63 lines
2.2 KiB
Markdown
63 lines
2.2 KiB
Markdown
|
---
|
||
|
|
tags:
|
||
|
|
- logic-gates
|
||
|
|
- binary
|
||
|
|
---
|
||
|
|
|
||
|
|
# Four-bit adder
|
||
|
|
|
||
|
|
A single
|
||
|
|
[half adder](Half_adder_and_full_adder.md#half-adder)
|
||
|
|
and
|
||
|
|
[full adder](Half_adder_and_full_adder.md#fufll-adder)
|
||
|
|
allows us to calculate the sum of two 1-bit numbers, but this is not much use in
|
||
|
|
practice. To approximate what is really happening at the circuit level in
|
||
|
|
computers we need to be able to add bigger binary numbers. We will demonstrate
|
||
|
|
how this can be achieved for a four-bit number (nibble) using repeated full
|
||
|
|
adders and half adders.
|
||
|
|
|
||
|
|
We want to be able to calculate the following sum:
|
||
|
|
|
||
|
|
```
|
||
|
|
0010
|
||
|
|
0011
|
||
|
|
____
|
||
|
|
0101
|
||
|
|
```
|
||
|
|
|
||
|
|
We will achieve this by using three full adders and one half adder, moving from
|
||
|
|
right to left. The half adder will be used at the beginning to calculate the
|
||
|
|
least significant bit as it will have no carry-in. The subsequent three bits
|
||
|
|
will all be added using a full adder.
|
||
|
|
|
||
|
|
Let's walk through the process:
|
||
|
|
|
||
|
|
>
|
||
|
|
|
||
|
|
1. HA receives the bits $0$ and $1$ as inputs. It outputs $1$ as the sum bit and
|
||
|
|
$0$ as the carry-out.
|
||
|
|
2. FA1 receives $0$ as the carry-in bit plus $1$ and $1$ as its input. This
|
||
|
|
means it has the following calculation to execute: $1 + 1 + 0$. This gives
|
||
|
|
$0$ as the sum bit and $1$ as the carry-out bit.
|
||
|
|
3. FA2 receives $1$ as the carry-in (the carry-out from FA1) plus $0$ and $0$ as
|
||
|
|
its own inputs. $0 + 0 + 1$ gives us $1$ as the sum bit and $0$ as the
|
||
|
|
carry-out.
|
||
|
|
4. Finally FA3 receives $0$ as its carry-in plus $0$ and $0$ as its inputs.
|
||
|
|
$0 + 0 + 0 = 0$, therefore its sum bit is $0$ and the overall sum of the four
|
||
|
|
bits is $0101$. There is no carry out as there are no more bits to add.
|
||
|
|
|
||
|
|
## Ripple carry adder
|
||
|
|
|
||
|
|
We notice that as the calculation proceeds, the carry bits propagate or _ripple_
|
||
|
|
through the circuit. For this reason, a four-bit adder constructed in the manner
|
||
|
|
of this circuit is called a **ripple carry adder**.
|
||
|
|
|
||
|
|
There are two important consequences to note about this type of circuit:
|
||
|
|
|
||
|
|
- Each carry bit that ripples to the next full adder introduces a small delay.
|
||
|
|
Therefore extending the circuit to handle more bits will make the circuit
|
||
|
|
slower overall.
|
||
|
|
- Because the value of the bits changes through the course of the ripple, the
|
||
|
|
output of the circuit will be innacurate until all the carry bits have had
|
||
|
|
time to propagate.
|