tema 1 introduction

8/13/2019 Tema 1 Introduction

1/12

Topic 1

Introduction to Digital Electronic Systems

Index

1.1. Introduction1.2. Design of Digital Systems

1.3. Embedded Systems

1.4. The Microprocessor

1.5 Internal Architecture of a microprocessor

1.5.1 Evolution of the microporcessor architecture.

1.6 Instructions.

1.7. Input / Output

1.8. Microprocessor Evolution

1.9. Developing Tools2

1.1. Introduction: What is a system?

3

1.1. Introduction

Interacts with the environment

Is divided into three stages:

1. Input

2. Processing

3. Output

The processing is based on:

Combinational logic and sequential circuits

Microprocessors

Microcontrollers

Digital Signal Processor (DSP)

Programmable logic circuits

Programmable Logic Controllers (PLCs)

4

1.1. Introduction

Programmable Devices

The behavior of a chip is ruled by a program

The same chip can make different actions depending of its program.

A program is...

Its a sequential set of instructions that describe, define or characterize

the execution of a task in a microprocessor

Instructions are sequentially executed.

5

A microcontroller (C) is a chip that includes:

the processor (P, CPU)

different types and amounts of memoryand

various input/output interfaces and peripherals

All of them interconnected by uni/bidirectional busses

The main difference between P and C:

The C includes in a single chip all elements needed in a

digital system, therefore achieving more integration and lower price.

1.1. Introduction


2/12

More differences between Pand C:

Lower time to marked when implementing a project in a C

In former Call operations in the ALU where developed using an accumulator

The accumulator output is connected to one of the ALUs inputs, being always

therefore one of the operands

Instructions with a single operand (c lear, increment, decrement, invert) operate with

the accumulator

1.1. Introduction 1.2. Design of Digital Systems

Combinational/Sequential integrated circuits

Big space, delays, difficult redesign.

Full-custom circuits

Very expensive design, slow process of debugging

Semi-custom circuits

ASIC (Application Specific Integrated Circuit)

Created from standard cells from the library, for a specific application.

Easy design modification.

8

Standard circuits

ASSP circuits (Application Specific Standard Part).

Originally designed for a client and then offered to the general public

Microcontrollers and Microprocessors

Generic processor with embedded devices for specific applications

DSPs

Device for Digital Signal Processing It has been designed for carrying out a small set of operations with a large amount of data.

9

1.2. Design of Digital Systems

Design with microcontroller

First, we have to know the problem to solve.

There are many microcontroller chips in the market. We should choose the proper

microcontroller to solve our problem,

Microcontroller with a particular subsystem

Microcontroller with sufficient storage dimmensions

Microcontroller with proper speed, dimensions,... PCB scheme is designed with the proper connections with other systems.

Program is designed

Taking into account the specifications required

Simulation

We test that our program solve the problem

Integration of program and hardware and test of the prototype.10

1.2. Design of Digital Systems

Based on programmable devices

(e.g. microcontrollers, DSPs.)

In general, they are real-time reactive systems:

They react to external events

They keep continuous interaction

They are continuously running

Their work is subjected to external time constraints

They do concurrently several tasks


11


SupercomputersServersWorkstationsPersonal computersCalculator

Specific computersPersonal Computers+ I/O boardsPLCsDigital regulators

Microprocessors-based boards+ I/O boards + VME busPC's + I/O boards + ISA busMicrocontrollers

Scientific calculationManagement (accounting, etc.)Databases

Industrial ControlFlight simulatorsRobotics

Appliances

AeronauticsMobile roboticsMobile phones

DO CONTROL / REAL TIME

EMBEDDED

NOT

EMBEDDED

DO NOT CONTROL

12


3/12

Embedded system examples

Consumer electronics CD players, HIFI, television, ...

Washer machines, fridges, dishwashers, ...

Automotive Speed control, air conditioning, etc.

ABS, ASR, Electronic fuel injection

Telecommunications Mobile phone

Avionics, Space Flying computers

Path-finder

Defense Bombs and intelligent missiles

Vehicles, ...

Instrumentation

13

1.3. Embedded Systems 1.3. Embbeded Systems: Features

Concurrency

All the tasks carried out by the system under control work simultaneously

The control system has to generate simultaneously the control actions

If there is only on processing unit: multiprocessing techniques are required, by using

real time operating systems

It there are several processing units, we are speaking about a multiprocessor solution

Reliability and safety

A fault in a control system can drive the controlled system to behave dangerously or

uneconomical

It is important to guarantee that if the control system fails, the system remains in a

safe controlled state: to anticipate potential design errors and exceptions

14

Low power

Many of these systems are powered by batteries.

Less consumption => greater autonomy

In many cases low voltage (3V) is needed

Low weight

Desirable feature in portable systems

It not only depends on t he onboard computer and its periphery but also on the sup Low prices

Related to consumer electronics and other devices with very competitive markets

Small size

The dimensions of an embedded system depend not only on i tself but also on the

space available on the system that controls and / or monitors

1.3. Embbeded Systems: Features

15

From the wired machine to the programmed one

WIRED:

Mainly simple logic gates

According to the inputs it is deterministic, very rigid with constant output

The technology was introducing more complex logic functions

Hard its implementation

A huge amount of integrated circuits is needed (multiplication, division, n order roots)

At the early 70 the PROGRAMMED MACHINE became only one electronic circuit:

the MICROPROCESSOR


16

General purpose circuit

It is a very versatile chip, easy to adapt a multiple applications, limited by the

designer:

Logic functions (AND, OR,...)

If a specific function is not implemented, it can be obtained by the combination of the

existing ones

Programmable: it works by means of a set of basic steps

Each step is called instruction

Each information element is called data

Program: it is a structured set of instructions

If the program is changed, the functionality changes accordingly

17


Advantages of the microprocessor (uP)

By changing the program a new function is obtained. In wired systems it is required

to change the design, and some electronic circuits

Much easier to debug and to optimize applications

Easy to update versions

Easy to adapt the design to a different application

Easy to design digital electronic circuits

It is possible to work with analog signals by using A/D and D/A converters

18



4/12

Elements of a P based digital system

PROGRAM

MEMORYCPUI/O

DATA

MEMORY

ADDRESS BUS

DATA BUS

CONTROL BUS


Unidirectional

Bidirectional

Power and clock Tri-state

19

The P needs memory enough to store both instructionsand data

(variables and constants) of a program

Instructions and constant data are stored in non-volatile memories (their contents are

preserved without power). Called program memory, it can only be read or

programmed (not written)

Variable data are stored in volatile memories(their contents are lost with power

down). Called data memory, it can be read and written

Big endian/little endian


20

Word size

It is the size of the a single data that the CPU considers

Byte

16 bits

32 bits.

Instruction execution time


21

Steps in the execution of an instruction

Obtaining the instruction to execute (operation code).

Calculation of the effective address

Search operands

Executing operation

Store the result

Preparing next instruction.

MEMORY

OPERATIONCODE

CPU

PC ADD.

OpcodeI.R.

Internal architecture

1.5. Internal architecture of a microprocessor


OPERATIONS THAT A CPU CAN PERFORM:

LOAD REGISTERS

READ OR WRITE MEMORY LOCATIONS

PERFORM ARITHMETIC OR LOGIC OPERATIONS

ROTATIONS AND SHIFTS


5/12


OPERATING SECTION

REGISTERS

STORE TEMPORARY DATA

ALU

LOGIC AND ARITHMETICSOPERATIONS, SHIFTS ANDROTATIONS

WITH OR WITHOUT

ACCUMULATOR

Internal architecture of a microprocessor1.5. Internal architecture of a microprocessor


CONTROL SECTIONLOAD INSTRUCTIONS

DECODE

GENERATES MICROORDERS

ENABLE THE TRANSFER BETWEENTHESE REGISTERS AND CONFIGURETHE ALU OPERATIONS


Internal architectureOTHERS REGISTERS

AUXILIARY REGISTERS

STATUS REGISTER

FLAGS: ZERO, SIGN CARRY,OVERFLOW

PROGRAM COUNTER

REGISTER

STACK POINTER REGISTER


Instruction register

It loads the operation code from the memory.

The decoder generates microorders depending on its content.

Program counter

It contains the address of the instruction to be executed.

After execution of an instruction PC is modified to point to the next

one.


Status register

It contains flags that provide information on the outcome of an

operation: sign, zero, carry, overflow.

It allows decisions conditional on the outcome of the previous

operation.

Stack pointer

It stores a memory address where data can be stored

automatically or manually.


Internal bus

Its a communication bus between the various components of the

microprocessor.

The number of bus lines is determined by the number of bits

processed in parallel in the ALU and its one of the parameters to

classify microprocessors.

This bus is related to the external data bus through the data bus

buffer.



6/12

Microprocessors Architectures: Von-Neuman

Two busses: Address (ADD) and Data (D).

Instructions and Data are accessed sequentially through the same

Data Bus.

1.5.1. Evolution of microprocessors architecture

PROGRAM

AND

DATA

MEMORY

CPU

ADDRESS BUS

DATA BUS

Microprocessors Architectures: Harvard

Two Address busses, a Data bus and an Instructions bus.

Fetch the next instruction and move a data at the same time.

DATA

MEMORYCPU

DATA ADD.

DATA BUS

PROGRAM

MEMORY

PROGRAM ADD.

INSTRUCTION BUS

Evolution of microprocessors architecture1.5.1. Evolution of microprocessors architecture

Pipelining

Electronics design methodology for processors to overlap the

execution of different instructions.

At each pipe stage one par t of the instruction is developed.

A processor with N stages can run N instructions at the same time,

increasing N times its efficiency.

The efficiency increase is obtained just reorganizing instructions, no

extra effort is needed in the CPU.

On the other hand the design of the Control Unit is more complex, in

order to get the best of the pipelining (do the best of instructions and

their parts reorganization).


Pipelining

The most used pipeline architecture in 90shad 5 stages: fetch (IF),

decode (ID), operand read (MEM), execution (EX) and operand write

back (WB).

Nowadays normal pipelines have a bigger number of stages, for

example 20 in Pentium IV.


Super-scalar processors

Electronics design methodology to overlap the execution of different

instructions using repeated functional units in the CPU.

Processor with scalar level L has L repeated functional units that can

thus run L instructions at the same time.

2.8. Evolution of microprocessors architecture1.5.1. Evolution of microprocessors architecture

Super-pipelining

Pipeline architecture in which slowest stages are divided in sub-stages

in order to increase their speed.

Two or more instructions can be execute different parts of the same

stage at a moment.

In a super-pipeline architecture the pipelining technology is applied

twice: in a global context and internally in its functional units.



7/12

f etc h decode ex ec ute

time

1

f et ch dec ode ex ecute

f etc h decode ex ec ute

2

3

in struction

3 stages pipeline: Hazards and risks

Fetch:The instruction is transferred from program memory and storedin the instructions pipeline

Decode:The instruction is decoded

Execute:ALU operations and data transferences

1 instruction 3 cycles

1 cycle 1 instruction


37

Multi-cycle instructions hazard

The instruction needs more cycles than the standard pipelinedinstructions

fetch ADD decode execute

time

1

fetch STR decode


2

3

data xfer

fetch SUB decode execute4

instruction

calc. add


38

Structural hazard

At a moment two instructions need the same hardware resources


time

1

fetch STR decode


2

3

data xfer

fetch ADD decode execute4

instruction


39

Control risk

In conditional jumps, the pipeline may need to be emptied

decode

ti me

. .

f etch ADD decode execute

1

2

3

4

instruction

fetch ADD decode execute1

fetch BCC decode execute

fetch SUB

2

3

4

5

fetch STR


40

1.6. Instructions. General concepts

What is a program?

Its a sequential set of instructions that describe, define or characterize

the execution of a task in a microprocessor

What is an instruction?

Its the CPU elementary command

Its a set of digital input signals to the decoder

The instruction decoder (sequential circuit) state evolves and it

changes its outputs sequentially based on these signals.

The output signals of the instruction decoder are the microorders.

41

Coding Instructions

The instructions are stored in memory

Binary format, set '1 'and '0', machine code

Possible generation of binary codes

Directly from a conversion table function / code

Using an assembly starting from an assembly language

Assembly language: set of mnemonics associated with the instructions

Using a compiler, starting from a high level language

Generates machine code from a program in high level language (eg C)

Instructions. General concepts


8/12


Example of Instruction

Add 2 to the contents of register 'A' (Historical solution) Using a conversion table it shows that this operation

corresponds to the code 0x23, followed by the value 0x02.

(Tedious solution) Using an assembler program, It would be written to a file: "add2, A", the result would be the same after you run the assembler (0x23, 0x02).

(Current solution) Using a high-level language: "A + = 2", and after compiling, theresult would be the same again (0x23, 0x02).


Instruction format

Opcode + [Source operand] + [Destination operand]

Access to the operands

Effective address: location of the operand

Addressing Modes: Different ways to express the effective address

Provide power to the device

Simplify programming


Steps in the execution of an instruction

Instruction fecth

PC ADDRESS BUS; READING OPCODE

Calculation of the effective address

Search operands

Executing operation

Store the result

Increased PC, new cycle

MEMORY

OPERATIONCODE

CPU

PC ADD.

OpcodeI.R.


Transfer instructions

To Copy information from registers, registers and memory, or between

memory.

Arithmetic instructions

They perform simple arithmetic operations: addition, subtraction, etc..

Logical Instructions

They perform logic operations on the operands: AND, OR, XOR, etc..

1.6. Instructions.Type of instructions

Bit manipulation instructions

They modify a single bit of the operand

Shift Instructions

They shift the contents of a register in both directions

Program Control Instructions

They change the flow of program execution

Types of instructions1.6. Instructions.Type of instructions


9/12

Mechanism for specifying the location of an operand.

A microprocessor has several addressing modes.

Objectives of addressing modes:

Facilitate the management of the data structures.

Allow relocation code.

Reduce the memory space occupied by the instructions.

1.6. Instructions. Addressing modes

Possible addressing modes:

Implicit Register direct

Immediate

Memory direct

Indirect

Indirect with offset

Indirect with Index

Indirect with pre-indexed

Indirect with post-indexed

Addressing modes1.6. Instructions. Addressing modes

1.7. Inputs / Outputs

Inputs / Outputs(I/O)

The reference for the interchange of information is the CPU.

I/O: the data or information that is passed into or out of a CPU

CPU takes an external data from the external system Input operation

CPU gives the data to the external system Output operation

P external access are carried out through the buses

I/O devices can be connected to the main system buses

Or can be a special subsystem depending on the main system.

51


Example: domestic heating system

P external access are carried out through the buses

I/O classification:

Directly with the environment under control (analog or digital I/O)

With the user (analog or digital I/O)

With other digital systems(this improves the capabilities of the whole system)

52


Devices classification

Output

Capability of holding data. Example: Latch

Input

They need tri-state outputs

Bidirectional

Both features in the same device

53

1.7. Inputs / Outputs. Interface structure

Interface circuits

They make possible transition between two different systems

Interface structure

In general, there are two options for the design.

General purpose circuits (standard circuits such as logic gates, latches, )

Special circuits designed for a specific function (e.g. USB adapter)

The CPU sees the interface circuits as a group of registers located

(mapped) in specific addresses.

54


10/12


Digital Interfaces

They only manage digital information.

In general, there are two kinds for transferring information:

Parallel data transfer:

A group of n bits are transferredsimultaneously.

Serial data transfer:

The bits are transferred one byone. It is needed time orsynchronism information forunderstanding the data withouterrors..

Advantages and disadvantages?

55


Digital interfaces

Asynchronous serial:

Synchronous serial:

8 b its

R elo j

In icio Fi n

T ran sm is o r

as n cro n o

B 0 B 1 B 2 B 3 B 4 B 5 B 7B 6

Asynchronous

Transmitter

Clock

Start End

8 b i t s

R el o j maes t r o

T r ans mi s o r / r ecep t o r

s ncr ono

8 b i t s

D a t o s

R e l o j

S en t i do de l a t r ans f er enc i a

T r ans mi s o r / r ecep t o r

s ncr ono

Clock

Synchronous

Transmitter/

Receiver

Synchronous

Transmitter/

ReceiverTransfer direction

Clock

Data

56


Analog interfaces

They are circuits that facilitate the conversion between numerical

codes (related to the digital system) and the multiplicity of instant

values of the analog signals

Circuits: Digital/Analog Converter (DAC, D/A)

and Analog/Digital Converter (ADC, A/D)

An important parameter related to A/D and D/A is the number of bits ofresolution and the sample frequency

57


I/O Interface and peripherals addressing

For the microprocessor the interface is a group of registers

For working with the interface, it is needed to read or write the registers

Depending on the microprocessor, Inputs and Outputs can be:

mapped in memory (standard accesses to the registers)

Out of the memory map (special instructions for accessing)

58


Buses accessing

Interconnection electrical problems

Unidirectional Buses. (fan-out, fan-in)

Bidirectional Buses. It is needed arbitrations techniques to prevent conflicts (whois using the bus and when).

Two or more devices connected to the same bus never can transfer simultaneously

information through the bus. Solutions:

Integrated circuits with high impedance controlled by the selection of the chip (CS,

chip select signal) or output enable (OE = Output Enable Signal).

Devices without high impedance outputs can use additional tri- state buffers with

enable logic. .

59


Evolution based on ideas, and not only on the technology

At the beginning the evolution depended only on the technology

Goal: to increase the clock frequency

Efficiency is affected for more parameters:

Memory accesses (bottleneck)

Input/Output operations

Compilation

Operating System Overload

CPU

60


11/12

Von Neumann's machine had a single functional unit

CISC processors:

Complex Instruction Set Computer Short programs with complex instructions (each instruction does many things)

Very sophisticated operations:

It reduces the gap between High level and ASM programming

A few number of instructions, but a lot of machine cycles

More complex functional units, based on microcode

Memory Operations (LOAD/STORE): Few internal registers

Simple compilers

61


Segmentation

Without finishing one instruction, another starts to be processed

62


New philosophy

New compilers

Machine cycles as fast as technology permits

More operations in one machine cycle

The simple decoding and the segmented execution are more important than the

code size (only one instruction format)

New words appear in the processor design:

Super-segmentationand superscalararchitectures

Goal: to increase the nu mber of simultaneous operation (PARALELISM)

63


RISC processors:

Reduced Instruction Set Computer

Long programs

Simple CPU

Small execution time of the instructions

Easily Pipelining / Segmenting (PARALELISM)

64


Simulator

It is a software that permits to simulate the behavior of a program. It imitates thebehavior of a microprocessor or a microcontroller.

It permits to visualize and modify the registers and the memory content

Real-time functionalities are not available

In general, physical and time aspects are no considered

The running speed is lower than the one with the real device

1.9. Developing Tools

65

On-chip debugging

It is needed a special software running in the device, and a communication port forconnecting with a computer.

Specific ports for this task:

JTAG (Joint Test Action Group) is a method of accessing memory and CPU resources

without having an application running on the target

BDM or background debugging mode also achieves this objective

The debugging application slows down the program

It is easy to visualize the real content of the memory, registers, etc.

Other features: step-by-step execution, break points, etc.

1.9. Developing Tools

66


12/12

This document has been made using material developed by the

Electronics Department personal.

1.10. Acknowledgments

67

tema 1 introduction

Documents