Control Data 3100 Reference Manual page 183

EXAMPLE 3
Convert floating point number
1774400 000 000 0008
to fixed octal
1. The floating point number is positive and
remains uncomplemented.
2. The biased exponent
<
20008; therefore, sub-
tract 17778 from the biased exponent to ob-
tain the true exponent of the number. 17748-
17778 =-3
3. Coefficient
=
400 000 000 0008
=
.1002.
Move binary point to the left three places.
Coefficient
=
.000 1002
4. The sign of the coefficient is positive because
the floating point number was not comple-
mented in step one.
5. Represent in fixed octal notation .
. 000 1002
=
.048
BINARY CODED DECIMAL (BCD) ARITHMETIC
General
The Binary Coded Decimal (BCD) option expands the arithmetic capabilities of a 3100, 3200,
or 3300 Computer by providing the necessary logic for loading, storing, shifting, adding and
subtracting binary coded decimal characters. A standard 24-bit data word is comprised of
four 6-bit BCD characters. The general format for a BCD word and the bit function within a
typical character are illustrated in Figure B-l. Tables B-3 and B-4 define the significance of
binary data within a character.
Figure B-2 depicts the ED register and the other digits displayed on the 3200 Console.
23 18 17 12 11 06 05 00
o
1 2
I
3
~
, .1
~
~Character
positions
~
BCD Character
Figure B-1. BCD Word and Character Format
B-16