Digital Logic Design

Lecture 18
 Announcements
• HW 6 up on webpage, due on Thursday, 11/6
 Agenda
• MSI Components
– Binary Adders and Subtracters (5.1, 5.1.1)
– Carry Lookahead Adders (5.1.2, 5.1.3)
– Decimal Adders (5.2)
– Comparators (5.3)
 Scale of Integration
• Scale of Integration = Complexity of the Chip
– SSI: small-scale integrated circuits, 1-10 gates
– MSI: medium-scale IC, 10-100 gates
– LSI: large scale IC, 100-1000 gates
– VLSI: very large-scale IC, 1000+ gates
– Today’s chip has millions of gates on it.
• MSI components: adder, subtracter,
comparator, decoder, encoder, multiplexer.
 Scale of Integration
• LSI technology introduced highly generalized circuit
structures known as programmable logic devices
(PLDs).
– Can consist of an array of and-gates and an array of or-
gates. Must be modified for a specific application.
– Modification involves specifying the connections using a
hardware procedure. Procedure is known as
programming.
• Three types of programmable logic devices:
– Programmable read-only memory (PROM)
– Programmable logic array (PLA)
– Programmable array logic (PAL)
MSI Components
 Binary Full Adder

𝒙𝒊 𝒚𝒊 𝒄𝒊 𝒄𝒊+𝟏 𝒔𝒊
0 0 0 0 0
0 0 1 0 1
0 1 0 0 1
0 1 1 1 0
1 0 0 0 1
1 0 1 1 0
1 1 0 1 0
1 1 1 1 1
 Finding a Simplified Circuit

0 1 0 1 0 0 1 0

1 0 1 0 0 1 1 1

Corresponding minimal sums:

We can simplify the sum for 𝑠𝑖 by using xor:

𝑐𝑖

𝑐𝑖+1
 What about many bits?
• Consider addition of two binary numbers,
each consisting of 𝑛 bits.
• Direct approach: Write a truth table with 22𝑛
rows corresponding to all the combinations of
values and specifying the values of the sum
bits. Then find a minimal combinational
network.
• This will be intractable.
Parallel (ripple) Binary Adder
𝑥3 𝑦3 𝑥2 𝑦2 𝑥1 𝑦1 𝑥0 𝑦0 𝑐𝑖𝑛

𝑐3 𝑐2 𝑐1

𝑥 𝑦 𝑐𝑖𝑛 𝑥 𝑦 𝑐𝑖𝑛 𝑥 𝑦 𝑐𝑖𝑛 𝑥 𝑦 𝑐𝑖𝑛

𝑐𝑜𝑢𝑡 𝑠 𝑐𝑜𝑢𝑡 𝑠 𝑐𝑜𝑢𝑡 𝑠 𝑐𝑜𝑢𝑡 𝑠

𝑐3 𝑐2 𝑐1

𝑐𝑜𝑢𝑡 𝑠3 𝑠2 𝑠1 𝑠0

Why is it called “ripple” adder?

𝒙𝒊 𝒚𝒊 𝒃𝒊 𝒃𝒊+𝟏 𝒅𝒊
0 0 0 0 0
0 0 1
0 1 0
0 1 1
1 0 0
1 0 1
1 1 0
1 1 1
 Binary Subtracters
Compute: 𝑥𝑖 − 𝑦𝑖 .
𝑏𝑖 is a borrow-in bit from previous bit-order position.
𝑏𝑖+1 is a borrow-out bit.

𝒙𝒊 𝒚𝒊 𝒃𝒊 𝒃𝒊+𝟏 𝒅𝒊
0 0 0 0 0
0 0 1 1 1
1
0 1 0 10
0 1 1 -0
1 0 0
1 0 1
1 1 0
1 1 1
 Binary Subtracters
Compute: 𝑥𝑖 − 𝑦𝑖 .
𝑏𝑖 is a borrow-in bit from previous bit-order position.
𝑏𝑖+1 is a borrow-out bit.

𝒙𝒊 𝒚𝒊 𝒃𝒊 𝒃𝒊+𝟏 𝒅𝒊
0 0 0 0 0
0 0 1 1 1
0 1 0 1 1 10
0 1 1 -1
1 0 0
1 0 1
1 1 0
1 1 1
 Binary Subtracters
Compute: 𝑥𝑖 − 𝑦𝑖 .
𝑏𝑖 is a borrow-in bit from previous bit-order position.
𝑏𝑖+1 is a borrow-out bit.

𝒙𝒊 𝒚𝒊 𝒃𝒊 𝒃𝒊+𝟏 𝒅𝒊
0 0 0 0 0
0 0 1 1 1
1
0 1 0 1 1 10
0 1 1 1 0 -1
1 0 0
1 0 1
1 1 0
1 1 1
 Binary Subtracters
Compute: 𝑥𝑖 − 𝑦𝑖 .
𝑏𝑖 is a borrow-in bit from previous bit-order position.
𝑏𝑖+1 is a borrow-out bit.

𝒙𝒊 𝒚𝒊 𝒃𝒊 𝒃𝒊+𝟏 𝒅𝒊
0 0 0 0 0
0 0 1 1 1
1
0 1 0 1 1
0 1 1 1 0
1 0 0 0 1
1 0 1 0 0
1 1 0 0 0
1 1 1 1 1
 Finding a Simplified Circuit
𝑑𝑖 = 𝑏𝑖 ⊕ 𝑥𝑖 ⊕ 𝑦𝑖 (Same as sum in adder)
𝑏𝑖+1 = 𝑥𝑖 𝑦𝑖 + 𝑥𝑖 𝑏𝑖 + 𝑦𝑖 𝑏𝑖
𝑥3 𝑦3 𝑥2 𝑦2 𝑥1 𝑦1 𝑥0 𝑦0
𝑏0

𝑏4
𝑑3 𝑑2 𝑑1 𝑑0
 A better approach using 2’s
complement
𝑦3 𝑦2 𝑦1 𝑦0

𝑥3 𝑥2 𝑥1 𝑥0 𝑐𝑖𝑛 = 1

𝑐3 𝑐2 𝑐1

𝑥 𝑦 𝑐𝑖𝑛 𝑥 𝑦 𝑐𝑖𝑛 𝑥 𝑦 𝑐𝑖𝑛 𝑥 𝑦 𝑐𝑖𝑛

𝑐𝑜𝑢𝑡 𝑠 𝑐𝑜𝑢𝑡 𝑠 𝑐𝑜𝑢𝑡 𝑠

𝑐3 𝑐2 𝑐1
𝑐𝑜𝑢𝑡 𝑠3 𝑠2 𝑠1 𝑠0
 Parallel Adder/Subtracter
𝑦3 𝑦2 𝑦1 𝑦0

𝑥3 𝑥2 𝑥1 𝑥0
 Carry Lookahead Adder
• Ripple effect:
– If a carry is generated in the least-significant-bit
the carry must propagate through all the
remaining stages.
– Assuming two-levels of logic are need to
propogate the carry through each of the next
higher-order stages. Delay is 2n.
• Must speed up propagation of the carries.
Adders designed with this consideration in
mind are called high-speed adders.
 Carry Lookahead Adder
• Consider 𝑐𝑖+1 = 𝑥𝑖 𝑦𝑖 + 𝑥𝑖 𝑐𝑖 + 𝑦𝑖 𝑐𝑖
= 𝑥𝑖 𝑦𝑖 + 𝑥𝑖 + 𝑦𝑖 𝑐𝑖
• The first term 𝑥𝑖 𝑦𝑖 is called the carry-generate
function since it corresponds to the formation of
a carry at the i-th stage.
• The second term 𝑥𝑖 + 𝑦𝑖 𝑐𝑖 corresponds to a
previously generated carry 𝑐𝑖 that must
propagate past the i-th stage to the next stage.
• The 𝑥𝑖 + 𝑦𝑖 part of this term is called the carry-
propagate function.
• Carry-generate function will be denoted by 𝑔𝑖 ,
carry-propagate function will be denoted by 𝑝𝑖 .
 Carry Lookahead Adder
𝑔𝑖 = 𝑥𝑖 𝑦𝑖
𝑝𝑖 = 𝑥𝑖 + 𝑦𝑖
𝑐𝑖+1 = 𝑔𝑖 + 𝑝𝑖 𝑐𝑖
Using this general result, the output carry at
each of the stages can be written in terms of the
𝑔’s, 𝑝’s and initial input carry 𝑐0 .
 Carry Lookahead Adder
𝑐1 = 𝑔0 + 𝑝0 𝑐0
𝑐2 = 𝑔1 + 𝑝1 𝑐1
= 𝑔1 + 𝑝1 𝑔0 + 𝑝0 𝑐0
= 𝑔1 + 𝑝1 𝑔0 + 𝑝1 𝑝0 𝑐0
𝑐3 = 𝑔2 + 𝑝2 𝑐2
= 𝑔2 + 𝑝2 (𝑔1 + 𝑝1 𝑔0 + 𝑝1 𝑝0 𝑐0 )
= 𝑔2 + 𝑝2 𝑔1 + 𝑝2 𝑝1 𝑔0 + 𝑝2 𝑝1 𝑝0 𝑐0

𝑐𝑖+1
= 𝑔𝑖 + 𝑝𝑖 𝑔𝑖−1 + 𝑝𝑖 𝑝𝑖−1 𝑔𝑖−2 + ⋯ + 𝑝𝑖 𝑝𝑖−1 ⋯ 𝑝1 𝑔0
+ 𝑝𝑖 𝑝𝑖−1 ⋯ 𝑝0 𝑐0
Why is this a good idea? Do we save on computation?
Carry Lookahead Adder
Carry Lookahead Adder
 Carry Lookahead Adder
• What is the delay?
– One level of logic to form g’s, p’s
– Two levels of logic to propagate through the carry
lookahead
– One level of logic to have the carry effect a sum
output.
– Total: 4 units of time.
 Large High-Speed Adders
• The carry lookahead network can very large as
the number of bits increases.
• Approach: Divide bits of the operands into
blocks, use carry lookahead adders for each
block. Cascade the adders for the blocks.
• Ripple carries occur between the cascaded
adders.
 Another Approach to Large High-
Speed Adders
• Carry lookahead generators that generate the
carry of an entire block.
• Assume 4-bit blocks.
• For each block, 4-bit carry lookahead
generator outputs:
𝐺 = 𝑔3 + 𝑝3 𝑔2 + 𝑝3 𝑝2 𝑔1 + 𝑝3 𝑝2 𝑝1 𝑔0
𝑃 = 𝑝3 𝑝2 𝑝1 𝑝0
Carry Lookahead Generator
Large High-Speed Adders
 Decimal Adders
8421 weighted coding scheme or BCD Code
Decimal Digit BCD
0 0000
1 0001
2 0010 Forbidden codes: 1010,
3 0011 1011, 1100, 1101, 1110,
1111
4 0100
5 0101
6 0110
7 0111
8 1000
9 1001
 Decimal Adder
• Inputs: 𝐴3 𝐴2 𝐴1 𝐴0 , 𝐵3 𝐵2 𝐵1 𝐵0 , 𝐶𝑖𝑛 from
previous decade.
• Output: 𝐶𝑜𝑢𝑡 (carry to next decade),
𝑍3 𝑍2 𝑍1 𝑍0 .
• Idea: Perform regular binary addition and
then apply a corrective procedure.