Faculty of Engineering and Technology

Electrical Engineering Department

EED220: Logic Design

Lecture 3
Signed Numbers & Binary Codes
Mostafa Salah, Ph.D.
[email protected]
Contents
• Signed Binary Number
• Number Representations
• Fixed Point
• Floating Point
• Binary Codes

2
 Signed Binary Number
If positive, 0 followed by magnitude
Signed magnitude 00001001
Signed-magnitude

If negative, 1 followed by magnitude

Signed magnitude 10001001

Number
Example: Number 9.
8-bit Representation If positive, leave number unchanged
Signed 1’s complement 00001001
Signed 2’s complement 00001001

Signed-complement

If negative, complement number

Signed 1’s complement 11110110
Signed 2’s complement 11110111

3
 Signed Binary Number
System Range
Unsigned 0 to 2n – 1
Signed Magnitude –(2(n – 1) – 1) to 2(n – 1) – 1
Two’s Complement –2n – 1 to 2(n – 1) – 1
One’s Complement –(2(n – 1) – 1) to 2(n – 1) – 1

For 4-bit binary encodings:

Unsigned

2’s Complement

Signed Magnitude

4
 Signed Binary Number
• The signed‐2’s‐complement system has
only one representation for 0, which is
always positive.
• The signed‐magnitude system is used in
ordinary arithmetic but is awkward
when employed in computer arithmetic
because of the separate handling of the
sign and the magnitude.
• Therefore, the signed‐complement
system is normally used.
• The 1’s complement imposes some
difficulties, and It is useful as a logical
operation, since the change of 1 to 0 or 0
to 1 is equivalent to a logical complement
operation.

5
 Arithmetic Addition
• The addition of two signed
binary numbers with +6→ 00000110 +6→ 00000110
negative numbers + +
+ 13→ 00001101 – 13→ 11110011
represented in -------------------------------- --------------------------------
+ 19 → 00010011 –7→ 11111001
signed‐2’s‐complement
form is obtained from the
addition of the two
numbers, including their –6→ 11111010 –6→ 11111010
sign bits. + +
– 13→ 11110011 + 13→ 00001101
• A carry out of the sign‐bit --------------------------------
– 19 → 11101101
--------------------------------
+7→ 00000111
position is discarded.

6
 Arithmetic Subtraction
• Subtraction of two signed binary numbers when negative numbers are in
2’s‐complement form is simple and can be stated as follows:
Take the 2’s complement of the subtrahend (including the sign bit) and add it to the
minuend (including the sign bit). A carry out of the sign‐bit position is discarded.

( A) – (+B) = ( A) + (–B)
( A) – (–B) = ( A) + (+B)

✓ To convert a positive number into a negative number, take the complement

Computers need only one common hardware

circuit to handle addition and subtraction
7
 Number Representations!
• The key to represent fractional numbers, like 26.5, is the concept of binary point. A
binary point is like the decimal point in a decimal system. It acts as a divider between
the integer and the fractional part of a number.

• The very same concept of decimal point can be applied to our binary representation,
making a "binary point".
• Digits (or bits) on the right of binary point carries a weight of 2–1, 2–2, 2–3, and so on.

Example: The bit pattern of 2.75 and 22 is 24 23 22 21 20 Binary Point 2–1 2–2 2–3
(2.75)10
exactly the same. The only difference, is the 0 0 0 1 0 . 1 1 0
position of binary point.
24 23 22 21 20 Binary Point 2–1 2–2 2–3
(22)10
1 0 1 1 0 . 0 0 0

8
 Number Representations! – Fixed Point Number
• Fixed Point Number Representation
▪ To represent a real number in computers (or any hardware in general), a
definition of a fixed-point number type simply done by implicitly fixing the binary
point to be at some position of a numeral.

▪ To define a fixed-point type conceptually, the notation fixed<w,b> is used,

▪ w denotes the number of bits used as a whole (the Width of a number),
▪ b denotes the position of binary point counting from the least significant bit (counting from 0).

Example: fixed<8,3> denotes an 8-bit fixed point number, of which 3 right most bits are fractional.

Example: Given the pattern 0 0 0 1 0 1 1 0

• Represents (0.6875)10 in fixed<8,5>

• Represents (5.5)10 in fixed<8,2>
9
 Number Representations! – Fixed Point Number
• Fixed Point Number Representation

Pros:
• Straight-forward and efficient as integers arithmetic in computers.
• Fixed point arithmetic comes for free on computers.

Cons:
• The loss of range and precision when compare with floating point number
representations.
• For example, in a fixed <8,1> representation, the fractional part is only precise to 0.5.
Number like 0.75 can’t represented. One can represent 0.75 with fixed<8,2>, but then
loose range on the integer part.

10
 Number Representations! – Floating Point Number
• Floating Point Number Representation
▪ Instead of using the fixed-point representation, which would require
many significant digits, it is better to use the floating-point
representation.
▪ Numbers are represented by a:
▪ Mantissa comprising the significant digits and
▪ Exponent of the radix R.

Number = Mantissa × RExponent

The numbers are often normalized, such that the radix point is placed to the right of
the first nonzero digit, as in 5.234 × 1043.

11
 Number Representations! – Floating Point Number
• Binary floating-point representation has been standardized by the Institute of
Electrical and Electronic Engineers (IEEE).
• Two sizes of formats are specified in this standard— a single-precision 32-bit format
and a double-precision 64-bit format.

Prof. Kahan was instrumental in creating

the IEEE 754-1985 standard for floating-
point computation

12
 Number Representations! – Floating Point Number
▪ Single-Precision Floating-Point Format
• Sign bit, S
• 23-bit Mantissa field, M.

• an 8-bit exponent field, E,

• the exponent can be either positive or negative to be able to represent both very large and very small
numbers.
• The IEEE standard specifies the exponent in the excess-127 format. In this format the value 127 is added
to the value of the actual exponent so that

Exponent = E − 127
• The range of E is 0 to 255. The extreme values of E = 0 and E = 255 are taken to denote the exact zero
and infinity, Therefore, the normal range of the exponent is −126 to 127, which is represented by the
values of E from 1 to 254.

13
 Number Representations! – Floating Point Number
▪ Single-Precision Floating-Point Format
• The IEEE standard calls for a normalized mantissa, which means that the most-
significant bit is always equal to 1.
• Thus, it is not necessary to include this bit explicitly in the mantissa field.
• Therefore, if M is the bit vector in the mantissa field, the actual value of the
mantissa is 1.M , which gives a 24-bit mantissa.
• Consequently, the floating-point format represents the number

Value = (–1)S ×1.M × 2E−127

• The size of the mantissa field allows the representation of numbers that have the
precision of about seven decimal digits. The exponent field range of 2−126 to 2127
corresponds to about 10±38.

14
 Number Representations! – Floating Point Number
▪ Single-Precision Floating-Point Format
Example: Convert 0xC0B40000 single precision floating point into decimal.

(C0B40000)16 = (1100 0000 1011 0100 0000 0000 0000 0000)2

Sign
E Mantissa

S = 1 → Negative number
E = (10000001)2 = 129 → Exponent = E − 127 = 2
M = 01101

Value = (–1)S ×1.M ×2E−127

Value = (–1)1 ×1.01101 × 2E−127 = –1 ×1.01101 × 22 = – (20 + 2–2 + 2–3 + 2–5)22 = –5.625
15
 Number Representations! – Floating Point Number
▪ Single-Precision Floating-Point Format
Example: Convert the decimal 329.390625 to single precision floating point.

(329.390625)10 = (101001001.011001)2 = 1.01001001011001 ×28

Positive number → S = 0

E = 8 → Exponent = 8 + 127 = 135 = (10000111)2

M = 01001001011001 remember the implied 1 of the mantissa

means we don't include the leading 1

single precision
floating point 0 10000111 01001001011001000000000

16
 Number Representations! – Floating Point Number
▪Single-Precision Floating-Point Format
More Examples:

• (– 0.75)10 → 1 01111110 10000000000000000000000

• (0xC1580000)16 → 1 10000010 10110000000000000000000 → (– 13.5)10

Special Cases:
Single Precision Exponent Mantissa Value
Normalized Number 1 to 245 Anything (–1)S ×1.M × 2E−127
Zero 0 0  0 (according to value of S)
Infinity 255 0  ꝏ (according to value of S)
NaN 255 Nonzero NaN

17
 Number Representations! – Floating Point Number
▪ Double-Precision Floating-Point Format
• Sign bit, S
• 52-bit Mantissa field, M.

• an 11-bit exponent field, E,

• The IEEE standard specifies the exponent in the excess-1023 format.

Exponent = E − 1023

18
 Number Representations! – Floating Point Number
• Single-Precision Floating-Point Boundaries
• Largest positive/negative number =  (2 – 2 –23) x2127   2 x1038

• Smallest positive/negative number =  1 x2 –126   2 x10 – 38

• Single-Precision Floating-Point Boundaries

• Largest positive/negative number = (2 – 2 –52) x21023   2 x10308

• Smallest positive/negative number =  1 x2 –1022   2 x10 –308

19
 Number Representations! – Floating Point Number
▪Computation Error
Example:
(0.6)10 = (0.1001 1001 1001 1001 1001…)2
=+2-1 x (1.001 1001 1001 1001 1001…)2
Then:
Positive number → S = 0
E = -1 → Exponent = -1 + 127 = 126 = (01111110)2
M = 001 1001 1001 1001 1001 1001
single precision
0 01111110 01110011001100110011001 → 0.59999996423
floating point

Error = 0.6 – 0.59999996423 ≈ 0

20
 Number Representations! – Fixed Vs Floating
Steve Wozniak is brilliant guy, he
The Technology giant Steve Jobs and Bill Gates wrote the best BASIC* on the
planet, it is perfect in every way
except for one thing; it is just use
fixed point, it is not floating point

Steve Wozniak
Co-founder of Apple Inc.

https://www.businessinsider.com/the-bill-gates-steve-jobs-feud-frenemies-2017-6
21
References
‒ M. Mano and M. Ciletti, Digital Design, with an introduction to the
Verilog HDL. 5th Ed. Pearson, 2013.
‒ John F. Wakerly, Digital Design: Principles and Practices. 4th Ed. Pearson,
2005.
‒ R. Katz and G. Boriello, Contemporary Logic Design. 2nd Ed. Pearson,
2005.
‒ S. Brown and Z. Vranesic , Fundamentals of Digital Logic with Verilog
Design. 3rd Ed. SEM, 2013.
‒ Machine Structure Course. Online:
https://inst.eecs.berkeley.edu/~cs61c/sp06/handout/
‒ B. Parhami, Computer Arithmetic: Algorithms and Hardware Designs. 2nd
Ed. Oxford U. Press, 2010.
‒ Floating-Point Arithmetic, IEEE standard 754, 2019.

22