraction
80 7F 7E
H H H
1 0 -1
b0
02 01 00
H H H
-125 -126 Not used