IDEC MICROSmart FC6A Series, Руководство пользователя

Страница: 424 / 578

8: I
NSTRUCTIONS
R
EFERENCE
8-8 FC6A S
ERIES
MICROS
MART
U
SER
’
S
M
ANUAL
FC9Y-B1722
Floating-Point Data Format
The FC6A Series MICROSmart can specify floating-point data (F) for advanced instructions. Like double word (D) and long integer
(L) data, floating-point data also uses two consecutive data registers to execute advanced instructions. The FC6A Series
MICROSmart supports the floating-point data based on the single storage format of the IEEE (The Institute of Electrical and
Electronics Engineers) Standard 754.
Single Storage Format
The IEEE single storage format consists of three fields: a 23-bit fraction, f; an 8-bit biased exponent, e; and 1-bit sign, s. These
fields are stored contiguously in one 32-bit word, as shown in the figure below. Bits 0:22 contain the 23-bit fraction, f, with bit 0
being the least significant bit of the fraction and bit 22 being the most significant; bits 23:30 contain the 8-bit biased exponent, e,
with bit 23 being the least significant bit of the biased exponent and bit 30 being the most significant; and the highest-order bit 31
contains the sign bit, s.
The table below shows the correspondence between the values of the three constituent fields s, e, and f and the value represented
by the single storage format bit pattern. When any value out of the bit pattern is entered to the advanced instruction or when
execution of advanced instructions, such as division by zero, has produced any value out of the bit pattern, a user program
execution error will result, turning on special internal relay M8004 and the ERR LED on the FC6A Series MICROSmart.
Carry and Borrow in Floating-Point Data Processing
When advanced instructions involving floating-point data are executed, special internal relay M8003 (carry and borrow) is updated.
Value Exponent Field e Fraction Field f Representation in WindLDR
±0 e=0 f=0 0.0
Denormal numbers e=0 f ≠ 0 -1.175494E-38 to 1.175494E-38
Normal numbers 0<e<255 Arbitrary
-3.38 to -1.175494E-38
1.175494E-38 to 3.38
± ∞ (± infinities)
e=255
f=0 INF
NaNs (Not a number values) f ≠ 0 NAN
s e[30:23] f[22:0]
31 30 23 22 0
23-bit fraction 8-bit biased exponent
Sign bit (0: positive, 1: negative)
M8003 Execution Result Value
1 Overflow (exceeds the range between –3.402823  10
38
and 3.402823  10
38
)
1 0 Not zero (within the range between –1.175495  10
–38
and 1.175495  10
–38
)
0 0 Zero
0

0
–1.175495

10
–38
M8003 1 1
1.175495

10
–38
0 –3.402823

10
38
0
Execution Result
1
Overflow
0 1
3.402823

10
38
Overflow
Not Zero