Motorola MVME166IG/D2 - Manual (Page 89)

Floating Point Support
MVME166IG/D2 4-19
4
Packed Decimal Real
This format would appear in memory as:
A packed decimal number takes 12 bytes in memory.
Scientific Notation
This format provides a convenient way to enter and display a floating point
decimal number. Internally, the number is assembled into a packed decimal
number and then converted into a number of the specified data type.
Entering data in this format requires the following fields:
An optional sign bit (+ or -).
One decimal digit followed by a decimal point.
Up to 17 decimal digits (at least one must be entered).
An optional Exponent field that consists of:
An optional underscore.
The Exponent field identifier, letter "E".
An optional Exponent sign (+, -).
From 1 to 3 decimal digits.
For more information about the MC68040 floating point unit, refer to the
M68040 Microprocessor User’s Manual.
4-bit sign field (4 binary digits)
16-bit exponent field (4 hex digits)
68-bit mantissa field (17 hex digits)