SOLID STATE VOICE RECORDER SYSTEM

By
Muhammed BEYAZ

Engineering Project Report

Yeditepe University
Faculty of Engineering and Architecture
Department of Computer Engineering
2009
 ii

SOLID STATE VOICE RECORDER SYSTEM

APPROVED BY:

Assoc. Prof. Dr. Mustafa Türkboyları …………………………

(Supervisor)
…………………………

………………………….

Prof. Dr. Şebnem Baydere …………………………

…………………………

………………………….

Assoc. Prof. Dr. Birol Aygün …………………………

…………………………

………………………….

DATE OF APPROVAL:
 iii

ACKNOWLEDGEMENTS

Many thanks to my supervisor, Mustafa Türkboyları, for spending his weeks and effort
to develop this project board, debug hardware and software and also for his financial support
on project.

Special thanks to Bilal Hatipoğlu for his guidance about project and his help to debug
software and hardware with me.
 iv

ABSTRACT

SOLID STATE VOICE RECORDER SYSTEM

In this project, a voice recording system that has been composed of a microcontroller
and a high density NAND flash was implemented. The microcontroller collects the voice
samples, passes through the ADPCM codec and stores on the NAND flash. A simple file
allocation table has been designed for the files located in the NAND flash. In addition to the
hardware part of this project a PC application has been developed. The purpose of the PC
application is to download and playback of the stored voice recordings on the NAND flash.
The PC application communicates to the microcontroller and transfers the stored voice
recordings on the NAND flash to the PC. The PC application converts the stored voice
recordings to a common audio format of WAV (WAVE).
 v

ÖZET

KATI HAL SES KAYIT SİSTEMİ

Bu projede, bir mikro denetleyici ve yüksek yoğunlukluklu NAND flaştan oluşan katı
hal ses kayıt sistemi gerçekleştirdik. Mikro denetleyici ses örneklerini toplayıp, bu örnekleri
ADPCM den geçirip çıkan sonucu NAND flaş üzerine kaydediyor. NAND flaşta bulunan ses
kayıtları içinse basit dosya yerleşim tablosu dizayn edildi. Projenin donanımsal kısmına ek
olarak bilgisayar uygulamasıda geliştirildi. Bilgisayar uygulamasının amacı ise NAND flaşta
bulunan ses kayıtlarını indirmek ve bu kayıtları çalmaktır. Bilgisayar uygulaması mikro
denetleyici ile haberleşip bu dosyaları bilgisayar üzerine transfer eder ve ses kayıtlarınının
formatını genel olarak bilinen bir format olan WAV (dalga) formatına dönüştürür.
 vi

TABLE OF CONTENT

ACKNOWLEDGEMENTS ....................................................................................................... iii

ABSTRACT............................................................................................................................... iv

ÖZET .......................................................................................................................................... v

TABLE OF CONTENT ............................................................................................................. vi

LIST OF FIGURES ................................................................................................................... ix

LIST OF TABLES ..................................................................................................................... xi

LIST OF SYMBOLS / ABBREVIATIONS ............................................................................. xii

1. INTRODUCTION .................................................................................................................. 1

2. BACKGROUND .................................................................................................................... 3

2.1 Sound Recording ........................................................................................................... 3

2.2 Types of Recording ........................................................................................................... 3

2.2.1 Analog recording ........................................................................................................ 3

2.2.2 Digital Recording ....................................................................................................... 4

2.3 History of Recording ......................................................................................................... 4

2.3.1 Phonograph ................................................................................................................. 4

2.3.2 Magnetic Tape ............................................................................................................ 6

3. DESIGN AND IMPLEMENTATION ................................................................................... 9

3.1 Development Board .......................................................................................................... 9

3.2 Microcontroller Technology ........................................................................................... 13

3.2.1 Key Features ............................................................................................................. 15

3.2.2 History of MSP430 ................................................................................................... 17

3.2.3 Pros ........................................................................................................................... 17

3.2.4 Cons .......................................................................................................................... 18

 vii

3.2.5 Why do we choose MSP430F2274?......................................................................... 18

3.3 Flash Memory ................................................................................................................. 20

3.3.1 Which one do we choose? ........................................................................................ 21

3.3.2 Key Features ............................................................................................................. 22

3.3.3 Programming Flash .................................................................................................. 24

3.4 File System ...................................................................................................................... 30

3.4.1 FAT........................................................................................................................... 31

3.4.2 Creating Your Own File System .............................................................................. 33

3.5 Sampling Theorem .......................................................................................................... 35

3.6 Quantization .................................................................................................................... 36

3.7 Coding ............................................................................................................................. 36

3.7.1 PCM .......................................................................................................................... 36

3.7.1.1 Modulation ........................................................................................................ 37

3.7.1.2 Demodulation.................................................................................................... 38

3.7.2 DPCM ....................................................................................................................... 39

3.7.3 ADPCM .................................................................................................................... 40

3.8 Microphone Circuitry ...................................................................................................... 43

3.9 Voice Recording Mechanism .......................................................................................... 46

3.10 Wav Format ................................................................................................................... 48

3.11 Communication ............................................................................................................. 50

3.11.1 RS-232 and Its Specifications ................................................................................ 52

3.11.1.1 RS-232 Specifications..................................................................................... 52

3.11.1.1.1 Electrical Characteristics ......................................................................... 53

3.11.1.1.2 Functional Characteristic ......................................................................... 54

3.11.1.1.3 Mechanical Interface Characteristics ....................................................... 55

 viii

3.11.2 Communication Principles of Flash Memory Based Voice Recorder System ....... 56

3.11.2.1 Communication Commands ........................................................................... 56

3.11.3 Guaranty Reliable Communication: Fletcher‘s Checksum Algorithm .................. 62

3.12 Finite State Machine ..................................................................................................... 63

3.13 Communicating with Real World: PC Application ...................................................... 65

4. ANALYSIS ........................................................................................................................... 69

4.1 Memory Usage ................................................................................................................ 69

4.2 Timing ............................................................................................................................. 70

4.2.1 CPU .......................................................................................................................... 70

4.2.2 Flash ......................................................................................................................... 71

4.3 Power Consumption ........................................................................................................ 72

4.3.1 CPU .......................................................................................................................... 72

4.3.2 Flash ......................................................................................................................... 73

5. CONCLUSION ..................................................................................................................... 74

5.1 Engineering Criteria ........................................................................................................ 74

5.2 Further Development ...................................................................................................... 76

REFERENCES ......................................................................................................................... 78
 ix

LIST OF FIGURES

Figure 2.1: Edison Phonograph .................................................................................................. 6

Figure 2.2: Magnetic Tape .......................................................................................................... 8
Figure 3.1: First development board ........................................................................................... 9
Figure 3.2: PCB of first development board ............................................................................. 10
Figure 3.3: Second development board .................................................................................... 11
Figure 3.4: PCB of second development board ........................................................................ 11
Figure 3.5: Third development board ....................................................................................... 12
Figure 3.6: PCB of third development board ............................................................................ 13
Figure 3.7: Block diagram of MSP430F2274 ........................................................................... 16
Figure 3.8: Comparison of NAND and NOR flash .................................................................. 21
Figure 3.9: Block diagram of flash ........................................................................................... 25
Figure 3.11: Sampling and quantization of a sine wave for 4-bit PCM. .................................. 37
Figure 3.12: Encode and decode mechanisms of DPCM ......................................................... 40
Figure 3.13: Encoder and decoder mechanisms of ADPCM .................................................... 42
Figure 3.14: General perspectives of sampling, quantization and coding ................................ 43
Figure 3.15: Microphone circuitry of the project ..................................................................... 45
Figure 3.16: Voice recording mechanism ................................................................................. 47
Figure 3.17: General format of wav file ................................................................................... 50
Figure 3.18: Synchronous data transmission format ................................................................ 51
Figure 3.19: Asynchronous data transmission format .............................................................. 51
Figure 3.20: Logic levels defined by RS-232 standard ............................................................ 53
Figure 3.21: Mechanical interfaces of RS-232 standard........................................................... 56
Figure 3.22: File table request .................................................................................................. 57
Figure 3.23: File request ........................................................................................................... 59
Figure 3.24: Setting time of MSP430 ....................................................................................... 60
Figure 3.25: Getting system internal time ................................................................................ 61
Figure 3.26: Resetting file system ............................................................................................ 62
Figure 3.27: Fletcher checksum algorithm is implemented in C programming language ........ 63
 x

Figure 3.28: Finite state machine of the project ....................................................................... 64

Figure 3.29: Connection menu of PC application..................................................................... 66
Figure 3.30: File menu of PC application ................................................................................. 67
Figure 3.31: Date menu of PC application ............................................................................... 68
Figure 3.32: Reset menu of PC application .............................................................................. 68
 xi

LIST OF TABLES

Table 3.1: Signal names of flash ............................................................................................... 26

Table 3.2: Pin description of bus operations............................................................................. 30
Table 3.3: Differences between different FAT file systems ..................................................... 32
Table 3.4: Wav audio formats................................................................................................... 49
Table 3.5: Electrical specification of the RS-232 standard ...................................................... 54
Table 3.6: Functional specification of the RS-232 standard ..................................................... 55
Table 4.1: Memory usage of whole code .................................................................................. 69
Table 4.2: Memory requirements of some modules ................................................................. 70
Table 4.3: CPU usage of key functions .................................................................................... 70
Table 4.4: Timing of Numonyx NAND flash ........................................................................... 71
Table 4.5: Current consumption of MSP430F2274 for different modes .................................. 72
Table 4.6: Current consumption of Numonyx NAND flash .................................................... 73
 xii

LIST OF SYMBOLS / ABBREVIATIONS

ADC Analog to Digital Converter

ADPCM Adaptive Differential Pulse Code Modulation
AEG Aallgemein Elektricitat Gesellschaft
AIFF Audio Interchange File Format
AL Address Latch
BASF Baden Aniline and Soda Factory
BSL Bootstrap Loader
CD Compact Disk
CL Command Latch
CPU Central Processing Unit
DAC Digital to Analog Converter
DB9S Data Bus 9 Pins
DCE Data Communication Equipment
DMA Direct Memory Access
DPCM Differential Pulse Code Modulation
DRAM Dynamic Random Access Memory
DS0 Digital Signal 0
DTE Data Terminal Equipment
DVD Digital Versatile Disc
DVI Digital Video Interface
DVR Digital Video Recorder
E Enable
EEPROM Electronically Erasable Programmable Read-Only Memory
EIA/TIA Electronic Industry Association and the Telecommunications Industry
Association
EPROM Erasable Programmable Read Only Memory
FAT File Allocation Table
 xiii

FSM Finite State Machine

GSM Global System for Mobile Communications
IBM International Business Machines
IFF Interchange File Format
IMA Integrated Media Association
I/O Input/Output
I2C Inter-Integrated Circuit
IrDA Infrared Data Association
ITU-R International Telecommunication Union Radio Communications Sector
LCD Liquid Crystal Display
JTAG Joint Test Action Group
LIN Local Interconnect Network
LPCM Linear Pulse Code Modulation
MCU Microcontroller Unit
MIPS Millions of Instruction Per Second
MP3 MPEG Audio Layer III
MS-DOS Microsoft Disk Operating System
NTFS New Technology File System
ONFI Open NAND Flash Interface
OPAMP Operational Amplifier
OS Operating System
PC Personal Computer
PCM Pulse Code Modulation
PCB Printed Circuit Board
PIC Programmable Intelligent Computer
P/E/R Program/Erase/Read
QFN Quad Flat No Leads
R Read
RAM Random Access Memory
RB Ready Busy
RIFF Resource Interchange File Format
 xiv

RISC Reduced Instruction Set Computer

ROM Read Only Memory
SDHC Secure Digital High Capacity
SPI Serial Peripheral Interface
TSSOP Thin Small Outline Package
TTL Transistor Transistor Logic
UART Universal Asynchronous Receiver/Transmitter
USART Universal Synchronous/Asynchronous Receiver/Transmitter
VOIP Voice over IP
W Write
WP Write Protect
 1

1. INTRODUCTION

Everyday a new technology product is coming into our houses. One of the most
promising technologies is solid state storage. Until now, we have seen MP3 players and video
players which are using this new technology. Also, in short term mass storage devices such as
harddrives are changing their structures to solid state technology. We want to adopt ourselves
to this new technology so we implemented a new solid state voice recording system. Our
system is recording on flash but playback function is implemented on PC by using Java
programming language.

We want to adopt ourselves to this new technology so we implemented a new solid

state voice recorder system. This system is based on a high density NAND flash which is used
to store records and a microcontroller which is processing samples, data and transmission of
records. First, ADC of microcontroller is capturing the samples by using a carbon microphone.
Second, CPU of microcontroller is processing samples and passing them through an ADPCM
encoder because raw samples need too many spaces and a recording system must use its space
in an intelligent way. Then, these encoded samples are stored on the NAND flash. Voice
samples are now on flash but they do not mean anything because a user cannot understand
which samples belong to which record so we develop a file system which gives the
information about record time, start address of the record and size of the record. Until now, we
get samples, record them on NAND flash and have the information about the record but this
means nothing to ordinary users and we must get back the records to real world so we develop
the PC application which can download and playback the records. To do so, it communicates
to MCU with RS-232 industrial standard by using our own communication protocol format
and it guarantees the reliability of communication by using Fletcher checksum algorithm.
Also, the PC application stores voice recordings as a common audio format WAV.

The organization of this report is as follows: Chapter 2 contains the information about
sound recording and its history. Chapter 3 is the design and implementation part of the project.
 2

At this chapter, hardware structure of the project, microcontroller technology, flash memory
technology, file system, PC application, sampling, quantization, coding, audio format,
communication protocol, microphone circuitry and finite state of project is expressed. Chapter
4 is analysis part of the project about memory usage, timing and power consumption.

Finally, we provided the results and conclusion in the conclusion chapter.

 3

2. BACKGROUND

2.1 Sound Recording

For almost 130 years the technology to preserve and replay sounds, that would
otherwise be lost the moment they occurred, have been used. In saving these brief sonic
events, people thought, to some extent, personal interaction in "real time" has to be replaced
with a one-way communication process that is independent of time. For example, people used
to either perform their own music or attend a public performance. Now people are more likely
to listen to recordings (or records) than gather around the piano or go to a concert. Music and,
to a lesser extent, the spoken word have become products that are consumed rather than an
activity that are shared. Thanks to the mass production of records and players, people now
have access to more diverse styles and more expertly performed music than ever before.
Ironically, few of us can now sing the songs our grandparents sang.

Sound recording and reproduction is an electrical or mechanical inscription and re-

creation of sound waves, such as spoken voice, singing, instrumental music, or sound effects.
The two main classes of sound recording technology are analog recording and digital
recording.

2.2 Types of Recording

2.2.1 Analog recording

Analog (or analogue) recording (Greek, ana is "according to" and logos "relationship")
is a technique used to store signals of audio or video information for later playback.
 4

In analog recording, original sound signal is modulated onto another physical medium
as a continual wave in or on the media. The wave might be stored as a physical texture on a
phonograph record, or a fluctuation in the field strength of a magnetic recording. This is
different from digital recording, which converts audio signals into discrete numbers and also
the physical quality in the medium (e.g., the intensity of the magnetic field or the path of a
record groove) is directly related, or analogous, to the physical properties of the original sound
(e.g., the amplitude, phase, etc.)

Analog recording methods store audio signals as a continual wave in or on the media.
The first successful demonstration of analog recording for audio was by Thomas Alva Edison.

2.2.2 Digital Recording

A digital recording is produced by converting the physical properties of the original

sound into a sequence of numbers, which can then be stored and played back for reproduction.
The accuracy of the conversion process depends on the sampling rate (how often the sound is
sampled and a related numerical value is created) and the sampling depth (how much
information each sample contains, which can also be described as the maximum numerical
size of each sampled value). However, unlike analog recording which depends critically on the
long-term durability of the fidelity of the waveforms recorded on the medium, the physical
medium storing digital samples is essentially immaterial in playback of the encoded
information so long as the original sequence of numbers can be recovered.

2.3 History of Recording

2.3.1 Phonograph

There was sound recording before the phonograph, but not sound reproduction. The
record player, phonograph or gramophone was the most common device for playing recorded
sound from the 1870s through the 1980s.
 5

The earliest known invention of a phonographic recording device was the

phonautograph, invented by Frenchman Edouard-Leon Scott de Martinville and patented on
March 25, 1857. It could transcribe sound to a visible medium, but had no means to play back
the sound after it was recorded. In 2008, phonautograph recordings were for the first time
played back as sound by American audio historians, using computers to decode the transcribed
waveforms. [1]

Charles Cros, a French scientist, produced a theory (April 18, 1877) concerning a
phonograph, but he did not manufacture a working model. His theory was submitted to the
French Academy of Sciences, and was read to the public in December 1877, by which time
Edison had produced a working model. Cros and Edison apparently discovered their theories
independently.

Thomas Alva Edison conceived the principle of recording and reproducing sound
between May and July 1877 as a byproduct of his efforts to "play back" recorded telegraph
messages and to automate speech sounds for transmission by telephone. [2][3] He announced
his invention of the first phonograph is shown in Figure 2.1, a device for recording and
replaying sound, on November 21, 1877, and he demonstrated the device for the first time on
November 29 (it was patented on February 19, 1878 as US Patent 200,521). Edison's early
phonographs recorded onto a tinfoil sheet phonograph cylinder using an up-down ("hill-and-
dale") motion of the stylus. [4][5]

The tinfoil sheet was wrapped around a grooved cylinder, and the sound was recorded as
indentations into the foil. Edison's early patents show that he also considered the idea that
sound could be recorded as a spiral onto a disc, but Edison concentrated his efforts on
cylinders, since the groove on the outside of a rotating cylinder provides a constant velocity to
the stylus in the groove, which Edison considered more "scientifically correct". [6]

Edison's patent specified that the audio recording be embossed, and it was not until 1886
that vertically modulated engraved recordings using wax coated cylinders were patented by
Chichester Bell and Charles Summer Tainter. They named their version the graphophone.
 6

Emile Berliner patented his gramophone in 1887. The gramophone involved a system of
recording using a lateral (back and forth) movement of the stylus as it traced a spiral onto a
zinc disc coated with a compound of beeswax in a solution of benzene. The zinc disc was
immersed in a bath of chromic acid; this etched the groove into the disc where the stylus had
removed the coating, after which the recording could be played.

Figure 2.1: Edison Phonograph

2.3.2 Magnetic Tape

Magnetic tape (in Figure 2.2) is a continuous plastic strip covered with magnetic oxide.
Nearly all recording tape is of this type, whether used for recording audio or video or for
computer data storage.

Practical magnetic tape recording originated with an Austrian inventor living in Dresden,
Germany, named Fritz Pfleumer. In the 1920s, sales of cigarettes boomed, and the
manufacturers wanted a cheap alternative to the expensive gold leaf band that decorated the
 7

tips of the most expensive cigarette brands. An expert in special paper and related processes
for industrial uses, Pfleumer succeeded in creating a process for striping a cheap gold-colored
bronze band on the cigarette paper. Pfleumer was interested in the magnetic recording process,
and reasoned that he could use his cigarette-paper manufacturing process to create a
magnetizable paper recording tape that would replace expensive and impractical wire and
steel.

The inventor glued pulverized iron particles onto a 16 millimeter-wide strip of paper,
creating the first magnetic tape. He also built the first tape recorder in the modern sense. For
his discovery, Pfleumer received German patent DE 500 900 on 31 January 1928, titled
―Lautschriftträger‖ (sound record carrier) and, amongst others, British patent GB 333,154,
―Improvements in or relating to Sound Records‖, on February 5, 1929. The audio tape
recorder produced poor quality sound that nonetheless offered great promise. The paper tape
tore readily, but could also easily be repaired with glue instead of the welding method of steel
and wire. In his promotion of the invention, Pfleumer pointed out that editing and assembling
the edited pieces of his recording tape were as easy as handling movie film, a well-established
technology. [7]

AEG realized that this invention was one of the most promising technology. AEG signed
contract with Pfleumer. The AEG engineers quickly found they were better at designing
electronics and tape transports than they were at experimenting with the chemical processes of
iron powder, glues, and coating methods. So they contacted the logical people to help them
with the job, the BASF division in Ludwigshafen of the giant chemical combine I.G.
Farbenindustrie. The two made a gentlemen‘s agreement‖, later formalized in a written
contract, that AEG would build the machines and BASF would make the tape, with their
development teams cooperating closely.

There was Second World War so this invention was kept in secret. It was only after the
war that Americans, particularly Jack Mullin, John Herbert Orr and Richard H. Ranger were
able to bring this technology out of Germany and develop it into commercially viable formats.
 8

A wide variety of recorders and formats have developed since, most significantly reel to
reel and Compact Cassette.

Figure 2.2: Magnetic Tape

 9

3. DESIGN AND IMPLEMENTATION

3.1 Development Board

Embedded system project needs a development board which is designed by

considering project requirement. We do not have such a board so I have learnt to draw PCB
(Printed Circuit Board) by using Mentor Graphics‘ PADS Logic and Layout program and then
I have designed three types of board in order to develop this project.

First board is based on MSP430F169 and NAND Flash but its features are not good
enough to solve problem. Its data pins are separate from each other around different ports and
this creates a big problem about timing of data.

Figure 3.1: First development board

 10

Figure 3.2: PCB of first development board

Second board is based on MSP430F2274 and NAND Flash but it has two big
problems. First problem is about flash‘s PCB design, flash does not have some open drain pull
up resistance. Second problem is about microphone circuitry. Microphones pins are connected
to ground so there is no a voltage difference between pins and we can not get any sample.
 11

Figure 3.3: Second development board

Figure 3.4: PCB of second development board

 12

Third board is also based on MSP430F2274 and NAND Flash. Its designed is like
same second board but third one has two microphone circuits, one of the circuits is active but I
can choose which one will be active. Third one has two more additional connectors, RJ10 and
RJ12. Open drain pull up resistances are also designed. It also has test points to test flash‘s pin
configuration.

Figure 3.5: Third development board

 13

Figure 3.6: PCB of third development board

Three boards have flash memory, UART module, buttons, microphone circuit and
external clocks. First board has MSP430F169, wireless module slot and LCD. The second and
third one have MSP430F2274 and operational amplifiers. We chose third one because it has
more functional units and does not have any problematic structure.

3.2 Microcontroller Technology

Microcontroller (also called as microcontroller unit) is a kind of small computer on a

single integrated circuit consisting of relatively small CPU compared with a normal PC and
also supported with crystal oscillator, timers, watchdog and analog I/O. It also has other
functionalities such as ROM or a flash which is small capacity and has simple read/write
operations. Microcontroller is used in small embedded systems. It usually waits for an input
from a sensor or an interrupt which is usually generated by pressing a button and can sleep if
there is no event. It operates at very low frequencies such as 32 KHz so it is used at low power
consumption, long life stable, battery used embedded applications such as wireless sensors,
 14

remote control, engine control, etc. There are many different well known microcontroller
families such as PIC, 8051 and MSP430. [8]

PIC (Programmable Intelligent Computer) is the first microcontroller unit which uses
Harvard architecture and is created by Microchip Company at middle of 1970s. Harvard
architecture is the computer architecture which has separate instruction memory and data
memory. PIC‘s 14 cores model is the most well known model at industry because it is good to
program and affordable but the separate memories makes it much slower than other structures.
[9]

8051 is another well known microcontroller unit which is created by Intel Company at
1980s. It is an ancient model but its basic structure is well organized so many different
companies (such as Atmel, Dallas Semiconductors and Philips) buy it and make their
enhancement. Because of these reasons, it has wide application areas. [10]

MSP430 is one of the most promising microcontroller families from Texas

Instruments. It is built around 16 bit CPU and designed for low cost, low battery consumption
embedded applications. MSP430 enables system designers to simultaneously interface to
analog signals, sensors and digital components while maintaining unmatched low power.
Typical applications include utility metering, portable instrumentation, intelligent sensing, and
consumer electronics. [36]

Its peripherals are usually watchdog timer, timers, DMA, ADC, DAC, flash memory,
I2C, SPI, USART, OPAMPs, brownout reset and LCD. Apart from some older EPROM
(PMS430E3xx) and high volume mask ROM (MSP430Cxxx) versions, all of the devices are
in-system programmable via JTAG or a built in bootstrap loader (BSL) using RS-232 but the
included peripherals vary from device to device.

MSP430 is a good choice to be used in low powered embedded applications. In idle

mode its power consumption is lower than 1 micro ampere. Its clock frequency varies and the
top speed is 25 MHz. MIPS and MHz do not mean the same thing. MIPS means millions of
 15

instructions per second and MHz is the number of clock signals generated by CPU divide by
one million, so sometimes lower clock frequencies have bigger MIPS rates.

Consumers usually need more than the system has but the designers need some
specifications and limitations to make it much simpler. MSP430 does not have external
memory bus so it has a limited flash memory from 10 KB to 32 KB so it is not suitable for
larger memory and buffer needed application.

MSP430 uses von Neumann architecture which means a single address space for data
and instructions. Its memory is byte addressable and pair of bytes are mapped as little endian
format to get a 16-bit word.

CPU has 16 bit registers and these registers are used as program counter, stack pointer,
constant generators and general purpose registers.

The instruction set is very simple; there are 27 instructions in three families. Most
instructions are available in 8-bit (byte) and 16-bit (word) versions, depending on the value of
a B/W bit - the bit is set to 1 for 8-bit and 0 for 16-bit. Byte operations to memory affect only
the addressed byte, while byte operations to registers clear the most significant byte. [11][12]

3.2.1 Key Features

Key features of the MSP430x2xx family include:

 Ultra low-power architecture extends battery life

o 0.1-μA RAM retention
o 0.8-μA real-time clock mode
o 250-μA / MIPS active
 16

 High-performance analog ideal for precision measurement

o Comparator-gated timers for measuring resistive elements

 16-bit RISC CPU enables new applications at a fraction of the code size.
o Large register file eliminates working file bottleneck
o Compact core design reduces power consumption and cost
o Optimized for modern high-level programming
o Only 27 core instructions and seven addressing modes
o Extensive vectored-interrupt capability

 In-system programmable Flash permits flexible code changes, field upgrades

and data logging

Details and specification of MSP430 is not included in this report. You can find very
detailed description and specifications of MSP430F2274 in MSP430x2xx Family User Guide,
MSP430F2274 Datasheet and MSP430F2274 Device Erratasheet. [11][12][13]

The block diagram of MSP430F2274 is shown in Figure 3.7:

Figure 3.7: Block diagram of MSP430F2274

 17

3.2.2 History of MSP430

MSP430 family is established approx. 10 years ago. In the beginnings there are no
UARTs and LCD-display. Many peripherals are not included, they are just in mind so it is not
attractive and there are no good development tools. At the end of nineties , the MSP430 family
and its development tools became very attractive because the cost of system has been lowered
dramatically and JTAG based programming & debugging in flash based devices made
development tools cheaper. Then MSP430 became a good competitor for 8051 and PIC for
low budget wide range industry.

3.2.3 Pros

MSP430 is used in a wide range application size from smaller to bigger and its current
competitors are usually PIC and 8051 architecture.

MSP430 uses von Neumann architecture so it does not have an external bus. It has
linear address space for 8 bit and 16 bit instructions. By the use of a stable well known linear
address space; programming from memory to register, register to memory, peripherals to
memory and vice versa have an appropriate way to communicate over an efficient protocol.
Single 16 bit transfer from memory to peripheral and vice versa changes the value of variables
by interrupts so usually C programmers make their biggest mistakes to think programming
MSP430 as usual C programming so they must consider about this before starting to program.

Harvard architecture was used more common in the past. Using separate address space
and data space made programming easier in old days. The needs for using larger memory are
starting to grow up rapidly, so then everything turns back because very large external memory
and flash are not easy to manage. It became complicated to program. Codes were not modular
and reusable. All of these reasons make MSP430 usage more common. MSP430‘s linear
address space creates good code utilization for programmers.
 18

3.2.4 Cons

MSP430s market usage is growing up rapidly but there are some small problems. In
past, many companies used PIC and 8051 so the development tools were created according to
them so MSP430 does not have sufficient number of programming environment. Another
problem is MSP430s development tools are based on C and assembly programming languages
but programmers want to have higher level more modular programming environment. These
small problems are being solved in a few years because of growing market rate.

3.2.5 Why do we choose MSP430F2274?

Other key features of MSP430F2274:

 Basic Clock Module Configurations:

o Internal Frequencies up to 16 MHz With
o Four Calibrated Frequencies to ±1%
o Internal Very-Low-Power Low-Frequency
o Oscillator
o 32-kHz Crystal
o High-Frequency Crystal up to 16 MHz
o Resonator
o External Digital Clock Source
o External Resistor

 16-Bit Timer_A With Three Capture/Compare Registers

 16-Bit Timer_B With Three Capture/Compare Registers

 19

 Universal Serial Communication Interface

o Enhanced UART Supporting
o Auto-Baudrate Detection (LIN)
o IrDA Encoder and Decoder
o Synchronous SPI
o I2C™

 10-Bit, 200-ksps A/D Converter With Internal Reference, Sample-and-Hold,

Autoscan, and Data Transfer Controller

 Two Configurable Operational Amplifiers

 Brownout Detector

 Serial Onboard Programming, No External Programming Voltage Needed

Programmable Code Protection by Security Fuse

 Bootstrap Loader

 On Chip Emulation Module

 1 KB RAM

 32 KB + 256 B Flash Memory

In out project, we need operational amplifiers for microphone circuit, ADC to convert
voice to digital pulse, timer_a for voice sampling, timer_b for time calculation, clock module
for 32 KHz crystal, enhanced UART support for PC communication RS-232, data transfer
controller to write voice to flash without usage of CPU and other features when we need.
MSP430F2274 has two types of package: one of is 40-pin QFN and the other one is 38-pin
 20

TSSOP. We choose 38-pin TSSOP package type because it is easy to be populated and
soldered. [11][12]

3.3 Flash Memory

Flash memory technology is a mix of EPROM and EEPROM technologies. The term
―flash‖ was chosen because a large chunk of memory could be erased at one time. The name,
therefore, distinguishes flash devices from EEPROMs, where each byte is erased individually.
Flash memory technology is today a mature technology. It is a strong competitor to other
nonvolatile memories such as EPROMs and EEPROMs, and to some DRAM applications.
Flash is a solid state device so for the rest of this thesis, we refer solid state to flash.

In the internal circuit configuration of NOR Flash, the individual memory cells are
connected in parallel, which enables the device to achieve random access. This configuration
enables the short read times required for the random access of microprocessor instructions.
NOR Flash is ideal for lower-density, high-speed read applications, which are mostly read
only, often referred to as code-storage applications.

NAND Flash was developed as an alternative optimized for high-density data storage,
giving up random access capability in a tradeoff to achieve a smaller cell size, which translates
to a smaller chip size and lower cost-per-bit. This was achieved by creating an array of eight
memory transistors connected in a series. Utilizing the NAND Flash architecture‘s high
storage density and smaller cell size, NAND Flash systems enable faster write and erase by
programming blocks of data. NAND Flash is ideal for low-cost, high-density, high-speed
program/erase applications, often referred to as data-storage applications. [14][15]
 21

Comparison of NOR and NAND Flash is shown in Figure 3.8:

Figure 3.8: Comparison of NAND and NOR flash

The communication between flash and MCU is divided into two categories; one is
serial and the other one is parallel. Serial communication needs SPI communication interface
so the clock rate is much more important. Parallel communication does not need any SPI
communication interface but reading is slower than serial one. Serial communication sends
data as bits but parallel communication usually sends data as byte or two bytes. [16]

3.3.1 Which one do we choose?

Flash memory based voice recording system depends on two key features. The first
feature is selection of MCU and we did it above. The second important feature is flash
memory technology.
 22

Voice recording is a data storage processing. Writing speed is very important at voice
recording system because the system needs no sample loss to get a good voice record and data
integrity. Reading speed is not so important for this development because reading from flash is
done by RS-232 communication in order to transfer data to PC application.

According to above properties, we choose parallel NAND flash memory. Parallel

communication is required to write samples at the same time. NAND technology is used as
data storage. The writing speed is much better in NAND and reading speed is not so important
but the sequential reading speed is also sufficient to carry communication busy and not to lose
connection.

The used product is Numonyx NAND02G-B2D.

3.3.2 Key Features

Numonyx NAND02G-B2D‘s key features are :

 High density NAND flash memory

o Up to 2 Gbits of memory array
o Cost-effective solution for mass storage applications

 NAND interface
o x8 or x16 bus width
o Multiplexed address/data

 Supply voltage: 1.8 V or 3 V device

 Page size
o x8 device: (2048 + 64 spare) bytes
o x16 device: (1024 + 32 spare) words
 23

 Block size
o x8 device: (128 K + 4 K spare) bytes
o x16 device: (64 K + 2 K spare) words

 Multiplane architecture
o Array split into two independent planes
o Program/erase operations can be performed on both planes at the same
time

 Page read/program
o Random access: 25 us (max)
o Sequential access: 25 ns (min)
o Page program time: 200 us (typ)
o Multiplane page program time (2 pages): 200 us (typ)

 Copy back program with automatic EDC (error detection code)

 Cache read mode

 Fast block erase

o Block erase time: 1.5 ms (typ)
o Multiblock erase time (2 blocks): 1.5 ms(typ)

 Status register

 Electronic signature

 Chip Enable ‗don‘t care‘

 Serial number option

 24

 Data protection:
o Hardware program/erase disabled during power transitions
o Non-volatile protection option

 ONFI 1.0 compliant command set

 Data integrity
o 100,000 program/erase cycles (with ECC)
o 10 years data retention

 ECOPACK® packages

We choose x8 device because MCU does not have enough pins (x8 uses 15 pins
andx16 uses 23pins) and x8 is enough to save voice samples. [17]

3.3.3 Programming Flash

NAND flash is a non volatile memory structure so programming NAND flash is not as
easy as programming a volatile memory.

NAND flash basically has two structures. First one is page which has 2048 byte data
and 64 byte spare area. Data part is the part where audio samples are stored and spare are is
the part where error correction and other stuff can done. Second structure is block which
consists of 64 pages. If error correction is implemented and on then each block can be
programmed and erased up to 100,000 cycles.

Each page can read and program individually. Up to 4 (included) different

programming attempts can be done to a page but if there is a need for fifth attempt then the
block ,which programming page is in, must be erased completely. [17]
 25

Used architecture is based on x8.Block diagram of flash is shown in Figure 3.9:

Figure 3.9: Block diagram of flash

 26

Signal names of flash are shown in Table 3.1:

Table 3.1: Signal names of flash

Signal description of flash is [17]

 Inputs/outputs (I/O0-I/O7)
Input/outputs 0 to 7 input the selected address, output the data during a
read operation, or input a command or data during a write operation. The inputs
are latched on the rising edge of Write Enable. I/O0-I/O7 are left floating when
the device is deselected or the outputs are disabled.

 Address Latch Enable (AL)

The Address Latch Enable activates the latching of the address inputs in
the command interface. When AL is High, the inputs are latched on the rising
edge of Write Enable.
 27

 Command Latch Enable (CL)

The Command Latch Enable activates the latching of the command
inputs in the command interface. When CL is High, the inputs are latched on
the rising edge of Write Enable.

 Chip Enable (E)

The Chip Enable input, E, activates the memory control logic, input
buffers, decoders and sense amplifiers. When Chip Enable is Low, V IL, the
device is selected. If Chip Enable goes High, V IH, while the device is busy, the
device remains selected and does not go into standby mode.

 Read Enable (R)

The Read Enable pin, R, controls the sequential data output during read
operations. Data is valid t after the falling edge of R. The falling edge of R also
increments the internal column address counter by one.

 Write Enable (W)

The Write Enable input, W, controls writing to the command interface,
input address, and data latches. Both addresses and data are latched on the
rising edge of Write Enable. During power-up and power-down a recovery time
of 10 us (min) is required before the command interface is ready to accept a
command. It is recommended to keep Write Enable high during the recovery
time.

 Write Protect (WP)

The Write Protect pin is an input that gives a hardware protection
against unwanted program or erase operations. When Write Protect is Low,
VIL, the device does not accept any program or erase operations. It is
recommended to keep the Write Protect pin Low, VIL, during power-up and
power-down.
 28

 Ready/Busy (RB)
The Ready/Busy output, RB, is an open-drain outputs that can be used
to identify if the P/E/R controller is currently active. When Ready/Busy is Low,
VOL, a read, program or erase operation is in progress. When the operation
completes, Ready/Busy goes High, VOH. The use of an open-drain output
allows the Ready/Busy pins from several memories to be connected to a single
pull-up resistor. A Low then indicates that one or more of the memories is/are
busy. During power-up and power-down a minimum recovery time of 10 us is
required before the command interface is ready to accept a command. During
this period the RB signal is Low.

 VDD supply voltage

VDD provides the power supply to the internal core of the memory
device. It is the main power supply for all operations (read, program and erase).
Each device in a system should have VDD decoupled with a 0.1 uF capacitor.
The PCB track widths should be sufficient to carry the required program and
erase currents.

 VSS ground
Ground, VSS, is the reference for the power supply. It must be connected
to the system ground.

There are six standard bus operations that control the memory:

 Command input
Command input bus operations give commands to the memory.
 29

 Address input
Address input bus operations input the memory addresses. Five bus
cycles are required to input the addresses.

 Data input
Data input bus operations input the data to be programmed.

 Data output
Data output bus operations read the data in the memory array, the status
register, the electronic signature, and the unique identifier.

 Write protect
Write protect bus operations are used to protect the memory against
program or erase operations. When the Write Protect signal is Low, the device
does not accept program or erase operations, and, therefore, the contents of the
memory array cannot be altered. The Write Protect signal is not latched by
Write Enable to ensure protection, even during power up.

 Standby
When Chip Enable is High the memory enters standby mode, the device
is deselected, outputs are disabled, and power consumption is reduced.
 30

Pin descriptions of bus operations are shown in below Table 3.10:

Table 3.2: Pin description of bus operations

Four types of commands are used in the project. First one is sequential which reads data
sequentially reading from flash memory. Second one is sequential programming page which
write to flash sequentially. Third one is block erase which erase whole pages in the given
block. The last one is reading signature of flash memory which reads properties of the flash.

3.4 File System

Most users of computers are roughly familiar with what a file system does, what a file
is, what a directory is, and so on. This knowledge is gained from direct experience with
computers. Instead of basing our discussion on prior experiences, which will vary from user to
user, we will start over again and think about the problem of storing information on a
computer, and then move forward from there.

The main purpose of computers is to create, manipulate, store, and retrieve data. A file
system provides the machinery to support these tasks. At the highest level a file system is a
way to organize, store, retrieve, and manage information on a permanent storage medium such
as a disk. File systems manage permanent storage and form an integral part of all operating
systems.
 31

There are many different approaches to the task of managing permanent storage. At
one end of the spectrum are simple file systems that impose enough restrictions to
inconvenience users and make using the file system difficult. At the other end of the spectrum
are persistent object stores and object-oriented databases that abstract the whole notion of
permanent storage so that the user and programmer never even need to be aware of it. The
problem of storing, retrieving, and manipulating information on a computer is of a general-
enough nature that there are many solutions to the problem.

There is no ―correct‖ way to write a file system. In deciding what type of filing system
is appropriate for a particular operating system, we must weigh the needs of the problem with
the other constraints of the project. For example, a flash-ROM card as used in some game
consoles has little need for an advanced query interface or support for attributes. Reliability of
data writes to the medium, however, are critical, and so a file system that supports journaling
may be a requirement. Likewise, a file system for a high-end mainframe computer needs
extremely fast throughput in many areas but little in the way of user-friendly features, and so
techniques that enable more transactions per second would gain favor over those that make it
easier for a user to locate obscure files. [18]

There are many different types of file systems such as: FAT16/FAT32, NTFS, etc. or
creating your own file system.

3.4.1 FAT

FAT (File Allocation Table) is a partially patented file system developed by Microsoft
for MS-DOS®. It involves the use of a table that centralizes the information about the areas
belonging to files. This information helps to identify which areas are free or usable, and where
each file is stored on the disk. It also identifies the areas that are unusable. To reduce the
management complexity, disk space is allocated to files in contiguous groups of hardware
sectors called as clusters. The maximum number of clusters possible that can be addressed by
a file system depends on the number of bits being used to identify a cluster in FAT.
 32

Currently, there are three types of FAT file systems: FAT12, FAT16, and FAT32. The
basic difference between them is the number of bits in the FAT entries on a disk. FAT12 uses
12-bit entries, FAT16 uses 16-bit entries, and FAT32 uses 28-bit entries. The FAT standard
has also evolved in several ways, preserving backward compatibility with existing software.
Table 3.11 shows basic differences between the different FAT file systems. [20][21]

Table 3.3: Differences between different FAT file systems

FAT is still a normal file system for removable media (with the exception of CDs and
DVDs), with FAT12 used on floppies and FAT16 used on most other removable media (such
as flash memory cards for digital cameras, mobile phones, and USB flash drives). Most
removable media are not yet large enough to benefit from FAT32, although some larger flash
drives such as Secure Digital High-Capacity (SDHC) cards use it. FAT16 is used on these
drives for reasons of compatibility and size overhead.

There are many commands such as: listing in current directory files and folders,
creating a directory, copying a file, reading a file, searching a file, navigating through sub-
folders in current directory, or deleting a file or directory.
 33

3.4.2 Creating Your Own File System

If a programmer creates its own file system, the reason to do so could be

 Known file systems are not suited well to the programmer‘s structure.
 Known file systems can be very large and needs too many disk space.
 Known file systems are very complicated to implement.
 User interactions are not permitted or system is not using external commands
which are given by users.

In the voice recording system, file system is not directly reachable by users. It just gets
the requests of the users and gives the right output by using UART communication. Recording
a file and putting its properties into the file system is done by MCU. In this system, user can
not generate its directory or file, he just requests a file and receive it. Because of these reasons,
we generate our own file system. In our system, there is a file system table and records (files).
In file system table there are three columns: first one is name column (also called as timestamp
because system uses timestamp as file name), the second one is start address column which
gives the start address of a record in external flash and the last one is size column which gives
the size of a record.

In file system, name column and size column are kept together and start address
column is kept in a different location. File system is implemented in internal flash. The reason
to do so is external flash has blocks and pages (each of which is subpart of a block), each
page can be programmed individually but system can only do four write attempts, then in
order to do fifth attempt the whole block need to be deleted. The size of a block is very big
and RAM of MCU is not enough to keep all data of a block. If we choose to use a small part
of a block then whole remaining part of the block is wasted. In internal flash, we have four
different chunks (big buffer) each of has 512 byte. We wrote encoded samples to internal
flash. If size of samples in a chunk reaches 512 byte then we write new encoded samples to
 34

next chunk. If all four chunks are full of encoded samples which mean that we have 2048 byte
encoded sample then we write 2048 byte to external flash as its page and empty big buffer.

The general structure of file system is shown in Figure 3.10:

Figure 3.10: General structure of our own file system

Each column has 256 byte total size. As we mentioned above, name column and file
size column are kept together because they are sent together to PC application but start address
column is different, there is no reason to send it to user because user does not use this
information. If there is a file request user chooses file name and send it to MCU, then MCU
finds corresponding start address of file name and sends the requested file to user. Also system
has just 1 KB RAM and there is not enough to space to keep all values if table is bigger so
system internally restricts the size of file system as 256 bytes for each column.

The timestamp of the system is kept as minute according to system internal epoch
01.01.2009 (DD.MM.YYYY) 00:00 (HH:MM) but PC application converts the minutes which
is based on our epoch to a human readable way like DD.MM.YYYY HH:MM.

Voice recoding is a sequential process so in external flash we write samples as page

one after another. Each record can be downloaded to PC by itself and downloaded record is
not deleted in external flash. User can download whatever he wants but when he wants to
delete then whole table must be deleted because file allocation system is implemented to do it
so there is no segmentation in external flash and all records are kept as one after another.
 35

3.5 Sampling Theorem

In our lives, continues information keeps to many space and time. Maybe the receiver
cannot handle too many information at a time or the required bandwidth to do this process is
not always maintained. To keep communication alive and handle all other problems system
does not want to process all information, it just wants some part of the information this
information is called as sample and getting samples is known as sampling so sampling means
picking some values of a continues signal and keeping the value of continues signal as discrete
quantities. In order to keep information related, sampling must be done in a given time period
at a determined number of samples over a continues signal.

Continues signals are frequency limited waveforms. If we are sampling a continues

bandwidth limited waveform, in order to reproduce it we need to follow a theorem which is
called Sampling Theorem. Sampling Theorem says that if a waveform is bandwidth-limited to
W Hz, then it can be represented exactly by its samples if they are taken at a rate of 2W
samples/second or more. If two times of frequency is bigger than sampling rate then there are
some aliasing frequencies and information is corrupted. In the world of video, several forms of
visual aliasing are possible. For example, temporal aliasing occurs where the wheel of a car
seems to turn backwards. [22]

The Nyquist-Shannon Sampling Theorem is a fundamental tenet in the field of

information theory, in particular telecommunications. This theorem states that, when
converting from an analog signal to digital(or otherwise sampling a signal at discrete
intervals), the sampling frequency must be greater than twice the highest frequency of the
input signal in order to be able to reconstruct the original perfectly from the sampled version.
If the sampling frequency is less than this limit, then frequencies in the original signal that are
above half the sampling rate will be ―aliased‖ and will appear in the resulting signal as lower
frequencies. The theorem was first formulated by Harry Nyquist in 1928 ("Certain topics in
telegraph transmission theory" reference book), but was only formally proved by Claude E.
Shannon in 1949 ("Communication in the presence of noise" reference book). Mathematically,
the theorem is formulated as a statement about the Fourier Transform. [23]
 36

In digital world, every sampling needed project is based on this theory. As an example,
the sampling rate used for CDs nowadays is 44,100 samples per second. That means when you
play a CD, the speakers in your stereo system are moved to a new position 44,100 times per
second, or once every 23 microseconds. Controlling a speaker this fast enables it to generate
any sound in the human hearing range because we cannot hear frequencies higher than around
20,000 cycles per second, and a sampling rate more than twice the highest frequency in the
sound guarantees that exact reconstruction is possible from the samples. [24]
In voice recorder project, the Nyquist-Shannon sampling theorem is also used. Sampling
frequency is 8 KHz, voices are passed through a low pass filter and their frequencies are
between 338.5 Hz to 3000 Hz. Sampling rate is bigger than twice of the highest frequency.

3.6 Quantization

A digital audio signal is usually represented by uniformly sampled values with a fixed
word length N, which means that each sample can have a value between – (2N-1) and (2N-1-1).
The digital sample value represents the signal amplitude at a specified instant (the sample
instant) as shown in figure 2. The number of samples per second is specified by the sampling
frequency fS. This technique is called linear quantization or LPCM (Linear Pulse Code
Modulation).

3.7 Coding

3.7.1 PCM

Pulse-code modulation (PCM) is a digital representation of an analog signal where the

magnitude of the signal is sampled regularly at uniform intervals, then quantized to a series of
symbols in a numeric (usually binary) code. PCM has been used in digital telephone systems
and 1980s-era electronic musical keyboards. It is also the standard form for digital audio in
computers and the compact disk red books format. It is also standard in digital video, for
example, using ITU-R BT-601. Uncompressed PCM is not typically used for video in standard
 37

definition consumer applications such as DVD or DVR because the bit rate required is far too
high. [25][26][27]

3.7.1.1 Modulation

Figure 3.11: Sampling and quantization of a sine wave for 4-bit PCM.

In Figure 3.11, a sin wave (red curve) is sampled and quantized for PCM. The sine
wave is sampled at regular intervals, shown as ticks on the x-axis. For each sample, one of the
available values (ticks on the y-axis) is chosen by some algorithm (in this case, the floor
function is used). This produces a fully discrete representation of the input signal (shaded
area) that can be easily encoded as digital data for storage or manipulation. For the sine wave
example at right, we can verify that the quantized values at the sampling moments are 7, 9, 11,
12, 13, 14, 14, 15, 15, 15, 14, etc. Encoding these values as binary numbers would result in the
 38

following set of nibbles: 0111, 1001, 1011, 1100, 1101, 1110, 1110, 1111, 1111, 1111, 1110,
etc. These digital values could then be further processed or analyzed by a purpose-specific
digital signal processor or general purpose CPU. Several Pulse Code Modulation streams
could also be multiplexed into a larger aggregate data stream, generally for transmission of
multiple streams over a single physical link. This technique is called time division
multiplexing, or TDM, and is widely used, notably in the modern public telephone system.

3.7.1.2 Demodulation

To produce output from the sampled data, the procedure of modulation is applied in
reverse. After each sampling period has passed, the next value is read and a signal is shifted to
the new value. As a result of these transitions, the signal will have a significant amount of
high-frequency energy. To smooth out the signal and remove these undesirable aliasing
frequencies, the signal would be passed through analog filters that suppress energy outside the
expected frequency range (that is, greater than the Nyquist frequency fs / 2). Some systems
use digital filtering to remove some of the aliasing, converting the signal from digital to analog
at a higher sample rate such that the analog filter required for anti alising is much simpler. In
some systems, no explicit filtering is done at all; as it's impossible for any system to reproduce
a signal with infinite bandwidth, inherent losses in the system compensate for the artifacts —
or the system simply does not require much precision. The sampling theorem suggests that
practical PCM devices, provided a sampling frequency that is sufficiently greater than that of
the input signal, can operate without introducing significant distortions within their designed
frequency bands.
 39

3.7.2 DPCM

Differential pulse code modulation (DPCM) is a procedure of converting an analog

into a digital signal in which an analog signal is sampled and then the difference between the
actual sample value and its predicted value (predicted value is based on previous sample or
samples) is quantized and then encoded forming a digital value.

DPCM code words represent differences between samples unlike PCM where code
words represented a sample value.

Basic concept of DPCM - coding a difference, is based on the fact that most source
signals show significant correlation between successive samples so encoding uses redundancy
in sample values which implies lower bit rate.

Realization of basic concept (described above) is based on a technique in which we

have to predict current sample value based upon previous samples (or sample) and we have to
encode the difference between actual value of sample and predicted value (the difference
between samples can be interpreted as prediction error).
Because it's necessary to predict sample value DPCM is form of predictive coding.

DPCM compression depends on the prediction technique, well-conducted prediction

techniques lead to good compression rates, in other cases DPCM could mean expansion
comparing to regular PCM encoding. [28]
 40

Figure 3.12: Encode and decode mechanisms of DPCM

3.7.3 ADPCM

Adaptive DPCM (ADPCM) is a variant of DPCM (differential pulse-code modulation)

that varies the size of the quantization step, to allow further reduction of the required
bandwidth for a given signal to noise ratio. Typically, the adaptation to signal statistics in
ADPCM consists simply of an adaptive scale factor before quantizing the difference in the
DPCM encoder.

ADPCM was developed in the early 1970s at Bell Labs for voice coding, by P.
Cummiskey, N. S. Jayant and James L. Flaganan.
 41

In telephony a standard audio signal for a single phone call is encoded as 8000 analog
samples per second, of 8 bits each, giving a 64 kbit/s digital signal known as DS0. The default
signal compression encoding on a DS0 is either mu-law PCM (North America and Japan) or A
law PCM (Europe and most of the rest of the world). These are logarithmic compression
systems where a 12 or 13 bit linear PCM sample number is mapped into an 8 bit value. This
system is described by international standard G.711. Where circuit costs are high and loss of
voice quality is acceptable, it sometimes makes sense to compress the voice signal even
further. An ADPCM algorithm is used to map a series of 8 bit u-law (or a-law) PCM samples
into a series of 4 bit ADPCM samples. In this way, the capacity of the line is doubled. The
technique is detailed in the G.726 standard. [29]

The Interactive Multimedia Association (IMA) recommended a standard for ADPCM-

codec in multimedia applications, known as IMA or DVI ADPCM, in the early 1990s. This
algorithm is now used in most cross-platform ADPCM-based audio applications. The
proposed IMA-standard, now widely used in multimedia applications, uses an adaptive
quantizer, but a fixed predictor to limit the computational complexity. The predictor is
identical to the one in the DPCM. The compression level is 4:1, which means the 16-bit
original signal is quantized to 4-bits. The stepsize of the quantization depends on the input
signal, thus making it adaptive.

ADPCM encoder and decoder mechanisms are shown in Figure 3.13:

 42

Figure 3.13: Encoder and decoder mechanisms of ADPCM

In this project, sampling rate is used as 8 KHz, quantization is done over 10 bit and used
coding is ADPCM IMA.

Sampling, quantization and coding is shown in Figure 3.14:

 43

Figure 3.14: General perspectives of sampling, quantization and coding

3.8 Microphone Circuitry

A microphone, sometimes referred to as a mic or mike, is an acoustic-to-electric

transducer or sensor that converts sound into an electrical signal. Microphones are used in
many applications such as telephones, tape recorders, hearing aids, motion picture production,
live and recorded audio engineering, in radio and television broadcasting and in computers for
recording voice, VOIP, and for non-acoustic purposes such as ultrasonic checking. There
 44

many different types of microphone systems such as dynamic microphones, carbon

microphones, laser microphones, etc. [30]

Carbon microphone is a capsule containing carbon granules pressed between two metal
plates. A voltage is applied to across metal plates. One of the plates is diaphragm; sound
waves vibrate the diaphragm and apply a varying pressure to carbon. Carbon granules are
deformed. The distances between granules are changed and this changes the resistance of
system. The electrical resistance is changed so this produces electrical signal.

Carbon microphones are very cheap systems and easily found in everywhere. It is also
used as an amplifier so I choose carbon microphone in my project.

Human can hear frequencies between 20 Hz and 20 KHz. Human usually use
frequencies between 300 Hz to 3500 Hz. To reduce the noise and have good quality of
recording , frequencies which are out of this range must be suppressed. In order to suppress
the frequencies I need to use filter.

Filter is a system which suppresses higher or lower frequencies than the cutoff
frequency of the system. There two basic types of filter: the first one is low pass filter which
suppresses the frequencies higher than cutoff frequency and the second one is high pass filter
which suppresses the frequencies lower than cutoff frequency. I need to use a high and a low
pass filter. The combination of this filter is called as bandpass filter. A bandpass filter is a
filter which has two different filters: low pass and high pass filter. The bandpass filter
suppresses the frequencies higher than the cutoff frequency of the low pass filter and also
suppresses the frequencies lower than the cutoff frequency of the high pass filter.

Calculation of cut of frequency of a simple filter has the formula of:

Frequency = 1
(2 xxCxR )
 45

R : resistance value of resistor in ohm

C : capacitance value of capacitor in F.

Human can hear frequencies between 20 Hz to 20 KHz. Used frequencies, while we

are talking, are between 300 Hz to 3 KHz. I try to suppress the other frequencies by using a
bandpass filter.

Our microphone circuitry is shown in below Figure 3.15:

Figure 3.15: Microphone circuitry of the project

The high pass filter has two elements. First one is the capacitor has the value of 470
nF and second one is the resistor has the value of 1 KOhm.
 46

The cutoff frequency of high pass filter is:

Frequency = 1
(2 xxCxR )
R : 1 Kohm
C : 470 nF
Frequency = 338.5 Hz

The cutoff frequency of low pass filter is:

Frequency = Frequency = 1
(2 xxCxR )
R : 150 Kohm
C : 350 pF
Frequency = 3030.3 Hz .

My bandpass filter passes frequencies between 338.5 Hz and 3030.3 Hz.

3.9 Voice Recording Mechanism

Our system recording mechanism is a digital recording mechanism. Digital recording

is converting the physical properties of the original sound into a sequence of numbers.
Accuracy of recording depends on two things

1. Sampling Rate
Sampling rate means how many sample we have in one second.

2. Sampling Depth
Sampling depth means how many bits we use in order to show original sound
of data in digital world.

Our system sampling rate is 8192 Hz and sampling depth (resolution) is also 10 bit.
Voice coding mechanism, is also mentioned before, is ADPCM encoding which is 32 kbps,
requires low computation and gives acceptable quality for voice.
 47

Voice recording mechanism is shown in Figure 3.16:

Figure 3.16: Voice recording mechanism

Microphone circuitry catches voice and sends it to ADC. ADC gets these samples and puts
them into a small buffer. If buffer gets 128 samples then it gives an interrupt. Interrupt is taken by
MCU and MCU sends the small buffer to encoding. ADPCM encoding gets buffer, processes it and
gives the output as 64 byte to a big buffer which is located in internal flash. Big buffer keeps these
encoded samples. If the big buffer‘s size reaches 2048 byte then these samples are written to external
flash so we write every 2048 byte to external flash because a page in external flash is 2048 byte and it
is the smallest unit which can be programmed at once in external flash.

A page is 2048 byte and it keeps 4096 samples in it. Our sampling rate is 8192 sample
so a page keeps a 0.5 second voice record in it.

Each record has a size field in file table. Size field is 4 byte which is equal to 32 bit.
Size filed of a record keeps how may pages it has in external flash so a record may have size
of 232 pages.

Maximum size of a record = 232 pages x 2048 byte/page = 243 bytes.

Maximum record time of a record = 232 pages x 0.5 second/page  231 second.

Our external flash‘s size is 2 Gbit (256 MB). Maximum record time of a record in our external
flash is equal to equation1.
 48

Equation1 = 256 MB x 1 second / 4KB x 1/60 minute/second x 1/60 hour/minute

 18.2 hours.

3.10 Wav Format

A waveform audio file, also known as a wave file, or simply .wav after its extension, is
a common type of sound file. Microsoft and IBM introduced the wav file in 1991 for use on
the Microsoft Windows 3.1 operation system (OS). It is the main format used on windows
systems for raw and typically uncompressed audio. The usual bitstream encoding is the Pulse
Code Modulation(PCM) format.

The wav format is based on the Resource Interchange File Format (RIFF), which stores
audio files in indexed ―chunks‖ and ―sub-chunks.‖ RIFF is in turn based on the earlier
Interchange File Format (IFF), established by Electronic Arts in 1985 for use in electronic
gaming. Apple‘s version, known as Audio Interchange File Format (AIFF), was released in
1988 for Macintosh computers. Due to the common roots of these various audio formats,
however, the audio files will play on any computer system, IBM or Apple.

The wav file had two very big things going for it when introduced. Firstly, it could
digitize sounds 100% faithful to the original source because it is a lossless format. ―Lossless‖
means that the wav file format does not compromise audio quality even when it holds
compressed data. Secondly, the wav file is very easy to edit and manipulate with software.
Luckily for audiophiles, free wav file editing software has been available nearly as long as
wav files themselves. A WAV file can hold compressed audio, the most common WAV
format contains uncompressed audio in the linear pulse code modulation (LPCM) format. The
standard audio file format for CDs, for example, is LPCM-encoded, containing two channels
of 44,100 samples per second, 16 bits per sample. Since LPCM uses an uncompressed,
lossless storage method, which keeps all the samples of an audio track, professional users or
audio experts may use the WAV format for maximum audio quality. [31]
 49

Wav files can be encoded with a variety of codecs to reduce the file size (for example
the GSM or MP3 codecs). Below picture explains the format, bitrate and one minute record
size of a wav file.

Well known wav audio formats are shown in Table 3.4:

Table 3.4: Wav audio formats

In the project we choose wav format because lossless storage important. In MCU, we
can use ADMPC to compress voice samples and wav file format support ADMPC coding. Our
 50

sample rate is 8192 Hz and every 10 bit ADC sample is converted into 4 bit sample by
ADMPC. The format we chose in the project is in Figure 3.17:

Figure 3.17: General format of wav file

3.11 Communication

A communication protocol is a convention for data transmission that includes such

functions as timing, control, formatting, and data presentation. There are two categories
depending on the clocking of the data on the serial link: [32]
 51

 Synchronous protocols: each successive datum in a stream of data is governed

by a master clock and appears at a specific interval in time.

Figure 3.18: Synchronous data transmission format

 Asynchronous protocols: successive data appear in the data stream at arbitrary

times, with no specific clock control governing the relative delays between
data.

Figure 3.19: Asynchronous data transmission format

 52

There are special IC chips made for serial data communications. These chips are called
UART (universal asynchronous receiver-transmitter) and USART (universal synchronous-
asynchronous receiver-transmitter).

Synchronous communication is usually used in small distance between two

communicating units. Synchronous communication types are SPI, I2C, etc.

Asynchronous communication is used in large distance and there is no clock need to

communicate. Asynchronous communication types are RS-232, etc.

In project, asynchronous communication is chosen with RS-232.

3.11.1 RS-232 and Its Specifications

The EIA/TIA-232-E standard was introduced in 1962 and has since been updated four
times to meet the evolving needs of serial communication applications. The letter "E" in the
standard's name indicates that this is the fifth revision of the standard. [33]

3.11.1.1 RS-232 Specifications

RS-232 is a complete standard. This means that the standard sets out to ensure
compatibility between the host and peripheral systems by specifying:

1. Common voltage and signal levels

2. Common pin-wiring configurations
3. A minimal amount of control information between the host and peripheral
systems.

Unlike many standards which simply specify the electrical characteristics of a given
interface, RS-232 specifies electrical, functional, and mechanical characteristics to meet the
above three criteria. Each of these aspects of the RS-232 standard is discussed below. [33]
 53

3.11.1.1.1 Electrical Characteristics

The electrical characteristics section of the RS-232 standard specifies voltage levels, rate
of change for signal levels, and line impedance.

As the original RS-232 standard was defined in 1962 and before the days of TTL logic,
it is no surprise that the standard does not use 5V and ground logic levels. Instead, a high level
for the driver output is defined as between +5V to +15V, and a low level for the driver output
is defined as between -5V and -15V. The receiver logic levels were defined to provide a 2V
noise margin. As such, a high level for the receiver is defined as between +3V to +15V, and a
low level is between -3V to -15V. Figure 3.20 illustrates the logic levels defined by the RS-
232 standard. It is necessary to note that, for RS-232 communication, a low level (-3V to -
15V) is defined as a logic 1 and is historically referred to as "marking." Similarly, a high level
(+3V to +15V) is defined as a logic 0 and is referred to as "spacing".

Figure 3.20: Logic levels defined by RS-232 standard

 54

Table 3.5 summarizes the electrical specifications in the current standard.

Table 3.5: Electrical specification of the RS-232 standard

3.11.1.1.2 Functional Characteristic

The standard also addresses the functional characteristics of the interface which is shown
in Table 3.6. This essentially means that RS-232 defines the function of the different signals
used in the interface. These signals are divided into four different categories: common, data,
control, and timing. See table below. The standard provides abundant control signals and
supports a primary and secondary communications channel. Fortunately few applications, if
any, require all these defined signals. For example, only eight signals are used for a typical
modem.
 55

Table 3.6: Functional specification of the RS-232 standard

3.11.1.1.3 Mechanical Interface Characteristics

The third area covered by RS-232 is the mechanical interface. Specifically, RS-232
specifies a 25-pin connector as the minimum connector size that can accommodate all the
signals defined in the functional portion of the standard. The pin assignment for this connector
is shown in Figure 3.14. The connector for DCE equipment is male for the connector housing
and female for the connection pins. Likewise, the DTE connector is a female housing with
male connection pins. Although RS-232 specifies a 25-position connector, this connector is
often not used. Most applications do not require all the defined signals, so a 25-pin connector
is larger than necessary. Consequently, other connector types are commonly used. Perhaps the
most popular connector is the 9-position DB9S connector, also illustrated in Figure 3.21. This
 56

9-position connector provides, for example, the means to transmit and receive the necessary
signals for modem applications.

Figure 3.21: Mechanical interfaces of RS-232 standard

3.11.2 Communication Principles of Flash Memory Based Voice Recorder System

In every communication there is standard between communication devices such as baud

rate, data bits, parity bit, stop bit and communication port. Communication port is chosen by
user. Baud rate is automatically set to 115200. Data bits are set to 8 bits. Parity bit is used as
none. Stop bit is used as one stop bit. In order to guaranty the reliability of communication
and reliable data transfer, Fletcher16 is used as checksum algorithm and calculated on the
given size of data.

3.11.2.1 Communication Commands

There are five different communication commands between MCU and PC application.
These are file system table request, file request, setting date of MCU, getting systems internal
time and reset file system.
 57

1. File System Table Request

This command gets the file system table of the system from internal flash to PC
application. First, PC application sends ―T‖ character to MCU. MCU receives ―T‖ character
and learn that it must send the file system table to PC application. It sends 256 bytes data,
calculates checksum of 256 bytes data and sends 2 bytes checksum. PC application receives
256 byte data, calculates checksum of it and compares the calculated checksum and received
checksum if they are equal then PC application sends ―C‖ character as continue, if they are not
equal then MCU sends 256 bytes data and 2 bytes checksum until comparison becomes equal.
MCU does these operations until all data of file system table is sent.

Figure 3.22: File table request

 58

2. File Request

This command gets the requested file from internal flash to PC application. First, PC
application sends ―F‖ character to MCU and then sends 4 bytes name (timestamp) of file and
its 2 byte checksum. MCU receives ―T‖ character, 4 bytes name of file and 2 bytes checksum
of name of file. Then MCU calculates checksum of name of file and compares checksum if
they are equal it sends ―C‖ character as continue, if they are not equal then MCU requests
timestamp one more time until comparison becomes equal. After they become equal, MCU
sends 256 bytes data of requested file, calculates checksum of 256 bytes data and sends 2
bytes checksum. PC application receives 256 byte data, calculates checksum of it and
compares the calculated checksum and received checksum if it is equal then PC application
sends ―C‖ character as continue, if it is not equal then MCU sends 256 bytes data and 2 bytes
checksum until comparison becomes equal. MCU does these operations until all data of file
system table is sent.
 59

Figure 3.23: File request

3. Set the Time of MCU

This command sets the time of MCU, time is given by user as format DD.MM.YYYY
HH:MM according to systems internal epoch 01.01.2009 00:00. First, PC application sends
―D‖ character to MCU and then sends 4 bytes timestamp, which is calculated as minutes based
on epoch of the system, and its 2 byte checksum. MCU receives ―D‖ character, 4 bytes
 60

timestamp and 2 bytes checksum of timestamp. Then MCU calculates checksum of timestamp
and compares checksum if they are equal it sends ―C‖ character as continue, if they are not
equal then MCU requests timestamp and checksum one more time until comparison becomes
equal. If they become equal, MCU sets its internal time.

Figure 3.24: Setting time of MSP430

4. Get System Internal Time

This command gets system internal time. First, PC application sends ―S‖ character to
MCU. MCU receives ―S‖ and then sends 4 bytes timestamp and its 2 byte checksum. MCU
receives ―D‖ character, 4 bytes timestamp and 2 bytes checksum of timestamp. Then MCU
calculates checksum of timestamp and compares checksum if they are equal PC application
sends ―C‖ character as continue, if they are not equal then MCU sends timestamp and
checksum one more time until comparison becomes equal. If they become equal, PC
application calculates real time based on epoch and shows it on the screen.
 61

Figure 3.25: Getting system internal time

5. Reset File System

This command resets file system. First, PC application sends ―R‖ character, ―ESET‖ as 4
bytes and 2 bytes checksum of ―ESET‖ to MCU. MCU receives ―R‖, 4 bytes ―ESET and
―ESET‖‘s 2 bytes checksum. Then MCU calculates checksum of ―ESET‖ and compares
checksum if they are equal PC application sends ―C‖ character as continue, if they are not
equal then MCU requests ―ESET and its checksum one more time until comparison becomes
equal. If they become equal, MCU resets file system.
 62

Figure 3.26: Resetting file system

3.11.3 Guaranty Reliable Communication: Fletcher’s Checksum Algorithm

In every communication there can be loss or change in data because of noise that are
generated by other factors so there must be a way to check there is a loss or not. The way to
do so is using a checksum algorithm which is implemented at both side of the communication.
If receiver part of communication gets data as wrong then sender part of communication
resends data. By the use of this way, it ensures data integrity and guaranties reliability of
connection.

Fletcher's checksum is one of several types of checksum algorithms, which are relatively
simple processes used by computers to check the integrity of data.

The implementation of the Fletcher-32 is very similar to the Adler-32 algorithm but
several differences should be noted. Fletcher wraps around at modulo 65535 while Adler
wraps at the prime 65521. In other words, Fletcher adds overflow bits (16-31) into its sum;
while Adler multiplies those bits by 15, and then adds the product into its sum. Fletcher-32
works on 16 bit data while Adler works on 8 bit data. [34][35]
 63

It is designed to overcome some of the inadequacies of simply summing all the bytes as
in the original checksum. Fletcher's checksum, unlike the original checksum, can detect the
inserting/deleting of zero value bytes, the reordering of bytes, and the incrementing and
decrementing of bytes in opposite directions. [34]

Fletcher-32 is slightly more reliable than Adler-32. Fletcher-16 is chosen because 8 bits
data is used in communication. The C implementation of Fletcher-16 is given in Figure 3.27:

Figure 3.27: Fletcher checksum algorithm is implemented in C programming language

3.12 Finite State Machine

We have listed hardware and other software properties above. Many different software
can be implemented using this system. Our project is about Flash Memory Based Voice
Recording, Flash Memory is known as Solid State.

All embedded application is also a finite state machine. System is implemented

according to below Figure 3.28:
 64

Figure 3.28: Finite state machine of the project

States and their properties are listed below:

1. Initial State

This state is the start state of the project and it is also the final return state of UART
Receive Handle and Record State. In this state; board, MCU, FSM and sampling initialized.
ADC is stopped. Button and UART receive are enabled. In this state, if record button is
pressed then system goes to record start state. If a communication is started with PC
application by using UART then system goes to UART receive handle state.

2. Record Start State

This state is the state which prepares system to record state. UART receive is
disabled. Button is enabled. ADC is initialized. Sampling is started. When an ADC interrupt is
occurred then this state goes to record state.
 65

3. Record State

This state is the state which all recording is handled. Voice samples, which are
prepared by ADC, is encoded and stored. If there is an ADC interrupt occurs then state returns
back to itself but if there is a button interrupt or EOF buffer is reached then system goes to
record stop state.

4. Record Stop State

This state is stop recording ADC is stopped and updates are done. When all
processes finish then it goes back to initial state.

5. UART Receive Handle State

This state does handle all UART communication protocol. In this state, UART
receive and button interrupt is disabled. UART commands are handled according to our
UART communication protocol properties.

In all states, if there is an unknown incident occurs then system goes to a null state and
blocks itself.

3.13 Communicating with Real World: PC Application

Most of embedded applications need a PC application in order to communicate with

real world. They sense world, communicate with other embedded systems, process and store
data. The raw data, is coming from MCU, is not perceptible to users. If a system does all stuff
and does not give any perceptible output to users then it does not mean anything. We develop
our own PC application to set MCU and get records.
 66

PC application is developed by using Java Programming Language, JDK 1.6.0_07 and

GiovynetSerialPort1.3 Library.

PC application is composed of a main window and a communication window.

Communication window appears when a new communication is wanted to establish by
choosing ―Communication‖ tab on menu and clicking ―New communication‖. New
communication window helps user to set communication‘s properties, system date and time.
MCU is configured to 115200 baud rate, 8 bit character, no parity bit and 1 stop bit so our
configuration is based on those values.

Figure 3.29: Connection menu of PC application

All other processes are done on main window. On menu, there are four tabs and three
of them are using in main window, which are File, Reset and Date tabs. There are also two
different buttons which are ―Download and Play‖ and ―Clear Tool Output and Info‖ buttons.

On file tab, there is only one item which is ―Get File List‖ and this item is used to get
records file table from MCU. If it is clicked then a communication is established and MCU
sends file table of records.
 67

Figure 3.30: File menu of PC application

If user wants to set and get time of MCU then he must choose ―Date‖ tab. If date tab is
clicked there are three options, one of them is ―Get MSP Date‖, second one is ―Set MSP Date‖
and third one is ―Set Date‖. Get MSP Date is used to learn the MCU internal time and if it is
clicked then a window appears and shows the date of MCU. Set Date gets the date from user,
to do so it shows a window to enter date and time setting. Set MSP Date sets MCU time
according to our epoch 01.01.2009 00:00.
 68

Figure 3.31: Date menu of PC application

On reset tab, there is only one item which is ―Reset MSP430‖ and this item is used to
reset file system of MCU. If it is clicked then a communication is established and MCU resets
its file system table.

Figure 3.32: Reset menu of PC application

 69

4. ANALYSIS

Analysis is an important part of developing an application. Especially embedded

applications need much more analysis then PC application because in embedded systems
memory, CPU speed and power is limited. In order to implement a good application, analysis
must be done by considering power consumption, CPU usage and code size of application.

IAR Embedded Workbench is a good compiler and debugger interface and it has a
good simulation module. By the use of this tool, CPU and memory analysis is done.

Power consumption of CPU is written on specific datasheet.

4.1 Memory Usage

According to linker output of IAR Compiler, the memory requirements are shown into
following Table 4.1.

Table 4.1: Memory usage of whole code

Code Memory Data Memory Constant Memory

No Optimized 3968 bytes 991 bytes 325 bytes
Size Optimized 3790 bytes 991 bytes 323 bytes
Speed Optimized 3804 bytes 991 bytes 323 bytes
Balanced Optimized 3794 bytes 991 bytes 323 bytes
 70

The code size of some required modules is shown Table 4.2 which is based on Size
Optimized Compilation:

Table 4.2: Memory requirements of some modules

Code Memory Data Memory Constant Memory

ADPCM Module 202 bytes 4 bytes 186 bytes
Flash Module 50 bytes 15 bytes -
Protocol Module 296 bytes - -
UART Module 200 bytes 4 bytes -
Main Module 1286 bytes 736 bytes 73 bytes

4.2 Timing

4.2.1 CPU

The listed CPU usage analysis is done by IAR Embedded Workbench Simulator. The
code is compiled at Size Optimized Mode.

The average values are calculated over 100 different executions in Table 4.3:

Table 4.3: CPU usage of key functions

Cycles Per Call Theoretical time spent

(for 12 MHz)
(Average)

Fletcher Checksum 2434 cycles 200 us

Encode 17024 cycles 1.5 ms
Erase Internal Flash 60 000 cycles 5 ms
Write Internal Flash 60 000 cycles 5 ms
 71

4.2.2 Flash

External flash timing diagram is taken from its datasheet and results are shown below
Table 4.4: [17]

Table 4.4: Timing of Numonyx NAND flash

 72

4.3 Power Consumption

4.3.1 CPU

According to datasheet the power consumption of MSP430 for different operating

modes is as follows: [12]

Table 4.5: Current consumption of MSP430F2274 for different modes

 73

4.3.2 Flash

Flash power consumption is taken from specific datasheet and it is shown below Table
4.6: [17]

Table 4.6: Current consumption of Numonyx NAND flash

 74

5. CONCLUSION

In the scope of this project, we developed a solid state voice recorder system. Also we
developed other useful hardware abstraction layer library in order to develop different project
by using our library.

We used MSP430F2274 and Numonyx NAND02G-B2D. They are good choices to use
because they both have low power mode and current consumption of devices are very small.
This is an important point to develop low-power embedded systems.

RS-232 is an industry standard and very well known serial communication protocol to
communicate embedded system with PC.

PC application is implemented by using Java Programming Language. Java is another

good choice of the project because it is an open source language and it is widely supported by
all types of computer.

IMA‘s ADPCM implementation is an optimum solution for this kind of limited

applications, providing good sound quality. MSP430 has a limited computational power but
this codec can be enhanced for future developments.

5.1 Engineering Criteria

MCU is one of the most important parts of the project. It must be chosen well. We
have two different MCUs of Texas Instrument Inc. One of the MCU is MSP430F2274 and
another one is MSP430F169. MSP430F169 has much more current consumption, almost
twice, than MSP430F2274. Wake up time from standby mode of MSP430F169 is six times
greater than MSP430F2274. MSP430F2274 has cycle time which is half of MSP430F169‘s.
Another important point is price of MCUs, MSP430F2274‘s price is around 5 dollar but
 75

MSP430F169‘s price is around 13 dollar. Because of these reasons we choose MSP430F2274

as our MCU.

Flash technology is the hearth of the project. There are NAND or NOR and serial or
parallel flashes. NOR flashes have higher read speeds but their write speed is smaller than
NAND flashes. Serial flashes have smaller pin needs but their write speed is also smaller and
for sequential access address of data must be set for every two bytes. Parallel flashes have
higher write speed and sequential access needs once address initialization. Parallel flashes also
divide two sub categories x8 and x16. X8 category has 8 data pins and x16 category need 16
data pins. Our MCU does not have too much pins to give the x16 parallel flashes. As a
consequence of this reasons, we choose x8 parallel NAND flash.

Voice samples needs too many spaces but we have a limited capacity of flash. In order
to keep maximum utilization space, we need to compress and encode samples. Many different
alternatives can be used to compress voice samples but MCU has a limited calculation
capacity. According to computation need and compression ratio, ADPCM is the optimum
solution.

Many embedded applications also need PC applications and our PC application is

developed by using Java programming language. There are many choices about programming
language such as C, C++, C# and Java. C and C++ are good choices about speed but they are
not good at visual programming and also platform dependent. C# is another good choice and
easy to program but it is also platform dependent. Java is the best solution because it is
platform independent, easy to program and widely supported open source system so we
choose Java programming language.

Our RAM and computation speed are limited. To have better system, we need to
optimize our program. We create our own functional libraries, use well organized efficient
algorithms and try to keep MCU as busy as possible between sample‘s times. Code is also
complied according to different optimization such as speed, size or balanced.
 76

File system is important for solid state voice recorder system. FAT and NTFS file
systems are analyzed but they need to much MCU cycles and it is not easy to implement them.
Our system is not reached directly from other systems. It gets commands, analyzes it and gives
the need output. Because of these reasons, we create our own file system to get a better CPU
utilization and keep records as safe from outsourced attacks.

PC application and MCU need a language to talk to each other in a proper way. We
design a communication protocol. This communication protocol must use minimum
commands and each command must need smaller time to be analyzed. There are too many
noises that change the sent data. To have a reliable communication, we need a checksum
algorithm. Cyclic Redundancy Check is on of the choices but it needs to many computation
and it is not easy to implement. Simple checksum algorithms like sum all of the data and
sending total sum is easy to implement but it needs also too many bytes as checksum. Fletcher
checksum algorithm is a usual checksum algorithm, it does not need to many computation and
its sum is just two bytes so we implement Fletcher checksum algorithm to keep our
communication reliable.

5.2 Further Development

This project can also be enhanced in some ways:

 CPU can be worked at higher frequencies and so computational depending

functions can changed to get better results but power consumption of system is
increased.
 Multiplane programming can be implemented on flash library and writing
speed is increased on average.
 Multiplane block erasing can be used to decrease erasing time of flash.
 UART communication protocol can use higher baud rates but better checksum
and also error correction codes must be implemented to decrease resending of
corrupted data.
 77

 External codec implemented chips can be added on board to decrease the work-
load of CPU.
 PC application may be enhanced to store voice records as different file format
and merge records into one file or split records into different file.
 78

REFERENCES

1. Jody Rosen, "Researchers Play Tune Recorded Before Edison", New York Times, 2008.

2. Patrick Feaster, "Speech Acoustics and the Keyboard Telephone: Rethinking Edison's
Discovery of the Phonograph Principle", ARSC Journal, Spring 2007.

3. Oliver Berliner and Patrick Feaster, "Letters to the Editor: Rethinking Edison's Discovery
of the Phonograph Principle", ARSC Journal, Fall 2007.

4. Gelatt Roland, "The Fabulous Phonograph: From Tin Foil to High Fidelity", Philadelphia
J. B. Lippincott Company, 1955.

5. Allen, "Edison Cylinder Records, 1889-1912", New York Stellar Production, 1969.

6. Leonard DeGraff., "Confronting the Mass Market: Thomas Edison and the Entertainment
Phonograph", 1995.

7. S. J. , "Magnetic Recording", New York Rinehart & Company, 1949.

8. Jan Axelson, "The Microcontroller Idea Book Circuits", 1994.

9. Han-Way Huang, "PIC microcontroller: an Introduction to Software and Hardware

Interfacing", 2005.

10. Kenneth J. Ayala , "The 8051 Microcontroller", 2004.

11. "MSP430x2xx Family Users Guide", Texas Instrument Inc., 2007.

 79

12. "MSP430F2274 Device Datasheet", Texas Instrument Inc., 2007.

13. "MSP430F2274 Device Erratasheet", Texas Instrument Inc., 2008.

14. "NAND vs. NOR Flash Memory Technology Overview", Toshiba Inc., 2006.

15. "Flash Memory Technology", Integrated Circuit Engineering Corporation.

16. "Introduction to Serial Flash Memory", Next Flash Technologies Inc., 2003.

17. "NAND02G-B2D Datasheet", Numonyx Inc., 2008.

18. Dominic Giampaolo, "Practical File System Design with the Be File System", 1999.

19. "Engineer-to-Engineer Note EE-329", 2007.

20. "Extensible Firmware Initiative FAT32 File System Specification", Microsoft Corp., 2000.

21. John Leis, "Disk File System Examples", 2006.

22. R.G. Gallager, "The Sampling Theorem".

23. C. E. Shannon, "Communication in the Presence of Noise", Proc. Institute of Radio

Engineers, Jan. 1949. Reprint as classic paper in: Proc. IEEE, Vol. 86, No. 2, (Feb 1998).

24. Dan Lavry, "Sampling Theory for Digital Audio", Lavry Engineering Inc.

25. "PCM Tutorial", L3 Telecommunications.

26. Yao Wang, "Source Coding Basics and Speech Coding", Polytechnic University,
Brooklyn.
 80

27. Nimrod Peleg, "Introduction to Speech Coding", 2009.

28. Bernd Girod, "EE368b Image and Video Compression DPCM Overview", Stanford
University.

29. "The Implementation of G.726 Adaptive Differential Pulse Code Modulation (ADPCM) on
TMS320C54x DSP", Texas Instrument Inc., 1997.

30. Mark Valentino, "Microphone Handbook", PCB Piezotronics Vibration Division.

31. Rob Ryan, "RIFF WAVE (.WAV) File Format", Brown University.

32. Ph.D. Suree Pumrin, "Introduction to Microprocessors Chapter 12 Serial

Communications".

33. "Application Note 83 Fundamentals of RS-232 Serial Communications", Maxim Integrated

Products Inc., 2001.

34. "RFC 1146 Fletcher Checksum Algorithm".

35. "RFC 905 - ISO Transport Protocol Specification".

36. Chris Nagy, "Embedded Systems Design Using the TI MSP430 Series", 2003.