Digital Communication Training Kit

Digital Communication Training Kit.1 Mohammad Sharawi and Husam Abu-Ajwah Electronics Engineering Princess Sumaya University College for Technology (Royal Scientific Society) Amman–Jordan August 1999

1

c by Mohammad Sharawi and Husam Abu-Ajwah All rights Reserved. 2000 No part of this report is to be copied, reproduced, or distributed in anyway, without written consent of the Author. Contact: [email protected] or [email protected]

1

To our Moms, and Dads.

Abstract The aim of this project is to demonstrate three of the Digital

Modulation schemes used these days in Communications, which are the FSK(Frequency Shift Keying), the PSK(Phase Shift Keying), and the QPSK(Quadrature Phase Shift Keying), and to design the electronic circuits for these systems on a single PCB(Printed Circuit Board). This board is to be used for educational purposes in the Communications Lab at Princess Sumaya University College for Technology. This project includes: 1. Analysis and description for the main features of each of the modulation schemes. 2. Design and implementation of the electronic circuits for each of the three systems. 3. Design and implementation of a variable repetitive 8-bit word generator. 4. Design of the PCB Layout that would have the three systems mounted on.

Contents 1 Introduction

5

2 Background 2.1 FSK Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 PSK Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 QPSK Modulation . . . . . . . . . . . . . . . . . . . . . . . . . .

7 7 11 13

3 Circuit Design 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . 3.2 Universal Clock and Word Generator . . . . . . . . 3.2.1 The Universal Clock . . . . . . . . . . . . . 3.2.2 The Word Generator . . . . . . . . . . . . . 3.3 FSK System . . . . . . . . . . . . . . . . . . . . . 3.3.1 FSK Transmitter . . . . . . . . . . . . . . . 3.3.2 FSK Receiver . . . . . . . . . . . . . . . . 3.4 The PSK Systems . . . . . . . . . . . . . . . . . . 3.4.1 The PSK Transmitter . . . . . . . . . . . . 3.4.2 The PSK Receiver . . . . . . . . . . . . . . 3.4.3 Balanced Modulator (MC1496 IC) . . . . . 3.5 The QPSK System . . . . . . . . . . . . . . . . . 3.5.1 The Phase Shifter . . . . . . . . . . . . . . 3.5.2 The Demultiplexer and Multiplexer circuits 3.5.3 The QPSK Transmitter . . . . . . . . . . 3.5.4 The QPSK Receiver . . . . . . . . . . . . 3.5.5 PCB Design and Final Circuit Schematic .

18 18 18 18 18 19 19 22 23 24 25 30 31 34 34 37 38 40

4 Results

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

46

5 Conclusion and Further enhancements 47 5.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 5.2 Further Enhancements . . . . . . . . . . . . . . . . . . . . . . . 47

2

List of Figures 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12

FSK signal Constellation. [Hay94] . . . . . . . . . . . . . . An FSK modulator block diagram. . . . . . . . . . . . . . . FSK demodulator block . . . . . . . . . . . . . . . . . . . . PLL block diagram. . . . . . . . . . . . . . . . . . . . . . . PSK signal Constellation. [Hay94] . . . . . . . . . . . . . . PSK Block diagram. . . . . . . . . . . . . . . . . . . . . . . PSK demodulator Block diagram. . . . . . . . . . . . . . . . The two possible QPSK signal Constellations. [Rap96] . . . QPSK signal constellation with its decision regions. [Hay94] QPSK Modulator Block diagram . . . . . . . . . . . . . . . QPSK Demodulator Block diagram . . . . . . . . . . . . . . Noise Performance in AWGN.[Hay94] . . . . . . . . . . . . .

. . . . . . . . . . . .

8 9 10 10 11 12 13 14 14 15 16 17

3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 3.15 3.16 3.17 3.18 3.19 3.20 3.21

Universal Clock Circuit. . . . . . . . . . . . . . . . . . . . . . . . Word Generator Circuit. . . . . . . . . . . . . . . . . . . . . . . . Word Signal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Voltage Divider at VCO I/P. . . . . . . . . . . . . . . . . . . . . DC Offset Circuit. . . . . . . . . . . . . . . . . . . . . . . . . . . VCO Circuit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FSK Output Signal. . . . . . . . . . . . . . . . . . . . . . . . . . PLL Circuit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FSK total Circuit. . . . . . . . . . . . . . . . . . . . . . . . . . . PSK Transmitter Circuit. . . . . . . . . . . . . . . . . . . . . . . PSK output signal. . . . . . . . . . . . . . . . . . . . . . . . . . . PSK Receiver Circuit. . . . . . . . . . . . . . . . . . . . . . . . . Sellen-Key Filter. . . . . . . . . . . . . . . . . . . . . . . . . . . . In/Out signals of the PSK receiver. . . . . . . . . . . . . . . . . . Leveling Circuit. . . . . . . . . . . . . . . . . . . . . . . . . . . . Gilbert Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MC1496 IC schematic . . . . . . . . . . . . . . . . . . . . . . . . Phase shift circuit . . . . . . . . . . . . . . . . . . . . . . . . . . Simulated Demux Circuit . . . . . . . . . . . . . . . . . . . . . . Implemented Demux circuit . . . . . . . . . . . . . . . . . . . . The CLK,CLK, Word, Flipflop1(even) and flipflop2(odd) outputs for the simulated demultiplexer. . . . . . . . . . . . . . . . . The Mux circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . Word in to demultiplexer, odd, even, word out from multiplexer streams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The QPSK Transmitter Circuit . . . . . . . . . . . . . . . . . . . The QPSK Transmitter Signal . . . . . . . . . . . . . . . . . . . The QPSK receiver circuit . . . . . . . . . . . . . . . . . . . . . .

19 20 20 21 21 22 23 24 25 26 26 27 28 29 29 32 33 34 35 35

3.22 3.23 3.24 3.25 3.26

3

. . . . . . . . . . . .

. . . . . . . . . . . .

36 36 36 37 38 39

4

LIST OF FIGURES 3.27 3.28 3.29 3.30 3.31 3.32

QPSK receiver signals . . . . . . . . . . The final output and the initial input . The Circuit Schematic of the Module. . The Circuit IC and Component Layout. The PCB Layout, Layer (1). . . . . . . . The PCB Layout, Layer (2). . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

39 40 42 43 44 45

Chapter 1

Introduction T

hrough ages, man always tried to find a way for long distance communication. Man used pigeons, and men riding horses to deliver mail. These two methods were inefficient; the first was limited to certain directions that the pigeons were trained to go to, and the latter took longer time for further distances. These methods were used till the 19th century, when Alexander Graham Bell succeeded in performing the first wired telephone call (transmitting and electrical signal) in 1875. This invention revolutionized world communication. Since the invention of the telephone, communications industry kept on growing and widening. Another important invention was accomplished by Guglielmo Marconi, who was the first to transmit electrical signals through the air, and then receive the transmitted signal using a receiver. The first transmission trough the Atlantic was in 1901 which opened the door for Radio broadcasting. In World War I&II, wireless radio was very important in command and control of troops. Analog transmitting schemes were used. The aim was to transmit information signals (voice) via the air. This was accomplished using Analog Modulation, which can be defined as: The process by which some characteristic of a carrier is varied in accordance with a modulating wave (information signal) [Hay94] The carrier frequency is usually a sinusoidal signal that has a frequency much higher than the highest frequency component of the information signal. Right after World War II, the idea of having Global Wireless information transmission arose. The developed countries started their rally of launching satellites to ease global communications, by covering larger areas that receive satellite transmission. Also the Digital Modulation technology started to develop. Nowadays, modern mobile communication systems use digital modulation techniques. Advancements in VLSI (Very Large Scale Integration) and DSP (Digital Signal Processing) made digital modulation more coast efficient than analog modulation. There are other advantages of digital modulation over analog modulation including greater noise immunity, robustness to channel impairment, easier multiplexing, and greater security. Furthermore digital 5

CHAPTER 1. INTRODUCTION

6

modulation accommodates error-correction codes, supports complex signal conditioning and processing techniques such as coding, encryption, and equalization to improve the overall performance of the communication channel [Rap96]. There are several factors that influence the choice of a digital modulation scheme. The desired scheme is the one that provides lowest bit error rates, performs well in multipath and fading conditions, occupies a minimum bandwidth, and is coast effective. In reality there are trade-offs when selecting a digital modulation scheme depending on the application. The two measures that determine the performance of the modulation scheme are Power efficiency and Bandwidth efficiency. Power efficiency describes the ability of a modulation technique to preserve the fidelity of the digital massage at low power levels. The power efficiency ηP is often expressed as the ratio of the signal energy per Eb required at the receiver input to have a bit to noise power spectral density N 0 certain probability of error (say 107 ) [Rap96]. Bandwidth efficiency describes the ability of a modulation scheme to accommodate data within a certain limited bandwidth. In general, increasing the data rate implies decreasing the width of each data bit, which increases the bandwidth of the signal. If R is the Data Rate in bits/sec, and B is the bandwidth occupied by the modulated RF (Radio Frequency) signal, the bandwidth efficiency is: R ηB = bits/Hz (1.1) B The upper bound on the bandwidth efficiency is given by the Shannon-Hartley Capacity Theorem: ’ “ S C = B log2 1 + (1.2) N Where C is the channel capacity (in bps), B is the channel bandwidth, and S/N is the signal to noise ratio. There is always a trade-off between Power efficiency and Bandwidth efficiency. We will investigate the bandwidth efficiency, power efficiencies, and the basic principles for each of the three digital modulation schemes encountered in our design, FSK, PSK, and QPSK later in this report.

Chapter 2

Background This chapter will mainly describe the three Digital Modulation/Demodulation schemes used in our design, FSK, PSK, QPSK. It will specify the basic parameters of each scheme, its equations, its block diagram, and its performance in AWGN(Additive White Gaussian Noise) channel.

2.1

FSK Modulation

The

Frequency Shift Keying (FSK) scheme is used in new commercial data communication. Because of its close behavior to the FM modulation technique, FSK is replacing FM in the new digital based radio receivers because of the benefits digital modulation has over analog ones, as we mentioned previously. As in FM, the amplitude of the FSK modulated signal is constant, but it’s the changes in the frequency that carries the signal information. In general, an FSK signal can be represented as: r Eb si = cos(2πfi t + θ)1 (2.1) (0 ≤ t ≤ Tb ), (i = 1, 2, 3, · · · , M ) Tb where the frequency term fi will have M discrete values, and the term θ is an arbitrary constant, Tb is the bit time duration, and Eb is the energy per bit [Hay88]. In our design we are using Binary FSK, that uses two levels of frequencies, one is for transmitting a binary 1, the other is for transmitting binary 0. The two signal equations in use are: r Eb s0 = cos(2πf0 t + θ) (2.2) Tb and,

r

Eb (2.3) cos(2πf1 t + θ) Tb where f0 is the frequency for transmitting a zero, and f1 is for transmitting a one. Such variations in the frequency due to the Digital input stream m(t) can be obtained using a VCO. The Signal Constellation for the Binary FSK is shown in Fig.(2.1). Going to bandwidth issues, the transmission bandwidth of an FSK signal is given by Carson’s rule: (2.4) BT = 2∆f + 2B s1 =

1E

b

=

1 2 A T 2 c b

, where Ac is the amplitude of the Carrier.

7

8

CHAPTER 2. BACKGROUND

Figure 2.1: FSK signal Constellation. [Hay94] were B is the bandwidth of the digital baseband signal (information signal), and ∆f is the frequency deviation. Assuming the use of rectangular pulses, this means that B=R (R is the data rate), hence Eq.(2.5) becomes: BT = 2(∆f + R)

(2.5)

A block diagram for an FSK modulator is shown in Fig.(2.2). The average Probability of symbol error, or equivalently, the bit error rate for coherent binary FSK is [Hay94]:  r ! 1 Eb 2 Pe = erf c (2.6) 2 2N0 Eb is the signal energy per to noise power spectral density. were N 0 An M-ary FSK signal consist of a set of Orthogonal3 set of M-frequency shifted signals. When these signals are detected coherently, the adjacent signals need 1 to be separated by a frequency of 2T so as to maintain orthogonality. Hence we can define the bandwidth of an M-FSK by:

B=

M 2T

(2.7)

For M-ary signals, the symbol duration is given by: T = Tb log2 M 2 Erfc

is the complementary error function, were erfc(u) = use tables to find its value. [Hay94] 3 Two

signals s1 , s2 are orthogonal if:

RT 0

(2.8) √2 π

s1 ∗ s2 = 0. [Skl88]

R∞ u

2 e(−z)

dz. Usually we

9

CHAPTER 2. BACKGROUND

Figure 2.2: An FSK modulator block diagram. where Tb is the bit duration. Now, substituting Eq.(2.8) in Eq.(2.7), we redefine the channel bandwidth for and M-ary FSK signal as: B=

Rb M 2 log2 M

(2.9)

yielding a Bandwidth efficiency of: 2 log2 M (2.10) M Table 2.1 illustrates the bandwidth efficiency and the power efficiency of the M-levels for FSK Modulation: [Rap96] ηB =

Eb N0

M ηB bits/Hz for a BER of (10−6 )

2 1 13.5

4 1 10.8

8 0.75 9.3

16 0.5 8.2

32 0.3125 7.5

64 0.1875 6.9

Table 2.1: The bandwidth efficiency and the power efficiency of the M-levels FSK. The tables tell us that the Bandwidth efficiency decreases with the increase in the number of levels. A typical coherent FSK demodulator is shown in Fig.(2.3). Here the input FSK signal is applied to two multipliers, one has the frequency of a the transmitted one, the other has the frequency of a zero. These frequencies should be as the ones used in the modulator circuit. The multiplier is followed by an integrator, then the output is summed, and applied to the decision device to determine if a zero or a one has been transmitted. In our design we built a non-coherent FSK system. We used a PLL to lock on the incoming frequency, and from the phase detector’s output we took our final output. This gave us a digital stream as the one being transmitted by the VCO, but this output needed some filtration to get a clear digital output. The block diagram in Fig.(2.4)demonstrates the way we built the FSK system.

CHAPTER 2. BACKGROUND

Figure 2.3: FSK demodulator block

Figure 2.4: PLL block diagram.

10

CHAPTER 2. BACKGROUND

11

Figure 2.5: PSK signal Constellation. [Hay94]

2.2

PSK Modulation

The Phase Shift Keying (PSK) Digital Modulation scheme is also widely used in modern communications , especially in satellite communications, where Mary PSK modulation techniques are used for their bandwidth efficiency. PSK is considered one of the linear modulation techniques that has constant envelope [Rap96]. In the BPSK modulation(Binary PSK, which is the one used here), the phase of the constant amplitude signal is switched between two values according to two possible signals m1 and m2 corresponding to binary 1 and 0 respectively. Usually the phases are 180◦ apart. The equations of the two modulated signals are: r 2Eb s1 = (2.11) cos(2πfc t + θ) Tb r 2Eb s1 = − cos(2πfc t + θ) (2.12) Tb The signal constellation for a BPSK is shown in Fig.(2.5): A block diagram of a BPSK system is shown in Fig.(2.6): Here the NRZ Encoder transforms the level of the binary input data into a +ve level for the binary 1, and a -ve level for the binary 0 in order to change the carrier phases. The average Probability of symbol error, or equivalently, the bit error rate for coherent binary PSK is: [Hay94]  r ! Eb 1 erf c (2.13) Pe = 2 N0

12

CHAPTER 2. BACKGROUND

Figure 2.6: PSK Block diagram. Eb is double the one in its FSK counterpart, which means that we Here the N 0 Eb have to double the N for the FSK signal in order to have the same average error 0 rate as the BPSK one. This is obvious when looking at the signal constellation of the BPSK and the BFSK. The distance between the two binary points in the FSK constellation is almost half the distance in the BPSK constellation, which means that the probability of false decision in the FSK is double its PSK Eb for the FSK to compensate counterpart. That’s why we need to double the N 0 for and have the same average error probability of its PSK counterpart. BPSK is considered bandwidth efficient and its bandwidth efficiency increases with the increase of the number of bits per symbol, this will affect its power efficiency. This is best illustrated in Table(2.1) [Rap96]. The channel bandwidth required to pass M-ary PSK signals is given by:

2 (2.14) T Where T is the symbol duration, which is related to bit duration by Eq.(2.8). Also using Rb = T1b , Eq.(2.14) becomes: B=

B=

2Rb log2 M

(2.15)

Equivalently we can use Eq.(1.1) to obtain the PSK bandwidth efficiency formula: log2 M ηB = (2.16) 2

Eb N0

M ηB bits/Hz for a BER of (10−6 )

2 0.5 10.5

4 1 10.5

8 1.5 14

16 2 18.5

32 2.5 23.4

64 3 28.5

Table 2.2: The bandwidth efficiency and the power efficiency of the M-levels for PSK Modulation. Comparing this table with Table(2.1), we notice that the BPSK is more bandwidth efficient than its FSK counterpart, but there is a trade off in its Power efficiency.

CHAPTER 2. BACKGROUND

13

Figure 2.7: PSK demodulator Block diagram. As for the Demodulation process, For coherent detection, the basic demodulation scheme is shown in Fig.(2.7): Here we have the incoming signal multiplied with a synchronized oscillator that has the same frequency of the modulation frequency. These two are multiplied together, and the output is applied to a LPF (Low Pass Filter) to remove noise and to perform integration on the incoming signal. The decision device follows giving the Binary output. In our design, we followed this basic modulator demodulator concept. The circuit design is shown in the coming chapter.

2.3

QPSK Modulation

The Quadrature Phase Shift Keying (QPSK) is a 4-ary PSK signal. The QPSK is a modulation scheme widely used in Satellite Communications. As we mentioned earlier, PSK signals are considered one of the linear modulation techniques, which are well known of their bandwidth efficiency. The phase of the carrier in the QPSK takes 1 of 4 equally spaced shifts, such as 0, π2 , π, 3π 2 , where each value of phase corresponds to a unique pair of message bits. The QPSK transmitted signal is defined by: r Eb π cos(2πfc t + (i − 1) ) (0 ≤ t ≤ Tb ), (i = 1, 2, 3, 4) (2.17) si = Tb 2 The signal Constellation of this equation is shown in fig (2.8a). Another way 5π 7π of having a QPSK modulation is by using the phase shifts π4 , 3π 4 , 4 , 4 . This method yields the signal constellation as the shown in Fig.(2.8b). Also a table that shows the signal space characterization of the QPSK signaling is shown in Table (2.3). As for the QPSK modulator (transmitter), the basic block diagram of such a system is shown in Fig.(2.10): In Fig.(2.10), the QPSK modulator consists of two streams of PSK modulators, one is fed with the Odd data sequence, and the other with the even

CHAPTER 2. BACKGROUND

Figure 2.8: The two possible QPSK signal Constellations. [Rap96]

Figure 2.9: QPSK signal constellation with its decision regions. [Hay94]

14

15

CHAPTER 2. BACKGROUND

Input Dibit(0 ≤ t ≤ T )

Phase of QPSK Signal(radians)

10

π 4

00

3π 4

01

5π 4

11

7π 4

Coordinates of message point Sq Sq i1 i2 E + 2 − E2 q q − E2 − E2 q q − E2 + E2 q q + E2 + E2

Table 2.3: The Binary Dibit and its corresponding phase.

Figure 2.10: QPSK Modulator Block diagram sequence. Also one stream is modulated with a cos wave and the other with a sine wave to have a π2 phase difference. As for the bandwidth and power efficiency measures, please refer to Table(2.2). The bandwidth efficiency of QPSK is twice that of PSK since we are transmitting two bits per signal. The Average probability of bit error of the QPSK signaling is: [Hay94]  r ! 1 Eb Pe = erf c (2.18) 2 N0 Surprisingly, the average probability of bit error of the QPSK is the same as that of the PSK in AWGN channel4 , while as much data can be sent in the same bandwidth. Thus compared to PSK, QPSK provides twice the spectral efficiency with exactly the same energy efficiency. 4 Refer to Communication Systems by Haykin for proof of P of both the PSK, and QPSK. e [Hay94]

CHAPTER 2. BACKGROUND

16

Figure 2.11: QPSK Demodulator Block diagram A QPSK receiver block diagram is shown in Fig.(2.11): As shown in the block diagram, the input QPSK stream is applied to two sinusoidal modulators that are 90◦ apart. The outputs from the multipliers are then applied to two integrators that perform both filtering and integration. After that the input is applied to a decision device that determines the level received. The outputs from the decision devices are applied to a multiplexer that performs parallel to serial data conversion, thus reconstructing the final output data stream. In the circuit design section, this block diagram was satisfied by substituting each of the blocks with a suitable circuit designed to perform its function. The circuit gave impressive results. Fig.(2.12) Illustrates an overall graphical comparison for the Noise performance of the three schemes, the FSK, PSK, and QPSK in AWGN:

CHAPTER 2. BACKGROUND

Figure 2.12: Noise Performance in AWGN.[Hay94]

17

Chapter 3

Circuit Design 3.1

Introduction

Our project is composed of five major circuits which are : FSK ,PSK ,Q-PSK modulators/demodulators, the Clock, and the Word generator. All mentioned circuits were simulated by the Electronic Work Bench (EWB) simulation software. This chapter will illustrate each circuit with further information needed to show how these circuits were designed and why. Also we needed additional circuits such as, Attenuators, Filters, Comparators, Amplifiers, Phase shifter, Summer, Multiplexer and Demultiplexer were designed according to the application requirements. The last part of this chapter shows the whole schematic of this project that helps the student to find the places of the components for designed circuits in this chapter(every component is labeled so anyone can trace the circuit easily).Also the PCB design is shown in this part of the chapter. Limitations, requirements and educational purposes of this design will be discussed to justify the reason for PCB board to appear the way it is.

3.2 3.2.1

Universal Clock and Word Generator The Universal Clock

We will begin with the Universal clock circuit which consists of IC 4520 dual counter as shown in Fig.(3.1): A leveling circuit was built at the input, and by applying the super position principle the output of the divider is 8V p − p (-4 V – 4 V), and the DC source causes an offset of 4 V so the signal to the 4520 counter is from 0–8 V. This 4520 counter will detect the 512 KHz input sine wave ,(gives logic high above its threshold which is 6 V and logic low bellow this level),and will give a CMOS square wave outputs with the frequencies of (2, 4, 8, 16, 64, 128, 256, 512 KHz) which are used in our system to control the data rate of the word coming out from the Word generator, and the clocks needed for the Multiplexer and the Demultiplexer circuits in the QPSK system.

3.2.2

The Word Generator

The word generator was implemented by using two ICs (4014 shift register,4516 4-bit counter), and by the referring to the data sheets for these two ICs, we were 18

CHAPTER 3. CIRCUIT DESIGN

19

Figure 3.1: Universal Clock Circuit. able to construct this logic circuit in order to have a controllable 8-bits repetitive stream of data (by 8-dip switches),with a controllable data rate obtained by changing the clocks feeding the two ICs. These clocks are 8 KHz for the FSK system, 64 KHz for the PSK system and 128 KHz for the Q-PSK system. As we said the input to the register that can be controlled in our module is by the dip switches, here the stream is 11011010 set in parallel by our dip switches and the same stream was obtained at the output but in a serial form, with a data rate specified according to the clock of the system we are operating as shown in Fig.(3.3)

3.3

FSK System

Our approach for designing the FSK modulator/demodulator was based on the fact that the output of a VCO (Voltage Controlled Oscillator), which is used as a modulator, is an FSK output, and can be detected by a PLL (Phase Locked Loop), which is used as the demodulator. The ICs used are LM566 for the VCO (U7) and LM565 for the PLL (U8). Referring to the data sheets related to these ICs, external circuits were designed according to the procedure bellow.

3.3.1

FSK Transmitter

The desired Free Running Frequency for this system is 1.024 MHz. An approach of design must be made for the external circuits connected to this IC according to the following: The word coming out from the word generator (which has a data rate of 8 KBPS for this system) has a CMOS level which is 12 V (0 V for the low level output and 12 V for the high level ), so this signal should be

CHAPTER 3. CIRCUIT DESIGN

Figure 3.2: Word Generator Circuit.

Figure 3.3: Word Signal.

20

21

CHAPTER 3. CIRCUIT DESIGN

Figure 3.4: Voltage Divider at VCO I/P.

Figure 3.5: DC Offset Circuit. attenuated to obtain an output varying between 0–0.8 V word, according to the data sheets of the LM566. The calculations for this divider are : “ ’ R12 Vo = V i (3.1) R11 + R12 substituting, 0.8 = 12

’

R12 R11 + R12

“

and by assuming R11 = 10KΩ so R12 should be 714Ω and by using the standard value 680Ω so Vo = 0.76V dc which is acceptable. Also the input should have a DC offset of at least 34 Vcc which is almost = 9 V dc to obtain the free running frequency, so a DC offset is supplied to this divider as shown in the Fig.(3.5) : For the lower divider which consists of the resistors R13 , R14 and the supply 12 V dc, the output is 9.84 V, and when implemented it was discovered that this offset would be attenuated due to loading effect so a buffer was used to eliminate this problem and the offset was added to the input word . Now the input word to the LM566 is a word that varies from 9.84 V for the low level and 10.6 V for the high level. The external circuits of the LM566 used

22

CHAPTER 3. CIRCUIT DESIGN

Figure 3.6: VCO Circuit. to obtain the desired free running frequency must be designed according to the following relationship and the former variables (input level and the DC offset) which are: “ ’ + V − V5 (3.2) fo = 2.4 RCV +

Our Free running frequency is = 1.024 MHz, V + is our supply which is = 12 V dc, V5 is 34 Vcc which is almost = 9 V dc, but it turned out that this formula does not apply for specific ranges such as the case here, so this formula must be verified. The values of R & C can be calculated and by assuming the value of C = 50 pF, R was found to be = 11.72KΩ which is wrong because when implemented results obtained were not accurate. By methodical tests and measurements, a resistor of 3.3KΩ(R16 ) in series with a pot of 1KΩ(P ot1 ) were put instead as shown in the Fig.(3.6). This leads us to verify this equation according to the frequency we are working at as in formula 3.2b: “ ’ + V −V5 fo = 0.82 (3.3) RCV + Note: The output of this IC which is an FSK output was DC blocked in our design by a small capacitor of 0.1µF (C2 ) as shown in the schematic for the transmitter and the receiver together fig (3.9). As mentioned before the output of this IC is an FSK output as shown in Fig.(3.7). We can notice from the FSK signal that the low frequency (0.905 MHz) represents the logic high while the high frequency (1.06 MHz) represents the logic low. The FSK amplitude is 6Vp−p .

3.3.2

FSK Receiver

The receiver (PLL) IC LM565 (U8 ) was used to extract the data from the FSK signal. Our PLL design was partially dependent on the formula given in the data sheets related to this IC, which is: fo =

0.3 Ro Co

(3.4)

23

CHAPTER 3. CIRCUIT DESIGN

Figure 3.7: FSK Output Signal. In our FSK system we know that the free running frequency is = 1.024 MHz, and by assuming Co = 50pF we got a value of Ro = 6.3KΩ which also turned to be not true when implemented, and by referring to the data sheets related to this IC especially to a figure entitled VCO frequency, we can see that this formula does not apply for frequencies that exceed 1 MHz. Again by methodical tests and measurements, a value of 1.1KΩ(R2 2 + P ot2 ) was put instead to obtain the lock range1 of the FSK frequencies we are working at as can be seen in Fig.(3.8).The formula can be verified as in the following relationship: fo =

0.056 Ro Co

(3.5)

Notice that the output of this circuit is applied to a high pass filter (composed of the capacitor 0.22µF (C4 ) and the resistor 100KΩ(R23 )), and by applying 1 for passive filters) which also acts as a differentiator the relation f = 2πRC with a cutoff frequency of almost 8 KHz (in our design about 7.2 KHz), this would output a remarkable inverted signal that detects the input signal, and by applying this output to a level detector(comparator U6/2 ) this gave us a signal identical to the one at the input of the modulator but inverted. This output is then applied to a low pass filter (R24 &C6 ) to get rid of high frequency components. In order to reconstruct a digital output the same as the one at the input with the same data rate and also for protection purposes, this signal is applied to a digital inverter (U9/4 ) . The whole circuit for the receiver and the transmitter is shown in Fig.(3.9):

3.4

The PSK Systems

The basic IC used in the PSK and Q-PSK systems is the Motorola MC1496 Balanced Modulator (U12 , U14 , U15 , U16 ). Here we will begin with our calculations, simulations, and the results obtained from these simulations. The multiplier is a built in component in the EWB simulation software, while in practice it’s an IC that needs design. This also makes some differences in the circuits accompanying this IC, such as leveling circuits and filters. For example, in our design 1 Lock

range: The total frequency range, within which phase lock is achieved.

CHAPTER 3. CIRCUIT DESIGN

24

Figure 3.8: PLL Circuit. the filters that worked with this IC were all passive, while in our simulations Sallen-Key active filters were essential to get an acceptable output at the receiver as shown in Fig.(3.12). In this section calculations, simulations, and the implementation will be discussed each in a separate part.

3.4.1

The PSK Transmitter

Referring to the block diagrams in the theory and with the knowledge of our data rate (which is 64 KBPS), and the carrier used for modulation and demodulation (which is 512 KHz) for coherent detection, we were able to simulate the following circuit for the PSK modulator as shown in Fig.(3.10) : This circuit consists of a simple non-inverting amplifier with a gain of 2 applied to the output of the word generator whose output varies between 0–5 V square wave, this would give a data word between 0–10 V (in our implementation the word level is already a CMOS level which is almost 10 V. Because of this, this amplifier can not be seen in the board). This output is applied to a leveling circuit (composed of -5 V dc supply which yields the word signal between -5–5 V, and then attenuated by half after the divider R59 &R60 so the output of this divider varies between -2.5–2.5 V, this can be found utilizing the superposition principle). After leveling is done, an amplification of 2 was needed to get the variation of -5–5 V word signal to be multiplied with the carrier. A digital

CHAPTER 3. CIRCUIT DESIGN

25

Figure 3.9: FSK total Circuit. input word signal (assumed here 11001010) and the PSK signal output from the multiplier are shown in Fig.(3.11): We can clearly see from Fig(3.11) the phase transitions (The PSK signal in the bottom) when the logic level goes either from high to low or from low to high. Note: In simulation, ideal components were used, and this would result in having ideal outputs which are not the case in practice, so extra circuits such as filters would be needed. Also the signals may have DC offset at the ICs’ outputs so coupling capacitors were added. Added to that, some ICs turned out to be current driven, which means that we have to add extra capacitors and resistors connected to them. That’s why the practical schematic is a little bit different than the simulated one as will be seen later.

3.4.2

The PSK Receiver

Now this output should be detected by another circuit (receiver) to be able to reconstruct the digital output again, so noise and errors should be reduced to minimum in order to obtain a clear and faithful signal as the one being transmitted. The circuit simulated for detecting the PSK signal output from the transmitter is shown in Fig.(3.12): This circuit consists of a multiplier, Sallen-Key low pass filter (in our board a passive filter consists of R89 , R90 , C21 ) and a simple comparator (U17/2 ) that

CHAPTER 3. CIRCUIT DESIGN

Figure 3.10: PSK Transmitter Circuit.

Figure 3.11: PSK output signal.

26

27

CHAPTER 3. CIRCUIT DESIGN

Figure 3.12: PSK Receiver Circuit. Gain R1 R2 R3 R4 C1

1 1.422 5.399 Open 0 0.33C

2 1.126 2.250 6.752 6.752 C

4 0.824 1.537 3.148 9.444 2C

6 0.617 2.051 3.203 16.012 2C

8 0.521 2.429 3.372 23.602 2C

10 0.462 2.742 3.560 32.038 2C

Table 3.1: Second order Low Pass Batterworth Filter design values (All resistor values are in KΩ ). works as a level detector to give a digital wave which is applied to digital gates as in the procedure done for the FSK system in order to obtain a digital level and for protection purposes. As we said before the carrier for our PSK system is 512 KHz and the data output from the demodulator should have the same data rate as the one at the input of the modulator (which is 64 KBPS), so the low pass filter at the output of this demodulator was designed as follows: Because the data rate for this system is 64 KBPS we chose 70 KHz to be the cutoff frequency. For this low pass filter to operate at this frequency, we selected a value for the capacitor of 1 nF in our design. Then, we chose the gain of the filter which is in our design a gain of 2. We applied the relationship [EM88]: 10−4 K= f Ca1 in order to calculate K. With the K constant calculated, and by referring to Table(3.1) we can calculate the values of the resistors of the filter. So by applying the above relationship K was found to be 1.428 so the values of the resistors was found as follows:

CHAPTER 3. CIRCUIT DESIGN

28

Figure 3.13: Sellen-Key Filter. Rc1 = K ∗ 1.126KΩ = 1.61KΩ Rd1 = K ∗ 2.250KΩ = 3.21KΩ Re1 = K ∗ 6.752KΩ = 9.64KΩ Re1 = K ∗ 6.752KΩ = 9.64KΩ C = Ca1 = 1nF. The original word transmitted is the same as the one obtained at the receiver with some delay and peak to peak difference, this in practice would be corrected by applying the stream to digital gates that would cause additional delay for one of the streams which would eliminate this effect, this is essential PSK and QPSK systems. The level output of these ICs is CMOS (0–12V). By comparing Fig.(3.11) and Fig.(3.14) we can notice that the input to the modulator and the output of the demodulator are identical. The same digital word 11001010 appeared at the output of the receiver with some delay . This indicates the efficiency of these two circuits. In our practical design the word is supplied from the word generator, so leveling circuit is needed as in the Fig.(3.15) to have the data wave varies between -2.5–2.5 V to the following amplifier : The -5 V dc supply was obtained from a (7905) regulator(RG1 ). Another regulator (7908) (RG2 ) was used to obtain -8 V dc, used to bias the 1496 balanced modulator. The output word would be supplied to a non-inverting amplifier with gain of 2 to give an output that varies between -5–5 V data wave connected to our modulator. A coupling capacitor is added at the output of the modulator (C15 ) to eliminate the dc component, a PSK signal is obtained at this point, then this signal is fed to another 1496 as shown in the schematic for the receiver part, then to another coupling capacitor (C10 ) and directly to a low pass filter(composed of the resistor R89 = 1.5KΩandtheC21 = 1nF ) with cutoff frequency of 106 KHz, a resistor of R90 = 100KΩ is put in parallel with the capacitor in order to increase the time constant (t), because during implementation it was found that the signal was having a decreasing dc level, so this resistor would cause slow discharging through the capacitor. The sig-

CHAPTER 3. CIRCUIT DESIGN

Figure 3.14: In/Out signals of the PSK receiver.

Figure 3.15: Leveling Circuit.

29

30

CHAPTER 3. CIRCUIT DESIGN

nal is then applied to a non-inverting amplifier (U17/1 ) with gain of 11, to a comparator(U17/2 ) which would detect the signal to give an output identical to the one at the input of the transmitter. The signal is then fed to a digital gate such as a D-flip flop(U19/1 ) to give a digital signal having the same bit duration as the one being transmitted. Another thing to mention here is that the carrier is taken directly from the function generator and should be attenuated to an appropriate level so that the balanced modulator would work properly, the level of sine wave carrier coming from the function generator is 12Vp−p , this is the level the function generator was set to from the beginning and can be noticed in figure (3.1).

3.4.3

Balanced Modulator (MC1496 IC)

This IC is used as the multiplier block of the PSK, and the QPSK systems. This IC performs modulation using the two inputs to it. One of them is the carrier signal, the other one is the information signal. The internal structure of this modulator is based on the Gilbert Cell multiplier circuit. The values of the external components connected to this IC are obtained using the equations specified in the data sheet of the MC1496. As for the signal gain for low frequencies of operation, the following equation describes the gain behavior(refer to the data sheet): As = a gain of 1 is needed, ⇒ 1 =

RL Vo = Vs Re + 2re

RL Re +2re ,

and re =

VT IE

(3.6) =

25m I5 .

For linear operation:

Vs ≤ I 5 R e

(3.7)

referring to Fig.(3.16), and assuming that I5 = I6 = I9 2 . and IB œ IC , then R5 =

V−−φ − 500 , where φ = 0.75V at 25◦ C I5

(3.8)

We will use V − as -8.2V, which is taken from the 7908 Regulator output. The input current is assumed to be 1.22 mA(since at least 1 mA should flow in the divider). Plugging these values in Eq.(3.8) yields: R5 =

8.2 − 0.75 − 500 = 5.6KΩ 1.22m

⇒ V5 = −6.8V = VBQ7 = VBQ8 ⇒ IQ7 = IQ8 = IQ9 For Biasing, the following equations must be satisfied: 30V ≥ [(V6 , V12 ) − (V7 , V8 )] ≥ 2V 30V ≥ [(V7 , V8 ) − (V1 , V4 )] ≥ 2.7V 30V ≥ [(V1 , V4 ) − (V5 )] ≥ 2.7V

(3.9)

here V6 ≈ V12 , V7 ≈ V8 , V1 ≈ V4 . If we use 2.7kΩ Load resistors at pins 6 and 12, then

2 For

VCQ6 = VCQ12 = 12V − (2.7k × 1.22m) = 8.706V

a better circuit diagram, refer to the accompanied data sheet for the MC1496 IC

31

CHAPTER 3. CIRCUIT DESIGN

Using a divider network between VCC Pin 8 and the Ground, the voltage at pin 8 will be: V8 = 12 × 0.5 = 6V Since pin 8 and pin 10 are connected together via a 51Ω resistor, then V 8 ≈ V 10. By this we satisfy the first equation in Eq.(3.9). To satisfy the second equation in Eq.(3.9), we have V8 ≈ V10 = 6V 51 = −0.017V V1 = V4 = −8.2 × 25k ⇒ second equation is satisfied. Finally to satisfy the third equation in Eq.(3.9), we have V5 = −1.22m × 5.6k = −6.8V Substituting in the third equation ⇒ 30V > [3 − −6.8] > 2.7V . Now as for C1 and C2 , their reactance should be no more than 5Ω at 1.024 MHz. 1 Zc1 = Zc2 = (3.10) 2πf C ⇒ C1 = C2 =

1 2π × 5 × 1024

here we chose C2 = C1 = 47nF . Going back to Eq.(3.6), we have 1 =

RL RE +2re ,

re =

25m I5

= 20.5Ω.

RE = RL − 2re = 2.7k − 41 ≈ 2.7kΩ . By this we come to the end of the design of the MC1496 Balanced Modulator. Now we can add these components to the IC to obtain the modulator we are looking for. See Fig.(3.17) for a complete circuit schematic. One more thing to mention is that the capacitor values were designed for a carrier of 1.024 MHz, but after we implemented and tested the performance of the circuit for the QPSK system, it turned out that a frequency of 512 KHz for the carrier will have a much better performance than the 1.024 MHz, so we fixed our carrier at 512 KHz, and the capacitor values were not changed. The circuit actually worked well.

3.5

The QPSK System

The Q-PSK system consists of two PSK systems with data input demultiplexed at the transmitter, carriers phase shifted between the two streams (odd and even one) by 90◦ at the transmitter and the receiver, and output multiplexed at the receiver. We used this fact and took advantage of the presence of our implemented PSK stream (considered as the odd stream), so another one was added (the even stream), and by using a switch we were able to choose either the PSK system or the Q-PSK system. Also to be mentioned that the input data stream which has the rate of 128 KBPS must be demultiplexed in a way to get odd and even streams each of 64 KBPS rate, at the input of modulator, and multiplexed at the output of the demodulator to get the 128 KBPS rate, which means that the word is reconstructed again. So we need additional circuits which would be discussed .

CHAPTER 3. CIRCUIT DESIGN

Figure 3.16: Gilbert Cell

32

CHAPTER 3. CIRCUIT DESIGN

Figure 3.17: MC1496 IC schematic

33

CHAPTER 3. CIRCUIT DESIGN

34

Figure 3.18: Phase shift circuit

3.5.1

The Phase Shifter

We will begin with the phase shifter, which consists of two consecutive low pass filters. The transfer function of the two consecutive low pass filters is given by: H(s) =

1 (1 + SR55 C12 )(1 + SR56 C13 )

(3.11)

This transfer function has two poles, the first one occurs at R551C12 and the other at R561C13 . A Bode Plot of a low pass filter illustrates a 45◦ phase shift between the input and the output at the pole. Since we have two consecutive low pass filters, this means that the output signal of such a combination will be 90◦ phase shifted from the input signal. We chose R55 = R56 = 680Ω, and C12 = C13 = 1nF , this gave us a double pole at 1.4 MHz. A remarkably attenuated sine wave of 2.1Vp−p with almost 90◦ phase shift was obtained. The output of this phase shifter was connected to one stream as is (the even stream), while the other carrier input was attenuated to a level (by the divider R57 = 1.5kΩ and R58 = 330Ω as shown on the schematic) almost 15 of the one from the function generator to get a carrier equal to the one at the output of this phase shifter, which would be then applied to the other stream (the odd stream). The carrier applied from the function generator was 12Vp − p, and the output was 2.1Vp−p after the phase shifter (when measured it was found 1.2V dc because of loading effect when connected to the multiplier). This was proved by the former simulation for the phase shifter, at 512 KHz the attenuation was almost 15 dB on the bode plotter while in calculations was as follows: dB = 20Log(

Vi 12 ) = 20Log( ) = 15.1dB Vo 2.1

so the voltage divider must output a signal almost identical to the one at the output of this phase shifter. In the Q-PSK simulation, there was no need to simulate phase shifting because we were able to apply two cosine waves with phase shift of 90◦ at the multiplier as can be shown in Figures (3.24) and (3.26).

3.5.2

The Demultiplexer and Multiplexer circuits

Two additional circuits were added to this system, one was the demultiplexer and the other was the multiplexer circuit. These two circuits were used at the transmitter and the receiver respectively. The main function of these circuits is for serial to parallel conversion at the transmitter of the incoming data stream to be applied to the separate paths Q-PSK system, and the other serves as the

CHAPTER 3. CIRCUIT DESIGN

35

Figure 3.19: Simulated Demux Circuit

Figure 3.20: Implemented Demux circuit parallel to serial conversion at the receiver .This can be illustrated from the demultiplexer circuit in Fig.(3.19) and the multiplexer circuit in Fig.(3.22) As we said before data for the Q-PSK system needs to be demultiplexed at the input of the transmitter. The input data rate is 128 KBPS for a word 01101000 assumed for simulation. The thing needed here is to have two streams even and odd one each of 64 KBPS as follows : The even stream is 1000 and the odd one is 0110. We can notice that the clock for the upper flip flop which is 64 KHz is the inverted clock of the lower one, this caused the input word to be separated into two different streams each of 64 KBPS. Simulated output proved that the even stream lags the odd one with a bit duration of 7.8µs (half a bit), and to have the first bit of the even stream arrive to the upper multiplier at the same instant the first bit of the odd one does, an additional flip flop with a clock of 128 KHz was added to the lower stream to correct this delay other wise incorrect results would be obtained at the output of the Q-PSK transmitter as can be seen in Fig.(3.19). Another thing, in our board the last flip flop’s clock was of 64 KHz added to the upper stream not to the lower one, this is because in practice the delay was found to be of bigger duration than simulation (delay of 1 bit) in the odd stream not in the even one(so the even stream should be delayed) as can be seen in Fig.(3.20) which might be caused by the components in the circuit which would never work ideally as in simulation. Notice from Fig.(3.21): 1. The first two outputs which are the clocks, each of 64 KHz with inverted

CHAPTER 3. CIRCUIT DESIGN

36

Figure 3.21: The CLK,CLK, Word, Flipflop1(even) and flipflop2(odd) outputs for the simulated demultiplexer.

Figure 3.22: The Mux circuit wave forms. 2. The third output which is the input word 01101000 of 128 KHz. 3. Also we can notice from the last two outputs that they both begin at the same instant of time (at the first positive edge of the inverted clock ). Before the addition of the delay flip flop to the lower stream, the flipflop2’s output (the fifth wave form) appeared half a bit earlier, (at the first positive edge of the clock) this is why this flip flop of clock 128 KHz was added to the lower one. The even stream output 1000 and the odd stream output 0110 each of 64 KBPS both begins now at the first positive edge of the inverted clock as desired to be transmitted by the transmitter. Now, these even and odd streams must be reconstructed again at the receiver after the demodulators as can be seen later in the receiver part, but let’s assume for the while that the same streams were obtained at these specific outputs (which are1000 for even and 0110 for odd) for illustration purposes, a circuit called multiplexer was used to mix these two streams into one (parallel to serial conversion) to obtain the original input serial word from the word generator which is 01101000 as can be seen in Fig.(3.22). We can notice from Fig.(3.23)

Figure 3.23: Word in to demultiplexer, odd, even, word out from multiplexer streams

CHAPTER 3. CIRCUIT DESIGN

37

Figure 3.24: The QPSK Transmitter Circuit that the input word at the input of the demultiplexer was reconstructed again at the output of the multiplexer. This proves the efficiency of the demultiplexer and multiplexer circuits. Also from Fig.(3.22) notice that a flip flop U22/1 with a clock of 128 KHz was added at the output of the demultiplexer, this was needed here to correct the bit duration and the wave form at the output. Also two additional flip flops U19/1 and U19/2 with clocks of 64 KHz were added at the output of each stream’s level detector before being multiplexed for the same purpose, this will be shown later in the Q-PSK receiver schematic Fig.(3.26).

3.5.3

The QPSK Transmitter

Now we can make use of our implemented PSK transmitter as we did for the odd stream in our implementation, and build another one for the even stream then to sum the outputs of the demodulators with the carriers 90◦ phase shifted between the two streams as shown in Fig.(3.24), at this point a Q-PSK signal is obtained as in Fig.(3.25). We can see that the Q-PSK consists of two different PSK systems demultiplexed by flip flops as illustrated in sub section 3.5.2 to have odd and even streams each modulated and summed by a simple summer as shown.(The opamp we used for the summing function in our module is a special IC the LM359 (U13 ) with high gain bandwidth product followed by a buffer (U11/2 ) to illuminate loading effect. Notice also that all the amplifiers used in this circuit are TL072 low noise operational amplifiers with higher gain bandwidth product than the LM741, but at this point a higher GBWP was needed). The output obtained is a Q-PSK output as shown in Fig.(3.25). Circuits

CHAPTER 3. CIRCUIT DESIGN

38

Figure 3.25: The QPSK Transmitter Signal used for the Q-PSK system after the demultiplexer are the same ones used for the PSK system, so there is no need to discuss them again. To be mentioned here is the difference in leveling circuits, in PSK simulations we amplified the signal coming out from the word generator (TTL word) to get a level of 10 V data word because in implementation the word level is also of CMOS level, here in the Q-PSK system the flip flops used (in simulation) for the demultiplexer are TTL flip flops so the output level is a TTL level which is 5 V, leveling should be made according to this output which is not the case in implementation where the output level of the flip flops used are of CMOS level . In other words in our implementation the DC level was -5 V dc needed after the CMOS flip flops which is the same in PSK circuits while in simulation for the Q-PSK system the level is -2.5 V dc needed after the TTL flip flops . By referring to the data sheets of the IC LM359 ,it was found that capacitors of 1 nF each should be connected at the inputs of the summer (C16 , C17 ), and during implementation a small capacitor of 100 pF (C18 ) was connected to ground at the output in order to voltage drive this output (current driven output) and not enough to drive the following buffer (U11/2 ).

3.5.4

The QPSK Receiver

The output of the summer was applied to two streams of PSK receivers whose carriers are also shifted by 90◦ , and then multiplexed at the output to be able to reconstruct the same digital word transmitted. The receiver circuit is illustrated in Fig.(3.26). Fig.(3.27) illustrates the Q-PSK signal at the input of receiver and the digital signal obtained at the output . Fig.(3.28) shows the input word at the transmitter and the output word from the receiver which are identical, this proves the efficiency of these circuits. As we said before circuits here do not need illustration again because they were all discussed in the PSK receiver and the multiplexer subsections. Two flip flops (U19/1 and U19/2 ) of 64 KHz were added at the output of the circuit shown in Fig.(3.26) before the multiplexer circuit to overcome bit duration and timing problems. An extra flip flop Uc of 64 KHz appears here in the schematic, this is because while simulation a delay of 1 bit to the even stream was needed to enable the odd stream arrive to the multiplexer circuit 1 bit earlier to mix together in a way that would reconstruct the input word again with the rate of 128 KBPS, (this Uc was not used when implemented because the odd stream was already delayed by 1 bit which is needed and inverted so in

CHAPTER 3. CIRCUIT DESIGN

Figure 3.26: The QPSK receiver circuit

Figure 3.27: QPSK receiver signals

39

CHAPTER 3. CIRCUIT DESIGN

40

Figure 3.28: The final output and the initial input implementation the odd output was inverted by U9/3 only). As can be seen from the former figures, the Q-PSK transmitter/receiver circuits were simulated successfully and also worked fine when implemented. The original input word transmitted at the input of the transmitter was the same received at the output of the receiver with a delay of four bits here for our approach of simulations.

3.5.5

PCB Design and Final Circuit Schematic

The PCB design was done using a software called PROTEL3 for Windows version 1.5. Some limitations were taken into consideration such as the size of our module and the position of supply tracks needed for the circuit which are 12 V,-12 V and the ground. This is because we had to design our module according to the power supply panel base available in the communications lab. Tests were made to insure this power supply is sufficient for the circuit. We connected our circuit to it, same results were obtained. We tried to organize our module in a way that each system is in a separate block, 7 sections for 7 major circuits can be easily noticed. Another thing to mention is that we designed a double side board, this is due to the number of ICs and the size we are limited to. The circuit schematic, circuit layout, and the PCB print out shown in Figures (3.29),(3.30) and (3.31,3.32) respectively will clearly illustrate this. Switches appears on the schematic are Sw1, Sw2, Sw3, Sw4, Sw5, Sw6. Switches 4,5,6 are for the option of connecting or disconnecting the receivers of the FSK, the odd, and the even streams respectively from their transmitters. Sw1 is an 8-bit dip switches which determines the word needed to be generated from the word generator. Sw2 is also an 8-bit dip switch whose outputs are the clocks generated from the universal clock. These outputs from the left are (2, 4, 8, 16, 256, 128, 64, 32 KHz). Sw3 is a 4-bit dip switches which determines which system would operate at the time and disconnects the other systems at this moment in the following combinations From Left : • Up, Down, Down, Down : This would enable the FSK system and disconnects the PSK input streams. • Down, Down, Down, Up : This would enable the PSK system (only the odd stream) and disconnects the FSK and the even input streams. 3 c

Protel PCB and Schematic are registered programs for Protel Technologies.

CHAPTER 3. CIRCUIT DESIGN

41

• Down, Up, Up, Down : This would enable the Q-PSK system (odd and even streams) and disconnects the FSK input stream.

CHAPTER 3. CIRCUIT DESIGN

Figure 3.29: The Circuit Schematic of the Module.

42

CHAPTER 3. CIRCUIT DESIGN

Figure 3.30: The Circuit IC and Component Layout.

43

CHAPTER 3. CIRCUIT DESIGN

Figure 3.31: The PCB Layout, Layer (1).

44

CHAPTER 3. CIRCUIT DESIGN

Figure 3.32: The PCB Layout, Layer (2).

45

Chapter 4

Results This chapter will demonstrate the final results and output waveforms obtained from the system we designed. This chapter will be divided into three sections, one will illustrate the output signals obtained from the FSK system, the other for the PSK, and the final one will demonstrate the QPSK waveforms. These graphs were taken using a Tektronix Digital Oscilloscope.

Results are not included in this report version. For results, please contact the Authors!

46

Chapter 5

Conclusion and Further enhancements 5.1

Conclusion

Our project was successfully completed with all the circuits functioning properly. The design demonstrated three of the most common digital modulation techniques used these days in communications. The circuits were first built and tested on a breadboard, then we designed a PCB and attached the components to it. It worked successfully. Our main aim was to demonstrate these techniques for educational purposes. We think that our design is reliable, and is intended to be used in the digital communication Lab for Educational purposes. That is based on the fact the circuits were built from commercial ICs that are familiar to all students. Also the PCB was designed in a way to be placed on an existing Power Supply base in the lab that is used for other modules, which mean that such a board will be easily set up in the lab. The PCB was designed with great care in order to have the least chance of having routing errors that might affect the system performance. Circuit design with the least amount of components was utilized. Designing this system involved many thoughts. These included researching ideas, and components, considering simplicity, efficiency, reliability, coast, and size. We think that the system as a whole was a total success!

5.2

Further Enhancements

Our design was based on the Coherent Detection scheme. That is the carrier is known in advance by the receiver. In real life systems, this is not the case. Non-coherent detection schemes are used in real life receivers. An enhancement on this project is to design non-coherent receivers for the three systems, and to put them on another PCB. This will allow the student to take a broader look at the no-ncoherent detection systems and circuit for Digital Modulation techniques. One other improvement on this project might be to design an efficient discrete multiplier circuit for the PSK and QPSK streams. In the early stages of our project we were searching and testing the applicability of designing a discrete Multiplier circuit. Unfortunately we didn’t have much time to conduct a 47

CHAPTER 5. CONCLUSION AND FURTHER ENHANCEMENTS

48

lot of tests, so we searched for an IC to do the job. This is another enhancement that might be applicable.

Bibliography [Bla93] R. Blake. Basic Electronic Communication. West Publishing, second edition, 1993. [Bla97] Richard Blahut. Introduction to Telecommunication Theory. University of Illinois at Urbana-Champaign, ECE 364 class notes, 1997. [Bog90] T. F. Bogart. Electronic Devices and Circuits. Merrill Publishing, second edition, 1990. [EM88] E. I. El-Masri. Analog Filter Design. Technical University of Nova Scotia, 1988. [Fra98] S. J. Franke. Radio Communication Circuits. University of Illinois at Urbana-Champaign, ECE 353 class notes, 1998. [Hay88] Simon Haykin. Digital Communications. John Wiley and Sons, second edition, 1988. [Hay94] Simon Haykin. Communication Systems. John Wiley and Sons, third edition, 1994. [Rap96] Theodore Rapparport. Wireless Communications, Principles and practice. Prentice–Hall Inc., third edition, 1996. [Raz98] Behzad Razavi. RF Microelectronics. Prentice–Hall Inc., first edition, 1998. [Skl88] Bernard Sklar. Digital Communications, Fundamentals and applications. Prentice–Hall Inc., third edition, 1988. [SS98]

A. Sedra and K. Smith. Microelectronic Circuits. Oxford University Press, third edition, 1998.

49