FPGA and VHDL

Department of Electronics
University Amar Telidji of Laghouat
Pr. Lahcene Merah
[email protected]
Programmable Logic Devices
Introduction

Logic circuits are the fundamental components that make up any

digital computer. A solid understanding of logic circuits is crucial
for today's electronics and computer engineers. These circuits are
the primary building blocks of a computer and are also used in
various other applications. They can be found in commonly used
products like digital watches, household appliances, CD players,
and electronic games, as well as in larger systems such as
equipment for telephone and television networks.
Evolution of Logic Circuits

• The technology used for the fabrication of digital hardware has

evolved dramatically over the past four decades.
• Until 1960, logic circuits were built using bulky components,
such as transistors and resistors that came as individual
pieces.
• The advent of integrated circuits allowed for multiple
transistors and, hence, an entire circuit to be placed on a
single chip. Initially, these circuits had only a few transistors,
but as technology improved, they became larger.
• Integrated circuit chips are manufactured on a silicon wafer,
as shown in the following figure. The wafer is cut to produce
individual chips, which are then placed inside a special type of
chip package.
 Wafer

• In 1970, it was possible to integrate all the circuits needed to create a microprocessor
on a single chip.

• By the late 1990s, it became possible to manufacture chips containing over 10 million
transistors. Currently, chips can have more than one billion transistors.
A snapshot from Wikipedia presenting actual statics
An integrated circuit (IC)
An integrated circuit is defined in contrast to a discrete circuit. In a
discrete circuit, each component or "circuit element" has its own
substrate, and the interconnection is done using external wires.
On the other hand, an integrated circuit corresponds to a single substrate
(hence the term "monolithic") made of semiconductor material—usually
silicon—on which elementary electronic components (transistors,
resistors, capacitors, inductors) and interconnection paths are implanted.
The electrical conductivity of a semiconductor falls between that of
metals (good conductors) and insulators.
Design of digital systems

For the design of digital systems, three main types of integrated

circuits can be used:

• Circuits standard (Standard Chips).

• Circuits personnalisées (Custom-Designed Chips)
généralement appelés ASIC (Application Specific Integrated
Circuit).
• Circuits logiques programmables (Programmable Logic
Devices).
Circuits standard (Standard Chips):

A widely used approach until the mid-1980s was to interconnect

small integrated circuits, each containing only a few logic gates.
An example that was highly recognized at the time is the 74XX
series.
Let's take an example of how a logic circuit is implemented using
the 7400 series family. Let's say our circuit performs the following
logic function:
• This example uses only a portion of the available gates on the
three chips, so the remaining gates can be used to implement
other functions.

• Due to their low logic capacity, standard circuits are rarely used
in practice today.

• The function provided by each circuit in the 74XX series is

fixed and cannot be tailored to a specific design situation.

• This, combined with the limitation that each circuit contains

only a few logic gates, makes them inefficient for building
large-scale logic circuits.
Programmable Logic Devices (PLD)

It is possible to manufacture circuits containing relatively large

amounts of logic circuits with a structure that is not fixed. These
circuits were introduced in the 1970s and are called
Programmable Logic Devices (PLD).

These devices are initially composed of a network of

interconnected logic gates, and this interconnection is
programmable.
Implementation principle of a function

 The fact is that to be programmable, we need a mechanism

that allows us to configure (program) a predefined silicon chip.
 We will start by examining a very simple programmable
function with two inputs called a and b and a single output y.
Fuse-Links Technology (FLT)

In this case, the device is manufactured with all the links in place,
where each link is referred to as a fuse:

• If an overvoltage occurs, then the fuses will be burned.

• The fuses in a silicon chip are formed using the same processes
that are used to create transistors and traces on the chip, so
they are microscopic in size.
• The fuses in the initial state are intact (initial state at 0).
Designers can selectively eliminate unwanted fuses by applying
relatively high voltage and current pulses to the device inputs.

This fuse elimination process is commonly referred to as "device

programming." Devices based on fuse-link technology are called
OTP (one-time programmable) or programmable only once, as
once the fuse is blown, it cannot be replaced, and there is no return
to its initial state.
Antifuse link technologies

In its unprogrammed state, the fuse exhibits high resistance, akin

to an open circuit (a break in the wire):
The link is a microscopic column of amorphous (non-crystalline)
silicon connecting two metal traces. In its unprogrammed state, the
amorphous silicon acts as an insulator with very high resistance,
exceeding one billion ohms.

 To program this element is to convert the insulating amorphous

silicon into conductive polycrystalline silicon.
 Programming can be achieved by applying a high voltage (16 V
for 1 ms) to the insulator between two metal areas, causing the
latter to diffuse into the insulator and render it conductive.
 Each cell occupies approximately 1,8 µm² (700 µm² for a fuse-
linked device).

 This technology allows high density of integration.

 Apart from the non-reprogrammability, it is the best technology

(speed and above all integration density.
Other interconnection links (Reminders on memories)

Memories

ROM RAM

ROM, PROM, EPROM SRAM, DRAM

EEPROM (Flash-ROM)
Read-Only Memory (ROM):

They are programmed using photomasks, which are used to

create transistors and metal tracks:
The PROMs (Programmable ROM):

The problem with mask programmed ROM is that:

• Their manufacturing cost is very expensive unless you
want to produce them in very large quantities.
• They are of little use in a development environment
where there is often a need to modify their contents.

So,
• In 1970, the development of the first PROMs took place.
• The technology was based on Nichrome fuse technology.
Programmable ROM (PROM):

In the unprogrammed state, as intended by the manufacturer, all

fuse links in the device are present. Programming is achieved by
applying an overvoltage to burn the fuses.
Erasable PROM (EPROM) :

A EPROM transistor has the same basic structure as a MOS-type transistor, but
with the addition of a second floating gate made of polysilicon, isolated by oxide
layers.

• In its unprogrammed state, the floating gate is discharged and does not affect the
normal operation of the control gate.
• By applying a high voltage (12V) between the control gate and the drain,
electrons force their way through the oxide into the floating gate in a process
known as hot electron injection
Erasable PROM (EPROM) :

• When the programming signal is removed, a negative

charge remains in the floating gate. This charge is highly
stable and will not dissipate for more than a decade
under normal operating conditions.

• The charge stored on the floating gate inhibits the

normal operation of the control gate and, therefore,
enables distinguishing between cells that have been
programmed and those that have not. This means that
we can use such a transistor to form a memory cell.
Erasing an EPROM memory

A first solution to erase the memory is to expose the circuit to ultraviolet

radiation for a few tens of minutes. These circuits are equipped with a quartz
window and encapsulated in a ceramic package to withstand the heating
Electrically EPROMs (EEPROM)

An EEPROM cell is 2.5 times larger than that of an EPROM

because it includes two transistors with space between them

The second transistor can be used to electrically erase the cell.

Random Access Memory (RAM)
Random Access Memory (RAM) is in contrast to Read-Only
Memory (ROM): the data stored in RAM is lost when the power is
turned off, whereas ROM retains its data in the absence of
electrical power.
There are two main versions of semiconductor RAM devices:
Dynamic Random Access Memory (DRAM) and Static Random
Access Memory (SRAM).

DRAM :
Each bit in a DRAM consists of a transistor and a capacitor:
The capacitor determines the value of the bit:

• The bit is 1 if the capacitor is charged, It is 0 otherwise.

• The transistor manages access to the capacitor.
The drawback of DRAM is that the capacitor has a natural tendency to
discharge. To maintain consistent information, it is necessary to perform
memory refresh every few milliseconds. This refresh involves reading and
rewriting the data.

The advantages of DRAM are:

- High integration density (as one bit is represented by one transistor).

- Low manufacturing cost.
- Low power consumption (between one-sixth to half that of SRAM).
SRAM (Static Random Access Memory):

Each cell is composed of 6 transistors (flip-flops) and is used to

store 1 bit.

SRAM is a volatile memory; the data is lost when the power is

turned off.

SRAM is faster and more expensive than DRAM; it is typically

used for cache and internal registers of a processor, while DRAM
is used for the main memory of a computer.
Memories and Programmable Logic Circuits:

EPROM

EEPROM/
FLASH

SRAM
SRAM Cell
The Origin of FPGAs (Field-Programmable Gate Arrays)

In order to get a good understanding of how FPGAs developed

and the reasons they emerged in the first place, it is best to
consider them in the context of other related technologies.
 The discovery of transistors in 1947.
 1958: The first integrated circuit.
 From the 1960s onward: the emergence of SRAM, DRAM,
and microprocessors.
 From 1970: the first IC (PLD) in the form of PROMs
appeared.
 By the end of the 1970s, tasks became more complex,
leading to the birth of CPLDs. Depending on the technology
and architecture used, several subfamilies of PLDs
emerged."
Programmable Logic Devices (PLD):
The SPLD PLA (Programmable Logic Array):

The PLA has a set of

programmable AND gate planes,
which link to a set of
programmable OR gate planes,
which can then be conditionally
complemented to produce an
output. It has 2N AND gates for N
input variables, and for M outputs
from PLA, there should be M OR
gates, each with programmable
inputs from all of the AND gates.
The SPLD PAL (Programmable Array Logic):

For PLAs, the fact that both AND and OR networks are programmable means that
signals take time to propagate through the circuit. Therefore, they are relatively slow. To
address the speed issues imposed by PLAs, PALs (Programmable Array Logic) emerged.

PALs have advantages over PLAs

because they are faster. This is due
to the fact that only one network is
programmable in PALs
GAL : Generic Array of Logic :

The PAL and PLA circuits are programmable once (OTP) based
on a PROM, so the PAL or PLA configuration cannot be
changed once it has been set. This limitation means that the
configured circuit should be discarded, and a new circuit is
required for reprogramming. The GAL, although similar in
architecture to PAL, uses EEPROM for configuration.
Examples of SPLD Programmers:
CPLD (Complex PLD):

• PLAs and PALs are useful for implementing a wide

variety of small digital circuits. T
• hese components are limited to relatively modest sizes,
typically supporting a combined number of inputs and
outputs of no more than 32.
• The early 1980s began to see the emergence of more
sophisticated PLD devices known as Complex PLDs
(CPLDs).
• The first CPLD was proposed by MMI (Monolithic
Memories Inc)."
• The CPLD represents a step up in complexity compared to the
SPLD; it builds upon the SPLD architecture and creates a
much broader design. Therefore, the SPLD can be used to
integrate the functions of several discrete digital ICs into a
single device, and the CPLD can be used to integrate the
functions of several SPLDs into a single device.

• A generic CPLD architecture is shown in the following figure:

• The CPLD consists of a number of logic blocks (sometimes
called functional blocks), each containing a macrocell and an
arrangement of PLA or PAL circuits. In this view, eight logic
blocks are displayed. The macrocell provides additional circuits
to accommodate registered (sequential) or unregistered
(combinatorial) outputs, as well as a signal polarity control.

• The actual number of logic blocks in a CPLD varies; the more

logic blocks available, the larger the design that can be
configured. At the center of the design is a programmable
global interconnection.