Binary PDF

Conversion of Binary, Octal and
Hexadecimal Numbers
From Binary to Octal
Starting at the binary point and working left, separate the bits into
groups of three and replace each group with the corresponding octal
digit.
100010112 = 010 001 011 = 2138
From Binary to Hexadecimal

Starting at the binary point and working left, separate the bits into
groups of four and replace each group with the corresponding
hexadecimal digit.
100010112 = 1000 1011 = 8B16
From Octal to Binary

Replace each octal digit with the corresponding 3-bit binary string.
2138 = 010 001 011 = 100010112
From Hexadecimal to Binary

Replace each hexadecimal digit with the corresponding 4-bit binary
string.
8B16 = 1000 1011 = 100010112
Conversion of Decimal Numbers

From Decimal to Binary
2 139
2 69
1
1
34
17
LSD
13910 = 100010112
MSD
From Binary to Decimal

100010112
= 127 + 026 + 025 + 024 + 123 + 022 + 121 + 120
= 128 + 8 + 2 + 1
Conversion of Fractions
Starting at the binary point, group the binary digits that lie to
the right into groups of three or four.
0.101112 = 0.101 110 = 0.568

0.101112 = 0.1011 1000 = 0.B816
Problems
Convert the following
Binary
10011010
Octal
Decimal
Hex
2705
2705
3BC
Binary
10011010
10111000101
101010010001
1110111100
Octal
232
2705
5221
1674
8 2705 1
8 338 2
8
42 2
5
Decimal
154
1477
2705
956
16 2705 1
16 169 9
10=A
Hex
9A
5C5
A91
3BC
Add
1
+ 1
Subtract
1
-
Multiply
normally
for implementation - add the shifted

multiplicands one at a time.
= 14
= 13
+ 1
+ 0
+ 1
+ 1
(8 bits)
Divide
1101
110
1111) 11000101|
1111
1001101|
1111
1101) 1011001|
1101
100101|
1101
10001|
1011|
0000
0000|
10001|
1111|
10
1001
1101) 1111001|
1101
10001|
0000
10001|
0000
10001|
1101|
100
1011
Sign-Magnitude
0 = positive
1 = negative
n bit range = -(2n-1-1) to +(2n-1-1)
4 bits range = -7 to +7
2 possible representation of zero.
2's Complement
flip bits and add one.
n bit range = -(2n-1) to +(2n-1-1)
4 bits range = -8 to +7
0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111
=0
=1
=2
=7
= -8
= -7
= -2
= -1
Example
1110
0001
0010
= 14
flip bits
add one WRONG this is not -14. Out of range. Need 5 bits
0 1 1 1 0 = 14
1 0 0 0 1 flip bits
1 0 0 1 0 add one. This is -14.
Sign Extend
add 0 for positive numbers
add 1 for negative numbers
Add 2's Complement

1110
= -2
+ 1101
11011
= -3
1110
= -2
+ 0011
=3
ignore carry = -5
10001
ignore carry = 1
Be careful of overflow errors. An addition overflow occurs whenever the sign of the sum if
different from the signs of both operands. Ex.
0100
=4
1100
= -4
+ 0101
=5
+ 1011
= -5
1001
= -7 WRONG
10111
ignore carry = 7 WRONG
Multiply 2's Complement
1110
= -2
1101
= -3
11111110
+
+
= -2
0011
=3
11111110
0000000
+1 1 1 1 1 1 0
11111110
111111010
111110
111110110
sign extend to 8 bits
1110
00010
100000110
ignore carry
negate -2 for sign bit
ignore carry = 6
10010
= -14
10011
= -13
1111110010
+ 111110010
11111010110
ignore carry
+ 00000000
1111010110
+ 0000000
1111010110
+ 001110
10010110110
negate -14 for sign bit

ignore carry = 182

ignore carry = -6
Floating-Point Numbers
mantissa x (radix)exponent
The floating-point representation always gives us more range and less precision than the
fixed-point representation when using the SAME number of digits.
Mantissa
Sign
sign
exponent
General format
0
1
9
Mantissa 8-bit excess-127
sign
characteristic
32-bit standard
Mantissa magnitude
31
23-bit normalized fraction
Implied binary point
0
1
12
63
Mantissa
11-bit excess
52-bit normalized fraction
sign
1023 charactstic
64-bit standard
Normalized fraction - the fraction always starts with a nonzero bit. e.g.
0.01 x 2e would be normalized to 0.1 x 2e-1
1.01 x 2e would be normalized to 0.101 x 2e+1
Since the only nonzero bit is 1, it is usually omitted in all computers today. Thus, the 23-bit
normalized fraction in reality has 24 bits.
The exponent is represented in a biased form.
If we take an m-bit exponent, there are 2m possible unsigned integer values.

Re-label these numbers: 0 to 2m-1 -2m-1 to 2m-1-1 by subtracting a constant value (or
bias) of 2m-1 (or sometimes 2m-1-1).
Ex. using m=3, the bias = 23-1 = 4. Thus the series 0,1,2,3,4,5,6,7 becomes -4,-3,-2,1,0,1,2,3. Therefore, the true exponent -4 is represented by 0 in the bias form and -3 by
+1, etc.
zero is represented by 0.0 x 20.
Ex. if n = 1010.1111, we normalize it to 0.10101111 x 24. The true exponent is +4. Using the
32-bit standard and a bias of 2m-1-1 = 28-1-1 = 127, the true exponent (+4) is stored as a biased
exponent of 4+127 = 131, or 10000011 in binary. Thus we have
0|10000011|01011110000000000000000
Notice that the first 1 in the normalized fraction is omitted.
The biased exponent representation is also called excess n, where n is 2m-1-1 (or 2m-1).

Binary PDF

Uploaded by

Copyright:

Available Formats

Binary PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Binary PDF

Uploaded by

Copyright:

Available Formats

Conversion of Binary, Octal and

100010112 = 010 001 011 = 2138

From Binary to Hexadecimal

100010112 = 1000 1011 = 8B16

From Octal to Binary

2138 = 010 001 011 = 100010112

From Hexadecimal to Binary

8B16 = 1000 1011 = 100010112

Conversion of Decimal Numbers

From Binary to Decimal

0.101112 = 0.101 110 = 0.568

Convert the following

for implementation - add the shifted

Add 2's Complement

ignore carry = 7 WRONG

Multiply 2's Complement

sign extend to 8 bits

sign extend to 10 bits

negate -14 for sign bit

sign extend to 8 bits

If we take an m-bit exponent, there are 2m possible unsigned integer values.

You might also like