User manual
Computer Gateway User Manual 4-9 5/96
4.1.3
4.1.3 Data Representations
NOTE
In any data-exchange message (except for History) you are required to specify the data type
for each data-point parameter being transferred. You also will need to know the set of value
assignments (both ordinal and character string) for any enumerations accessed.
REALSโReal values consist of 32 bits. The format seen at the host processor depends on
configuration choice and will be one of the following:
IEEE 754
Word 1
0 00000000 1111111
๎1 23456789 0123456
fraction
sign exponent
1112222222222333
7890123456789012
Word 2
In the IEEE 754 format, the exponent is represented by 2
E
where E = value in the
exponent field -127. The fraction is left justified such that a hidden 1 bit exists. A
positive or negative value with an exponent of 255 and a non-zero fraction is used to
indicate a Bad Value (also referred to as Not a Number or NaN). TDC 3000
X
most
commonly generates the hexadecimal value 7F800001 to represent a Bad Value.
DPS6
Word 1
0000000 0 01111111
๎1234567 8 90123456
fractionsignexponent
1112222222222333
7890123456789012
Word 2
IBM 370
Word 1
0 0000000 01111111
๎1 2345678 90123456
fraction
sign exponent
1112222222222333
7890123456789012
Word 2
In the DPS6 and IBM formats, the exponent is represented by 16
E
where E = value in
the exponent field -64. The fraction is composed of six binary coded hexadecimal
digits, normalized so that the first digit is not zero. The form -0 (sign bit only set) is
used to indicate a Bad Value, also referred to as Not a Number (NaN).