Mitsubishi Electric MELSEC iQ-R Series Programming Manual page 51

Configuration of double-precision real number data
Double-precision real number data consists of a sign, mantissa, and exponent, and is expressed as shown below.
1.
Sign
The following figure shows the bit configuration of the internal expression of double-precision real number data and the
meaning of each part.
b63 b62 ∙∙∙ b52 b51
Sign Exponent
■Sign (1 bit)
This bit represents the positive or negative sign of a numerical value. "0" indicates a positive number or 0. "1" Indicates a
negative number.
■Mantissa (52 bits)
A mantissa means XXXXX of 1.XXXXX2
■Exponent (11 bits)
An exponent means N of 1.XXXXX2
relationships between the exponent value and N of a single-precision real number.
Exponent (b52 to b62) 7FFH
N Not used
Precautions
■When setting an input value of single-precision real number from the engineering tool
The number of significant digits is about 7 because the engineering tool processes single precision real number data in 32-bit
single precision.
When the input value of single-precision real number data exceeds 7 digits, the 8th digit is rounded off.
Therefore, if the rounded-off value goes out of the range from -2147483648 to 2147483647, it will not be an intended value.
Ex.
When "2147483647" is set as an input value, it is handled as "2147484000" because 8th digit "6" is rounded off.
Ex.
When "E1.1754943562" is set as an input value, it is handled as "E1.175494" because 8th digit "3" is rounded off.
■When setting an input value of double-precision real number from the engineering tool
The number of significant digits is about 15 because the engineering tool processes double precision real number data in 64-
bit double precision.
When the input value of double-precision real number data exceeds 15 digits, the 16th digit is rounded off.
Therefore, if the rounded-off value goes out of the range from -2147483648 to 2147483647, it will not be an intended value.
Ex.
When "2147483646.12345678" is set as an input value, it is handled as "2147483646.12346" because 16th digit "6" is
rounded off.
Ex.
When "E1.7976931348623157+307" is set as an input value, it is handled as "E1.79769313486232+307" because 16th digit
"5" is rounded off.
× 2
Mantissa
∙∙∙ b16 b15
Mantissa
N
representing a single-precision real number in binary.
N
representing a single-precision real number in binary. The following table shows the

7FEH 7FDH 401H

1023 1022 2
Exponent
∙∙∙ b0
400H 3FFH 3FEH
1 0 -1
1.2 Data Specification Method

02H 01H 00H

-1021 -1022 Not used
1 OVERVIEW
1
49